diff --git "a/checkpoint-47000/trainer_state.json" "b/checkpoint-47000/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-47000/trainer_state.json" @@ -0,0 +1,329033 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.47, + "eval_steps": 500, + "global_step": 47000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 1e-05, + "grad_norm": 1.2723733186721802, + "learning_rate": 3e-06, + "loss": 10.8324, + "step": 1 + }, + { + "epoch": 2e-05, + "grad_norm": 1.2627283334732056, + "learning_rate": 6e-06, + "loss": 10.8325, + "step": 2 + }, + { + "epoch": 3e-05, + "grad_norm": 1.2797267436981201, + "learning_rate": 9e-06, + "loss": 10.8328, + "step": 3 + }, + { + "epoch": 4e-05, + "grad_norm": 1.2568743228912354, + "learning_rate": 1.2e-05, + "loss": 10.8307, + "step": 4 + }, + { + "epoch": 5e-05, + "grad_norm": 1.26583731174469, + "learning_rate": 1.5e-05, + "loss": 10.8273, + "step": 5 + }, + { + "epoch": 6e-05, + "grad_norm": 1.268788456916809, + "learning_rate": 1.8e-05, + "loss": 10.8268, + "step": 6 + }, + { + "epoch": 7e-05, + "grad_norm": 1.215290904045105, + "learning_rate": 2.1000000000000002e-05, + "loss": 10.8151, + "step": 7 + }, + { + "epoch": 8e-05, + "grad_norm": 1.1221675872802734, + "learning_rate": 2.4e-05, + "loss": 10.794, + "step": 8 + }, + { + "epoch": 9e-05, + "grad_norm": 1.0951769351959229, + "learning_rate": 2.7e-05, + "loss": 10.7888, + "step": 9 + }, + { + "epoch": 0.0001, + "grad_norm": 1.105851650238037, + "learning_rate": 3e-05, + "loss": 10.7765, + "step": 10 + }, + { + "epoch": 0.00011, + "grad_norm": 1.0884467363357544, + "learning_rate": 3.2999999999999996e-05, + "loss": 10.763, + "step": 11 + }, + { + "epoch": 0.00012, + "grad_norm": 1.0829719305038452, + "learning_rate": 3.6e-05, + "loss": 10.7508, + "step": 12 + }, + { + "epoch": 0.00013, + "grad_norm": 1.0506291389465332, + "learning_rate": 3.9e-05, + "loss": 10.7324, + "step": 13 + }, + { + "epoch": 0.00014, + "grad_norm": 1.037864089012146, + "learning_rate": 4.2000000000000004e-05, + "loss": 10.7208, + "step": 14 + }, + { + "epoch": 0.00015, + "grad_norm": 1.0115288496017456, + "learning_rate": 4.4999999999999996e-05, + "loss": 10.7117, + "step": 15 + }, + { + "epoch": 0.00016, + "grad_norm": 0.9676744341850281, + "learning_rate": 4.8e-05, + "loss": 10.6952, + "step": 16 + }, + { + "epoch": 0.00017, + "grad_norm": 0.9355509877204895, + "learning_rate": 5.1000000000000006e-05, + "loss": 10.6792, + "step": 17 + }, + { + "epoch": 0.00018, + "grad_norm": 0.9286826848983765, + "learning_rate": 5.4e-05, + "loss": 10.6649, + "step": 18 + }, + { + "epoch": 0.00019, + "grad_norm": 0.9110698699951172, + "learning_rate": 5.7e-05, + "loss": 10.6512, + "step": 19 + }, + { + "epoch": 0.0002, + "grad_norm": 0.9136782288551331, + "learning_rate": 6e-05, + "loss": 10.6404, + "step": 20 + }, + { + "epoch": 0.00021, + "grad_norm": 0.8971966505050659, + "learning_rate": 6.3e-05, + "loss": 10.6274, + "step": 21 + }, + { + "epoch": 0.00022, + "grad_norm": 0.8972620368003845, + "learning_rate": 6.599999999999999e-05, + "loss": 10.6136, + "step": 22 + }, + { + "epoch": 0.00023, + "grad_norm": 0.8984483480453491, + "learning_rate": 6.9e-05, + "loss": 10.6016, + "step": 23 + }, + { + "epoch": 0.00024, + "grad_norm": 0.8967456817626953, + "learning_rate": 7.2e-05, + "loss": 10.5894, + "step": 24 + }, + { + "epoch": 0.00025, + "grad_norm": 0.8972211480140686, + "learning_rate": 7.500000000000001e-05, + "loss": 10.5752, + "step": 25 + }, + { + "epoch": 0.00026, + "grad_norm": 0.9028114080429077, + "learning_rate": 7.8e-05, + "loss": 10.5611, + "step": 26 + }, + { + "epoch": 0.00027, + "grad_norm": 0.8926876783370972, + "learning_rate": 8.1e-05, + "loss": 10.5491, + "step": 27 + }, + { + "epoch": 0.00028, + "grad_norm": 0.8921052813529968, + "learning_rate": 8.400000000000001e-05, + "loss": 10.536, + "step": 28 + }, + { + "epoch": 0.00029, + "grad_norm": 0.8942669034004211, + "learning_rate": 8.7e-05, + "loss": 10.5219, + "step": 29 + }, + { + "epoch": 0.0003, + "grad_norm": 0.9005073308944702, + "learning_rate": 8.999999999999999e-05, + "loss": 10.5056, + "step": 30 + }, + { + "epoch": 0.00031, + "grad_norm": 0.8994124531745911, + "learning_rate": 9.3e-05, + "loss": 10.491, + "step": 31 + }, + { + "epoch": 0.00032, + "grad_norm": 0.8968571424484253, + "learning_rate": 9.6e-05, + "loss": 10.4763, + "step": 32 + }, + { + "epoch": 0.00033, + "grad_norm": 0.8976972103118896, + "learning_rate": 9.900000000000001e-05, + "loss": 10.4597, + "step": 33 + }, + { + "epoch": 0.00034, + "grad_norm": 0.8977769017219543, + "learning_rate": 0.00010200000000000001, + "loss": 10.4427, + "step": 34 + }, + { + "epoch": 0.00035, + "grad_norm": 0.902169406414032, + "learning_rate": 0.00010500000000000002, + "loss": 10.4252, + "step": 35 + }, + { + "epoch": 0.00036, + "grad_norm": 0.8990501165390015, + "learning_rate": 0.000108, + "loss": 10.4079, + "step": 36 + }, + { + "epoch": 0.00037, + "grad_norm": 0.8933607935905457, + "learning_rate": 0.000111, + "loss": 10.39, + "step": 37 + }, + { + "epoch": 0.00038, + "grad_norm": 0.8925058245658875, + "learning_rate": 0.000114, + "loss": 10.3704, + "step": 38 + }, + { + "epoch": 0.00039, + "grad_norm": 0.8942745923995972, + "learning_rate": 0.000117, + "loss": 10.3512, + "step": 39 + }, + { + "epoch": 0.0004, + "grad_norm": 0.8984111547470093, + "learning_rate": 0.00012, + "loss": 10.3286, + "step": 40 + }, + { + "epoch": 0.00041, + "grad_norm": 0.8943851590156555, + "learning_rate": 0.000123, + "loss": 10.3097, + "step": 41 + }, + { + "epoch": 0.00042, + "grad_norm": 0.8935915231704712, + "learning_rate": 0.000126, + "loss": 10.2894, + "step": 42 + }, + { + "epoch": 0.00043, + "grad_norm": 0.8975799679756165, + "learning_rate": 0.000129, + "loss": 10.2654, + "step": 43 + }, + { + "epoch": 0.00044, + "grad_norm": 0.8982045650482178, + "learning_rate": 0.00013199999999999998, + "loss": 10.2433, + "step": 44 + }, + { + "epoch": 0.00045, + "grad_norm": 0.9000449180603027, + "learning_rate": 0.000135, + "loss": 10.2204, + "step": 45 + }, + { + "epoch": 0.00046, + "grad_norm": 0.8900250792503357, + "learning_rate": 0.000138, + "loss": 10.1983, + "step": 46 + }, + { + "epoch": 0.00047, + "grad_norm": 0.8965498805046082, + "learning_rate": 0.000141, + "loss": 10.1723, + "step": 47 + }, + { + "epoch": 0.00048, + "grad_norm": 0.8975719213485718, + "learning_rate": 0.000144, + "loss": 10.149, + "step": 48 + }, + { + "epoch": 0.00049, + "grad_norm": 0.8933398127555847, + "learning_rate": 0.000147, + "loss": 10.1239, + "step": 49 + }, + { + "epoch": 0.0005, + "grad_norm": 0.8988479375839233, + "learning_rate": 0.00015000000000000001, + "loss": 10.0972, + "step": 50 + }, + { + "epoch": 0.00051, + "grad_norm": 0.8983374834060669, + "learning_rate": 0.000153, + "loss": 10.0711, + "step": 51 + }, + { + "epoch": 0.00052, + "grad_norm": 0.8959178328514099, + "learning_rate": 0.000156, + "loss": 10.0437, + "step": 52 + }, + { + "epoch": 0.00053, + "grad_norm": 0.8871957063674927, + "learning_rate": 0.000159, + "loss": 10.0204, + "step": 53 + }, + { + "epoch": 0.00054, + "grad_norm": 0.9051761627197266, + "learning_rate": 0.000162, + "loss": 9.9878, + "step": 54 + }, + { + "epoch": 0.00055, + "grad_norm": 0.8952219486236572, + "learning_rate": 0.000165, + "loss": 9.963, + "step": 55 + }, + { + "epoch": 0.00056, + "grad_norm": 0.890164315700531, + "learning_rate": 0.00016800000000000002, + "loss": 9.9341, + "step": 56 + }, + { + "epoch": 0.00057, + "grad_norm": 0.8922548890113831, + "learning_rate": 0.000171, + "loss": 9.9069, + "step": 57 + }, + { + "epoch": 0.00058, + "grad_norm": 0.8901249766349792, + "learning_rate": 0.000174, + "loss": 9.882, + "step": 58 + }, + { + "epoch": 0.00059, + "grad_norm": 0.8989579677581787, + "learning_rate": 0.000177, + "loss": 9.8502, + "step": 59 + }, + { + "epoch": 0.0006, + "grad_norm": 0.8829832673072815, + "learning_rate": 0.00017999999999999998, + "loss": 9.8242, + "step": 60 + }, + { + "epoch": 0.00061, + "grad_norm": 0.8862175345420837, + "learning_rate": 0.000183, + "loss": 9.7955, + "step": 61 + }, + { + "epoch": 0.00062, + "grad_norm": 0.8893216848373413, + "learning_rate": 0.000186, + "loss": 9.7648, + "step": 62 + }, + { + "epoch": 0.00063, + "grad_norm": 0.8881028294563293, + "learning_rate": 0.000189, + "loss": 9.7373, + "step": 63 + }, + { + "epoch": 0.00064, + "grad_norm": 0.8868633508682251, + "learning_rate": 0.000192, + "loss": 9.7068, + "step": 64 + }, + { + "epoch": 0.00065, + "grad_norm": 0.8924434185028076, + "learning_rate": 0.00019500000000000002, + "loss": 9.6743, + "step": 65 + }, + { + "epoch": 0.00066, + "grad_norm": 0.8872416019439697, + "learning_rate": 0.00019800000000000002, + "loss": 9.6503, + "step": 66 + }, + { + "epoch": 0.00067, + "grad_norm": 0.8866419196128845, + "learning_rate": 0.000201, + "loss": 9.62, + "step": 67 + }, + { + "epoch": 0.00068, + "grad_norm": 0.8931677937507629, + "learning_rate": 0.00020400000000000003, + "loss": 9.5881, + "step": 68 + }, + { + "epoch": 0.00069, + "grad_norm": 0.879610002040863, + "learning_rate": 0.00020700000000000002, + "loss": 9.5597, + "step": 69 + }, + { + "epoch": 0.0007, + "grad_norm": 0.8912403583526611, + "learning_rate": 0.00021000000000000004, + "loss": 9.5236, + "step": 70 + }, + { + "epoch": 0.00071, + "grad_norm": 0.8847392797470093, + "learning_rate": 0.00021299999999999997, + "loss": 9.4965, + "step": 71 + }, + { + "epoch": 0.00072, + "grad_norm": 0.8860267996788025, + "learning_rate": 0.000216, + "loss": 9.4698, + "step": 72 + }, + { + "epoch": 0.00073, + "grad_norm": 0.886963963508606, + "learning_rate": 0.00021899999999999998, + "loss": 9.4354, + "step": 73 + }, + { + "epoch": 0.00074, + "grad_norm": 0.8829045295715332, + "learning_rate": 0.000222, + "loss": 9.4089, + "step": 74 + }, + { + "epoch": 0.00075, + "grad_norm": 0.8822581768035889, + "learning_rate": 0.000225, + "loss": 9.3746, + "step": 75 + }, + { + "epoch": 0.00076, + "grad_norm": 0.8918945789337158, + "learning_rate": 0.000228, + "loss": 9.3485, + "step": 76 + }, + { + "epoch": 0.00077, + "grad_norm": 0.8851014375686646, + "learning_rate": 0.000231, + "loss": 9.3206, + "step": 77 + }, + { + "epoch": 0.00078, + "grad_norm": 0.8782386183738708, + "learning_rate": 0.000234, + "loss": 9.2909, + "step": 78 + }, + { + "epoch": 0.00079, + "grad_norm": 0.8847852349281311, + "learning_rate": 0.00023700000000000001, + "loss": 9.2539, + "step": 79 + }, + { + "epoch": 0.0008, + "grad_norm": 0.8866625428199768, + "learning_rate": 0.00024, + "loss": 9.2184, + "step": 80 + }, + { + "epoch": 0.00081, + "grad_norm": 0.8906494975090027, + "learning_rate": 0.00024300000000000002, + "loss": 9.1857, + "step": 81 + }, + { + "epoch": 0.00082, + "grad_norm": 0.8960816264152527, + "learning_rate": 0.000246, + "loss": 9.1611, + "step": 82 + }, + { + "epoch": 0.00083, + "grad_norm": 0.8946034908294678, + "learning_rate": 0.00024900000000000004, + "loss": 9.1247, + "step": 83 + }, + { + "epoch": 0.00084, + "grad_norm": 0.8957815170288086, + "learning_rate": 0.000252, + "loss": 9.0939, + "step": 84 + }, + { + "epoch": 0.00085, + "grad_norm": 0.8891732692718506, + "learning_rate": 0.000255, + "loss": 9.0702, + "step": 85 + }, + { + "epoch": 0.00086, + "grad_norm": 0.8951762318611145, + "learning_rate": 0.000258, + "loss": 9.0346, + "step": 86 + }, + { + "epoch": 0.00087, + "grad_norm": 0.8848313093185425, + "learning_rate": 0.000261, + "loss": 9.0125, + "step": 87 + }, + { + "epoch": 0.00088, + "grad_norm": 0.8860167860984802, + "learning_rate": 0.00026399999999999997, + "loss": 8.9751, + "step": 88 + }, + { + "epoch": 0.00089, + "grad_norm": 0.8798444271087646, + "learning_rate": 0.000267, + "loss": 8.9486, + "step": 89 + }, + { + "epoch": 0.0009, + "grad_norm": 0.8837233185768127, + "learning_rate": 0.00027, + "loss": 8.9221, + "step": 90 + }, + { + "epoch": 0.00091, + "grad_norm": 0.879225492477417, + "learning_rate": 0.000273, + "loss": 8.8916, + "step": 91 + }, + { + "epoch": 0.00092, + "grad_norm": 0.8843948245048523, + "learning_rate": 0.000276, + "loss": 8.8596, + "step": 92 + }, + { + "epoch": 0.00093, + "grad_norm": 0.882838785648346, + "learning_rate": 0.000279, + "loss": 8.8297, + "step": 93 + }, + { + "epoch": 0.00094, + "grad_norm": 0.8830418586730957, + "learning_rate": 0.000282, + "loss": 8.8034, + "step": 94 + }, + { + "epoch": 0.00095, + "grad_norm": 0.8770962357521057, + "learning_rate": 0.000285, + "loss": 8.7743, + "step": 95 + }, + { + "epoch": 0.00096, + "grad_norm": 0.8804563879966736, + "learning_rate": 0.000288, + "loss": 8.7444, + "step": 96 + }, + { + "epoch": 0.00097, + "grad_norm": 0.8753241300582886, + "learning_rate": 0.000291, + "loss": 8.7211, + "step": 97 + }, + { + "epoch": 0.00098, + "grad_norm": 0.8762865662574768, + "learning_rate": 0.000294, + "loss": 8.6826, + "step": 98 + }, + { + "epoch": 0.00099, + "grad_norm": 0.8762408494949341, + "learning_rate": 0.000297, + "loss": 8.6601, + "step": 99 + }, + { + "epoch": 0.001, + "grad_norm": 0.8741625547409058, + "learning_rate": 0.00030000000000000003, + "loss": 8.6324, + "step": 100 + }, + { + "epoch": 0.00101, + "grad_norm": 0.8789051175117493, + "learning_rate": 0.00030300000000000005, + "loss": 8.5981, + "step": 101 + }, + { + "epoch": 0.00102, + "grad_norm": 0.8656250834465027, + "learning_rate": 0.000306, + "loss": 8.5772, + "step": 102 + }, + { + "epoch": 0.00103, + "grad_norm": 0.8743636012077332, + "learning_rate": 0.000309, + "loss": 8.5519, + "step": 103 + }, + { + "epoch": 0.00104, + "grad_norm": 0.8708215951919556, + "learning_rate": 0.000312, + "loss": 8.5298, + "step": 104 + }, + { + "epoch": 0.00105, + "grad_norm": 0.8859707713127136, + "learning_rate": 0.000315, + "loss": 8.5033, + "step": 105 + }, + { + "epoch": 0.00106, + "grad_norm": 0.9068981409072876, + "learning_rate": 0.000318, + "loss": 8.4736, + "step": 106 + }, + { + "epoch": 0.00107, + "grad_norm": 0.9245584011077881, + "learning_rate": 0.000321, + "loss": 8.4478, + "step": 107 + }, + { + "epoch": 0.00108, + "grad_norm": 0.9128947257995605, + "learning_rate": 0.000324, + "loss": 8.4188, + "step": 108 + }, + { + "epoch": 0.00109, + "grad_norm": 0.853670060634613, + "learning_rate": 0.000327, + "loss": 8.3963, + "step": 109 + }, + { + "epoch": 0.0011, + "grad_norm": 0.8760496377944946, + "learning_rate": 0.00033, + "loss": 8.3734, + "step": 110 + }, + { + "epoch": 0.00111, + "grad_norm": 0.9078761339187622, + "learning_rate": 0.000333, + "loss": 8.3444, + "step": 111 + }, + { + "epoch": 0.00112, + "grad_norm": 0.866322934627533, + "learning_rate": 0.00033600000000000004, + "loss": 8.3207, + "step": 112 + }, + { + "epoch": 0.00113, + "grad_norm": 0.8490086197853088, + "learning_rate": 0.000339, + "loss": 8.2796, + "step": 113 + }, + { + "epoch": 0.00114, + "grad_norm": 0.8713237047195435, + "learning_rate": 0.000342, + "loss": 8.2694, + "step": 114 + }, + { + "epoch": 0.00115, + "grad_norm": 0.8709179162979126, + "learning_rate": 0.00034500000000000004, + "loss": 8.2404, + "step": 115 + }, + { + "epoch": 0.00116, + "grad_norm": 0.8300504684448242, + "learning_rate": 0.000348, + "loss": 8.214, + "step": 116 + }, + { + "epoch": 0.00117, + "grad_norm": 0.8302497863769531, + "learning_rate": 0.000351, + "loss": 8.1818, + "step": 117 + }, + { + "epoch": 0.00118, + "grad_norm": 0.8652266263961792, + "learning_rate": 0.000354, + "loss": 8.1647, + "step": 118 + }, + { + "epoch": 0.00119, + "grad_norm": 0.9069057106971741, + "learning_rate": 0.000357, + "loss": 8.1514, + "step": 119 + }, + { + "epoch": 0.0012, + "grad_norm": 0.9525896906852722, + "learning_rate": 0.00035999999999999997, + "loss": 8.1239, + "step": 120 + }, + { + "epoch": 0.00121, + "grad_norm": 0.9400830268859863, + "learning_rate": 0.000363, + "loss": 8.095, + "step": 121 + }, + { + "epoch": 0.00122, + "grad_norm": 0.8189533948898315, + "learning_rate": 0.000366, + "loss": 8.0717, + "step": 122 + }, + { + "epoch": 0.00123, + "grad_norm": 0.8198633790016174, + "learning_rate": 0.000369, + "loss": 8.0459, + "step": 123 + }, + { + "epoch": 0.00124, + "grad_norm": 0.9460069537162781, + "learning_rate": 0.000372, + "loss": 8.0261, + "step": 124 + }, + { + "epoch": 0.00125, + "grad_norm": 1.0734294652938843, + "learning_rate": 0.000375, + "loss": 8.002, + "step": 125 + }, + { + "epoch": 0.00126, + "grad_norm": 0.9635769724845886, + "learning_rate": 0.000378, + "loss": 7.9871, + "step": 126 + }, + { + "epoch": 0.00127, + "grad_norm": 0.7875692844390869, + "learning_rate": 0.000381, + "loss": 7.9581, + "step": 127 + }, + { + "epoch": 0.00128, + "grad_norm": 0.8865201473236084, + "learning_rate": 0.000384, + "loss": 7.9374, + "step": 128 + }, + { + "epoch": 0.00129, + "grad_norm": 0.9998716115951538, + "learning_rate": 0.00038700000000000003, + "loss": 7.9265, + "step": 129 + }, + { + "epoch": 0.0013, + "grad_norm": 0.8098431825637817, + "learning_rate": 0.00039000000000000005, + "loss": 7.8932, + "step": 130 + }, + { + "epoch": 0.00131, + "grad_norm": 0.8202345967292786, + "learning_rate": 0.000393, + "loss": 7.8568, + "step": 131 + }, + { + "epoch": 0.00132, + "grad_norm": 0.9445962905883789, + "learning_rate": 0.00039600000000000003, + "loss": 7.8481, + "step": 132 + }, + { + "epoch": 0.00133, + "grad_norm": 0.8225625157356262, + "learning_rate": 0.00039900000000000005, + "loss": 7.8198, + "step": 133 + }, + { + "epoch": 0.00134, + "grad_norm": 0.8087729811668396, + "learning_rate": 0.000402, + "loss": 7.8072, + "step": 134 + }, + { + "epoch": 0.00135, + "grad_norm": 0.7232753038406372, + "learning_rate": 0.00040500000000000003, + "loss": 7.7727, + "step": 135 + }, + { + "epoch": 0.00136, + "grad_norm": 0.8383559584617615, + "learning_rate": 0.00040800000000000005, + "loss": 7.7588, + "step": 136 + }, + { + "epoch": 0.00137, + "grad_norm": 0.8329391479492188, + "learning_rate": 0.000411, + "loss": 7.738, + "step": 137 + }, + { + "epoch": 0.00138, + "grad_norm": 0.8072112202644348, + "learning_rate": 0.00041400000000000003, + "loss": 7.7102, + "step": 138 + }, + { + "epoch": 0.00139, + "grad_norm": 0.7906625270843506, + "learning_rate": 0.00041700000000000005, + "loss": 7.6947, + "step": 139 + }, + { + "epoch": 0.0014, + "grad_norm": 0.7997886538505554, + "learning_rate": 0.00042000000000000007, + "loss": 7.6753, + "step": 140 + }, + { + "epoch": 0.00141, + "grad_norm": 0.9642479419708252, + "learning_rate": 0.000423, + "loss": 7.6647, + "step": 141 + }, + { + "epoch": 0.00142, + "grad_norm": 0.8051616549491882, + "learning_rate": 0.00042599999999999995, + "loss": 7.6463, + "step": 142 + }, + { + "epoch": 0.00143, + "grad_norm": 0.7901502251625061, + "learning_rate": 0.00042899999999999997, + "loss": 7.6185, + "step": 143 + }, + { + "epoch": 0.00144, + "grad_norm": 0.6530913710594177, + "learning_rate": 0.000432, + "loss": 7.5947, + "step": 144 + }, + { + "epoch": 0.00145, + "grad_norm": 0.6823164820671082, + "learning_rate": 0.000435, + "loss": 7.5887, + "step": 145 + }, + { + "epoch": 0.00146, + "grad_norm": 0.7043561935424805, + "learning_rate": 0.00043799999999999997, + "loss": 7.5506, + "step": 146 + }, + { + "epoch": 0.00147, + "grad_norm": 0.660875141620636, + "learning_rate": 0.000441, + "loss": 7.5403, + "step": 147 + }, + { + "epoch": 0.00148, + "grad_norm": 0.6651095747947693, + "learning_rate": 0.000444, + "loss": 7.527, + "step": 148 + }, + { + "epoch": 0.00149, + "grad_norm": 0.594466507434845, + "learning_rate": 0.00044699999999999997, + "loss": 7.5177, + "step": 149 + }, + { + "epoch": 0.0015, + "grad_norm": 0.640634298324585, + "learning_rate": 0.00045, + "loss": 7.4872, + "step": 150 + }, + { + "epoch": 0.00151, + "grad_norm": 0.6399310231208801, + "learning_rate": 0.000453, + "loss": 7.4669, + "step": 151 + }, + { + "epoch": 0.00152, + "grad_norm": 0.6032711863517761, + "learning_rate": 0.000456, + "loss": 7.4595, + "step": 152 + }, + { + "epoch": 0.00153, + "grad_norm": 0.8105739951133728, + "learning_rate": 0.000459, + "loss": 7.4504, + "step": 153 + }, + { + "epoch": 0.00154, + "grad_norm": 0.9096337556838989, + "learning_rate": 0.000462, + "loss": 7.4401, + "step": 154 + }, + { + "epoch": 0.00155, + "grad_norm": 0.8464334607124329, + "learning_rate": 0.000465, + "loss": 7.4179, + "step": 155 + }, + { + "epoch": 0.00156, + "grad_norm": 0.8338698148727417, + "learning_rate": 0.000468, + "loss": 7.392, + "step": 156 + }, + { + "epoch": 0.00157, + "grad_norm": 0.6862301230430603, + "learning_rate": 0.000471, + "loss": 7.3898, + "step": 157 + }, + { + "epoch": 0.00158, + "grad_norm": 0.6174972057342529, + "learning_rate": 0.00047400000000000003, + "loss": 7.3642, + "step": 158 + }, + { + "epoch": 0.00159, + "grad_norm": 0.7215908765792847, + "learning_rate": 0.000477, + "loss": 7.3548, + "step": 159 + }, + { + "epoch": 0.0016, + "grad_norm": 0.7243570685386658, + "learning_rate": 0.00048, + "loss": 7.3363, + "step": 160 + }, + { + "epoch": 0.00161, + "grad_norm": 0.5998020172119141, + "learning_rate": 0.00048300000000000003, + "loss": 7.3105, + "step": 161 + }, + { + "epoch": 0.00162, + "grad_norm": 0.6509896516799927, + "learning_rate": 0.00048600000000000005, + "loss": 7.3157, + "step": 162 + }, + { + "epoch": 0.00163, + "grad_norm": 0.645023763179779, + "learning_rate": 0.0004890000000000001, + "loss": 7.2995, + "step": 163 + }, + { + "epoch": 0.00164, + "grad_norm": 0.47143545746803284, + "learning_rate": 0.000492, + "loss": 7.2948, + "step": 164 + }, + { + "epoch": 0.00165, + "grad_norm": 0.6931191086769104, + "learning_rate": 0.000495, + "loss": 7.2744, + "step": 165 + }, + { + "epoch": 0.00166, + "grad_norm": 0.5705839991569519, + "learning_rate": 0.0004980000000000001, + "loss": 7.2462, + "step": 166 + }, + { + "epoch": 0.00167, + "grad_norm": 0.5678435564041138, + "learning_rate": 0.000501, + "loss": 7.2388, + "step": 167 + }, + { + "epoch": 0.00168, + "grad_norm": 0.4481411278247833, + "learning_rate": 0.000504, + "loss": 7.2152, + "step": 168 + }, + { + "epoch": 0.00169, + "grad_norm": 0.5297079086303711, + "learning_rate": 0.0005070000000000001, + "loss": 7.2111, + "step": 169 + }, + { + "epoch": 0.0017, + "grad_norm": 0.5522683262825012, + "learning_rate": 0.00051, + "loss": 7.1957, + "step": 170 + }, + { + "epoch": 0.00171, + "grad_norm": 0.4941532611846924, + "learning_rate": 0.000513, + "loss": 7.1764, + "step": 171 + }, + { + "epoch": 0.00172, + "grad_norm": 0.40358924865722656, + "learning_rate": 0.000516, + "loss": 7.1752, + "step": 172 + }, + { + "epoch": 0.00173, + "grad_norm": 0.43254554271698, + "learning_rate": 0.0005189999999999999, + "loss": 7.1608, + "step": 173 + }, + { + "epoch": 0.00174, + "grad_norm": 0.49833717942237854, + "learning_rate": 0.000522, + "loss": 7.1779, + "step": 174 + }, + { + "epoch": 0.00175, + "grad_norm": 0.469341903924942, + "learning_rate": 0.000525, + "loss": 7.1406, + "step": 175 + }, + { + "epoch": 0.00176, + "grad_norm": 0.46195128560066223, + "learning_rate": 0.0005279999999999999, + "loss": 7.1165, + "step": 176 + }, + { + "epoch": 0.00177, + "grad_norm": 0.4399634003639221, + "learning_rate": 0.000531, + "loss": 7.1206, + "step": 177 + }, + { + "epoch": 0.00178, + "grad_norm": 0.42299556732177734, + "learning_rate": 0.000534, + "loss": 7.1134, + "step": 178 + }, + { + "epoch": 0.00179, + "grad_norm": 0.3987540602684021, + "learning_rate": 0.000537, + "loss": 7.1017, + "step": 179 + }, + { + "epoch": 0.0018, + "grad_norm": 0.37395715713500977, + "learning_rate": 0.00054, + "loss": 7.074, + "step": 180 + }, + { + "epoch": 0.00181, + "grad_norm": 0.3351408541202545, + "learning_rate": 0.000543, + "loss": 7.0613, + "step": 181 + }, + { + "epoch": 0.00182, + "grad_norm": 0.3589305281639099, + "learning_rate": 0.000546, + "loss": 7.0589, + "step": 182 + }, + { + "epoch": 0.00183, + "grad_norm": 0.38710907101631165, + "learning_rate": 0.000549, + "loss": 7.0486, + "step": 183 + }, + { + "epoch": 0.00184, + "grad_norm": 0.4560106098651886, + "learning_rate": 0.000552, + "loss": 7.0384, + "step": 184 + }, + { + "epoch": 0.00185, + "grad_norm": 0.8466277718544006, + "learning_rate": 0.000555, + "loss": 7.0199, + "step": 185 + }, + { + "epoch": 0.00186, + "grad_norm": 1.4107517004013062, + "learning_rate": 0.000558, + "loss": 7.0664, + "step": 186 + }, + { + "epoch": 0.00187, + "grad_norm": 0.5632089376449585, + "learning_rate": 0.000561, + "loss": 7.004, + "step": 187 + }, + { + "epoch": 0.00188, + "grad_norm": 1.07405686378479, + "learning_rate": 0.000564, + "loss": 6.9945, + "step": 188 + }, + { + "epoch": 0.00189, + "grad_norm": 1.1665420532226562, + "learning_rate": 0.000567, + "loss": 7.0095, + "step": 189 + }, + { + "epoch": 0.0019, + "grad_norm": 0.4235672950744629, + "learning_rate": 0.00057, + "loss": 6.962, + "step": 190 + }, + { + "epoch": 0.00191, + "grad_norm": 1.2953448295593262, + "learning_rate": 0.000573, + "loss": 6.9853, + "step": 191 + }, + { + "epoch": 0.00192, + "grad_norm": 0.5110867023468018, + "learning_rate": 0.000576, + "loss": 6.9512, + "step": 192 + }, + { + "epoch": 0.00193, + "grad_norm": 0.7966066002845764, + "learning_rate": 0.000579, + "loss": 6.9656, + "step": 193 + }, + { + "epoch": 0.00194, + "grad_norm": 0.5008851289749146, + "learning_rate": 0.000582, + "loss": 6.9458, + "step": 194 + }, + { + "epoch": 0.00195, + "grad_norm": 0.714582085609436, + "learning_rate": 0.000585, + "loss": 6.9325, + "step": 195 + }, + { + "epoch": 0.00196, + "grad_norm": 0.48010018467903137, + "learning_rate": 0.000588, + "loss": 6.917, + "step": 196 + }, + { + "epoch": 0.00197, + "grad_norm": 0.5283955335617065, + "learning_rate": 0.000591, + "loss": 6.9149, + "step": 197 + }, + { + "epoch": 0.00198, + "grad_norm": 0.5033705830574036, + "learning_rate": 0.000594, + "loss": 6.9041, + "step": 198 + }, + { + "epoch": 0.00199, + "grad_norm": 0.40711161494255066, + "learning_rate": 0.0005970000000000001, + "loss": 6.8819, + "step": 199 + }, + { + "epoch": 0.002, + "grad_norm": 0.4253259003162384, + "learning_rate": 0.0006000000000000001, + "loss": 6.8842, + "step": 200 + }, + { + "epoch": 0.00201, + "grad_norm": 0.347766637802124, + "learning_rate": 0.000603, + "loss": 6.8566, + "step": 201 + }, + { + "epoch": 0.00202, + "grad_norm": 0.4044833779335022, + "learning_rate": 0.0006060000000000001, + "loss": 6.8448, + "step": 202 + }, + { + "epoch": 0.00203, + "grad_norm": 0.3598291575908661, + "learning_rate": 0.0006090000000000001, + "loss": 6.8489, + "step": 203 + }, + { + "epoch": 0.00204, + "grad_norm": 0.35803648829460144, + "learning_rate": 0.000612, + "loss": 6.8337, + "step": 204 + }, + { + "epoch": 0.00205, + "grad_norm": 0.3630695044994354, + "learning_rate": 0.000615, + "loss": 6.82, + "step": 205 + }, + { + "epoch": 0.00206, + "grad_norm": 0.3439967930316925, + "learning_rate": 0.000618, + "loss": 6.8125, + "step": 206 + }, + { + "epoch": 0.00207, + "grad_norm": 0.3219742476940155, + "learning_rate": 0.000621, + "loss": 6.8026, + "step": 207 + }, + { + "epoch": 0.00208, + "grad_norm": 0.42236459255218506, + "learning_rate": 0.000624, + "loss": 6.8006, + "step": 208 + }, + { + "epoch": 0.00209, + "grad_norm": 0.44003602862358093, + "learning_rate": 0.000627, + "loss": 6.7802, + "step": 209 + }, + { + "epoch": 0.0021, + "grad_norm": 0.6629877686500549, + "learning_rate": 0.00063, + "loss": 6.7737, + "step": 210 + }, + { + "epoch": 0.00211, + "grad_norm": 0.9569016695022583, + "learning_rate": 0.000633, + "loss": 6.7754, + "step": 211 + }, + { + "epoch": 0.00212, + "grad_norm": 1.1293443441390991, + "learning_rate": 0.000636, + "loss": 6.7746, + "step": 212 + }, + { + "epoch": 0.00213, + "grad_norm": 0.43011096119880676, + "learning_rate": 0.000639, + "loss": 6.7385, + "step": 213 + }, + { + "epoch": 0.00214, + "grad_norm": 0.6478229761123657, + "learning_rate": 0.000642, + "loss": 6.7415, + "step": 214 + }, + { + "epoch": 0.00215, + "grad_norm": 0.6323032975196838, + "learning_rate": 0.000645, + "loss": 6.7384, + "step": 215 + }, + { + "epoch": 0.00216, + "grad_norm": 0.441693514585495, + "learning_rate": 0.000648, + "loss": 6.7285, + "step": 216 + }, + { + "epoch": 0.00217, + "grad_norm": 0.5594473481178284, + "learning_rate": 0.000651, + "loss": 6.7033, + "step": 217 + }, + { + "epoch": 0.00218, + "grad_norm": 0.5135915279388428, + "learning_rate": 0.000654, + "loss": 6.7073, + "step": 218 + }, + { + "epoch": 0.00219, + "grad_norm": 0.4307027757167816, + "learning_rate": 0.000657, + "loss": 6.6782, + "step": 219 + }, + { + "epoch": 0.0022, + "grad_norm": 0.45137229561805725, + "learning_rate": 0.00066, + "loss": 6.6953, + "step": 220 + }, + { + "epoch": 0.00221, + "grad_norm": 0.4729914963245392, + "learning_rate": 0.0006630000000000001, + "loss": 6.6737, + "step": 221 + }, + { + "epoch": 0.00222, + "grad_norm": 0.47246506810188293, + "learning_rate": 0.000666, + "loss": 6.6615, + "step": 222 + }, + { + "epoch": 0.00223, + "grad_norm": 0.36316192150115967, + "learning_rate": 0.000669, + "loss": 6.6543, + "step": 223 + }, + { + "epoch": 0.00224, + "grad_norm": 0.4332623779773712, + "learning_rate": 0.0006720000000000001, + "loss": 6.6525, + "step": 224 + }, + { + "epoch": 0.00225, + "grad_norm": 0.41814228892326355, + "learning_rate": 0.000675, + "loss": 6.639, + "step": 225 + }, + { + "epoch": 0.00226, + "grad_norm": 0.42956992983818054, + "learning_rate": 0.000678, + "loss": 6.634, + "step": 226 + }, + { + "epoch": 0.00227, + "grad_norm": 0.40267884731292725, + "learning_rate": 0.0006810000000000001, + "loss": 6.6072, + "step": 227 + }, + { + "epoch": 0.00228, + "grad_norm": 0.4361991882324219, + "learning_rate": 0.000684, + "loss": 6.6099, + "step": 228 + }, + { + "epoch": 0.00229, + "grad_norm": 0.47655290365219116, + "learning_rate": 0.000687, + "loss": 6.6086, + "step": 229 + }, + { + "epoch": 0.0023, + "grad_norm": 0.5011177659034729, + "learning_rate": 0.0006900000000000001, + "loss": 6.5846, + "step": 230 + }, + { + "epoch": 0.00231, + "grad_norm": 0.5389447212219238, + "learning_rate": 0.000693, + "loss": 6.5719, + "step": 231 + }, + { + "epoch": 0.00232, + "grad_norm": 0.5394959449768066, + "learning_rate": 0.000696, + "loss": 6.5809, + "step": 232 + }, + { + "epoch": 0.00233, + "grad_norm": 0.49784839153289795, + "learning_rate": 0.0006990000000000001, + "loss": 6.5675, + "step": 233 + }, + { + "epoch": 0.00234, + "grad_norm": 0.42049404978752136, + "learning_rate": 0.000702, + "loss": 6.5581, + "step": 234 + }, + { + "epoch": 0.00235, + "grad_norm": 0.5810425281524658, + "learning_rate": 0.000705, + "loss": 6.5463, + "step": 235 + }, + { + "epoch": 0.00236, + "grad_norm": 0.6429721117019653, + "learning_rate": 0.000708, + "loss": 6.5306, + "step": 236 + }, + { + "epoch": 0.00237, + "grad_norm": 0.6626091003417969, + "learning_rate": 0.0007109999999999999, + "loss": 6.5402, + "step": 237 + }, + { + "epoch": 0.00238, + "grad_norm": 0.5873957872390747, + "learning_rate": 0.000714, + "loss": 6.5299, + "step": 238 + }, + { + "epoch": 0.00239, + "grad_norm": 0.4890768527984619, + "learning_rate": 0.000717, + "loss": 6.5085, + "step": 239 + }, + { + "epoch": 0.0024, + "grad_norm": 0.6101468801498413, + "learning_rate": 0.0007199999999999999, + "loss": 6.5266, + "step": 240 + }, + { + "epoch": 0.00241, + "grad_norm": 0.5837545394897461, + "learning_rate": 0.000723, + "loss": 6.5125, + "step": 241 + }, + { + "epoch": 0.00242, + "grad_norm": 0.38101619482040405, + "learning_rate": 0.000726, + "loss": 6.481, + "step": 242 + }, + { + "epoch": 0.00243, + "grad_norm": 0.5516716837882996, + "learning_rate": 0.000729, + "loss": 6.4712, + "step": 243 + }, + { + "epoch": 0.00244, + "grad_norm": 0.6402163505554199, + "learning_rate": 0.000732, + "loss": 6.4787, + "step": 244 + }, + { + "epoch": 0.00245, + "grad_norm": 0.661472737789154, + "learning_rate": 0.000735, + "loss": 6.4666, + "step": 245 + }, + { + "epoch": 0.00246, + "grad_norm": 0.8242950439453125, + "learning_rate": 0.000738, + "loss": 6.457, + "step": 246 + }, + { + "epoch": 0.00247, + "grad_norm": 0.8979260921478271, + "learning_rate": 0.000741, + "loss": 6.4532, + "step": 247 + }, + { + "epoch": 0.00248, + "grad_norm": 0.7822521924972534, + "learning_rate": 0.000744, + "loss": 6.4595, + "step": 248 + }, + { + "epoch": 0.00249, + "grad_norm": 0.5830389261245728, + "learning_rate": 0.000747, + "loss": 6.4336, + "step": 249 + }, + { + "epoch": 0.0025, + "grad_norm": 0.5053834319114685, + "learning_rate": 0.00075, + "loss": 6.436, + "step": 250 + }, + { + "epoch": 0.00251, + "grad_norm": 0.6169653534889221, + "learning_rate": 0.000753, + "loss": 6.4018, + "step": 251 + }, + { + "epoch": 0.00252, + "grad_norm": 0.693821370601654, + "learning_rate": 0.000756, + "loss": 6.411, + "step": 252 + }, + { + "epoch": 0.00253, + "grad_norm": 0.7529676556587219, + "learning_rate": 0.000759, + "loss": 6.4053, + "step": 253 + }, + { + "epoch": 0.00254, + "grad_norm": 0.7882714867591858, + "learning_rate": 0.000762, + "loss": 6.3993, + "step": 254 + }, + { + "epoch": 0.00255, + "grad_norm": 0.8540387153625488, + "learning_rate": 0.0007650000000000001, + "loss": 6.399, + "step": 255 + }, + { + "epoch": 0.00256, + "grad_norm": 0.8460838198661804, + "learning_rate": 0.000768, + "loss": 6.3805, + "step": 256 + }, + { + "epoch": 0.00257, + "grad_norm": 0.7798411250114441, + "learning_rate": 0.000771, + "loss": 6.383, + "step": 257 + }, + { + "epoch": 0.00258, + "grad_norm": 0.7431406378746033, + "learning_rate": 0.0007740000000000001, + "loss": 6.362, + "step": 258 + }, + { + "epoch": 0.00259, + "grad_norm": 0.9195736646652222, + "learning_rate": 0.000777, + "loss": 6.3634, + "step": 259 + }, + { + "epoch": 0.0026, + "grad_norm": 0.9989089965820312, + "learning_rate": 0.0007800000000000001, + "loss": 6.3605, + "step": 260 + }, + { + "epoch": 0.00261, + "grad_norm": 0.8553241491317749, + "learning_rate": 0.0007830000000000001, + "loss": 6.3627, + "step": 261 + }, + { + "epoch": 0.00262, + "grad_norm": 0.8126974701881409, + "learning_rate": 0.000786, + "loss": 6.3448, + "step": 262 + }, + { + "epoch": 0.00263, + "grad_norm": 0.5453911423683167, + "learning_rate": 0.0007890000000000001, + "loss": 6.3416, + "step": 263 + }, + { + "epoch": 0.00264, + "grad_norm": 0.5339396595954895, + "learning_rate": 0.0007920000000000001, + "loss": 6.3252, + "step": 264 + }, + { + "epoch": 0.00265, + "grad_norm": 0.561769425868988, + "learning_rate": 0.000795, + "loss": 6.3181, + "step": 265 + }, + { + "epoch": 0.00266, + "grad_norm": 0.47888532280921936, + "learning_rate": 0.0007980000000000001, + "loss": 6.3105, + "step": 266 + }, + { + "epoch": 0.00267, + "grad_norm": 0.4935484528541565, + "learning_rate": 0.0008010000000000001, + "loss": 6.3049, + "step": 267 + }, + { + "epoch": 0.00268, + "grad_norm": 0.4065157175064087, + "learning_rate": 0.000804, + "loss": 6.2969, + "step": 268 + }, + { + "epoch": 0.00269, + "grad_norm": 0.5361817479133606, + "learning_rate": 0.0008070000000000001, + "loss": 6.2818, + "step": 269 + }, + { + "epoch": 0.0027, + "grad_norm": 0.6360214352607727, + "learning_rate": 0.0008100000000000001, + "loss": 6.2858, + "step": 270 + }, + { + "epoch": 0.00271, + "grad_norm": 0.6580653190612793, + "learning_rate": 0.000813, + "loss": 6.2898, + "step": 271 + }, + { + "epoch": 0.00272, + "grad_norm": 0.719866931438446, + "learning_rate": 0.0008160000000000001, + "loss": 6.2764, + "step": 272 + }, + { + "epoch": 0.00273, + "grad_norm": 0.726635754108429, + "learning_rate": 0.0008190000000000001, + "loss": 6.2654, + "step": 273 + }, + { + "epoch": 0.00274, + "grad_norm": 0.5728192925453186, + "learning_rate": 0.000822, + "loss": 6.2437, + "step": 274 + }, + { + "epoch": 0.00275, + "grad_norm": 0.469969242811203, + "learning_rate": 0.0008250000000000001, + "loss": 6.2475, + "step": 275 + }, + { + "epoch": 0.00276, + "grad_norm": 0.5783148407936096, + "learning_rate": 0.0008280000000000001, + "loss": 6.2405, + "step": 276 + }, + { + "epoch": 0.00277, + "grad_norm": 0.5995691418647766, + "learning_rate": 0.0008310000000000001, + "loss": 6.2356, + "step": 277 + }, + { + "epoch": 0.00278, + "grad_norm": 0.5173709988594055, + "learning_rate": 0.0008340000000000001, + "loss": 6.2181, + "step": 278 + }, + { + "epoch": 0.00279, + "grad_norm": 0.47013920545578003, + "learning_rate": 0.0008370000000000001, + "loss": 6.2175, + "step": 279 + }, + { + "epoch": 0.0028, + "grad_norm": 0.38072702288627625, + "learning_rate": 0.0008400000000000001, + "loss": 6.1988, + "step": 280 + }, + { + "epoch": 0.00281, + "grad_norm": 0.44907790422439575, + "learning_rate": 0.0008430000000000001, + "loss": 6.1893, + "step": 281 + }, + { + "epoch": 0.00282, + "grad_norm": 0.40965142846107483, + "learning_rate": 0.000846, + "loss": 6.1869, + "step": 282 + }, + { + "epoch": 0.00283, + "grad_norm": 0.48822489380836487, + "learning_rate": 0.0008489999999999999, + "loss": 6.1783, + "step": 283 + }, + { + "epoch": 0.00284, + "grad_norm": 0.726660966873169, + "learning_rate": 0.0008519999999999999, + "loss": 6.1771, + "step": 284 + }, + { + "epoch": 0.00285, + "grad_norm": 1.0991517305374146, + "learning_rate": 0.000855, + "loss": 6.1879, + "step": 285 + }, + { + "epoch": 0.00286, + "grad_norm": 0.9898068904876709, + "learning_rate": 0.0008579999999999999, + "loss": 6.1694, + "step": 286 + }, + { + "epoch": 0.00287, + "grad_norm": 0.9177366495132446, + "learning_rate": 0.000861, + "loss": 6.1687, + "step": 287 + }, + { + "epoch": 0.00288, + "grad_norm": 1.3172835111618042, + "learning_rate": 0.000864, + "loss": 6.175, + "step": 288 + }, + { + "epoch": 0.00289, + "grad_norm": 1.0531185865402222, + "learning_rate": 0.0008669999999999999, + "loss": 6.1733, + "step": 289 + }, + { + "epoch": 0.0029, + "grad_norm": 0.9814063310623169, + "learning_rate": 0.00087, + "loss": 6.1665, + "step": 290 + }, + { + "epoch": 0.00291, + "grad_norm": 0.7696391344070435, + "learning_rate": 0.000873, + "loss": 6.1455, + "step": 291 + }, + { + "epoch": 0.00292, + "grad_norm": 0.8531065583229065, + "learning_rate": 0.0008759999999999999, + "loss": 6.15, + "step": 292 + }, + { + "epoch": 0.00293, + "grad_norm": 0.7760049700737, + "learning_rate": 0.000879, + "loss": 6.1273, + "step": 293 + }, + { + "epoch": 0.00294, + "grad_norm": 0.7517282366752625, + "learning_rate": 0.000882, + "loss": 6.1432, + "step": 294 + }, + { + "epoch": 0.00295, + "grad_norm": 0.5758442282676697, + "learning_rate": 0.0008849999999999999, + "loss": 6.106, + "step": 295 + }, + { + "epoch": 0.00296, + "grad_norm": 0.5470280647277832, + "learning_rate": 0.000888, + "loss": 6.1152, + "step": 296 + }, + { + "epoch": 0.00297, + "grad_norm": 0.46315857768058777, + "learning_rate": 0.000891, + "loss": 6.0848, + "step": 297 + }, + { + "epoch": 0.00298, + "grad_norm": 0.52577805519104, + "learning_rate": 0.0008939999999999999, + "loss": 6.0996, + "step": 298 + }, + { + "epoch": 0.00299, + "grad_norm": 0.5289214253425598, + "learning_rate": 0.000897, + "loss": 6.0753, + "step": 299 + }, + { + "epoch": 0.003, + "grad_norm": 0.39721718430519104, + "learning_rate": 0.0009, + "loss": 6.0729, + "step": 300 + }, + { + "epoch": 0.00301, + "grad_norm": 0.42188870906829834, + "learning_rate": 0.0009029999999999999, + "loss": 6.0649, + "step": 301 + }, + { + "epoch": 0.00302, + "grad_norm": 0.49844104051589966, + "learning_rate": 0.000906, + "loss": 6.0504, + "step": 302 + }, + { + "epoch": 0.00303, + "grad_norm": 0.5113502144813538, + "learning_rate": 0.000909, + "loss": 6.0629, + "step": 303 + }, + { + "epoch": 0.00304, + "grad_norm": 0.6390882730484009, + "learning_rate": 0.000912, + "loss": 6.0593, + "step": 304 + }, + { + "epoch": 0.00305, + "grad_norm": 0.8851528763771057, + "learning_rate": 0.000915, + "loss": 6.0668, + "step": 305 + }, + { + "epoch": 0.00306, + "grad_norm": 0.9017530083656311, + "learning_rate": 0.000918, + "loss": 6.0552, + "step": 306 + }, + { + "epoch": 0.00307, + "grad_norm": 0.563444197177887, + "learning_rate": 0.000921, + "loss": 6.0386, + "step": 307 + }, + { + "epoch": 0.00308, + "grad_norm": 0.6002116203308105, + "learning_rate": 0.000924, + "loss": 6.0248, + "step": 308 + }, + { + "epoch": 0.00309, + "grad_norm": 0.7035393118858337, + "learning_rate": 0.000927, + "loss": 6.0412, + "step": 309 + }, + { + "epoch": 0.0031, + "grad_norm": 1.01050865650177, + "learning_rate": 0.00093, + "loss": 6.0314, + "step": 310 + }, + { + "epoch": 0.00311, + "grad_norm": 1.16908860206604, + "learning_rate": 0.000933, + "loss": 6.0397, + "step": 311 + }, + { + "epoch": 0.00312, + "grad_norm": 0.6785458922386169, + "learning_rate": 0.000936, + "loss": 6.0006, + "step": 312 + }, + { + "epoch": 0.00313, + "grad_norm": 0.7975029349327087, + "learning_rate": 0.0009390000000000001, + "loss": 6.0267, + "step": 313 + }, + { + "epoch": 0.00314, + "grad_norm": 0.6784241795539856, + "learning_rate": 0.000942, + "loss": 6.0253, + "step": 314 + }, + { + "epoch": 0.00315, + "grad_norm": 0.5287242531776428, + "learning_rate": 0.000945, + "loss": 5.9989, + "step": 315 + }, + { + "epoch": 0.00316, + "grad_norm": 0.5889810919761658, + "learning_rate": 0.0009480000000000001, + "loss": 5.9738, + "step": 316 + }, + { + "epoch": 0.00317, + "grad_norm": 0.8596201539039612, + "learning_rate": 0.000951, + "loss": 5.9963, + "step": 317 + }, + { + "epoch": 0.00318, + "grad_norm": 1.220719575881958, + "learning_rate": 0.000954, + "loss": 6.0052, + "step": 318 + }, + { + "epoch": 0.00319, + "grad_norm": 0.7490801215171814, + "learning_rate": 0.0009570000000000001, + "loss": 5.9868, + "step": 319 + }, + { + "epoch": 0.0032, + "grad_norm": 0.6210083365440369, + "learning_rate": 0.00096, + "loss": 5.9801, + "step": 320 + }, + { + "epoch": 0.00321, + "grad_norm": 0.5945920944213867, + "learning_rate": 0.000963, + "loss": 5.97, + "step": 321 + }, + { + "epoch": 0.00322, + "grad_norm": 0.6791667342185974, + "learning_rate": 0.0009660000000000001, + "loss": 5.9544, + "step": 322 + }, + { + "epoch": 0.00323, + "grad_norm": 0.9637515544891357, + "learning_rate": 0.000969, + "loss": 5.9627, + "step": 323 + }, + { + "epoch": 0.00324, + "grad_norm": 1.1402119398117065, + "learning_rate": 0.0009720000000000001, + "loss": 5.972, + "step": 324 + }, + { + "epoch": 0.00325, + "grad_norm": 1.0057023763656616, + "learning_rate": 0.0009750000000000001, + "loss": 5.9423, + "step": 325 + }, + { + "epoch": 0.00326, + "grad_norm": 0.5953328609466553, + "learning_rate": 0.0009780000000000001, + "loss": 5.942, + "step": 326 + }, + { + "epoch": 0.00327, + "grad_norm": 0.7124008536338806, + "learning_rate": 0.000981, + "loss": 5.9465, + "step": 327 + }, + { + "epoch": 0.00328, + "grad_norm": 0.7318410277366638, + "learning_rate": 0.000984, + "loss": 5.9241, + "step": 328 + }, + { + "epoch": 0.00329, + "grad_norm": 0.6503687500953674, + "learning_rate": 0.000987, + "loss": 5.922, + "step": 329 + }, + { + "epoch": 0.0033, + "grad_norm": 0.6151977181434631, + "learning_rate": 0.00099, + "loss": 5.9159, + "step": 330 + }, + { + "epoch": 0.00331, + "grad_norm": 0.49422070384025574, + "learning_rate": 0.0009930000000000002, + "loss": 5.9144, + "step": 331 + }, + { + "epoch": 0.00332, + "grad_norm": 0.563934326171875, + "learning_rate": 0.0009960000000000001, + "loss": 5.9008, + "step": 332 + }, + { + "epoch": 0.00333, + "grad_norm": 0.5146680474281311, + "learning_rate": 0.000999, + "loss": 5.8812, + "step": 333 + }, + { + "epoch": 0.00334, + "grad_norm": 0.5699781775474548, + "learning_rate": 0.001002, + "loss": 5.8922, + "step": 334 + }, + { + "epoch": 0.00335, + "grad_norm": 0.628279983997345, + "learning_rate": 0.001005, + "loss": 5.8934, + "step": 335 + }, + { + "epoch": 0.00336, + "grad_norm": 0.638155996799469, + "learning_rate": 0.001008, + "loss": 5.8854, + "step": 336 + }, + { + "epoch": 0.00337, + "grad_norm": 0.5850276947021484, + "learning_rate": 0.0010110000000000002, + "loss": 5.8631, + "step": 337 + }, + { + "epoch": 0.00338, + "grad_norm": 0.5985286831855774, + "learning_rate": 0.0010140000000000001, + "loss": 5.879, + "step": 338 + }, + { + "epoch": 0.00339, + "grad_norm": 0.9502546787261963, + "learning_rate": 0.0010170000000000001, + "loss": 5.8893, + "step": 339 + }, + { + "epoch": 0.0034, + "grad_norm": 1.3471951484680176, + "learning_rate": 0.00102, + "loss": 5.8789, + "step": 340 + }, + { + "epoch": 0.00341, + "grad_norm": 0.6621291041374207, + "learning_rate": 0.001023, + "loss": 5.8449, + "step": 341 + }, + { + "epoch": 0.00342, + "grad_norm": 0.8053567409515381, + "learning_rate": 0.001026, + "loss": 5.8551, + "step": 342 + }, + { + "epoch": 0.00343, + "grad_norm": 1.0873545408248901, + "learning_rate": 0.0010290000000000002, + "loss": 5.8656, + "step": 343 + }, + { + "epoch": 0.00344, + "grad_norm": 1.1315717697143555, + "learning_rate": 0.001032, + "loss": 5.8613, + "step": 344 + }, + { + "epoch": 0.00345, + "grad_norm": 0.7909812927246094, + "learning_rate": 0.001035, + "loss": 5.8449, + "step": 345 + }, + { + "epoch": 0.00346, + "grad_norm": 0.8379076719284058, + "learning_rate": 0.0010379999999999999, + "loss": 5.8518, + "step": 346 + }, + { + "epoch": 0.00347, + "grad_norm": 0.5857303738594055, + "learning_rate": 0.001041, + "loss": 5.8133, + "step": 347 + }, + { + "epoch": 0.00348, + "grad_norm": 0.6455392241477966, + "learning_rate": 0.001044, + "loss": 5.8262, + "step": 348 + }, + { + "epoch": 0.00349, + "grad_norm": 0.6315843462944031, + "learning_rate": 0.001047, + "loss": 5.8046, + "step": 349 + }, + { + "epoch": 0.0035, + "grad_norm": 0.6185011863708496, + "learning_rate": 0.00105, + "loss": 5.8282, + "step": 350 + }, + { + "epoch": 0.00351, + "grad_norm": 0.49840763211250305, + "learning_rate": 0.001053, + "loss": 5.8234, + "step": 351 + }, + { + "epoch": 0.00352, + "grad_norm": 0.43951740860939026, + "learning_rate": 0.0010559999999999999, + "loss": 5.7984, + "step": 352 + }, + { + "epoch": 0.00353, + "grad_norm": 0.4452185034751892, + "learning_rate": 0.001059, + "loss": 5.7993, + "step": 353 + }, + { + "epoch": 0.00354, + "grad_norm": 0.43185603618621826, + "learning_rate": 0.001062, + "loss": 5.7703, + "step": 354 + }, + { + "epoch": 0.00355, + "grad_norm": 0.4053448736667633, + "learning_rate": 0.001065, + "loss": 5.7848, + "step": 355 + }, + { + "epoch": 0.00356, + "grad_norm": 0.48363247513771057, + "learning_rate": 0.001068, + "loss": 5.777, + "step": 356 + }, + { + "epoch": 0.00357, + "grad_norm": 0.5601730346679688, + "learning_rate": 0.001071, + "loss": 5.7693, + "step": 357 + }, + { + "epoch": 0.00358, + "grad_norm": 0.7239671349525452, + "learning_rate": 0.001074, + "loss": 5.7638, + "step": 358 + }, + { + "epoch": 0.00359, + "grad_norm": 0.8531132936477661, + "learning_rate": 0.001077, + "loss": 5.76, + "step": 359 + }, + { + "epoch": 0.0036, + "grad_norm": 1.0332695245742798, + "learning_rate": 0.00108, + "loss": 5.7701, + "step": 360 + }, + { + "epoch": 0.00361, + "grad_norm": 1.3098387718200684, + "learning_rate": 0.001083, + "loss": 5.7755, + "step": 361 + }, + { + "epoch": 0.00362, + "grad_norm": 0.7546947598457336, + "learning_rate": 0.001086, + "loss": 5.7474, + "step": 362 + }, + { + "epoch": 0.00363, + "grad_norm": 0.6924042701721191, + "learning_rate": 0.001089, + "loss": 5.7394, + "step": 363 + }, + { + "epoch": 0.00364, + "grad_norm": 0.87959223985672, + "learning_rate": 0.001092, + "loss": 5.7603, + "step": 364 + }, + { + "epoch": 0.00365, + "grad_norm": 1.037275791168213, + "learning_rate": 0.001095, + "loss": 5.7581, + "step": 365 + }, + { + "epoch": 0.00366, + "grad_norm": 1.056171178817749, + "learning_rate": 0.001098, + "loss": 5.7464, + "step": 366 + }, + { + "epoch": 0.00367, + "grad_norm": 1.0635496377944946, + "learning_rate": 0.001101, + "loss": 5.7478, + "step": 367 + }, + { + "epoch": 0.00368, + "grad_norm": 0.9701796174049377, + "learning_rate": 0.001104, + "loss": 5.7478, + "step": 368 + }, + { + "epoch": 0.00369, + "grad_norm": 0.7430213689804077, + "learning_rate": 0.001107, + "loss": 5.745, + "step": 369 + }, + { + "epoch": 0.0037, + "grad_norm": 0.7348084449768066, + "learning_rate": 0.00111, + "loss": 5.7264, + "step": 370 + }, + { + "epoch": 0.00371, + "grad_norm": 0.8778790831565857, + "learning_rate": 0.001113, + "loss": 5.6986, + "step": 371 + }, + { + "epoch": 0.00372, + "grad_norm": 1.160132884979248, + "learning_rate": 0.001116, + "loss": 5.7343, + "step": 372 + }, + { + "epoch": 0.00373, + "grad_norm": 0.8288450241088867, + "learning_rate": 0.001119, + "loss": 5.7145, + "step": 373 + }, + { + "epoch": 0.00374, + "grad_norm": 0.7645081281661987, + "learning_rate": 0.001122, + "loss": 5.7142, + "step": 374 + }, + { + "epoch": 0.00375, + "grad_norm": 0.8342962265014648, + "learning_rate": 0.0011250000000000001, + "loss": 5.7215, + "step": 375 + }, + { + "epoch": 0.00376, + "grad_norm": 0.8966416716575623, + "learning_rate": 0.001128, + "loss": 5.7142, + "step": 376 + }, + { + "epoch": 0.00377, + "grad_norm": 1.1411352157592773, + "learning_rate": 0.001131, + "loss": 5.7202, + "step": 377 + }, + { + "epoch": 0.00378, + "grad_norm": 1.0639731884002686, + "learning_rate": 0.001134, + "loss": 5.7166, + "step": 378 + }, + { + "epoch": 0.00379, + "grad_norm": 1.0386251211166382, + "learning_rate": 0.001137, + "loss": 5.701, + "step": 379 + }, + { + "epoch": 0.0038, + "grad_norm": 0.8551567196846008, + "learning_rate": 0.00114, + "loss": 5.7166, + "step": 380 + }, + { + "epoch": 0.00381, + "grad_norm": 1.171457290649414, + "learning_rate": 0.0011430000000000001, + "loss": 5.693, + "step": 381 + }, + { + "epoch": 0.00382, + "grad_norm": 0.9382472634315491, + "learning_rate": 0.001146, + "loss": 5.6768, + "step": 382 + }, + { + "epoch": 0.00383, + "grad_norm": 1.011130452156067, + "learning_rate": 0.001149, + "loss": 5.6945, + "step": 383 + }, + { + "epoch": 0.00384, + "grad_norm": 0.7897657155990601, + "learning_rate": 0.001152, + "loss": 5.6883, + "step": 384 + }, + { + "epoch": 0.00385, + "grad_norm": 0.6210044622421265, + "learning_rate": 0.001155, + "loss": 5.673, + "step": 385 + }, + { + "epoch": 0.00386, + "grad_norm": 0.5515205264091492, + "learning_rate": 0.001158, + "loss": 5.6723, + "step": 386 + }, + { + "epoch": 0.00387, + "grad_norm": 0.5881383419036865, + "learning_rate": 0.0011610000000000001, + "loss": 5.6493, + "step": 387 + }, + { + "epoch": 0.00388, + "grad_norm": 0.4862520396709442, + "learning_rate": 0.001164, + "loss": 5.6435, + "step": 388 + }, + { + "epoch": 0.00389, + "grad_norm": 0.4490566551685333, + "learning_rate": 0.001167, + "loss": 5.649, + "step": 389 + }, + { + "epoch": 0.0039, + "grad_norm": 0.4499252140522003, + "learning_rate": 0.00117, + "loss": 5.6299, + "step": 390 + }, + { + "epoch": 0.00391, + "grad_norm": 0.3959794044494629, + "learning_rate": 0.001173, + "loss": 5.6342, + "step": 391 + }, + { + "epoch": 0.00392, + "grad_norm": 0.41570809483528137, + "learning_rate": 0.001176, + "loss": 5.6441, + "step": 392 + }, + { + "epoch": 0.00393, + "grad_norm": 0.4579019248485565, + "learning_rate": 0.0011790000000000001, + "loss": 5.615, + "step": 393 + }, + { + "epoch": 0.00394, + "grad_norm": 0.5718971490859985, + "learning_rate": 0.001182, + "loss": 5.6031, + "step": 394 + }, + { + "epoch": 0.00395, + "grad_norm": 0.8492469191551208, + "learning_rate": 0.001185, + "loss": 5.5961, + "step": 395 + }, + { + "epoch": 0.00396, + "grad_norm": 1.173663854598999, + "learning_rate": 0.001188, + "loss": 5.6242, + "step": 396 + }, + { + "epoch": 0.00397, + "grad_norm": 0.828730046749115, + "learning_rate": 0.001191, + "loss": 5.6071, + "step": 397 + }, + { + "epoch": 0.00398, + "grad_norm": 1.2022807598114014, + "learning_rate": 0.0011940000000000002, + "loss": 5.6278, + "step": 398 + }, + { + "epoch": 0.00399, + "grad_norm": 0.9577529430389404, + "learning_rate": 0.0011970000000000001, + "loss": 5.6234, + "step": 399 + }, + { + "epoch": 0.004, + "grad_norm": 1.0115303993225098, + "learning_rate": 0.0012000000000000001, + "loss": 5.6174, + "step": 400 + }, + { + "epoch": 0.00401, + "grad_norm": 0.9447324872016907, + "learning_rate": 0.001203, + "loss": 5.594, + "step": 401 + }, + { + "epoch": 0.00402, + "grad_norm": 0.8248692154884338, + "learning_rate": 0.001206, + "loss": 5.5987, + "step": 402 + }, + { + "epoch": 0.00403, + "grad_norm": 1.0206302404403687, + "learning_rate": 0.001209, + "loss": 5.6026, + "step": 403 + }, + { + "epoch": 0.00404, + "grad_norm": 1.090654730796814, + "learning_rate": 0.0012120000000000002, + "loss": 5.6036, + "step": 404 + }, + { + "epoch": 0.00405, + "grad_norm": 0.84739750623703, + "learning_rate": 0.0012150000000000002, + "loss": 5.5901, + "step": 405 + }, + { + "epoch": 0.00406, + "grad_norm": 0.9469770193099976, + "learning_rate": 0.0012180000000000001, + "loss": 5.5817, + "step": 406 + }, + { + "epoch": 0.00407, + "grad_norm": 0.9283969402313232, + "learning_rate": 0.0012209999999999999, + "loss": 5.5746, + "step": 407 + }, + { + "epoch": 0.00408, + "grad_norm": 0.8949937224388123, + "learning_rate": 0.001224, + "loss": 5.5962, + "step": 408 + }, + { + "epoch": 0.00409, + "grad_norm": 0.8844306468963623, + "learning_rate": 0.001227, + "loss": 5.573, + "step": 409 + }, + { + "epoch": 0.0041, + "grad_norm": 0.8062122464179993, + "learning_rate": 0.00123, + "loss": 5.5618, + "step": 410 + }, + { + "epoch": 0.00411, + "grad_norm": 0.8730618357658386, + "learning_rate": 0.001233, + "loss": 5.5872, + "step": 411 + }, + { + "epoch": 0.00412, + "grad_norm": 0.665547788143158, + "learning_rate": 0.001236, + "loss": 5.5644, + "step": 412 + }, + { + "epoch": 0.00413, + "grad_norm": 0.8426138758659363, + "learning_rate": 0.0012389999999999999, + "loss": 5.5674, + "step": 413 + }, + { + "epoch": 0.00414, + "grad_norm": 1.051952838897705, + "learning_rate": 0.001242, + "loss": 5.5591, + "step": 414 + }, + { + "epoch": 0.00415, + "grad_norm": 0.7785534858703613, + "learning_rate": 0.001245, + "loss": 5.5542, + "step": 415 + }, + { + "epoch": 0.00416, + "grad_norm": 0.5227160453796387, + "learning_rate": 0.001248, + "loss": 5.5279, + "step": 416 + }, + { + "epoch": 0.00417, + "grad_norm": 0.5770328044891357, + "learning_rate": 0.001251, + "loss": 5.5611, + "step": 417 + }, + { + "epoch": 0.00418, + "grad_norm": 0.4929839074611664, + "learning_rate": 0.001254, + "loss": 5.5305, + "step": 418 + }, + { + "epoch": 0.00419, + "grad_norm": 0.4660792648792267, + "learning_rate": 0.0012569999999999999, + "loss": 5.5168, + "step": 419 + }, + { + "epoch": 0.0042, + "grad_norm": 0.5160586833953857, + "learning_rate": 0.00126, + "loss": 5.5326, + "step": 420 + }, + { + "epoch": 0.00421, + "grad_norm": 0.5846797823905945, + "learning_rate": 0.001263, + "loss": 5.5249, + "step": 421 + }, + { + "epoch": 0.00422, + "grad_norm": 0.6270997524261475, + "learning_rate": 0.001266, + "loss": 5.5159, + "step": 422 + }, + { + "epoch": 0.00423, + "grad_norm": 0.6081735491752625, + "learning_rate": 0.001269, + "loss": 5.5118, + "step": 423 + }, + { + "epoch": 0.00424, + "grad_norm": 0.557420551776886, + "learning_rate": 0.001272, + "loss": 5.5105, + "step": 424 + }, + { + "epoch": 0.00425, + "grad_norm": 0.821638286113739, + "learning_rate": 0.001275, + "loss": 5.5176, + "step": 425 + }, + { + "epoch": 0.00426, + "grad_norm": 1.0497279167175293, + "learning_rate": 0.001278, + "loss": 5.5294, + "step": 426 + }, + { + "epoch": 0.00427, + "grad_norm": 0.8568355441093445, + "learning_rate": 0.001281, + "loss": 5.5072, + "step": 427 + }, + { + "epoch": 0.00428, + "grad_norm": 0.9392327070236206, + "learning_rate": 0.001284, + "loss": 5.488, + "step": 428 + }, + { + "epoch": 0.00429, + "grad_norm": 0.8972091674804688, + "learning_rate": 0.001287, + "loss": 5.5041, + "step": 429 + }, + { + "epoch": 0.0043, + "grad_norm": 0.8478754162788391, + "learning_rate": 0.00129, + "loss": 5.4993, + "step": 430 + }, + { + "epoch": 0.00431, + "grad_norm": 1.0406945943832397, + "learning_rate": 0.001293, + "loss": 5.4959, + "step": 431 + }, + { + "epoch": 0.00432, + "grad_norm": 0.8937470316886902, + "learning_rate": 0.001296, + "loss": 5.5023, + "step": 432 + }, + { + "epoch": 0.00433, + "grad_norm": 0.7159745693206787, + "learning_rate": 0.001299, + "loss": 5.4892, + "step": 433 + }, + { + "epoch": 0.00434, + "grad_norm": 0.6872638463973999, + "learning_rate": 0.001302, + "loss": 5.4897, + "step": 434 + }, + { + "epoch": 0.00435, + "grad_norm": 0.7336323857307434, + "learning_rate": 0.001305, + "loss": 5.4832, + "step": 435 + }, + { + "epoch": 0.00436, + "grad_norm": 0.727497935295105, + "learning_rate": 0.001308, + "loss": 5.4615, + "step": 436 + }, + { + "epoch": 0.00437, + "grad_norm": 0.8729015588760376, + "learning_rate": 0.001311, + "loss": 5.4821, + "step": 437 + }, + { + "epoch": 0.00438, + "grad_norm": 1.0639538764953613, + "learning_rate": 0.001314, + "loss": 5.4733, + "step": 438 + }, + { + "epoch": 0.00439, + "grad_norm": 0.9613514542579651, + "learning_rate": 0.001317, + "loss": 5.5005, + "step": 439 + }, + { + "epoch": 0.0044, + "grad_norm": 0.7993902564048767, + "learning_rate": 0.00132, + "loss": 5.4683, + "step": 440 + }, + { + "epoch": 0.00441, + "grad_norm": 0.6981948018074036, + "learning_rate": 0.001323, + "loss": 5.4617, + "step": 441 + }, + { + "epoch": 0.00442, + "grad_norm": 0.5610657334327698, + "learning_rate": 0.0013260000000000001, + "loss": 5.4493, + "step": 442 + }, + { + "epoch": 0.00443, + "grad_norm": 0.6162020564079285, + "learning_rate": 0.001329, + "loss": 5.4545, + "step": 443 + }, + { + "epoch": 0.00444, + "grad_norm": 0.5797529816627502, + "learning_rate": 0.001332, + "loss": 5.4538, + "step": 444 + }, + { + "epoch": 0.00445, + "grad_norm": 0.551798939704895, + "learning_rate": 0.001335, + "loss": 5.4358, + "step": 445 + }, + { + "epoch": 0.00446, + "grad_norm": 0.48300743103027344, + "learning_rate": 0.001338, + "loss": 5.4369, + "step": 446 + }, + { + "epoch": 0.00447, + "grad_norm": 0.5713039040565491, + "learning_rate": 0.001341, + "loss": 5.4366, + "step": 447 + }, + { + "epoch": 0.00448, + "grad_norm": 0.7566826939582825, + "learning_rate": 0.0013440000000000001, + "loss": 5.4234, + "step": 448 + }, + { + "epoch": 0.00449, + "grad_norm": 1.1563501358032227, + "learning_rate": 0.001347, + "loss": 5.4418, + "step": 449 + }, + { + "epoch": 0.0045, + "grad_norm": 1.132352590560913, + "learning_rate": 0.00135, + "loss": 5.4433, + "step": 450 + }, + { + "epoch": 0.00451, + "grad_norm": 1.0986182689666748, + "learning_rate": 0.001353, + "loss": 5.4341, + "step": 451 + }, + { + "epoch": 0.00452, + "grad_norm": 1.066072702407837, + "learning_rate": 0.001356, + "loss": 5.4212, + "step": 452 + }, + { + "epoch": 0.00453, + "grad_norm": 0.9297358989715576, + "learning_rate": 0.001359, + "loss": 5.4103, + "step": 453 + }, + { + "epoch": 0.00454, + "grad_norm": 0.9204379320144653, + "learning_rate": 0.0013620000000000001, + "loss": 5.4294, + "step": 454 + }, + { + "epoch": 0.00455, + "grad_norm": 1.0156644582748413, + "learning_rate": 0.0013650000000000001, + "loss": 5.4283, + "step": 455 + }, + { + "epoch": 0.00456, + "grad_norm": 1.070080041885376, + "learning_rate": 0.001368, + "loss": 5.4307, + "step": 456 + }, + { + "epoch": 0.00457, + "grad_norm": 0.7985509634017944, + "learning_rate": 0.001371, + "loss": 5.4131, + "step": 457 + }, + { + "epoch": 0.00458, + "grad_norm": 0.8899184465408325, + "learning_rate": 0.001374, + "loss": 5.4253, + "step": 458 + }, + { + "epoch": 0.00459, + "grad_norm": 1.0247424840927124, + "learning_rate": 0.0013770000000000002, + "loss": 5.4196, + "step": 459 + }, + { + "epoch": 0.0046, + "grad_norm": 0.8818691968917847, + "learning_rate": 0.0013800000000000002, + "loss": 5.4023, + "step": 460 + }, + { + "epoch": 0.00461, + "grad_norm": 0.9398977160453796, + "learning_rate": 0.0013830000000000001, + "loss": 5.4028, + "step": 461 + }, + { + "epoch": 0.00462, + "grad_norm": 0.9095609784126282, + "learning_rate": 0.001386, + "loss": 5.4141, + "step": 462 + }, + { + "epoch": 0.00463, + "grad_norm": 0.7167875170707703, + "learning_rate": 0.001389, + "loss": 5.3816, + "step": 463 + }, + { + "epoch": 0.00464, + "grad_norm": 0.8864797353744507, + "learning_rate": 0.001392, + "loss": 5.4115, + "step": 464 + }, + { + "epoch": 0.00465, + "grad_norm": 0.8739372491836548, + "learning_rate": 0.0013950000000000002, + "loss": 5.3819, + "step": 465 + }, + { + "epoch": 0.00466, + "grad_norm": 0.808113157749176, + "learning_rate": 0.0013980000000000002, + "loss": 5.3947, + "step": 466 + }, + { + "epoch": 0.00467, + "grad_norm": 0.878140389919281, + "learning_rate": 0.0014010000000000001, + "loss": 5.3861, + "step": 467 + }, + { + "epoch": 0.00468, + "grad_norm": 1.0618577003479004, + "learning_rate": 0.001404, + "loss": 5.3874, + "step": 468 + }, + { + "epoch": 0.00469, + "grad_norm": 0.8678603172302246, + "learning_rate": 0.001407, + "loss": 5.3793, + "step": 469 + }, + { + "epoch": 0.0047, + "grad_norm": 1.0598393678665161, + "learning_rate": 0.00141, + "loss": 5.397, + "step": 470 + }, + { + "epoch": 0.00471, + "grad_norm": 0.9507164359092712, + "learning_rate": 0.001413, + "loss": 5.3744, + "step": 471 + }, + { + "epoch": 0.00472, + "grad_norm": 0.891309916973114, + "learning_rate": 0.001416, + "loss": 5.3876, + "step": 472 + }, + { + "epoch": 0.00473, + "grad_norm": 0.9032427072525024, + "learning_rate": 0.001419, + "loss": 5.3933, + "step": 473 + }, + { + "epoch": 0.00474, + "grad_norm": 1.2588310241699219, + "learning_rate": 0.0014219999999999999, + "loss": 5.3882, + "step": 474 + }, + { + "epoch": 0.00475, + "grad_norm": 0.8014562129974365, + "learning_rate": 0.001425, + "loss": 5.367, + "step": 475 + }, + { + "epoch": 0.00476, + "grad_norm": 0.7612058520317078, + "learning_rate": 0.001428, + "loss": 5.3988, + "step": 476 + }, + { + "epoch": 0.00477, + "grad_norm": 0.6699860095977783, + "learning_rate": 0.001431, + "loss": 5.3462, + "step": 477 + }, + { + "epoch": 0.00478, + "grad_norm": 0.7476372718811035, + "learning_rate": 0.001434, + "loss": 5.3761, + "step": 478 + }, + { + "epoch": 0.00479, + "grad_norm": 0.7134982347488403, + "learning_rate": 0.001437, + "loss": 5.3522, + "step": 479 + }, + { + "epoch": 0.0048, + "grad_norm": 0.5948371887207031, + "learning_rate": 0.0014399999999999999, + "loss": 5.343, + "step": 480 + }, + { + "epoch": 0.00481, + "grad_norm": 0.4980184733867645, + "learning_rate": 0.001443, + "loss": 5.3421, + "step": 481 + }, + { + "epoch": 0.00482, + "grad_norm": 0.49235713481903076, + "learning_rate": 0.001446, + "loss": 5.3164, + "step": 482 + }, + { + "epoch": 0.00483, + "grad_norm": 0.5064442157745361, + "learning_rate": 0.001449, + "loss": 5.327, + "step": 483 + }, + { + "epoch": 0.00484, + "grad_norm": 0.506096601486206, + "learning_rate": 0.001452, + "loss": 5.3265, + "step": 484 + }, + { + "epoch": 0.00485, + "grad_norm": 0.5634677410125732, + "learning_rate": 0.001455, + "loss": 5.3337, + "step": 485 + }, + { + "epoch": 0.00486, + "grad_norm": 0.5971659421920776, + "learning_rate": 0.001458, + "loss": 5.306, + "step": 486 + }, + { + "epoch": 0.00487, + "grad_norm": 0.5582761168479919, + "learning_rate": 0.001461, + "loss": 5.3211, + "step": 487 + }, + { + "epoch": 0.00488, + "grad_norm": 0.5640081763267517, + "learning_rate": 0.001464, + "loss": 5.2971, + "step": 488 + }, + { + "epoch": 0.00489, + "grad_norm": 0.6793756484985352, + "learning_rate": 0.001467, + "loss": 5.2905, + "step": 489 + }, + { + "epoch": 0.0049, + "grad_norm": 0.7497550249099731, + "learning_rate": 0.00147, + "loss": 5.3082, + "step": 490 + }, + { + "epoch": 0.00491, + "grad_norm": 0.7264507412910461, + "learning_rate": 0.001473, + "loss": 5.3062, + "step": 491 + }, + { + "epoch": 0.00492, + "grad_norm": 0.5965330004692078, + "learning_rate": 0.001476, + "loss": 5.3011, + "step": 492 + }, + { + "epoch": 0.00493, + "grad_norm": 0.8212659358978271, + "learning_rate": 0.001479, + "loss": 5.3006, + "step": 493 + }, + { + "epoch": 0.00494, + "grad_norm": 1.2706849575042725, + "learning_rate": 0.001482, + "loss": 5.3221, + "step": 494 + }, + { + "epoch": 0.00495, + "grad_norm": 0.9726585149765015, + "learning_rate": 0.001485, + "loss": 5.3189, + "step": 495 + }, + { + "epoch": 0.00496, + "grad_norm": 1.0229647159576416, + "learning_rate": 0.001488, + "loss": 5.3095, + "step": 496 + }, + { + "epoch": 0.00497, + "grad_norm": 0.9751450419425964, + "learning_rate": 0.001491, + "loss": 5.294, + "step": 497 + }, + { + "epoch": 0.00498, + "grad_norm": 0.9788212776184082, + "learning_rate": 0.001494, + "loss": 5.3219, + "step": 498 + }, + { + "epoch": 0.00499, + "grad_norm": 0.897365391254425, + "learning_rate": 0.001497, + "loss": 5.2943, + "step": 499 + }, + { + "epoch": 0.005, + "grad_norm": 0.8972038626670837, + "learning_rate": 0.0015, + "loss": 5.2998, + "step": 500 + }, + { + "epoch": 0.00501, + "grad_norm": 1.123322606086731, + "learning_rate": 0.001503, + "loss": 5.3107, + "step": 501 + }, + { + "epoch": 0.00502, + "grad_norm": 1.085119366645813, + "learning_rate": 0.001506, + "loss": 5.3147, + "step": 502 + }, + { + "epoch": 0.00503, + "grad_norm": 0.9612423181533813, + "learning_rate": 0.0015090000000000001, + "loss": 5.3162, + "step": 503 + }, + { + "epoch": 0.00504, + "grad_norm": 1.0992624759674072, + "learning_rate": 0.001512, + "loss": 5.3083, + "step": 504 + }, + { + "epoch": 0.00505, + "grad_norm": 0.9857581257820129, + "learning_rate": 0.001515, + "loss": 5.2946, + "step": 505 + }, + { + "epoch": 0.00506, + "grad_norm": 1.1049542427062988, + "learning_rate": 0.001518, + "loss": 5.3003, + "step": 506 + }, + { + "epoch": 0.00507, + "grad_norm": 0.8998914957046509, + "learning_rate": 0.001521, + "loss": 5.298, + "step": 507 + }, + { + "epoch": 0.00508, + "grad_norm": 0.6991334557533264, + "learning_rate": 0.001524, + "loss": 5.2862, + "step": 508 + }, + { + "epoch": 0.00509, + "grad_norm": 0.7653549313545227, + "learning_rate": 0.0015270000000000001, + "loss": 5.2723, + "step": 509 + }, + { + "epoch": 0.0051, + "grad_norm": 0.7315691113471985, + "learning_rate": 0.0015300000000000001, + "loss": 5.288, + "step": 510 + }, + { + "epoch": 0.00511, + "grad_norm": 0.7975103855133057, + "learning_rate": 0.001533, + "loss": 5.2648, + "step": 511 + }, + { + "epoch": 0.00512, + "grad_norm": 0.9781049489974976, + "learning_rate": 0.001536, + "loss": 5.2672, + "step": 512 + }, + { + "epoch": 0.00513, + "grad_norm": 1.084666132926941, + "learning_rate": 0.001539, + "loss": 5.2896, + "step": 513 + }, + { + "epoch": 0.00514, + "grad_norm": 0.9010921120643616, + "learning_rate": 0.001542, + "loss": 5.2706, + "step": 514 + }, + { + "epoch": 0.00515, + "grad_norm": 0.96586012840271, + "learning_rate": 0.0015450000000000001, + "loss": 5.2764, + "step": 515 + }, + { + "epoch": 0.00516, + "grad_norm": 0.9655681848526001, + "learning_rate": 0.0015480000000000001, + "loss": 5.2769, + "step": 516 + }, + { + "epoch": 0.00517, + "grad_norm": 0.8448560833930969, + "learning_rate": 0.001551, + "loss": 5.2634, + "step": 517 + }, + { + "epoch": 0.00518, + "grad_norm": 0.7809770703315735, + "learning_rate": 0.001554, + "loss": 5.2515, + "step": 518 + }, + { + "epoch": 0.00519, + "grad_norm": 0.913107693195343, + "learning_rate": 0.001557, + "loss": 5.2572, + "step": 519 + }, + { + "epoch": 0.0052, + "grad_norm": 0.9221185445785522, + "learning_rate": 0.0015600000000000002, + "loss": 5.2658, + "step": 520 + }, + { + "epoch": 0.00521, + "grad_norm": 0.8511962294578552, + "learning_rate": 0.0015630000000000002, + "loss": 5.2423, + "step": 521 + }, + { + "epoch": 0.00522, + "grad_norm": 0.7266805171966553, + "learning_rate": 0.0015660000000000001, + "loss": 5.2383, + "step": 522 + }, + { + "epoch": 0.00523, + "grad_norm": 0.8215247392654419, + "learning_rate": 0.001569, + "loss": 5.2547, + "step": 523 + }, + { + "epoch": 0.00524, + "grad_norm": 0.8742693066596985, + "learning_rate": 0.001572, + "loss": 5.2653, + "step": 524 + }, + { + "epoch": 0.00525, + "grad_norm": 0.6882407069206238, + "learning_rate": 0.001575, + "loss": 5.2501, + "step": 525 + }, + { + "epoch": 0.00526, + "grad_norm": 0.7224147319793701, + "learning_rate": 0.0015780000000000002, + "loss": 5.2254, + "step": 526 + }, + { + "epoch": 0.00527, + "grad_norm": 0.6562958359718323, + "learning_rate": 0.0015810000000000002, + "loss": 5.2385, + "step": 527 + }, + { + "epoch": 0.00528, + "grad_norm": 0.6051112413406372, + "learning_rate": 0.0015840000000000001, + "loss": 5.2165, + "step": 528 + }, + { + "epoch": 0.00529, + "grad_norm": 0.647803008556366, + "learning_rate": 0.001587, + "loss": 5.2244, + "step": 529 + }, + { + "epoch": 0.0053, + "grad_norm": 0.6608071327209473, + "learning_rate": 0.00159, + "loss": 5.2339, + "step": 530 + }, + { + "epoch": 0.00531, + "grad_norm": 0.6765715479850769, + "learning_rate": 0.001593, + "loss": 5.2181, + "step": 531 + }, + { + "epoch": 0.00532, + "grad_norm": 0.7380223870277405, + "learning_rate": 0.0015960000000000002, + "loss": 5.2256, + "step": 532 + }, + { + "epoch": 0.00533, + "grad_norm": 0.7506837248802185, + "learning_rate": 0.0015990000000000002, + "loss": 5.2236, + "step": 533 + }, + { + "epoch": 0.00534, + "grad_norm": 0.577497661113739, + "learning_rate": 0.0016020000000000001, + "loss": 5.2066, + "step": 534 + }, + { + "epoch": 0.00535, + "grad_norm": 0.4974853992462158, + "learning_rate": 0.001605, + "loss": 5.2096, + "step": 535 + }, + { + "epoch": 0.00536, + "grad_norm": 0.5754765272140503, + "learning_rate": 0.001608, + "loss": 5.2121, + "step": 536 + }, + { + "epoch": 0.00537, + "grad_norm": 0.6681102514266968, + "learning_rate": 0.0016110000000000002, + "loss": 5.2067, + "step": 537 + }, + { + "epoch": 0.00538, + "grad_norm": 0.8286970257759094, + "learning_rate": 0.0016140000000000002, + "loss": 5.1882, + "step": 538 + }, + { + "epoch": 0.00539, + "grad_norm": 1.0212045907974243, + "learning_rate": 0.0016170000000000002, + "loss": 5.193, + "step": 539 + }, + { + "epoch": 0.0054, + "grad_norm": 1.0495171546936035, + "learning_rate": 0.0016200000000000001, + "loss": 5.2061, + "step": 540 + }, + { + "epoch": 0.00541, + "grad_norm": 0.9756328463554382, + "learning_rate": 0.001623, + "loss": 5.1953, + "step": 541 + }, + { + "epoch": 0.00542, + "grad_norm": 1.024538278579712, + "learning_rate": 0.001626, + "loss": 5.2247, + "step": 542 + }, + { + "epoch": 0.00543, + "grad_norm": 1.005081057548523, + "learning_rate": 0.0016290000000000002, + "loss": 5.1883, + "step": 543 + }, + { + "epoch": 0.00544, + "grad_norm": 1.1540062427520752, + "learning_rate": 0.0016320000000000002, + "loss": 5.2159, + "step": 544 + }, + { + "epoch": 0.00545, + "grad_norm": 1.0779460668563843, + "learning_rate": 0.0016350000000000002, + "loss": 5.2271, + "step": 545 + }, + { + "epoch": 0.00546, + "grad_norm": 0.9231882691383362, + "learning_rate": 0.0016380000000000001, + "loss": 5.1917, + "step": 546 + }, + { + "epoch": 0.00547, + "grad_norm": 0.9463688731193542, + "learning_rate": 0.001641, + "loss": 5.1943, + "step": 547 + }, + { + "epoch": 0.00548, + "grad_norm": 0.8151195645332336, + "learning_rate": 0.001644, + "loss": 5.1898, + "step": 548 + }, + { + "epoch": 0.00549, + "grad_norm": 0.6324855089187622, + "learning_rate": 0.0016470000000000002, + "loss": 5.1802, + "step": 549 + }, + { + "epoch": 0.0055, + "grad_norm": 0.7012510299682617, + "learning_rate": 0.0016500000000000002, + "loss": 5.1963, + "step": 550 + }, + { + "epoch": 0.00551, + "grad_norm": 0.6119561195373535, + "learning_rate": 0.0016530000000000002, + "loss": 5.1867, + "step": 551 + }, + { + "epoch": 0.00552, + "grad_norm": 0.6565516591072083, + "learning_rate": 0.0016560000000000001, + "loss": 5.1814, + "step": 552 + }, + { + "epoch": 0.00553, + "grad_norm": 0.60141921043396, + "learning_rate": 0.001659, + "loss": 5.1716, + "step": 553 + }, + { + "epoch": 0.00554, + "grad_norm": 0.5632038116455078, + "learning_rate": 0.0016620000000000003, + "loss": 5.1664, + "step": 554 + }, + { + "epoch": 0.00555, + "grad_norm": 0.7644810080528259, + "learning_rate": 0.0016650000000000002, + "loss": 5.1782, + "step": 555 + }, + { + "epoch": 0.00556, + "grad_norm": 0.8742493987083435, + "learning_rate": 0.0016680000000000002, + "loss": 5.1659, + "step": 556 + }, + { + "epoch": 0.00557, + "grad_norm": 0.8544741272926331, + "learning_rate": 0.0016710000000000002, + "loss": 5.1502, + "step": 557 + }, + { + "epoch": 0.00558, + "grad_norm": 0.9127110838890076, + "learning_rate": 0.0016740000000000001, + "loss": 5.182, + "step": 558 + }, + { + "epoch": 0.00559, + "grad_norm": 0.9648676514625549, + "learning_rate": 0.001677, + "loss": 5.1678, + "step": 559 + }, + { + "epoch": 0.0056, + "grad_norm": 0.8821786642074585, + "learning_rate": 0.0016800000000000003, + "loss": 5.1427, + "step": 560 + }, + { + "epoch": 0.00561, + "grad_norm": 0.891691267490387, + "learning_rate": 0.0016830000000000003, + "loss": 5.1733, + "step": 561 + }, + { + "epoch": 0.00562, + "grad_norm": 0.9146907925605774, + "learning_rate": 0.0016860000000000002, + "loss": 5.1636, + "step": 562 + }, + { + "epoch": 0.00563, + "grad_norm": 0.9244825839996338, + "learning_rate": 0.001689, + "loss": 5.1722, + "step": 563 + }, + { + "epoch": 0.00564, + "grad_norm": 0.7839826345443726, + "learning_rate": 0.001692, + "loss": 5.1431, + "step": 564 + }, + { + "epoch": 0.00565, + "grad_norm": 1.062433123588562, + "learning_rate": 0.001695, + "loss": 5.1564, + "step": 565 + }, + { + "epoch": 0.00566, + "grad_norm": 0.8471325635910034, + "learning_rate": 0.0016979999999999999, + "loss": 5.1512, + "step": 566 + }, + { + "epoch": 0.00567, + "grad_norm": 0.8534058928489685, + "learning_rate": 0.0017009999999999998, + "loss": 5.1473, + "step": 567 + }, + { + "epoch": 0.00568, + "grad_norm": 0.8670461177825928, + "learning_rate": 0.0017039999999999998, + "loss": 5.1567, + "step": 568 + }, + { + "epoch": 0.00569, + "grad_norm": 0.9165869951248169, + "learning_rate": 0.001707, + "loss": 5.1375, + "step": 569 + }, + { + "epoch": 0.0057, + "grad_norm": 1.1377770900726318, + "learning_rate": 0.00171, + "loss": 5.1664, + "step": 570 + }, + { + "epoch": 0.00571, + "grad_norm": 0.8107508420944214, + "learning_rate": 0.001713, + "loss": 5.1589, + "step": 571 + }, + { + "epoch": 0.00572, + "grad_norm": 0.7913006544113159, + "learning_rate": 0.0017159999999999999, + "loss": 5.1345, + "step": 572 + }, + { + "epoch": 0.00573, + "grad_norm": 0.7625595927238464, + "learning_rate": 0.0017189999999999998, + "loss": 5.1388, + "step": 573 + }, + { + "epoch": 0.00574, + "grad_norm": 0.8708691596984863, + "learning_rate": 0.001722, + "loss": 5.1477, + "step": 574 + }, + { + "epoch": 0.00575, + "grad_norm": 0.8358116149902344, + "learning_rate": 0.001725, + "loss": 5.126, + "step": 575 + }, + { + "epoch": 0.00576, + "grad_norm": 0.7503964900970459, + "learning_rate": 0.001728, + "loss": 5.1165, + "step": 576 + }, + { + "epoch": 0.00577, + "grad_norm": 0.743698000907898, + "learning_rate": 0.001731, + "loss": 5.1434, + "step": 577 + }, + { + "epoch": 0.00578, + "grad_norm": 0.7937496900558472, + "learning_rate": 0.0017339999999999999, + "loss": 5.1222, + "step": 578 + }, + { + "epoch": 0.00579, + "grad_norm": 0.7887423038482666, + "learning_rate": 0.0017369999999999998, + "loss": 5.1084, + "step": 579 + }, + { + "epoch": 0.0058, + "grad_norm": 0.7772188186645508, + "learning_rate": 0.00174, + "loss": 5.1272, + "step": 580 + }, + { + "epoch": 0.00581, + "grad_norm": 0.8483501076698303, + "learning_rate": 0.001743, + "loss": 5.1251, + "step": 581 + }, + { + "epoch": 0.00582, + "grad_norm": 1.0748672485351562, + "learning_rate": 0.001746, + "loss": 5.1133, + "step": 582 + }, + { + "epoch": 0.00583, + "grad_norm": 0.9845912456512451, + "learning_rate": 0.001749, + "loss": 5.1338, + "step": 583 + }, + { + "epoch": 0.00584, + "grad_norm": 1.0171496868133545, + "learning_rate": 0.0017519999999999999, + "loss": 5.1328, + "step": 584 + }, + { + "epoch": 0.00585, + "grad_norm": 0.932063102722168, + "learning_rate": 0.0017549999999999998, + "loss": 5.1125, + "step": 585 + }, + { + "epoch": 0.00586, + "grad_norm": 1.0053131580352783, + "learning_rate": 0.001758, + "loss": 5.1277, + "step": 586 + }, + { + "epoch": 0.00587, + "grad_norm": 1.0553542375564575, + "learning_rate": 0.001761, + "loss": 5.1113, + "step": 587 + }, + { + "epoch": 0.00588, + "grad_norm": 0.9641870260238647, + "learning_rate": 0.001764, + "loss": 5.1209, + "step": 588 + }, + { + "epoch": 0.00589, + "grad_norm": 0.7223602533340454, + "learning_rate": 0.001767, + "loss": 5.1139, + "step": 589 + }, + { + "epoch": 0.0059, + "grad_norm": 0.7580032348632812, + "learning_rate": 0.0017699999999999999, + "loss": 5.1016, + "step": 590 + }, + { + "epoch": 0.00591, + "grad_norm": 0.7553709149360657, + "learning_rate": 0.001773, + "loss": 5.097, + "step": 591 + }, + { + "epoch": 0.00592, + "grad_norm": 0.7395292520523071, + "learning_rate": 0.001776, + "loss": 5.122, + "step": 592 + }, + { + "epoch": 0.00593, + "grad_norm": 0.6007040739059448, + "learning_rate": 0.001779, + "loss": 5.1118, + "step": 593 + }, + { + "epoch": 0.00594, + "grad_norm": 0.6126047372817993, + "learning_rate": 0.001782, + "loss": 5.1081, + "step": 594 + }, + { + "epoch": 0.00595, + "grad_norm": 0.607721745967865, + "learning_rate": 0.001785, + "loss": 5.0957, + "step": 595 + }, + { + "epoch": 0.00596, + "grad_norm": 0.6556451916694641, + "learning_rate": 0.0017879999999999999, + "loss": 5.0948, + "step": 596 + }, + { + "epoch": 0.00597, + "grad_norm": 0.7172878384590149, + "learning_rate": 0.001791, + "loss": 5.0729, + "step": 597 + }, + { + "epoch": 0.00598, + "grad_norm": 0.6043835878372192, + "learning_rate": 0.001794, + "loss": 5.0805, + "step": 598 + }, + { + "epoch": 0.00599, + "grad_norm": 0.5666232109069824, + "learning_rate": 0.001797, + "loss": 5.0796, + "step": 599 + }, + { + "epoch": 0.006, + "grad_norm": 0.5673431754112244, + "learning_rate": 0.0018, + "loss": 5.0764, + "step": 600 + }, + { + "epoch": 0.00601, + "grad_norm": 0.5798671841621399, + "learning_rate": 0.001803, + "loss": 5.0708, + "step": 601 + }, + { + "epoch": 0.00602, + "grad_norm": 0.520574152469635, + "learning_rate": 0.0018059999999999999, + "loss": 5.0544, + "step": 602 + }, + { + "epoch": 0.00603, + "grad_norm": 0.5210540294647217, + "learning_rate": 0.001809, + "loss": 5.0732, + "step": 603 + }, + { + "epoch": 0.00604, + "grad_norm": 0.6759857535362244, + "learning_rate": 0.001812, + "loss": 5.0576, + "step": 604 + }, + { + "epoch": 0.00605, + "grad_norm": 0.7568657994270325, + "learning_rate": 0.001815, + "loss": 5.0701, + "step": 605 + }, + { + "epoch": 0.00606, + "grad_norm": 0.7632762789726257, + "learning_rate": 0.001818, + "loss": 5.0711, + "step": 606 + }, + { + "epoch": 0.00607, + "grad_norm": 0.788451611995697, + "learning_rate": 0.001821, + "loss": 5.073, + "step": 607 + }, + { + "epoch": 0.00608, + "grad_norm": 0.763842761516571, + "learning_rate": 0.001824, + "loss": 5.0575, + "step": 608 + }, + { + "epoch": 0.00609, + "grad_norm": 0.825861930847168, + "learning_rate": 0.001827, + "loss": 5.0665, + "step": 609 + }, + { + "epoch": 0.0061, + "grad_norm": 1.0143935680389404, + "learning_rate": 0.00183, + "loss": 5.0488, + "step": 610 + }, + { + "epoch": 0.00611, + "grad_norm": 1.1116124391555786, + "learning_rate": 0.001833, + "loss": 5.0562, + "step": 611 + }, + { + "epoch": 0.00612, + "grad_norm": 1.0579830408096313, + "learning_rate": 0.001836, + "loss": 5.0621, + "step": 612 + }, + { + "epoch": 0.00613, + "grad_norm": 1.2180272340774536, + "learning_rate": 0.001839, + "loss": 5.069, + "step": 613 + }, + { + "epoch": 0.00614, + "grad_norm": 0.7525346875190735, + "learning_rate": 0.001842, + "loss": 5.0757, + "step": 614 + }, + { + "epoch": 0.00615, + "grad_norm": 0.833108127117157, + "learning_rate": 0.001845, + "loss": 5.0659, + "step": 615 + }, + { + "epoch": 0.00616, + "grad_norm": 0.7170072793960571, + "learning_rate": 0.001848, + "loss": 5.0473, + "step": 616 + }, + { + "epoch": 0.00617, + "grad_norm": 0.718910276889801, + "learning_rate": 0.001851, + "loss": 5.0608, + "step": 617 + }, + { + "epoch": 0.00618, + "grad_norm": 0.6572015881538391, + "learning_rate": 0.001854, + "loss": 5.0425, + "step": 618 + }, + { + "epoch": 0.00619, + "grad_norm": 0.7236103415489197, + "learning_rate": 0.001857, + "loss": 5.0504, + "step": 619 + }, + { + "epoch": 0.0062, + "grad_norm": 0.8058017492294312, + "learning_rate": 0.00186, + "loss": 5.0386, + "step": 620 + }, + { + "epoch": 0.00621, + "grad_norm": 0.861880898475647, + "learning_rate": 0.001863, + "loss": 5.0409, + "step": 621 + }, + { + "epoch": 0.00622, + "grad_norm": 1.0328248739242554, + "learning_rate": 0.001866, + "loss": 5.0489, + "step": 622 + }, + { + "epoch": 0.00623, + "grad_norm": 0.8494102358818054, + "learning_rate": 0.001869, + "loss": 5.0318, + "step": 623 + }, + { + "epoch": 0.00624, + "grad_norm": 0.9827755093574524, + "learning_rate": 0.001872, + "loss": 5.0476, + "step": 624 + }, + { + "epoch": 0.00625, + "grad_norm": 1.0741342306137085, + "learning_rate": 0.001875, + "loss": 5.0686, + "step": 625 + }, + { + "epoch": 0.00626, + "grad_norm": 0.7305473685264587, + "learning_rate": 0.0018780000000000001, + "loss": 5.0304, + "step": 626 + }, + { + "epoch": 0.00627, + "grad_norm": 0.8084409832954407, + "learning_rate": 0.001881, + "loss": 5.0255, + "step": 627 + }, + { + "epoch": 0.00628, + "grad_norm": 1.0902513265609741, + "learning_rate": 0.001884, + "loss": 5.0666, + "step": 628 + }, + { + "epoch": 0.00629, + "grad_norm": 1.036152958869934, + "learning_rate": 0.001887, + "loss": 5.0439, + "step": 629 + }, + { + "epoch": 0.0063, + "grad_norm": 1.0308260917663574, + "learning_rate": 0.00189, + "loss": 5.0356, + "step": 630 + }, + { + "epoch": 0.00631, + "grad_norm": 0.984308123588562, + "learning_rate": 0.0018930000000000002, + "loss": 5.0356, + "step": 631 + }, + { + "epoch": 0.00632, + "grad_norm": 0.8649469017982483, + "learning_rate": 0.0018960000000000001, + "loss": 5.0479, + "step": 632 + }, + { + "epoch": 0.00633, + "grad_norm": 0.8776420950889587, + "learning_rate": 0.001899, + "loss": 5.0459, + "step": 633 + }, + { + "epoch": 0.00634, + "grad_norm": 0.8304409980773926, + "learning_rate": 0.001902, + "loss": 5.037, + "step": 634 + }, + { + "epoch": 0.00635, + "grad_norm": 0.7966147065162659, + "learning_rate": 0.001905, + "loss": 5.0227, + "step": 635 + }, + { + "epoch": 0.00636, + "grad_norm": 0.9172542095184326, + "learning_rate": 0.001908, + "loss": 5.0337, + "step": 636 + }, + { + "epoch": 0.00637, + "grad_norm": 1.3219475746154785, + "learning_rate": 0.0019110000000000002, + "loss": 5.0715, + "step": 637 + }, + { + "epoch": 0.00638, + "grad_norm": 0.7924789190292358, + "learning_rate": 0.0019140000000000001, + "loss": 5.0347, + "step": 638 + }, + { + "epoch": 0.00639, + "grad_norm": 0.8469759225845337, + "learning_rate": 0.001917, + "loss": 5.0139, + "step": 639 + }, + { + "epoch": 0.0064, + "grad_norm": 0.9544380307197571, + "learning_rate": 0.00192, + "loss": 5.0191, + "step": 640 + }, + { + "epoch": 0.00641, + "grad_norm": 1.0876184701919556, + "learning_rate": 0.001923, + "loss": 5.0379, + "step": 641 + }, + { + "epoch": 0.00642, + "grad_norm": 0.8299298286437988, + "learning_rate": 0.001926, + "loss": 5.0146, + "step": 642 + }, + { + "epoch": 0.00643, + "grad_norm": 0.9603999257087708, + "learning_rate": 0.0019290000000000002, + "loss": 5.0399, + "step": 643 + }, + { + "epoch": 0.00644, + "grad_norm": 0.7977001070976257, + "learning_rate": 0.0019320000000000001, + "loss": 5.0094, + "step": 644 + }, + { + "epoch": 0.00645, + "grad_norm": 0.7244200706481934, + "learning_rate": 0.001935, + "loss": 5.0161, + "step": 645 + }, + { + "epoch": 0.00646, + "grad_norm": 0.7832402586936951, + "learning_rate": 0.001938, + "loss": 5.0193, + "step": 646 + }, + { + "epoch": 0.00647, + "grad_norm": 0.8586620092391968, + "learning_rate": 0.001941, + "loss": 5.0299, + "step": 647 + }, + { + "epoch": 0.00648, + "grad_norm": 0.8153418302536011, + "learning_rate": 0.0019440000000000002, + "loss": 5.0163, + "step": 648 + }, + { + "epoch": 0.00649, + "grad_norm": 0.766000509262085, + "learning_rate": 0.0019470000000000002, + "loss": 4.9982, + "step": 649 + }, + { + "epoch": 0.0065, + "grad_norm": 0.7875446677207947, + "learning_rate": 0.0019500000000000001, + "loss": 5.0238, + "step": 650 + }, + { + "epoch": 0.00651, + "grad_norm": 0.7245673537254333, + "learning_rate": 0.001953, + "loss": 5.0037, + "step": 651 + }, + { + "epoch": 0.00652, + "grad_norm": 0.7840576767921448, + "learning_rate": 0.0019560000000000003, + "loss": 4.9987, + "step": 652 + }, + { + "epoch": 0.00653, + "grad_norm": 0.8112754225730896, + "learning_rate": 0.0019590000000000002, + "loss": 4.9969, + "step": 653 + }, + { + "epoch": 0.00654, + "grad_norm": 0.6785946488380432, + "learning_rate": 0.001962, + "loss": 4.9786, + "step": 654 + }, + { + "epoch": 0.00655, + "grad_norm": 0.6141355633735657, + "learning_rate": 0.001965, + "loss": 4.9897, + "step": 655 + }, + { + "epoch": 0.00656, + "grad_norm": 0.647098958492279, + "learning_rate": 0.001968, + "loss": 4.982, + "step": 656 + }, + { + "epoch": 0.00657, + "grad_norm": 0.6601396203041077, + "learning_rate": 0.001971, + "loss": 4.9729, + "step": 657 + }, + { + "epoch": 0.00658, + "grad_norm": 0.5589995384216309, + "learning_rate": 0.001974, + "loss": 4.9927, + "step": 658 + }, + { + "epoch": 0.00659, + "grad_norm": 0.551213264465332, + "learning_rate": 0.001977, + "loss": 4.972, + "step": 659 + }, + { + "epoch": 0.0066, + "grad_norm": 0.5106261968612671, + "learning_rate": 0.00198, + "loss": 4.963, + "step": 660 + }, + { + "epoch": 0.00661, + "grad_norm": 0.49552980065345764, + "learning_rate": 0.001983, + "loss": 4.9449, + "step": 661 + }, + { + "epoch": 0.00662, + "grad_norm": 0.4502437114715576, + "learning_rate": 0.0019860000000000004, + "loss": 4.9438, + "step": 662 + }, + { + "epoch": 0.00663, + "grad_norm": 0.4717095196247101, + "learning_rate": 0.0019890000000000003, + "loss": 4.9599, + "step": 663 + }, + { + "epoch": 0.00664, + "grad_norm": 0.5179165005683899, + "learning_rate": 0.0019920000000000003, + "loss": 4.9431, + "step": 664 + }, + { + "epoch": 0.00665, + "grad_norm": 0.6156288385391235, + "learning_rate": 0.0019950000000000002, + "loss": 4.9222, + "step": 665 + }, + { + "epoch": 0.00666, + "grad_norm": 0.6403276920318604, + "learning_rate": 0.001998, + "loss": 4.9456, + "step": 666 + }, + { + "epoch": 0.00667, + "grad_norm": 0.6612551808357239, + "learning_rate": 0.002001, + "loss": 4.9436, + "step": 667 + }, + { + "epoch": 0.00668, + "grad_norm": 0.6370317339897156, + "learning_rate": 0.002004, + "loss": 4.9618, + "step": 668 + }, + { + "epoch": 0.00669, + "grad_norm": 0.6632084250450134, + "learning_rate": 0.002007, + "loss": 4.9618, + "step": 669 + }, + { + "epoch": 0.0067, + "grad_norm": 0.6647160649299622, + "learning_rate": 0.00201, + "loss": 4.9534, + "step": 670 + }, + { + "epoch": 0.00671, + "grad_norm": 0.6171524524688721, + "learning_rate": 0.002013, + "loss": 4.9269, + "step": 671 + }, + { + "epoch": 0.00672, + "grad_norm": 0.6804357171058655, + "learning_rate": 0.002016, + "loss": 4.9362, + "step": 672 + }, + { + "epoch": 0.00673, + "grad_norm": 0.7436751127243042, + "learning_rate": 0.002019, + "loss": 4.9351, + "step": 673 + }, + { + "epoch": 0.00674, + "grad_norm": 0.8610292077064514, + "learning_rate": 0.0020220000000000004, + "loss": 4.9433, + "step": 674 + }, + { + "epoch": 0.00675, + "grad_norm": 1.046964168548584, + "learning_rate": 0.0020250000000000003, + "loss": 4.9648, + "step": 675 + }, + { + "epoch": 0.00676, + "grad_norm": 0.9578864574432373, + "learning_rate": 0.0020280000000000003, + "loss": 4.9492, + "step": 676 + }, + { + "epoch": 0.00677, + "grad_norm": 1.1337146759033203, + "learning_rate": 0.0020310000000000003, + "loss": 4.9695, + "step": 677 + }, + { + "epoch": 0.00678, + "grad_norm": 1.0358091592788696, + "learning_rate": 0.0020340000000000002, + "loss": 4.9763, + "step": 678 + }, + { + "epoch": 0.00679, + "grad_norm": 1.230987787246704, + "learning_rate": 0.002037, + "loss": 4.9735, + "step": 679 + }, + { + "epoch": 0.0068, + "grad_norm": 0.9104715585708618, + "learning_rate": 0.00204, + "loss": 4.9643, + "step": 680 + }, + { + "epoch": 0.00681, + "grad_norm": 1.1940183639526367, + "learning_rate": 0.002043, + "loss": 4.9659, + "step": 681 + }, + { + "epoch": 0.00682, + "grad_norm": 1.0067143440246582, + "learning_rate": 0.002046, + "loss": 4.9735, + "step": 682 + }, + { + "epoch": 0.00683, + "grad_norm": 1.224305510520935, + "learning_rate": 0.002049, + "loss": 4.9612, + "step": 683 + }, + { + "epoch": 0.00684, + "grad_norm": 0.9917755126953125, + "learning_rate": 0.002052, + "loss": 4.9694, + "step": 684 + }, + { + "epoch": 0.00685, + "grad_norm": 0.9187195897102356, + "learning_rate": 0.0020550000000000004, + "loss": 4.964, + "step": 685 + }, + { + "epoch": 0.00686, + "grad_norm": 1.035937786102295, + "learning_rate": 0.0020580000000000004, + "loss": 4.9577, + "step": 686 + }, + { + "epoch": 0.00687, + "grad_norm": 1.0923206806182861, + "learning_rate": 0.0020610000000000003, + "loss": 4.9579, + "step": 687 + }, + { + "epoch": 0.00688, + "grad_norm": 0.8355166912078857, + "learning_rate": 0.002064, + "loss": 4.9524, + "step": 688 + }, + { + "epoch": 0.00689, + "grad_norm": 0.7577447891235352, + "learning_rate": 0.002067, + "loss": 4.9417, + "step": 689 + }, + { + "epoch": 0.0069, + "grad_norm": 0.7621678709983826, + "learning_rate": 0.00207, + "loss": 4.9421, + "step": 690 + }, + { + "epoch": 0.00691, + "grad_norm": 0.6061983108520508, + "learning_rate": 0.0020729999999999998, + "loss": 4.9082, + "step": 691 + }, + { + "epoch": 0.00692, + "grad_norm": 0.591027021408081, + "learning_rate": 0.0020759999999999997, + "loss": 4.9292, + "step": 692 + }, + { + "epoch": 0.00693, + "grad_norm": 0.5834758281707764, + "learning_rate": 0.0020789999999999997, + "loss": 4.9222, + "step": 693 + }, + { + "epoch": 0.00694, + "grad_norm": 0.7208871841430664, + "learning_rate": 0.002082, + "loss": 4.9071, + "step": 694 + }, + { + "epoch": 0.00695, + "grad_norm": 0.8771729469299316, + "learning_rate": 0.002085, + "loss": 4.9171, + "step": 695 + }, + { + "epoch": 0.00696, + "grad_norm": 0.9051836133003235, + "learning_rate": 0.002088, + "loss": 4.9182, + "step": 696 + }, + { + "epoch": 0.00697, + "grad_norm": 1.1665294170379639, + "learning_rate": 0.002091, + "loss": 4.9339, + "step": 697 + }, + { + "epoch": 0.00698, + "grad_norm": 1.092050313949585, + "learning_rate": 0.002094, + "loss": 4.9261, + "step": 698 + }, + { + "epoch": 0.00699, + "grad_norm": 0.8412545323371887, + "learning_rate": 0.002097, + "loss": 4.9128, + "step": 699 + }, + { + "epoch": 0.007, + "grad_norm": 0.6720849871635437, + "learning_rate": 0.0021, + "loss": 4.9176, + "step": 700 + }, + { + "epoch": 0.00701, + "grad_norm": 0.563408613204956, + "learning_rate": 0.002103, + "loss": 4.8944, + "step": 701 + }, + { + "epoch": 0.00702, + "grad_norm": 0.5357261300086975, + "learning_rate": 0.002106, + "loss": 4.893, + "step": 702 + }, + { + "epoch": 0.00703, + "grad_norm": 0.5667074918746948, + "learning_rate": 0.0021089999999999998, + "loss": 4.8766, + "step": 703 + }, + { + "epoch": 0.00704, + "grad_norm": 0.597253143787384, + "learning_rate": 0.0021119999999999997, + "loss": 4.8825, + "step": 704 + }, + { + "epoch": 0.00705, + "grad_norm": 0.6211616396903992, + "learning_rate": 0.002115, + "loss": 4.8877, + "step": 705 + }, + { + "epoch": 0.00706, + "grad_norm": 0.687544047832489, + "learning_rate": 0.002118, + "loss": 4.8929, + "step": 706 + }, + { + "epoch": 0.00707, + "grad_norm": 0.7621776461601257, + "learning_rate": 0.002121, + "loss": 4.8688, + "step": 707 + }, + { + "epoch": 0.00708, + "grad_norm": 0.9027195572853088, + "learning_rate": 0.002124, + "loss": 4.8567, + "step": 708 + }, + { + "epoch": 0.00709, + "grad_norm": 0.9446965456008911, + "learning_rate": 0.002127, + "loss": 4.8935, + "step": 709 + }, + { + "epoch": 0.0071, + "grad_norm": 0.8206554055213928, + "learning_rate": 0.00213, + "loss": 4.8706, + "step": 710 + }, + { + "epoch": 0.00711, + "grad_norm": 0.8629757165908813, + "learning_rate": 0.002133, + "loss": 4.8565, + "step": 711 + }, + { + "epoch": 0.00712, + "grad_norm": 0.7438434362411499, + "learning_rate": 0.002136, + "loss": 4.869, + "step": 712 + }, + { + "epoch": 0.00713, + "grad_norm": 0.7951372861862183, + "learning_rate": 0.002139, + "loss": 4.8846, + "step": 713 + }, + { + "epoch": 0.00714, + "grad_norm": 0.9020676016807556, + "learning_rate": 0.002142, + "loss": 4.8773, + "step": 714 + }, + { + "epoch": 0.00715, + "grad_norm": 1.064352035522461, + "learning_rate": 0.0021449999999999998, + "loss": 4.8624, + "step": 715 + }, + { + "epoch": 0.00716, + "grad_norm": 0.7318432927131653, + "learning_rate": 0.002148, + "loss": 4.8409, + "step": 716 + }, + { + "epoch": 0.00717, + "grad_norm": 0.886417031288147, + "learning_rate": 0.002151, + "loss": 4.8705, + "step": 717 + }, + { + "epoch": 0.00718, + "grad_norm": 0.8509985208511353, + "learning_rate": 0.002154, + "loss": 4.8473, + "step": 718 + }, + { + "epoch": 0.00719, + "grad_norm": 0.8979188203811646, + "learning_rate": 0.002157, + "loss": 4.8656, + "step": 719 + }, + { + "epoch": 0.0072, + "grad_norm": 1.0766488313674927, + "learning_rate": 0.00216, + "loss": 4.8622, + "step": 720 + }, + { + "epoch": 0.00721, + "grad_norm": 1.0330792665481567, + "learning_rate": 0.002163, + "loss": 4.8716, + "step": 721 + }, + { + "epoch": 0.00722, + "grad_norm": 1.097432255744934, + "learning_rate": 0.002166, + "loss": 4.8541, + "step": 722 + }, + { + "epoch": 0.00723, + "grad_norm": 1.0046334266662598, + "learning_rate": 0.002169, + "loss": 4.8596, + "step": 723 + }, + { + "epoch": 0.00724, + "grad_norm": 1.0316451787948608, + "learning_rate": 0.002172, + "loss": 4.8595, + "step": 724 + }, + { + "epoch": 0.00725, + "grad_norm": 0.7518482804298401, + "learning_rate": 0.002175, + "loss": 4.8464, + "step": 725 + }, + { + "epoch": 0.00726, + "grad_norm": 0.8229779005050659, + "learning_rate": 0.002178, + "loss": 4.8512, + "step": 726 + }, + { + "epoch": 0.00727, + "grad_norm": 0.9577723741531372, + "learning_rate": 0.0021809999999999998, + "loss": 4.841, + "step": 727 + }, + { + "epoch": 0.00728, + "grad_norm": 0.9698926210403442, + "learning_rate": 0.002184, + "loss": 4.8594, + "step": 728 + }, + { + "epoch": 0.00729, + "grad_norm": 1.1789532899856567, + "learning_rate": 0.002187, + "loss": 4.8599, + "step": 729 + }, + { + "epoch": 0.0073, + "grad_norm": 1.0923309326171875, + "learning_rate": 0.00219, + "loss": 4.8656, + "step": 730 + }, + { + "epoch": 0.00731, + "grad_norm": 1.0404242277145386, + "learning_rate": 0.002193, + "loss": 4.8544, + "step": 731 + }, + { + "epoch": 0.00732, + "grad_norm": 0.8924814462661743, + "learning_rate": 0.002196, + "loss": 4.8537, + "step": 732 + }, + { + "epoch": 0.00733, + "grad_norm": 0.7918877005577087, + "learning_rate": 0.002199, + "loss": 4.8326, + "step": 733 + }, + { + "epoch": 0.00734, + "grad_norm": 0.8231402635574341, + "learning_rate": 0.002202, + "loss": 4.8485, + "step": 734 + }, + { + "epoch": 0.00735, + "grad_norm": 1.1126084327697754, + "learning_rate": 0.002205, + "loss": 4.8661, + "step": 735 + }, + { + "epoch": 0.00736, + "grad_norm": 1.1928813457489014, + "learning_rate": 0.002208, + "loss": 4.857, + "step": 736 + }, + { + "epoch": 0.00737, + "grad_norm": 0.9305274486541748, + "learning_rate": 0.002211, + "loss": 4.8183, + "step": 737 + }, + { + "epoch": 0.00738, + "grad_norm": 1.2922184467315674, + "learning_rate": 0.002214, + "loss": 4.8313, + "step": 738 + }, + { + "epoch": 0.00739, + "grad_norm": 1.0224477052688599, + "learning_rate": 0.0022170000000000002, + "loss": 4.8302, + "step": 739 + }, + { + "epoch": 0.0074, + "grad_norm": 0.8254541158676147, + "learning_rate": 0.00222, + "loss": 4.8183, + "step": 740 + }, + { + "epoch": 0.00741, + "grad_norm": 0.8494399785995483, + "learning_rate": 0.002223, + "loss": 4.7936, + "step": 741 + }, + { + "epoch": 0.00742, + "grad_norm": 0.8097528219223022, + "learning_rate": 0.002226, + "loss": 4.8203, + "step": 742 + }, + { + "epoch": 0.00743, + "grad_norm": 0.7318201065063477, + "learning_rate": 0.002229, + "loss": 4.782, + "step": 743 + }, + { + "epoch": 0.00744, + "grad_norm": 0.770041286945343, + "learning_rate": 0.002232, + "loss": 4.7897, + "step": 744 + }, + { + "epoch": 0.00745, + "grad_norm": 0.8400176167488098, + "learning_rate": 0.002235, + "loss": 4.793, + "step": 745 + }, + { + "epoch": 0.00746, + "grad_norm": 0.8187500834465027, + "learning_rate": 0.002238, + "loss": 4.7927, + "step": 746 + }, + { + "epoch": 0.00747, + "grad_norm": 0.8396742343902588, + "learning_rate": 0.002241, + "loss": 4.7802, + "step": 747 + }, + { + "epoch": 0.00748, + "grad_norm": 0.8425725698471069, + "learning_rate": 0.002244, + "loss": 4.7863, + "step": 748 + }, + { + "epoch": 0.00749, + "grad_norm": 0.9793819189071655, + "learning_rate": 0.002247, + "loss": 4.771, + "step": 749 + }, + { + "epoch": 0.0075, + "grad_norm": 0.9097406268119812, + "learning_rate": 0.0022500000000000003, + "loss": 4.8036, + "step": 750 + }, + { + "epoch": 0.00751, + "grad_norm": 1.1564972400665283, + "learning_rate": 0.0022530000000000002, + "loss": 4.8008, + "step": 751 + }, + { + "epoch": 0.00752, + "grad_norm": 1.1269639730453491, + "learning_rate": 0.002256, + "loss": 4.7666, + "step": 752 + }, + { + "epoch": 0.00753, + "grad_norm": 0.7554891705513, + "learning_rate": 0.002259, + "loss": 4.7895, + "step": 753 + }, + { + "epoch": 0.00754, + "grad_norm": 0.6434245705604553, + "learning_rate": 0.002262, + "loss": 4.7936, + "step": 754 + }, + { + "epoch": 0.00755, + "grad_norm": 0.6667845249176025, + "learning_rate": 0.002265, + "loss": 4.7629, + "step": 755 + }, + { + "epoch": 0.00756, + "grad_norm": 0.6454316973686218, + "learning_rate": 0.002268, + "loss": 4.7431, + "step": 756 + }, + { + "epoch": 0.00757, + "grad_norm": 0.5992982387542725, + "learning_rate": 0.002271, + "loss": 4.7549, + "step": 757 + }, + { + "epoch": 0.00758, + "grad_norm": 0.5403345227241516, + "learning_rate": 0.002274, + "loss": 4.7325, + "step": 758 + }, + { + "epoch": 0.00759, + "grad_norm": 0.6695131063461304, + "learning_rate": 0.002277, + "loss": 4.754, + "step": 759 + }, + { + "epoch": 0.0076, + "grad_norm": 0.8114665150642395, + "learning_rate": 0.00228, + "loss": 4.7449, + "step": 760 + }, + { + "epoch": 0.00761, + "grad_norm": 0.8682258129119873, + "learning_rate": 0.002283, + "loss": 4.7184, + "step": 761 + }, + { + "epoch": 0.00762, + "grad_norm": 0.9619300961494446, + "learning_rate": 0.0022860000000000003, + "loss": 4.7088, + "step": 762 + }, + { + "epoch": 0.00763, + "grad_norm": 1.0561281442642212, + "learning_rate": 0.0022890000000000002, + "loss": 4.7527, + "step": 763 + }, + { + "epoch": 0.00764, + "grad_norm": 1.0605075359344482, + "learning_rate": 0.002292, + "loss": 4.7609, + "step": 764 + }, + { + "epoch": 0.00765, + "grad_norm": 0.8591569066047668, + "learning_rate": 0.002295, + "loss": 4.7468, + "step": 765 + }, + { + "epoch": 0.00766, + "grad_norm": 0.8702619075775146, + "learning_rate": 0.002298, + "loss": 4.7378, + "step": 766 + }, + { + "epoch": 0.00767, + "grad_norm": 0.9817199110984802, + "learning_rate": 0.002301, + "loss": 4.7481, + "step": 767 + }, + { + "epoch": 0.00768, + "grad_norm": 1.0741162300109863, + "learning_rate": 0.002304, + "loss": 4.7612, + "step": 768 + }, + { + "epoch": 0.00769, + "grad_norm": 0.8188871145248413, + "learning_rate": 0.002307, + "loss": 4.7592, + "step": 769 + }, + { + "epoch": 0.0077, + "grad_norm": 0.7658451795578003, + "learning_rate": 0.00231, + "loss": 4.7148, + "step": 770 + }, + { + "epoch": 0.00771, + "grad_norm": 0.8762615323066711, + "learning_rate": 0.002313, + "loss": 4.754, + "step": 771 + }, + { + "epoch": 0.00772, + "grad_norm": 0.9279188513755798, + "learning_rate": 0.002316, + "loss": 4.7389, + "step": 772 + }, + { + "epoch": 0.00773, + "grad_norm": 0.8405673503875732, + "learning_rate": 0.0023190000000000003, + "loss": 4.737, + "step": 773 + }, + { + "epoch": 0.00774, + "grad_norm": 0.777439296245575, + "learning_rate": 0.0023220000000000003, + "loss": 4.7452, + "step": 774 + }, + { + "epoch": 0.00775, + "grad_norm": 0.8980410099029541, + "learning_rate": 0.0023250000000000002, + "loss": 4.7436, + "step": 775 + }, + { + "epoch": 0.00776, + "grad_norm": 1.0412962436676025, + "learning_rate": 0.002328, + "loss": 4.7539, + "step": 776 + }, + { + "epoch": 0.00777, + "grad_norm": 0.9281516075134277, + "learning_rate": 0.002331, + "loss": 4.7268, + "step": 777 + }, + { + "epoch": 0.00778, + "grad_norm": 0.7459467649459839, + "learning_rate": 0.002334, + "loss": 4.7271, + "step": 778 + }, + { + "epoch": 0.00779, + "grad_norm": 0.7607200145721436, + "learning_rate": 0.002337, + "loss": 4.689, + "step": 779 + }, + { + "epoch": 0.0078, + "grad_norm": 0.7529038190841675, + "learning_rate": 0.00234, + "loss": 4.706, + "step": 780 + }, + { + "epoch": 0.00781, + "grad_norm": 0.7776694297790527, + "learning_rate": 0.002343, + "loss": 4.7072, + "step": 781 + }, + { + "epoch": 0.00782, + "grad_norm": 0.7648219466209412, + "learning_rate": 0.002346, + "loss": 4.6998, + "step": 782 + }, + { + "epoch": 0.00783, + "grad_norm": 0.8795627355575562, + "learning_rate": 0.002349, + "loss": 4.6893, + "step": 783 + }, + { + "epoch": 0.00784, + "grad_norm": 0.9756646156311035, + "learning_rate": 0.002352, + "loss": 4.706, + "step": 784 + }, + { + "epoch": 0.00785, + "grad_norm": 1.03944993019104, + "learning_rate": 0.0023550000000000003, + "loss": 4.7094, + "step": 785 + }, + { + "epoch": 0.00786, + "grad_norm": 0.9842208027839661, + "learning_rate": 0.0023580000000000003, + "loss": 4.7094, + "step": 786 + }, + { + "epoch": 0.00787, + "grad_norm": 0.9913274049758911, + "learning_rate": 0.0023610000000000003, + "loss": 4.7124, + "step": 787 + }, + { + "epoch": 0.00788, + "grad_norm": 1.0203496217727661, + "learning_rate": 0.002364, + "loss": 4.711, + "step": 788 + }, + { + "epoch": 0.00789, + "grad_norm": 1.0524691343307495, + "learning_rate": 0.002367, + "loss": 4.7436, + "step": 789 + }, + { + "epoch": 0.0079, + "grad_norm": 0.8448042273521423, + "learning_rate": 0.00237, + "loss": 4.7139, + "step": 790 + }, + { + "epoch": 0.00791, + "grad_norm": 0.737777054309845, + "learning_rate": 0.002373, + "loss": 4.6607, + "step": 791 + }, + { + "epoch": 0.00792, + "grad_norm": 0.8730551600456238, + "learning_rate": 0.002376, + "loss": 4.6988, + "step": 792 + }, + { + "epoch": 0.00793, + "grad_norm": 1.0680596828460693, + "learning_rate": 0.002379, + "loss": 4.6938, + "step": 793 + }, + { + "epoch": 0.00794, + "grad_norm": 0.876390814781189, + "learning_rate": 0.002382, + "loss": 4.7016, + "step": 794 + }, + { + "epoch": 0.00795, + "grad_norm": 0.8010908961296082, + "learning_rate": 0.002385, + "loss": 4.6743, + "step": 795 + }, + { + "epoch": 0.00796, + "grad_norm": 0.8452677130699158, + "learning_rate": 0.0023880000000000004, + "loss": 4.6712, + "step": 796 + }, + { + "epoch": 0.00797, + "grad_norm": 0.7743445038795471, + "learning_rate": 0.0023910000000000003, + "loss": 4.7081, + "step": 797 + }, + { + "epoch": 0.00798, + "grad_norm": 0.7820720076560974, + "learning_rate": 0.0023940000000000003, + "loss": 4.6744, + "step": 798 + }, + { + "epoch": 0.00799, + "grad_norm": 0.8602663278579712, + "learning_rate": 0.0023970000000000003, + "loss": 4.703, + "step": 799 + }, + { + "epoch": 0.008, + "grad_norm": 0.83580082654953, + "learning_rate": 0.0024000000000000002, + "loss": 4.6615, + "step": 800 + }, + { + "epoch": 0.00801, + "grad_norm": 0.8171262145042419, + "learning_rate": 0.002403, + "loss": 4.6877, + "step": 801 + }, + { + "epoch": 0.00802, + "grad_norm": 0.8578851819038391, + "learning_rate": 0.002406, + "loss": 4.6772, + "step": 802 + }, + { + "epoch": 0.00803, + "grad_norm": 0.9747350215911865, + "learning_rate": 0.002409, + "loss": 4.662, + "step": 803 + }, + { + "epoch": 0.00804, + "grad_norm": 0.9363577961921692, + "learning_rate": 0.002412, + "loss": 4.6595, + "step": 804 + }, + { + "epoch": 0.00805, + "grad_norm": 0.9166631102561951, + "learning_rate": 0.002415, + "loss": 4.6601, + "step": 805 + }, + { + "epoch": 0.00806, + "grad_norm": 0.860556960105896, + "learning_rate": 0.002418, + "loss": 4.681, + "step": 806 + }, + { + "epoch": 0.00807, + "grad_norm": 0.8737242817878723, + "learning_rate": 0.0024210000000000004, + "loss": 4.6779, + "step": 807 + }, + { + "epoch": 0.00808, + "grad_norm": 0.8125926852226257, + "learning_rate": 0.0024240000000000004, + "loss": 4.6555, + "step": 808 + }, + { + "epoch": 0.00809, + "grad_norm": 0.9335473775863647, + "learning_rate": 0.0024270000000000003, + "loss": 4.6575, + "step": 809 + }, + { + "epoch": 0.0081, + "grad_norm": 0.7656557559967041, + "learning_rate": 0.0024300000000000003, + "loss": 4.6752, + "step": 810 + }, + { + "epoch": 0.00811, + "grad_norm": 0.8129197359085083, + "learning_rate": 0.0024330000000000003, + "loss": 4.6545, + "step": 811 + }, + { + "epoch": 0.00812, + "grad_norm": 0.7713532447814941, + "learning_rate": 0.0024360000000000002, + "loss": 4.6279, + "step": 812 + }, + { + "epoch": 0.00813, + "grad_norm": 0.8149270415306091, + "learning_rate": 0.0024389999999999998, + "loss": 4.6613, + "step": 813 + }, + { + "epoch": 0.00814, + "grad_norm": 0.866010308265686, + "learning_rate": 0.0024419999999999997, + "loss": 4.6696, + "step": 814 + }, + { + "epoch": 0.00815, + "grad_norm": 0.7402296662330627, + "learning_rate": 0.0024449999999999997, + "loss": 4.6746, + "step": 815 + }, + { + "epoch": 0.00816, + "grad_norm": 0.6451212167739868, + "learning_rate": 0.002448, + "loss": 4.6403, + "step": 816 + }, + { + "epoch": 0.00817, + "grad_norm": 0.7935466170310974, + "learning_rate": 0.002451, + "loss": 4.6538, + "step": 817 + }, + { + "epoch": 0.00818, + "grad_norm": 1.0368677377700806, + "learning_rate": 0.002454, + "loss": 4.6444, + "step": 818 + }, + { + "epoch": 0.00819, + "grad_norm": 1.1921635866165161, + "learning_rate": 0.002457, + "loss": 4.6695, + "step": 819 + }, + { + "epoch": 0.0082, + "grad_norm": 0.9146779179573059, + "learning_rate": 0.00246, + "loss": 4.6473, + "step": 820 + }, + { + "epoch": 0.00821, + "grad_norm": 0.8097479939460754, + "learning_rate": 0.002463, + "loss": 4.6216, + "step": 821 + }, + { + "epoch": 0.00822, + "grad_norm": 0.8611756563186646, + "learning_rate": 0.002466, + "loss": 4.6241, + "step": 822 + }, + { + "epoch": 0.00823, + "grad_norm": 0.9131811857223511, + "learning_rate": 0.002469, + "loss": 4.6307, + "step": 823 + }, + { + "epoch": 0.00824, + "grad_norm": 0.9472024440765381, + "learning_rate": 0.002472, + "loss": 4.6551, + "step": 824 + }, + { + "epoch": 0.00825, + "grad_norm": 1.1296700239181519, + "learning_rate": 0.0024749999999999998, + "loss": 4.6738, + "step": 825 + }, + { + "epoch": 0.00826, + "grad_norm": 0.9912372827529907, + "learning_rate": 0.0024779999999999997, + "loss": 4.6298, + "step": 826 + }, + { + "epoch": 0.00827, + "grad_norm": 0.9669742584228516, + "learning_rate": 0.002481, + "loss": 4.6558, + "step": 827 + }, + { + "epoch": 0.00828, + "grad_norm": 0.7501344680786133, + "learning_rate": 0.002484, + "loss": 4.6425, + "step": 828 + }, + { + "epoch": 0.00829, + "grad_norm": 0.6985933184623718, + "learning_rate": 0.002487, + "loss": 4.6351, + "step": 829 + }, + { + "epoch": 0.0083, + "grad_norm": 0.6681656837463379, + "learning_rate": 0.00249, + "loss": 4.6406, + "step": 830 + }, + { + "epoch": 0.00831, + "grad_norm": 0.7142512202262878, + "learning_rate": 0.002493, + "loss": 4.6218, + "step": 831 + }, + { + "epoch": 0.00832, + "grad_norm": 0.7265256643295288, + "learning_rate": 0.002496, + "loss": 4.6212, + "step": 832 + }, + { + "epoch": 0.00833, + "grad_norm": 0.7133427262306213, + "learning_rate": 0.002499, + "loss": 4.6125, + "step": 833 + }, + { + "epoch": 0.00834, + "grad_norm": 0.7037473917007446, + "learning_rate": 0.002502, + "loss": 4.6151, + "step": 834 + }, + { + "epoch": 0.00835, + "grad_norm": 0.9004167914390564, + "learning_rate": 0.002505, + "loss": 4.6154, + "step": 835 + }, + { + "epoch": 0.00836, + "grad_norm": 0.9459953904151917, + "learning_rate": 0.002508, + "loss": 4.6497, + "step": 836 + }, + { + "epoch": 0.00837, + "grad_norm": 0.9802148342132568, + "learning_rate": 0.0025109999999999998, + "loss": 4.5975, + "step": 837 + }, + { + "epoch": 0.00838, + "grad_norm": 0.8809778690338135, + "learning_rate": 0.0025139999999999997, + "loss": 4.6166, + "step": 838 + }, + { + "epoch": 0.00839, + "grad_norm": 0.792102038860321, + "learning_rate": 0.002517, + "loss": 4.6155, + "step": 839 + }, + { + "epoch": 0.0084, + "grad_norm": 0.7744638323783875, + "learning_rate": 0.00252, + "loss": 4.5999, + "step": 840 + }, + { + "epoch": 0.00841, + "grad_norm": 0.791641116142273, + "learning_rate": 0.002523, + "loss": 4.5943, + "step": 841 + }, + { + "epoch": 0.00842, + "grad_norm": 0.8285142183303833, + "learning_rate": 0.002526, + "loss": 4.5937, + "step": 842 + }, + { + "epoch": 0.00843, + "grad_norm": 0.9900093078613281, + "learning_rate": 0.002529, + "loss": 4.6057, + "step": 843 + }, + { + "epoch": 0.00844, + "grad_norm": 0.984833300113678, + "learning_rate": 0.002532, + "loss": 4.6269, + "step": 844 + }, + { + "epoch": 0.00845, + "grad_norm": 0.8697680234909058, + "learning_rate": 0.002535, + "loss": 4.5878, + "step": 845 + }, + { + "epoch": 0.00846, + "grad_norm": 0.8182123899459839, + "learning_rate": 0.002538, + "loss": 4.6097, + "step": 846 + }, + { + "epoch": 0.00847, + "grad_norm": 0.7307525277137756, + "learning_rate": 0.002541, + "loss": 4.5888, + "step": 847 + }, + { + "epoch": 0.00848, + "grad_norm": 0.7021674513816833, + "learning_rate": 0.002544, + "loss": 4.5844, + "step": 848 + }, + { + "epoch": 0.00849, + "grad_norm": 0.6287952661514282, + "learning_rate": 0.002547, + "loss": 4.5759, + "step": 849 + }, + { + "epoch": 0.0085, + "grad_norm": 0.5557106733322144, + "learning_rate": 0.00255, + "loss": 4.5834, + "step": 850 + }, + { + "epoch": 0.00851, + "grad_norm": 0.5636370778083801, + "learning_rate": 0.002553, + "loss": 4.5977, + "step": 851 + }, + { + "epoch": 0.00852, + "grad_norm": 0.6509172320365906, + "learning_rate": 0.002556, + "loss": 4.5892, + "step": 852 + }, + { + "epoch": 0.00853, + "grad_norm": 0.6835383176803589, + "learning_rate": 0.002559, + "loss": 4.5743, + "step": 853 + }, + { + "epoch": 0.00854, + "grad_norm": 0.6499077081680298, + "learning_rate": 0.002562, + "loss": 4.5629, + "step": 854 + }, + { + "epoch": 0.00855, + "grad_norm": 0.5777466297149658, + "learning_rate": 0.002565, + "loss": 4.5388, + "step": 855 + }, + { + "epoch": 0.00856, + "grad_norm": 0.6060221195220947, + "learning_rate": 0.002568, + "loss": 4.5724, + "step": 856 + }, + { + "epoch": 0.00857, + "grad_norm": 0.6906097531318665, + "learning_rate": 0.002571, + "loss": 4.5615, + "step": 857 + }, + { + "epoch": 0.00858, + "grad_norm": 0.7318242788314819, + "learning_rate": 0.002574, + "loss": 4.5575, + "step": 858 + }, + { + "epoch": 0.00859, + "grad_norm": 0.6517086029052734, + "learning_rate": 0.002577, + "loss": 4.5498, + "step": 859 + }, + { + "epoch": 0.0086, + "grad_norm": 0.6514325737953186, + "learning_rate": 0.00258, + "loss": 4.5756, + "step": 860 + }, + { + "epoch": 0.00861, + "grad_norm": 0.7899180054664612, + "learning_rate": 0.0025830000000000002, + "loss": 4.594, + "step": 861 + }, + { + "epoch": 0.00862, + "grad_norm": 0.874433696269989, + "learning_rate": 0.002586, + "loss": 4.5519, + "step": 862 + }, + { + "epoch": 0.00863, + "grad_norm": 0.8605120182037354, + "learning_rate": 0.002589, + "loss": 4.5772, + "step": 863 + }, + { + "epoch": 0.00864, + "grad_norm": 0.9631415009498596, + "learning_rate": 0.002592, + "loss": 4.5403, + "step": 864 + }, + { + "epoch": 0.00865, + "grad_norm": 1.1293022632598877, + "learning_rate": 0.002595, + "loss": 4.5658, + "step": 865 + }, + { + "epoch": 0.00866, + "grad_norm": 1.0691903829574585, + "learning_rate": 0.002598, + "loss": 4.5962, + "step": 866 + }, + { + "epoch": 0.00867, + "grad_norm": 0.9885998368263245, + "learning_rate": 0.002601, + "loss": 4.5818, + "step": 867 + }, + { + "epoch": 0.00868, + "grad_norm": 1.3688061237335205, + "learning_rate": 0.002604, + "loss": 4.5938, + "step": 868 + }, + { + "epoch": 0.00869, + "grad_norm": 0.8156008124351501, + "learning_rate": 0.002607, + "loss": 4.5787, + "step": 869 + }, + { + "epoch": 0.0087, + "grad_norm": 0.8631764054298401, + "learning_rate": 0.00261, + "loss": 4.5676, + "step": 870 + }, + { + "epoch": 0.00871, + "grad_norm": 0.9996145963668823, + "learning_rate": 0.002613, + "loss": 4.5932, + "step": 871 + }, + { + "epoch": 0.00872, + "grad_norm": 0.972501814365387, + "learning_rate": 0.002616, + "loss": 4.5764, + "step": 872 + }, + { + "epoch": 0.00873, + "grad_norm": 1.0140340328216553, + "learning_rate": 0.0026190000000000002, + "loss": 4.5878, + "step": 873 + }, + { + "epoch": 0.00874, + "grad_norm": 0.9380632042884827, + "learning_rate": 0.002622, + "loss": 4.6105, + "step": 874 + }, + { + "epoch": 0.00875, + "grad_norm": 0.9407688975334167, + "learning_rate": 0.002625, + "loss": 4.5588, + "step": 875 + }, + { + "epoch": 0.00876, + "grad_norm": 1.0034210681915283, + "learning_rate": 0.002628, + "loss": 4.5949, + "step": 876 + }, + { + "epoch": 0.00877, + "grad_norm": 1.2963709831237793, + "learning_rate": 0.002631, + "loss": 4.6268, + "step": 877 + }, + { + "epoch": 0.00878, + "grad_norm": 0.7438150644302368, + "learning_rate": 0.002634, + "loss": 4.537, + "step": 878 + }, + { + "epoch": 0.00879, + "grad_norm": 0.8438600301742554, + "learning_rate": 0.002637, + "loss": 4.5729, + "step": 879 + }, + { + "epoch": 0.0088, + "grad_norm": 0.7277801036834717, + "learning_rate": 0.00264, + "loss": 4.5704, + "step": 880 + }, + { + "epoch": 0.00881, + "grad_norm": 0.9043455123901367, + "learning_rate": 0.002643, + "loss": 4.5794, + "step": 881 + }, + { + "epoch": 0.00882, + "grad_norm": 0.9441054463386536, + "learning_rate": 0.002646, + "loss": 4.5807, + "step": 882 + }, + { + "epoch": 0.00883, + "grad_norm": 0.8490539193153381, + "learning_rate": 0.002649, + "loss": 4.5704, + "step": 883 + }, + { + "epoch": 0.00884, + "grad_norm": 0.8639208078384399, + "learning_rate": 0.0026520000000000003, + "loss": 4.555, + "step": 884 + }, + { + "epoch": 0.00885, + "grad_norm": 0.8000319600105286, + "learning_rate": 0.0026550000000000002, + "loss": 4.5531, + "step": 885 + }, + { + "epoch": 0.00886, + "grad_norm": 0.8000409007072449, + "learning_rate": 0.002658, + "loss": 4.5702, + "step": 886 + }, + { + "epoch": 0.00887, + "grad_norm": 0.5944011807441711, + "learning_rate": 0.002661, + "loss": 4.5499, + "step": 887 + }, + { + "epoch": 0.00888, + "grad_norm": 0.5328640341758728, + "learning_rate": 0.002664, + "loss": 4.5494, + "step": 888 + }, + { + "epoch": 0.00889, + "grad_norm": 0.5338924527168274, + "learning_rate": 0.002667, + "loss": 4.5018, + "step": 889 + }, + { + "epoch": 0.0089, + "grad_norm": 0.4938536286354065, + "learning_rate": 0.00267, + "loss": 4.545, + "step": 890 + }, + { + "epoch": 0.00891, + "grad_norm": 0.4882456064224243, + "learning_rate": 0.002673, + "loss": 4.5301, + "step": 891 + }, + { + "epoch": 0.00892, + "grad_norm": 0.4257451891899109, + "learning_rate": 0.002676, + "loss": 4.5393, + "step": 892 + }, + { + "epoch": 0.00893, + "grad_norm": 0.5505130290985107, + "learning_rate": 0.002679, + "loss": 4.5171, + "step": 893 + }, + { + "epoch": 0.00894, + "grad_norm": 0.6718395352363586, + "learning_rate": 0.002682, + "loss": 4.5132, + "step": 894 + }, + { + "epoch": 0.00895, + "grad_norm": 0.7553327679634094, + "learning_rate": 0.0026850000000000003, + "loss": 4.5223, + "step": 895 + }, + { + "epoch": 0.00896, + "grad_norm": 0.7149863243103027, + "learning_rate": 0.0026880000000000003, + "loss": 4.5119, + "step": 896 + }, + { + "epoch": 0.00897, + "grad_norm": 0.8281179070472717, + "learning_rate": 0.0026910000000000002, + "loss": 4.5328, + "step": 897 + }, + { + "epoch": 0.00898, + "grad_norm": 0.9519450068473816, + "learning_rate": 0.002694, + "loss": 4.5342, + "step": 898 + }, + { + "epoch": 0.00899, + "grad_norm": 0.9227859377861023, + "learning_rate": 0.002697, + "loss": 4.5445, + "step": 899 + }, + { + "epoch": 0.009, + "grad_norm": 1.040459156036377, + "learning_rate": 0.0027, + "loss": 4.5756, + "step": 900 + }, + { + "epoch": 0.00901, + "grad_norm": 0.8912736773490906, + "learning_rate": 0.002703, + "loss": 4.5234, + "step": 901 + }, + { + "epoch": 0.00902, + "grad_norm": 0.768046498298645, + "learning_rate": 0.002706, + "loss": 4.5317, + "step": 902 + }, + { + "epoch": 0.00903, + "grad_norm": 0.698384702205658, + "learning_rate": 0.002709, + "loss": 4.5227, + "step": 903 + }, + { + "epoch": 0.00904, + "grad_norm": 0.7553470134735107, + "learning_rate": 0.002712, + "loss": 4.5046, + "step": 904 + }, + { + "epoch": 0.00905, + "grad_norm": 0.9005333185195923, + "learning_rate": 0.002715, + "loss": 4.5273, + "step": 905 + }, + { + "epoch": 0.00906, + "grad_norm": 0.9150082468986511, + "learning_rate": 0.002718, + "loss": 4.54, + "step": 906 + }, + { + "epoch": 0.00907, + "grad_norm": 0.8349012732505798, + "learning_rate": 0.0027210000000000003, + "loss": 4.5543, + "step": 907 + }, + { + "epoch": 0.00908, + "grad_norm": 0.7856695055961609, + "learning_rate": 0.0027240000000000003, + "loss": 4.5219, + "step": 908 + }, + { + "epoch": 0.00909, + "grad_norm": 0.8433918952941895, + "learning_rate": 0.0027270000000000003, + "loss": 4.5423, + "step": 909 + }, + { + "epoch": 0.0091, + "grad_norm": 0.9650413393974304, + "learning_rate": 0.0027300000000000002, + "loss": 4.5577, + "step": 910 + }, + { + "epoch": 0.00911, + "grad_norm": 0.9894043803215027, + "learning_rate": 0.002733, + "loss": 4.514, + "step": 911 + }, + { + "epoch": 0.00912, + "grad_norm": 0.9117268919944763, + "learning_rate": 0.002736, + "loss": 4.5245, + "step": 912 + }, + { + "epoch": 0.00913, + "grad_norm": 1.020936369895935, + "learning_rate": 0.002739, + "loss": 4.5244, + "step": 913 + }, + { + "epoch": 0.00914, + "grad_norm": 1.1810083389282227, + "learning_rate": 0.002742, + "loss": 4.5723, + "step": 914 + }, + { + "epoch": 0.00915, + "grad_norm": 0.8899980187416077, + "learning_rate": 0.002745, + "loss": 4.5252, + "step": 915 + }, + { + "epoch": 0.00916, + "grad_norm": 0.9896888136863708, + "learning_rate": 0.002748, + "loss": 4.5028, + "step": 916 + }, + { + "epoch": 0.00917, + "grad_norm": 1.3935941457748413, + "learning_rate": 0.002751, + "loss": 4.5471, + "step": 917 + }, + { + "epoch": 0.00918, + "grad_norm": 1.1309375762939453, + "learning_rate": 0.0027540000000000004, + "loss": 4.562, + "step": 918 + }, + { + "epoch": 0.00919, + "grad_norm": 0.8358169198036194, + "learning_rate": 0.0027570000000000003, + "loss": 4.5255, + "step": 919 + }, + { + "epoch": 0.0092, + "grad_norm": 0.8860310316085815, + "learning_rate": 0.0027600000000000003, + "loss": 4.5423, + "step": 920 + }, + { + "epoch": 0.00921, + "grad_norm": 1.0510568618774414, + "learning_rate": 0.0027630000000000003, + "loss": 4.5336, + "step": 921 + }, + { + "epoch": 0.00922, + "grad_norm": 1.1311016082763672, + "learning_rate": 0.0027660000000000002, + "loss": 4.5483, + "step": 922 + }, + { + "epoch": 0.00923, + "grad_norm": 1.1111584901809692, + "learning_rate": 0.002769, + "loss": 4.5339, + "step": 923 + }, + { + "epoch": 0.00924, + "grad_norm": 0.9117261171340942, + "learning_rate": 0.002772, + "loss": 4.5412, + "step": 924 + }, + { + "epoch": 0.00925, + "grad_norm": 0.9757253527641296, + "learning_rate": 0.002775, + "loss": 4.5263, + "step": 925 + }, + { + "epoch": 0.00926, + "grad_norm": 0.907768189907074, + "learning_rate": 0.002778, + "loss": 4.5422, + "step": 926 + }, + { + "epoch": 0.00927, + "grad_norm": 0.8181371092796326, + "learning_rate": 0.002781, + "loss": 4.4999, + "step": 927 + }, + { + "epoch": 0.00928, + "grad_norm": 0.7728373408317566, + "learning_rate": 0.002784, + "loss": 4.5461, + "step": 928 + }, + { + "epoch": 0.00929, + "grad_norm": 0.624686598777771, + "learning_rate": 0.0027870000000000004, + "loss": 4.5192, + "step": 929 + }, + { + "epoch": 0.0093, + "grad_norm": 0.6143611073493958, + "learning_rate": 0.0027900000000000004, + "loss": 4.4968, + "step": 930 + }, + { + "epoch": 0.00931, + "grad_norm": 0.5353983044624329, + "learning_rate": 0.0027930000000000003, + "loss": 4.5041, + "step": 931 + }, + { + "epoch": 0.00932, + "grad_norm": 0.5034843683242798, + "learning_rate": 0.0027960000000000003, + "loss": 4.5081, + "step": 932 + }, + { + "epoch": 0.00933, + "grad_norm": 0.43557336926460266, + "learning_rate": 0.0027990000000000003, + "loss": 4.47, + "step": 933 + }, + { + "epoch": 0.00934, + "grad_norm": 0.42429375648498535, + "learning_rate": 0.0028020000000000002, + "loss": 4.4927, + "step": 934 + }, + { + "epoch": 0.00935, + "grad_norm": 0.4439206123352051, + "learning_rate": 0.002805, + "loss": 4.5015, + "step": 935 + }, + { + "epoch": 0.00936, + "grad_norm": 0.4363570511341095, + "learning_rate": 0.002808, + "loss": 4.4989, + "step": 936 + }, + { + "epoch": 0.00937, + "grad_norm": 0.4996969401836395, + "learning_rate": 0.002811, + "loss": 4.5097, + "step": 937 + }, + { + "epoch": 0.00938, + "grad_norm": 0.5577415227890015, + "learning_rate": 0.002814, + "loss": 4.4882, + "step": 938 + }, + { + "epoch": 0.00939, + "grad_norm": 0.6116971373558044, + "learning_rate": 0.002817, + "loss": 4.4823, + "step": 939 + }, + { + "epoch": 0.0094, + "grad_norm": 0.675699770450592, + "learning_rate": 0.00282, + "loss": 4.4838, + "step": 940 + }, + { + "epoch": 0.00941, + "grad_norm": 0.8027246594429016, + "learning_rate": 0.002823, + "loss": 4.4671, + "step": 941 + }, + { + "epoch": 0.00942, + "grad_norm": 0.9442744851112366, + "learning_rate": 0.002826, + "loss": 4.5129, + "step": 942 + }, + { + "epoch": 0.00943, + "grad_norm": 0.963239312171936, + "learning_rate": 0.002829, + "loss": 4.4975, + "step": 943 + }, + { + "epoch": 0.00944, + "grad_norm": 0.8721352815628052, + "learning_rate": 0.002832, + "loss": 4.493, + "step": 944 + }, + { + "epoch": 0.00945, + "grad_norm": 1.0316184759140015, + "learning_rate": 0.002835, + "loss": 4.5084, + "step": 945 + }, + { + "epoch": 0.00946, + "grad_norm": 0.7907041907310486, + "learning_rate": 0.002838, + "loss": 4.5093, + "step": 946 + }, + { + "epoch": 0.00947, + "grad_norm": 0.8508433699607849, + "learning_rate": 0.0028409999999999998, + "loss": 4.4859, + "step": 947 + }, + { + "epoch": 0.00948, + "grad_norm": 0.8942288756370544, + "learning_rate": 0.0028439999999999997, + "loss": 4.4805, + "step": 948 + }, + { + "epoch": 0.00949, + "grad_norm": 0.8804354071617126, + "learning_rate": 0.002847, + "loss": 4.5178, + "step": 949 + }, + { + "epoch": 0.0095, + "grad_norm": 0.9102524518966675, + "learning_rate": 0.00285, + "loss": 4.5104, + "step": 950 + }, + { + "epoch": 0.00951, + "grad_norm": 0.9140040278434753, + "learning_rate": 0.002853, + "loss": 4.4913, + "step": 951 + }, + { + "epoch": 0.00952, + "grad_norm": 0.996699869632721, + "learning_rate": 0.002856, + "loss": 4.5192, + "step": 952 + }, + { + "epoch": 0.00953, + "grad_norm": 1.0743249654769897, + "learning_rate": 0.002859, + "loss": 4.4742, + "step": 953 + }, + { + "epoch": 0.00954, + "grad_norm": 0.9619385004043579, + "learning_rate": 0.002862, + "loss": 4.532, + "step": 954 + }, + { + "epoch": 0.00955, + "grad_norm": 0.9820901155471802, + "learning_rate": 0.002865, + "loss": 4.5178, + "step": 955 + }, + { + "epoch": 0.00956, + "grad_norm": 1.0036413669586182, + "learning_rate": 0.002868, + "loss": 4.501, + "step": 956 + }, + { + "epoch": 0.00957, + "grad_norm": 0.8994410634040833, + "learning_rate": 0.002871, + "loss": 4.5266, + "step": 957 + }, + { + "epoch": 0.00958, + "grad_norm": 0.910679817199707, + "learning_rate": 0.002874, + "loss": 4.5103, + "step": 958 + }, + { + "epoch": 0.00959, + "grad_norm": 0.7567980885505676, + "learning_rate": 0.002877, + "loss": 4.4958, + "step": 959 + }, + { + "epoch": 0.0096, + "grad_norm": 0.7841255068778992, + "learning_rate": 0.0028799999999999997, + "loss": 4.4967, + "step": 960 + }, + { + "epoch": 0.00961, + "grad_norm": 0.6947323083877563, + "learning_rate": 0.002883, + "loss": 4.4665, + "step": 961 + }, + { + "epoch": 0.00962, + "grad_norm": 0.5831562280654907, + "learning_rate": 0.002886, + "loss": 4.4887, + "step": 962 + }, + { + "epoch": 0.00963, + "grad_norm": 0.5601492524147034, + "learning_rate": 0.002889, + "loss": 4.4846, + "step": 963 + }, + { + "epoch": 0.00964, + "grad_norm": 0.474755197763443, + "learning_rate": 0.002892, + "loss": 4.47, + "step": 964 + }, + { + "epoch": 0.00965, + "grad_norm": 0.5203716158866882, + "learning_rate": 0.002895, + "loss": 4.4451, + "step": 965 + }, + { + "epoch": 0.00966, + "grad_norm": 0.4714201092720032, + "learning_rate": 0.002898, + "loss": 4.4572, + "step": 966 + }, + { + "epoch": 0.00967, + "grad_norm": 0.4852457344532013, + "learning_rate": 0.002901, + "loss": 4.4924, + "step": 967 + }, + { + "epoch": 0.00968, + "grad_norm": 0.5026896595954895, + "learning_rate": 0.002904, + "loss": 4.4559, + "step": 968 + }, + { + "epoch": 0.00969, + "grad_norm": 0.5138400793075562, + "learning_rate": 0.002907, + "loss": 4.4843, + "step": 969 + }, + { + "epoch": 0.0097, + "grad_norm": 0.46154963970184326, + "learning_rate": 0.00291, + "loss": 4.4402, + "step": 970 + }, + { + "epoch": 0.00971, + "grad_norm": 0.4250973165035248, + "learning_rate": 0.002913, + "loss": 4.4437, + "step": 971 + }, + { + "epoch": 0.00972, + "grad_norm": 0.46983814239501953, + "learning_rate": 0.002916, + "loss": 4.45, + "step": 972 + }, + { + "epoch": 0.00973, + "grad_norm": 0.578143835067749, + "learning_rate": 0.002919, + "loss": 4.4478, + "step": 973 + }, + { + "epoch": 0.00974, + "grad_norm": 0.7184126377105713, + "learning_rate": 0.002922, + "loss": 4.4366, + "step": 974 + }, + { + "epoch": 0.00975, + "grad_norm": 0.8322773575782776, + "learning_rate": 0.002925, + "loss": 4.4683, + "step": 975 + }, + { + "epoch": 0.00976, + "grad_norm": 0.7910879850387573, + "learning_rate": 0.002928, + "loss": 4.4497, + "step": 976 + }, + { + "epoch": 0.00977, + "grad_norm": 0.7662490606307983, + "learning_rate": 0.002931, + "loss": 4.4868, + "step": 977 + }, + { + "epoch": 0.00978, + "grad_norm": 0.6879754066467285, + "learning_rate": 0.002934, + "loss": 4.4568, + "step": 978 + }, + { + "epoch": 0.00979, + "grad_norm": 0.7417387366294861, + "learning_rate": 0.002937, + "loss": 4.4491, + "step": 979 + }, + { + "epoch": 0.0098, + "grad_norm": 0.6087706089019775, + "learning_rate": 0.00294, + "loss": 4.4405, + "step": 980 + }, + { + "epoch": 0.00981, + "grad_norm": 0.5661059617996216, + "learning_rate": 0.002943, + "loss": 4.4516, + "step": 981 + }, + { + "epoch": 0.00982, + "grad_norm": 0.5970472097396851, + "learning_rate": 0.002946, + "loss": 4.4435, + "step": 982 + }, + { + "epoch": 0.00983, + "grad_norm": 0.6391454935073853, + "learning_rate": 0.0029490000000000002, + "loss": 4.4444, + "step": 983 + }, + { + "epoch": 0.00984, + "grad_norm": 0.5948253870010376, + "learning_rate": 0.002952, + "loss": 4.4303, + "step": 984 + }, + { + "epoch": 0.00985, + "grad_norm": 0.5908463597297668, + "learning_rate": 0.002955, + "loss": 4.4509, + "step": 985 + }, + { + "epoch": 0.00986, + "grad_norm": 0.6963019371032715, + "learning_rate": 0.002958, + "loss": 4.4177, + "step": 986 + }, + { + "epoch": 0.00987, + "grad_norm": 0.7601437568664551, + "learning_rate": 0.002961, + "loss": 4.4332, + "step": 987 + }, + { + "epoch": 0.00988, + "grad_norm": 0.8287732005119324, + "learning_rate": 0.002964, + "loss": 4.4707, + "step": 988 + }, + { + "epoch": 0.00989, + "grad_norm": 0.9172109961509705, + "learning_rate": 0.002967, + "loss": 4.4168, + "step": 989 + }, + { + "epoch": 0.0099, + "grad_norm": 0.9710732102394104, + "learning_rate": 0.00297, + "loss": 4.4601, + "step": 990 + }, + { + "epoch": 0.00991, + "grad_norm": 0.9308454990386963, + "learning_rate": 0.002973, + "loss": 4.4596, + "step": 991 + }, + { + "epoch": 0.00992, + "grad_norm": 0.8575865030288696, + "learning_rate": 0.002976, + "loss": 4.4567, + "step": 992 + }, + { + "epoch": 0.00993, + "grad_norm": 0.8846513032913208, + "learning_rate": 0.002979, + "loss": 4.4578, + "step": 993 + }, + { + "epoch": 0.00994, + "grad_norm": 1.0507467985153198, + "learning_rate": 0.002982, + "loss": 4.4695, + "step": 994 + }, + { + "epoch": 0.00995, + "grad_norm": 0.9225064516067505, + "learning_rate": 0.0029850000000000002, + "loss": 4.472, + "step": 995 + }, + { + "epoch": 0.00996, + "grad_norm": 1.0122137069702148, + "learning_rate": 0.002988, + "loss": 4.4787, + "step": 996 + }, + { + "epoch": 0.00997, + "grad_norm": 1.1055110692977905, + "learning_rate": 0.002991, + "loss": 4.4822, + "step": 997 + }, + { + "epoch": 0.00998, + "grad_norm": 1.0749611854553223, + "learning_rate": 0.002994, + "loss": 4.4866, + "step": 998 + }, + { + "epoch": 0.00999, + "grad_norm": 1.0997530221939087, + "learning_rate": 0.002997, + "loss": 4.469, + "step": 999 + }, + { + "epoch": 0.01, + "grad_norm": 0.853782057762146, + "learning_rate": 0.003, + "loss": 4.4844, + "step": 1000 + }, + { + "epoch": 0.01001, + "grad_norm": 0.7132817506790161, + "learning_rate": 0.003, + "loss": 4.4776, + "step": 1001 + }, + { + "epoch": 0.01002, + "grad_norm": 0.8123744130134583, + "learning_rate": 0.003, + "loss": 4.4523, + "step": 1002 + }, + { + "epoch": 0.01003, + "grad_norm": 0.7343231439590454, + "learning_rate": 0.003, + "loss": 4.4688, + "step": 1003 + }, + { + "epoch": 0.01004, + "grad_norm": 0.7312922477722168, + "learning_rate": 0.003, + "loss": 4.4567, + "step": 1004 + }, + { + "epoch": 0.01005, + "grad_norm": 0.7205970287322998, + "learning_rate": 0.003, + "loss": 4.4254, + "step": 1005 + }, + { + "epoch": 0.01006, + "grad_norm": 0.8032294511795044, + "learning_rate": 0.003, + "loss": 4.4272, + "step": 1006 + }, + { + "epoch": 0.01007, + "grad_norm": 0.8357327580451965, + "learning_rate": 0.003, + "loss": 4.4775, + "step": 1007 + }, + { + "epoch": 0.01008, + "grad_norm": 0.8112847208976746, + "learning_rate": 0.003, + "loss": 4.4864, + "step": 1008 + }, + { + "epoch": 0.01009, + "grad_norm": 0.9153414964675903, + "learning_rate": 0.003, + "loss": 4.46, + "step": 1009 + }, + { + "epoch": 0.0101, + "grad_norm": 0.9501859545707703, + "learning_rate": 0.003, + "loss": 4.4762, + "step": 1010 + }, + { + "epoch": 0.01011, + "grad_norm": 0.9173474907875061, + "learning_rate": 0.003, + "loss": 4.4819, + "step": 1011 + }, + { + "epoch": 0.01012, + "grad_norm": 0.8518911600112915, + "learning_rate": 0.003, + "loss": 4.4635, + "step": 1012 + }, + { + "epoch": 0.01013, + "grad_norm": 0.8621218204498291, + "learning_rate": 0.003, + "loss": 4.4943, + "step": 1013 + }, + { + "epoch": 0.01014, + "grad_norm": 0.6589127779006958, + "learning_rate": 0.003, + "loss": 4.419, + "step": 1014 + }, + { + "epoch": 0.01015, + "grad_norm": 0.7594727277755737, + "learning_rate": 0.003, + "loss": 4.4653, + "step": 1015 + }, + { + "epoch": 0.01016, + "grad_norm": 0.800542414188385, + "learning_rate": 0.003, + "loss": 4.4749, + "step": 1016 + }, + { + "epoch": 0.01017, + "grad_norm": 0.7398179769515991, + "learning_rate": 0.003, + "loss": 4.4124, + "step": 1017 + }, + { + "epoch": 0.01018, + "grad_norm": 0.6518093347549438, + "learning_rate": 0.003, + "loss": 4.4645, + "step": 1018 + }, + { + "epoch": 0.01019, + "grad_norm": 0.6830618381500244, + "learning_rate": 0.003, + "loss": 4.4391, + "step": 1019 + }, + { + "epoch": 0.0102, + "grad_norm": 0.8045121431350708, + "learning_rate": 0.003, + "loss": 4.4426, + "step": 1020 + }, + { + "epoch": 0.01021, + "grad_norm": 0.6937596201896667, + "learning_rate": 0.003, + "loss": 4.4245, + "step": 1021 + }, + { + "epoch": 0.01022, + "grad_norm": 0.5872735977172852, + "learning_rate": 0.003, + "loss": 4.4273, + "step": 1022 + }, + { + "epoch": 0.01023, + "grad_norm": 0.63226717710495, + "learning_rate": 0.003, + "loss": 4.4308, + "step": 1023 + }, + { + "epoch": 0.01024, + "grad_norm": 0.7018114924430847, + "learning_rate": 0.003, + "loss": 4.3934, + "step": 1024 + }, + { + "epoch": 0.01025, + "grad_norm": 0.5812709927558899, + "learning_rate": 0.003, + "loss": 4.4216, + "step": 1025 + }, + { + "epoch": 0.01026, + "grad_norm": 0.4411616921424866, + "learning_rate": 0.003, + "loss": 4.4039, + "step": 1026 + }, + { + "epoch": 0.01027, + "grad_norm": 0.6123212575912476, + "learning_rate": 0.003, + "loss": 4.4333, + "step": 1027 + }, + { + "epoch": 0.01028, + "grad_norm": 0.6239144802093506, + "learning_rate": 0.003, + "loss": 4.4176, + "step": 1028 + }, + { + "epoch": 0.01029, + "grad_norm": 0.46944838762283325, + "learning_rate": 0.003, + "loss": 4.386, + "step": 1029 + }, + { + "epoch": 0.0103, + "grad_norm": 0.5337734818458557, + "learning_rate": 0.003, + "loss": 4.4157, + "step": 1030 + }, + { + "epoch": 0.01031, + "grad_norm": 0.6015392541885376, + "learning_rate": 0.003, + "loss": 4.4386, + "step": 1031 + }, + { + "epoch": 0.01032, + "grad_norm": 0.49139612913131714, + "learning_rate": 0.003, + "loss": 4.3989, + "step": 1032 + }, + { + "epoch": 0.01033, + "grad_norm": 0.6846477389335632, + "learning_rate": 0.003, + "loss": 4.444, + "step": 1033 + }, + { + "epoch": 0.01034, + "grad_norm": 0.619234025478363, + "learning_rate": 0.003, + "loss": 4.4088, + "step": 1034 + }, + { + "epoch": 0.01035, + "grad_norm": 0.567406952381134, + "learning_rate": 0.003, + "loss": 4.4003, + "step": 1035 + }, + { + "epoch": 0.01036, + "grad_norm": 0.7047313451766968, + "learning_rate": 0.003, + "loss": 4.4001, + "step": 1036 + }, + { + "epoch": 0.01037, + "grad_norm": 0.6020769476890564, + "learning_rate": 0.003, + "loss": 4.4235, + "step": 1037 + }, + { + "epoch": 0.01038, + "grad_norm": 0.498977929353714, + "learning_rate": 0.003, + "loss": 4.4148, + "step": 1038 + }, + { + "epoch": 0.01039, + "grad_norm": 0.5447039008140564, + "learning_rate": 0.003, + "loss": 4.4085, + "step": 1039 + }, + { + "epoch": 0.0104, + "grad_norm": 0.5632253289222717, + "learning_rate": 0.003, + "loss": 4.4204, + "step": 1040 + }, + { + "epoch": 0.01041, + "grad_norm": 0.5558527708053589, + "learning_rate": 0.003, + "loss": 4.4009, + "step": 1041 + }, + { + "epoch": 0.01042, + "grad_norm": 0.5142215490341187, + "learning_rate": 0.003, + "loss": 4.4233, + "step": 1042 + }, + { + "epoch": 0.01043, + "grad_norm": 0.4451257884502411, + "learning_rate": 0.003, + "loss": 4.3651, + "step": 1043 + }, + { + "epoch": 0.01044, + "grad_norm": 0.41156554222106934, + "learning_rate": 0.003, + "loss": 4.4008, + "step": 1044 + }, + { + "epoch": 0.01045, + "grad_norm": 0.42055779695510864, + "learning_rate": 0.003, + "loss": 4.4203, + "step": 1045 + }, + { + "epoch": 0.01046, + "grad_norm": 0.5353807806968689, + "learning_rate": 0.003, + "loss": 4.3928, + "step": 1046 + }, + { + "epoch": 0.01047, + "grad_norm": 0.6298825144767761, + "learning_rate": 0.003, + "loss": 4.4015, + "step": 1047 + }, + { + "epoch": 0.01048, + "grad_norm": 0.7191411256790161, + "learning_rate": 0.003, + "loss": 4.3988, + "step": 1048 + }, + { + "epoch": 0.01049, + "grad_norm": 0.7613799571990967, + "learning_rate": 0.003, + "loss": 4.3808, + "step": 1049 + }, + { + "epoch": 0.0105, + "grad_norm": 0.7375987768173218, + "learning_rate": 0.003, + "loss": 4.4222, + "step": 1050 + }, + { + "epoch": 0.01051, + "grad_norm": 0.6425654888153076, + "learning_rate": 0.003, + "loss": 4.3959, + "step": 1051 + }, + { + "epoch": 0.01052, + "grad_norm": 0.7071349620819092, + "learning_rate": 0.003, + "loss": 4.4073, + "step": 1052 + }, + { + "epoch": 0.01053, + "grad_norm": 0.838892936706543, + "learning_rate": 0.003, + "loss": 4.4246, + "step": 1053 + }, + { + "epoch": 0.01054, + "grad_norm": 0.8081632852554321, + "learning_rate": 0.003, + "loss": 4.4168, + "step": 1054 + }, + { + "epoch": 0.01055, + "grad_norm": 1.0391567945480347, + "learning_rate": 0.003, + "loss": 4.4056, + "step": 1055 + }, + { + "epoch": 0.01056, + "grad_norm": 0.9966610074043274, + "learning_rate": 0.003, + "loss": 4.4166, + "step": 1056 + }, + { + "epoch": 0.01057, + "grad_norm": 0.8638164401054382, + "learning_rate": 0.003, + "loss": 4.4191, + "step": 1057 + }, + { + "epoch": 0.01058, + "grad_norm": 0.8240249156951904, + "learning_rate": 0.003, + "loss": 4.4115, + "step": 1058 + }, + { + "epoch": 0.01059, + "grad_norm": 0.8680912256240845, + "learning_rate": 0.003, + "loss": 4.4127, + "step": 1059 + }, + { + "epoch": 0.0106, + "grad_norm": 0.8156179785728455, + "learning_rate": 0.003, + "loss": 4.4084, + "step": 1060 + }, + { + "epoch": 0.01061, + "grad_norm": 0.7857629656791687, + "learning_rate": 0.003, + "loss": 4.4047, + "step": 1061 + }, + { + "epoch": 0.01062, + "grad_norm": 0.8633149862289429, + "learning_rate": 0.003, + "loss": 4.398, + "step": 1062 + }, + { + "epoch": 0.01063, + "grad_norm": 0.9093345403671265, + "learning_rate": 0.003, + "loss": 4.4015, + "step": 1063 + }, + { + "epoch": 0.01064, + "grad_norm": 0.8469107151031494, + "learning_rate": 0.003, + "loss": 4.4193, + "step": 1064 + }, + { + "epoch": 0.01065, + "grad_norm": 0.8231899738311768, + "learning_rate": 0.003, + "loss": 4.4189, + "step": 1065 + }, + { + "epoch": 0.01066, + "grad_norm": 0.8297374248504639, + "learning_rate": 0.003, + "loss": 4.423, + "step": 1066 + }, + { + "epoch": 0.01067, + "grad_norm": 0.678811252117157, + "learning_rate": 0.003, + "loss": 4.4286, + "step": 1067 + }, + { + "epoch": 0.01068, + "grad_norm": 0.7318910360336304, + "learning_rate": 0.003, + "loss": 4.4099, + "step": 1068 + }, + { + "epoch": 0.01069, + "grad_norm": 0.7908743023872375, + "learning_rate": 0.003, + "loss": 4.4229, + "step": 1069 + }, + { + "epoch": 0.0107, + "grad_norm": 0.8041807413101196, + "learning_rate": 0.003, + "loss": 4.4771, + "step": 1070 + }, + { + "epoch": 0.01071, + "grad_norm": 0.6532490849494934, + "learning_rate": 0.003, + "loss": 4.3932, + "step": 1071 + }, + { + "epoch": 0.01072, + "grad_norm": 0.6556645035743713, + "learning_rate": 0.003, + "loss": 4.426, + "step": 1072 + }, + { + "epoch": 0.01073, + "grad_norm": 0.8516754508018494, + "learning_rate": 0.003, + "loss": 4.4268, + "step": 1073 + }, + { + "epoch": 0.01074, + "grad_norm": 1.0397266149520874, + "learning_rate": 0.003, + "loss": 4.4077, + "step": 1074 + }, + { + "epoch": 0.01075, + "grad_norm": 0.8689488172531128, + "learning_rate": 0.003, + "loss": 4.4416, + "step": 1075 + }, + { + "epoch": 0.01076, + "grad_norm": 0.9471074342727661, + "learning_rate": 0.003, + "loss": 4.4417, + "step": 1076 + }, + { + "epoch": 0.01077, + "grad_norm": 0.9231577515602112, + "learning_rate": 0.003, + "loss": 4.4357, + "step": 1077 + }, + { + "epoch": 0.01078, + "grad_norm": 0.8509425520896912, + "learning_rate": 0.003, + "loss": 4.3906, + "step": 1078 + }, + { + "epoch": 0.01079, + "grad_norm": 0.7320951819419861, + "learning_rate": 0.003, + "loss": 4.4158, + "step": 1079 + }, + { + "epoch": 0.0108, + "grad_norm": 0.6985662579536438, + "learning_rate": 0.003, + "loss": 4.3942, + "step": 1080 + }, + { + "epoch": 0.01081, + "grad_norm": 0.5917143225669861, + "learning_rate": 0.003, + "loss": 4.396, + "step": 1081 + }, + { + "epoch": 0.01082, + "grad_norm": 0.567608654499054, + "learning_rate": 0.003, + "loss": 4.4109, + "step": 1082 + }, + { + "epoch": 0.01083, + "grad_norm": 0.5785332322120667, + "learning_rate": 0.003, + "loss": 4.3961, + "step": 1083 + }, + { + "epoch": 0.01084, + "grad_norm": 0.5440964102745056, + "learning_rate": 0.003, + "loss": 4.3992, + "step": 1084 + }, + { + "epoch": 0.01085, + "grad_norm": 0.49031350016593933, + "learning_rate": 0.003, + "loss": 4.3852, + "step": 1085 + }, + { + "epoch": 0.01086, + "grad_norm": 0.5103932619094849, + "learning_rate": 0.003, + "loss": 4.3604, + "step": 1086 + }, + { + "epoch": 0.01087, + "grad_norm": 0.5229532718658447, + "learning_rate": 0.003, + "loss": 4.3777, + "step": 1087 + }, + { + "epoch": 0.01088, + "grad_norm": 0.6494709253311157, + "learning_rate": 0.003, + "loss": 4.4131, + "step": 1088 + }, + { + "epoch": 0.01089, + "grad_norm": 0.6600701808929443, + "learning_rate": 0.003, + "loss": 4.3871, + "step": 1089 + }, + { + "epoch": 0.0109, + "grad_norm": 0.4794403612613678, + "learning_rate": 0.003, + "loss": 4.4042, + "step": 1090 + }, + { + "epoch": 0.01091, + "grad_norm": 0.4167538285255432, + "learning_rate": 0.003, + "loss": 4.369, + "step": 1091 + }, + { + "epoch": 0.01092, + "grad_norm": 0.4487409293651581, + "learning_rate": 0.003, + "loss": 4.3678, + "step": 1092 + }, + { + "epoch": 0.01093, + "grad_norm": 0.39760273694992065, + "learning_rate": 0.003, + "loss": 4.3829, + "step": 1093 + }, + { + "epoch": 0.01094, + "grad_norm": 0.3819560706615448, + "learning_rate": 0.003, + "loss": 4.3691, + "step": 1094 + }, + { + "epoch": 0.01095, + "grad_norm": 0.4009959101676941, + "learning_rate": 0.003, + "loss": 4.3966, + "step": 1095 + }, + { + "epoch": 0.01096, + "grad_norm": 0.4385271668434143, + "learning_rate": 0.003, + "loss": 4.3677, + "step": 1096 + }, + { + "epoch": 0.01097, + "grad_norm": 0.5154523253440857, + "learning_rate": 0.003, + "loss": 4.3585, + "step": 1097 + }, + { + "epoch": 0.01098, + "grad_norm": 0.6741756200790405, + "learning_rate": 0.003, + "loss": 4.3816, + "step": 1098 + }, + { + "epoch": 0.01099, + "grad_norm": 0.782984733581543, + "learning_rate": 0.003, + "loss": 4.3628, + "step": 1099 + }, + { + "epoch": 0.011, + "grad_norm": 0.7672849297523499, + "learning_rate": 0.003, + "loss": 4.3633, + "step": 1100 + }, + { + "epoch": 0.01101, + "grad_norm": 0.7932085394859314, + "learning_rate": 0.003, + "loss": 4.377, + "step": 1101 + }, + { + "epoch": 0.01102, + "grad_norm": 0.8302851915359497, + "learning_rate": 0.003, + "loss": 4.3891, + "step": 1102 + }, + { + "epoch": 0.01103, + "grad_norm": 0.747067391872406, + "learning_rate": 0.003, + "loss": 4.3737, + "step": 1103 + }, + { + "epoch": 0.01104, + "grad_norm": 0.789983332157135, + "learning_rate": 0.003, + "loss": 4.3715, + "step": 1104 + }, + { + "epoch": 0.01105, + "grad_norm": 1.0034455060958862, + "learning_rate": 0.003, + "loss": 4.4029, + "step": 1105 + }, + { + "epoch": 0.01106, + "grad_norm": 1.1279704570770264, + "learning_rate": 0.003, + "loss": 4.3962, + "step": 1106 + }, + { + "epoch": 0.01107, + "grad_norm": 0.916431725025177, + "learning_rate": 0.003, + "loss": 4.3914, + "step": 1107 + }, + { + "epoch": 0.01108, + "grad_norm": 0.9773505330085754, + "learning_rate": 0.003, + "loss": 4.3731, + "step": 1108 + }, + { + "epoch": 0.01109, + "grad_norm": 0.8878449201583862, + "learning_rate": 0.003, + "loss": 4.3883, + "step": 1109 + }, + { + "epoch": 0.0111, + "grad_norm": 0.7733376622200012, + "learning_rate": 0.003, + "loss": 4.4082, + "step": 1110 + }, + { + "epoch": 0.01111, + "grad_norm": 0.7688936591148376, + "learning_rate": 0.003, + "loss": 4.3824, + "step": 1111 + }, + { + "epoch": 0.01112, + "grad_norm": 0.8010363578796387, + "learning_rate": 0.003, + "loss": 4.4064, + "step": 1112 + }, + { + "epoch": 0.01113, + "grad_norm": 0.7642234563827515, + "learning_rate": 0.003, + "loss": 4.3689, + "step": 1113 + }, + { + "epoch": 0.01114, + "grad_norm": 0.6956503987312317, + "learning_rate": 0.003, + "loss": 4.386, + "step": 1114 + }, + { + "epoch": 0.01115, + "grad_norm": 0.5809863805770874, + "learning_rate": 0.003, + "loss": 4.3985, + "step": 1115 + }, + { + "epoch": 0.01116, + "grad_norm": 0.5753639936447144, + "learning_rate": 0.003, + "loss": 4.3898, + "step": 1116 + }, + { + "epoch": 0.01117, + "grad_norm": 0.5193636417388916, + "learning_rate": 0.003, + "loss": 4.3734, + "step": 1117 + }, + { + "epoch": 0.01118, + "grad_norm": 0.5058582425117493, + "learning_rate": 0.003, + "loss": 4.3742, + "step": 1118 + }, + { + "epoch": 0.01119, + "grad_norm": 0.4714462161064148, + "learning_rate": 0.003, + "loss": 4.3802, + "step": 1119 + }, + { + "epoch": 0.0112, + "grad_norm": 0.5877813100814819, + "learning_rate": 0.003, + "loss": 4.3643, + "step": 1120 + }, + { + "epoch": 0.01121, + "grad_norm": 0.6811574697494507, + "learning_rate": 0.003, + "loss": 4.3858, + "step": 1121 + }, + { + "epoch": 0.01122, + "grad_norm": 0.7246198058128357, + "learning_rate": 0.003, + "loss": 4.3577, + "step": 1122 + }, + { + "epoch": 0.01123, + "grad_norm": 0.6985503435134888, + "learning_rate": 0.003, + "loss": 4.3459, + "step": 1123 + }, + { + "epoch": 0.01124, + "grad_norm": 0.6198428273200989, + "learning_rate": 0.003, + "loss": 4.3921, + "step": 1124 + }, + { + "epoch": 0.01125, + "grad_norm": 0.6425508856773376, + "learning_rate": 0.003, + "loss": 4.3601, + "step": 1125 + }, + { + "epoch": 0.01126, + "grad_norm": 0.7215856313705444, + "learning_rate": 0.003, + "loss": 4.3688, + "step": 1126 + }, + { + "epoch": 0.01127, + "grad_norm": 0.614211916923523, + "learning_rate": 0.003, + "loss": 4.3955, + "step": 1127 + }, + { + "epoch": 0.01128, + "grad_norm": 0.5611268877983093, + "learning_rate": 0.003, + "loss": 4.3725, + "step": 1128 + }, + { + "epoch": 0.01129, + "grad_norm": 0.5580106973648071, + "learning_rate": 0.003, + "loss": 4.3639, + "step": 1129 + }, + { + "epoch": 0.0113, + "grad_norm": 0.5985743403434753, + "learning_rate": 0.003, + "loss": 4.3456, + "step": 1130 + }, + { + "epoch": 0.01131, + "grad_norm": 0.6220662593841553, + "learning_rate": 0.003, + "loss": 4.3584, + "step": 1131 + }, + { + "epoch": 0.01132, + "grad_norm": 0.6651334166526794, + "learning_rate": 0.003, + "loss": 4.3835, + "step": 1132 + }, + { + "epoch": 0.01133, + "grad_norm": 0.6686156392097473, + "learning_rate": 0.003, + "loss": 4.3408, + "step": 1133 + }, + { + "epoch": 0.01134, + "grad_norm": 0.6207416653633118, + "learning_rate": 0.003, + "loss": 4.3549, + "step": 1134 + }, + { + "epoch": 0.01135, + "grad_norm": 0.4418866038322449, + "learning_rate": 0.003, + "loss": 4.3765, + "step": 1135 + }, + { + "epoch": 0.01136, + "grad_norm": 0.496660441160202, + "learning_rate": 0.003, + "loss": 4.3534, + "step": 1136 + }, + { + "epoch": 0.01137, + "grad_norm": 0.516294002532959, + "learning_rate": 0.003, + "loss": 4.3812, + "step": 1137 + }, + { + "epoch": 0.01138, + "grad_norm": 0.7070192098617554, + "learning_rate": 0.003, + "loss": 4.3409, + "step": 1138 + }, + { + "epoch": 0.01139, + "grad_norm": 0.9681088924407959, + "learning_rate": 0.003, + "loss": 4.3765, + "step": 1139 + }, + { + "epoch": 0.0114, + "grad_norm": 1.0279847383499146, + "learning_rate": 0.003, + "loss": 4.361, + "step": 1140 + }, + { + "epoch": 0.01141, + "grad_norm": 0.8630406260490417, + "learning_rate": 0.003, + "loss": 4.3777, + "step": 1141 + }, + { + "epoch": 0.01142, + "grad_norm": 0.8118616938591003, + "learning_rate": 0.003, + "loss": 4.3623, + "step": 1142 + }, + { + "epoch": 0.01143, + "grad_norm": 0.6431841254234314, + "learning_rate": 0.003, + "loss": 4.3605, + "step": 1143 + }, + { + "epoch": 0.01144, + "grad_norm": 0.8104698061943054, + "learning_rate": 0.003, + "loss": 4.3374, + "step": 1144 + }, + { + "epoch": 0.01145, + "grad_norm": 0.8287402987480164, + "learning_rate": 0.003, + "loss": 4.3806, + "step": 1145 + }, + { + "epoch": 0.01146, + "grad_norm": 0.6933982968330383, + "learning_rate": 0.003, + "loss": 4.358, + "step": 1146 + }, + { + "epoch": 0.01147, + "grad_norm": 0.7871519327163696, + "learning_rate": 0.003, + "loss": 4.3521, + "step": 1147 + }, + { + "epoch": 0.01148, + "grad_norm": 0.879456639289856, + "learning_rate": 0.003, + "loss": 4.3655, + "step": 1148 + }, + { + "epoch": 0.01149, + "grad_norm": 0.973317563533783, + "learning_rate": 0.003, + "loss": 4.3825, + "step": 1149 + }, + { + "epoch": 0.0115, + "grad_norm": 0.9118475914001465, + "learning_rate": 0.003, + "loss": 4.3787, + "step": 1150 + }, + { + "epoch": 0.01151, + "grad_norm": 0.8108662962913513, + "learning_rate": 0.003, + "loss": 4.3725, + "step": 1151 + }, + { + "epoch": 0.01152, + "grad_norm": 0.6577885150909424, + "learning_rate": 0.003, + "loss": 4.3738, + "step": 1152 + }, + { + "epoch": 0.01153, + "grad_norm": 0.7330801486968994, + "learning_rate": 0.003, + "loss": 4.3918, + "step": 1153 + }, + { + "epoch": 0.01154, + "grad_norm": 0.6542471051216125, + "learning_rate": 0.003, + "loss": 4.3669, + "step": 1154 + }, + { + "epoch": 0.01155, + "grad_norm": 0.6441750526428223, + "learning_rate": 0.003, + "loss": 4.3518, + "step": 1155 + }, + { + "epoch": 0.01156, + "grad_norm": 0.6419921517372131, + "learning_rate": 0.003, + "loss": 4.3556, + "step": 1156 + }, + { + "epoch": 0.01157, + "grad_norm": 0.6812348365783691, + "learning_rate": 0.003, + "loss": 4.3858, + "step": 1157 + }, + { + "epoch": 0.01158, + "grad_norm": 0.6114307045936584, + "learning_rate": 0.003, + "loss": 4.3497, + "step": 1158 + }, + { + "epoch": 0.01159, + "grad_norm": 0.5222328901290894, + "learning_rate": 0.003, + "loss": 4.3411, + "step": 1159 + }, + { + "epoch": 0.0116, + "grad_norm": 0.5184376239776611, + "learning_rate": 0.003, + "loss": 4.3556, + "step": 1160 + }, + { + "epoch": 0.01161, + "grad_norm": 0.48943445086479187, + "learning_rate": 0.003, + "loss": 4.3274, + "step": 1161 + }, + { + "epoch": 0.01162, + "grad_norm": 0.44027477502822876, + "learning_rate": 0.003, + "loss": 4.3376, + "step": 1162 + }, + { + "epoch": 0.01163, + "grad_norm": 0.40750086307525635, + "learning_rate": 0.003, + "loss": 4.3549, + "step": 1163 + }, + { + "epoch": 0.01164, + "grad_norm": 0.3570636808872223, + "learning_rate": 0.003, + "loss": 4.3421, + "step": 1164 + }, + { + "epoch": 0.01165, + "grad_norm": 0.37404096126556396, + "learning_rate": 0.003, + "loss": 4.3343, + "step": 1165 + }, + { + "epoch": 0.01166, + "grad_norm": 0.3607056736946106, + "learning_rate": 0.003, + "loss": 4.3358, + "step": 1166 + }, + { + "epoch": 0.01167, + "grad_norm": 0.4395747482776642, + "learning_rate": 0.003, + "loss": 4.3038, + "step": 1167 + }, + { + "epoch": 0.01168, + "grad_norm": 0.5243505239486694, + "learning_rate": 0.003, + "loss": 4.3153, + "step": 1168 + }, + { + "epoch": 0.01169, + "grad_norm": 0.8133231997489929, + "learning_rate": 0.003, + "loss": 4.3628, + "step": 1169 + }, + { + "epoch": 0.0117, + "grad_norm": 0.8480315208435059, + "learning_rate": 0.003, + "loss": 4.3697, + "step": 1170 + }, + { + "epoch": 0.01171, + "grad_norm": 0.6388808488845825, + "learning_rate": 0.003, + "loss": 4.3357, + "step": 1171 + }, + { + "epoch": 0.01172, + "grad_norm": 0.6913546919822693, + "learning_rate": 0.003, + "loss": 4.3289, + "step": 1172 + }, + { + "epoch": 0.01173, + "grad_norm": 0.656560480594635, + "learning_rate": 0.003, + "loss": 4.3667, + "step": 1173 + }, + { + "epoch": 0.01174, + "grad_norm": 0.5804395079612732, + "learning_rate": 0.003, + "loss": 4.3352, + "step": 1174 + }, + { + "epoch": 0.01175, + "grad_norm": 0.7265805602073669, + "learning_rate": 0.003, + "loss": 4.3269, + "step": 1175 + }, + { + "epoch": 0.01176, + "grad_norm": 0.753591001033783, + "learning_rate": 0.003, + "loss": 4.3628, + "step": 1176 + }, + { + "epoch": 0.01177, + "grad_norm": 0.7016688585281372, + "learning_rate": 0.003, + "loss": 4.3288, + "step": 1177 + }, + { + "epoch": 0.01178, + "grad_norm": 0.7714430689811707, + "learning_rate": 0.003, + "loss": 4.3516, + "step": 1178 + }, + { + "epoch": 0.01179, + "grad_norm": 0.7303088903427124, + "learning_rate": 0.003, + "loss": 4.3309, + "step": 1179 + }, + { + "epoch": 0.0118, + "grad_norm": 0.6078006029129028, + "learning_rate": 0.003, + "loss": 4.3881, + "step": 1180 + }, + { + "epoch": 0.01181, + "grad_norm": 0.6842349767684937, + "learning_rate": 0.003, + "loss": 4.3776, + "step": 1181 + }, + { + "epoch": 0.01182, + "grad_norm": 0.6943366527557373, + "learning_rate": 0.003, + "loss": 4.3406, + "step": 1182 + }, + { + "epoch": 0.01183, + "grad_norm": 0.8091237545013428, + "learning_rate": 0.003, + "loss": 4.3348, + "step": 1183 + }, + { + "epoch": 0.01184, + "grad_norm": 1.061568021774292, + "learning_rate": 0.003, + "loss": 4.3914, + "step": 1184 + }, + { + "epoch": 0.01185, + "grad_norm": 0.8616968989372253, + "learning_rate": 0.003, + "loss": 4.3565, + "step": 1185 + }, + { + "epoch": 0.01186, + "grad_norm": 0.9476875066757202, + "learning_rate": 0.003, + "loss": 4.3829, + "step": 1186 + }, + { + "epoch": 0.01187, + "grad_norm": 0.8803635239601135, + "learning_rate": 0.003, + "loss": 4.3482, + "step": 1187 + }, + { + "epoch": 0.01188, + "grad_norm": 0.8929046988487244, + "learning_rate": 0.003, + "loss": 4.3663, + "step": 1188 + }, + { + "epoch": 0.01189, + "grad_norm": 1.0426416397094727, + "learning_rate": 0.003, + "loss": 4.3669, + "step": 1189 + }, + { + "epoch": 0.0119, + "grad_norm": 0.8442360758781433, + "learning_rate": 0.003, + "loss": 4.3513, + "step": 1190 + }, + { + "epoch": 0.01191, + "grad_norm": 0.6745409965515137, + "learning_rate": 0.003, + "loss": 4.3234, + "step": 1191 + }, + { + "epoch": 0.01192, + "grad_norm": 0.6114994883537292, + "learning_rate": 0.003, + "loss": 4.3525, + "step": 1192 + }, + { + "epoch": 0.01193, + "grad_norm": 0.545036792755127, + "learning_rate": 0.003, + "loss": 4.3526, + "step": 1193 + }, + { + "epoch": 0.01194, + "grad_norm": 0.556999921798706, + "learning_rate": 0.003, + "loss": 4.3504, + "step": 1194 + }, + { + "epoch": 0.01195, + "grad_norm": 0.5564072728157043, + "learning_rate": 0.003, + "loss": 4.3299, + "step": 1195 + }, + { + "epoch": 0.01196, + "grad_norm": 0.5982546210289001, + "learning_rate": 0.003, + "loss": 4.337, + "step": 1196 + }, + { + "epoch": 0.01197, + "grad_norm": 0.6563606858253479, + "learning_rate": 0.003, + "loss": 4.3029, + "step": 1197 + }, + { + "epoch": 0.01198, + "grad_norm": 0.7455101609230042, + "learning_rate": 0.003, + "loss": 4.379, + "step": 1198 + }, + { + "epoch": 0.01199, + "grad_norm": 0.7808868885040283, + "learning_rate": 0.003, + "loss": 4.3568, + "step": 1199 + }, + { + "epoch": 0.012, + "grad_norm": 0.6572416424751282, + "learning_rate": 0.003, + "loss": 4.3439, + "step": 1200 + }, + { + "epoch": 0.01201, + "grad_norm": 0.588565468788147, + "learning_rate": 0.003, + "loss": 4.3678, + "step": 1201 + }, + { + "epoch": 0.01202, + "grad_norm": 0.6117038726806641, + "learning_rate": 0.003, + "loss": 4.3358, + "step": 1202 + }, + { + "epoch": 0.01203, + "grad_norm": 0.5589160323143005, + "learning_rate": 0.003, + "loss": 4.343, + "step": 1203 + }, + { + "epoch": 0.01204, + "grad_norm": 0.537390410900116, + "learning_rate": 0.003, + "loss": 4.3383, + "step": 1204 + }, + { + "epoch": 0.01205, + "grad_norm": 0.6048082709312439, + "learning_rate": 0.003, + "loss": 4.3289, + "step": 1205 + }, + { + "epoch": 0.01206, + "grad_norm": 0.5843051671981812, + "learning_rate": 0.003, + "loss": 4.3719, + "step": 1206 + }, + { + "epoch": 0.01207, + "grad_norm": 0.5032393932342529, + "learning_rate": 0.003, + "loss": 4.2929, + "step": 1207 + }, + { + "epoch": 0.01208, + "grad_norm": 0.4879417419433594, + "learning_rate": 0.003, + "loss": 4.351, + "step": 1208 + }, + { + "epoch": 0.01209, + "grad_norm": 0.5636802911758423, + "learning_rate": 0.003, + "loss": 4.3346, + "step": 1209 + }, + { + "epoch": 0.0121, + "grad_norm": 0.6080469489097595, + "learning_rate": 0.003, + "loss": 4.3373, + "step": 1210 + }, + { + "epoch": 0.01211, + "grad_norm": 0.5971558094024658, + "learning_rate": 0.003, + "loss": 4.3219, + "step": 1211 + }, + { + "epoch": 0.01212, + "grad_norm": 0.5976467132568359, + "learning_rate": 0.003, + "loss": 4.3467, + "step": 1212 + }, + { + "epoch": 0.01213, + "grad_norm": 0.5944254994392395, + "learning_rate": 0.003, + "loss": 4.3244, + "step": 1213 + }, + { + "epoch": 0.01214, + "grad_norm": 0.5843701958656311, + "learning_rate": 0.003, + "loss": 4.306, + "step": 1214 + }, + { + "epoch": 0.01215, + "grad_norm": 0.6481133103370667, + "learning_rate": 0.003, + "loss": 4.3176, + "step": 1215 + }, + { + "epoch": 0.01216, + "grad_norm": 0.819490909576416, + "learning_rate": 0.003, + "loss": 4.3212, + "step": 1216 + }, + { + "epoch": 0.01217, + "grad_norm": 0.8658471703529358, + "learning_rate": 0.003, + "loss": 4.3388, + "step": 1217 + }, + { + "epoch": 0.01218, + "grad_norm": 1.0888832807540894, + "learning_rate": 0.003, + "loss": 4.3466, + "step": 1218 + }, + { + "epoch": 0.01219, + "grad_norm": 1.1137385368347168, + "learning_rate": 0.003, + "loss": 4.359, + "step": 1219 + }, + { + "epoch": 0.0122, + "grad_norm": 0.8392791152000427, + "learning_rate": 0.003, + "loss": 4.341, + "step": 1220 + }, + { + "epoch": 0.01221, + "grad_norm": 0.842366635799408, + "learning_rate": 0.003, + "loss": 4.355, + "step": 1221 + }, + { + "epoch": 0.01222, + "grad_norm": 0.9087112545967102, + "learning_rate": 0.003, + "loss": 4.3425, + "step": 1222 + }, + { + "epoch": 0.01223, + "grad_norm": 0.8507757186889648, + "learning_rate": 0.003, + "loss": 4.3381, + "step": 1223 + }, + { + "epoch": 0.01224, + "grad_norm": 0.745599091053009, + "learning_rate": 0.003, + "loss": 4.3236, + "step": 1224 + }, + { + "epoch": 0.01225, + "grad_norm": 0.6612358093261719, + "learning_rate": 0.003, + "loss": 4.3414, + "step": 1225 + }, + { + "epoch": 0.01226, + "grad_norm": 0.7157174348831177, + "learning_rate": 0.003, + "loss": 4.3359, + "step": 1226 + }, + { + "epoch": 0.01227, + "grad_norm": 0.7675558924674988, + "learning_rate": 0.003, + "loss": 4.3648, + "step": 1227 + }, + { + "epoch": 0.01228, + "grad_norm": 0.7858232855796814, + "learning_rate": 0.003, + "loss": 4.3493, + "step": 1228 + }, + { + "epoch": 0.01229, + "grad_norm": 0.750853419303894, + "learning_rate": 0.003, + "loss": 4.3338, + "step": 1229 + }, + { + "epoch": 0.0123, + "grad_norm": 0.7338488101959229, + "learning_rate": 0.003, + "loss": 4.3678, + "step": 1230 + }, + { + "epoch": 0.01231, + "grad_norm": 0.7138859033584595, + "learning_rate": 0.003, + "loss": 4.3349, + "step": 1231 + }, + { + "epoch": 0.01232, + "grad_norm": 0.6423472762107849, + "learning_rate": 0.003, + "loss": 4.3101, + "step": 1232 + }, + { + "epoch": 0.01233, + "grad_norm": 0.6176342964172363, + "learning_rate": 0.003, + "loss": 4.3143, + "step": 1233 + }, + { + "epoch": 0.01234, + "grad_norm": 0.5909737348556519, + "learning_rate": 0.003, + "loss": 4.3319, + "step": 1234 + }, + { + "epoch": 0.01235, + "grad_norm": 0.5610330700874329, + "learning_rate": 0.003, + "loss": 4.3196, + "step": 1235 + }, + { + "epoch": 0.01236, + "grad_norm": 0.5141494870185852, + "learning_rate": 0.003, + "loss": 4.3229, + "step": 1236 + }, + { + "epoch": 0.01237, + "grad_norm": 0.5010937452316284, + "learning_rate": 0.003, + "loss": 4.3305, + "step": 1237 + }, + { + "epoch": 0.01238, + "grad_norm": 0.5260263085365295, + "learning_rate": 0.003, + "loss": 4.3262, + "step": 1238 + }, + { + "epoch": 0.01239, + "grad_norm": 0.6261786222457886, + "learning_rate": 0.003, + "loss": 4.3109, + "step": 1239 + }, + { + "epoch": 0.0124, + "grad_norm": 0.6283076405525208, + "learning_rate": 0.003, + "loss": 4.294, + "step": 1240 + }, + { + "epoch": 0.01241, + "grad_norm": 0.520620584487915, + "learning_rate": 0.003, + "loss": 4.3044, + "step": 1241 + }, + { + "epoch": 0.01242, + "grad_norm": 0.45525336265563965, + "learning_rate": 0.003, + "loss": 4.3077, + "step": 1242 + }, + { + "epoch": 0.01243, + "grad_norm": 0.5128642320632935, + "learning_rate": 0.003, + "loss": 4.2904, + "step": 1243 + }, + { + "epoch": 0.01244, + "grad_norm": 0.6018971800804138, + "learning_rate": 0.003, + "loss": 4.3061, + "step": 1244 + }, + { + "epoch": 0.01245, + "grad_norm": 0.8249465227127075, + "learning_rate": 0.003, + "loss": 4.3067, + "step": 1245 + }, + { + "epoch": 0.01246, + "grad_norm": 0.983553946018219, + "learning_rate": 0.003, + "loss": 4.3098, + "step": 1246 + }, + { + "epoch": 0.01247, + "grad_norm": 0.8006371259689331, + "learning_rate": 0.003, + "loss": 4.3508, + "step": 1247 + }, + { + "epoch": 0.01248, + "grad_norm": 0.7486335635185242, + "learning_rate": 0.003, + "loss": 4.3351, + "step": 1248 + }, + { + "epoch": 0.01249, + "grad_norm": 0.731898307800293, + "learning_rate": 0.003, + "loss": 4.3269, + "step": 1249 + }, + { + "epoch": 0.0125, + "grad_norm": 0.7205458283424377, + "learning_rate": 0.003, + "loss": 4.3015, + "step": 1250 + }, + { + "epoch": 0.01251, + "grad_norm": 0.9905809760093689, + "learning_rate": 0.003, + "loss": 4.3425, + "step": 1251 + }, + { + "epoch": 0.01252, + "grad_norm": 0.7830891013145447, + "learning_rate": 0.003, + "loss": 4.319, + "step": 1252 + }, + { + "epoch": 0.01253, + "grad_norm": 0.6952295303344727, + "learning_rate": 0.003, + "loss": 4.2965, + "step": 1253 + }, + { + "epoch": 0.01254, + "grad_norm": 0.7467588782310486, + "learning_rate": 0.003, + "loss": 4.3427, + "step": 1254 + }, + { + "epoch": 0.01255, + "grad_norm": 0.7031508684158325, + "learning_rate": 0.003, + "loss": 4.2927, + "step": 1255 + }, + { + "epoch": 0.01256, + "grad_norm": 0.6603983640670776, + "learning_rate": 0.003, + "loss": 4.3194, + "step": 1256 + }, + { + "epoch": 0.01257, + "grad_norm": 0.6309519410133362, + "learning_rate": 0.003, + "loss": 4.3008, + "step": 1257 + }, + { + "epoch": 0.01258, + "grad_norm": 0.6506064534187317, + "learning_rate": 0.003, + "loss": 4.3107, + "step": 1258 + }, + { + "epoch": 0.01259, + "grad_norm": 0.5962826013565063, + "learning_rate": 0.003, + "loss": 4.3105, + "step": 1259 + }, + { + "epoch": 0.0126, + "grad_norm": 0.5265083909034729, + "learning_rate": 0.003, + "loss": 4.3316, + "step": 1260 + }, + { + "epoch": 0.01261, + "grad_norm": 0.5279543995857239, + "learning_rate": 0.003, + "loss": 4.3275, + "step": 1261 + }, + { + "epoch": 0.01262, + "grad_norm": 0.5171265006065369, + "learning_rate": 0.003, + "loss": 4.3175, + "step": 1262 + }, + { + "epoch": 0.01263, + "grad_norm": 0.5120646953582764, + "learning_rate": 0.003, + "loss": 4.3111, + "step": 1263 + }, + { + "epoch": 0.01264, + "grad_norm": 0.5407173037528992, + "learning_rate": 0.003, + "loss": 4.3, + "step": 1264 + }, + { + "epoch": 0.01265, + "grad_norm": 0.5656223893165588, + "learning_rate": 0.003, + "loss": 4.3113, + "step": 1265 + }, + { + "epoch": 0.01266, + "grad_norm": 0.4160173237323761, + "learning_rate": 0.003, + "loss": 4.3033, + "step": 1266 + }, + { + "epoch": 0.01267, + "grad_norm": 0.4453507661819458, + "learning_rate": 0.003, + "loss": 4.3244, + "step": 1267 + }, + { + "epoch": 0.01268, + "grad_norm": 0.4457267224788666, + "learning_rate": 0.003, + "loss": 4.3093, + "step": 1268 + }, + { + "epoch": 0.01269, + "grad_norm": 0.38706403970718384, + "learning_rate": 0.003, + "loss": 4.2734, + "step": 1269 + }, + { + "epoch": 0.0127, + "grad_norm": 0.38581886887550354, + "learning_rate": 0.003, + "loss": 4.3082, + "step": 1270 + }, + { + "epoch": 0.01271, + "grad_norm": 0.40513476729393005, + "learning_rate": 0.003, + "loss": 4.2869, + "step": 1271 + }, + { + "epoch": 0.01272, + "grad_norm": 0.522612988948822, + "learning_rate": 0.003, + "loss": 4.2863, + "step": 1272 + }, + { + "epoch": 0.01273, + "grad_norm": 0.8000910878181458, + "learning_rate": 0.003, + "loss": 4.321, + "step": 1273 + }, + { + "epoch": 0.01274, + "grad_norm": 1.0203778743743896, + "learning_rate": 0.003, + "loss": 4.3505, + "step": 1274 + }, + { + "epoch": 0.01275, + "grad_norm": 0.8743080496788025, + "learning_rate": 0.003, + "loss": 4.3003, + "step": 1275 + }, + { + "epoch": 0.01276, + "grad_norm": 0.9019266963005066, + "learning_rate": 0.003, + "loss": 4.3144, + "step": 1276 + }, + { + "epoch": 0.01277, + "grad_norm": 0.822109580039978, + "learning_rate": 0.003, + "loss": 4.3254, + "step": 1277 + }, + { + "epoch": 0.01278, + "grad_norm": 0.6606391668319702, + "learning_rate": 0.003, + "loss": 4.325, + "step": 1278 + }, + { + "epoch": 0.01279, + "grad_norm": 0.7756208777427673, + "learning_rate": 0.003, + "loss": 4.3104, + "step": 1279 + }, + { + "epoch": 0.0128, + "grad_norm": 0.7292629480361938, + "learning_rate": 0.003, + "loss": 4.311, + "step": 1280 + }, + { + "epoch": 0.01281, + "grad_norm": 0.6916735768318176, + "learning_rate": 0.003, + "loss": 4.2953, + "step": 1281 + }, + { + "epoch": 0.01282, + "grad_norm": 0.6080433130264282, + "learning_rate": 0.003, + "loss": 4.3294, + "step": 1282 + }, + { + "epoch": 0.01283, + "grad_norm": 0.6106524467468262, + "learning_rate": 0.003, + "loss": 4.301, + "step": 1283 + }, + { + "epoch": 0.01284, + "grad_norm": 0.5725691318511963, + "learning_rate": 0.003, + "loss": 4.2991, + "step": 1284 + }, + { + "epoch": 0.01285, + "grad_norm": 0.6401168704032898, + "learning_rate": 0.003, + "loss": 4.3022, + "step": 1285 + }, + { + "epoch": 0.01286, + "grad_norm": 0.620028555393219, + "learning_rate": 0.003, + "loss": 4.3246, + "step": 1286 + }, + { + "epoch": 0.01287, + "grad_norm": 0.662833034992218, + "learning_rate": 0.003, + "loss": 4.2902, + "step": 1287 + }, + { + "epoch": 0.01288, + "grad_norm": 0.6968905925750732, + "learning_rate": 0.003, + "loss": 4.307, + "step": 1288 + }, + { + "epoch": 0.01289, + "grad_norm": 0.5818215608596802, + "learning_rate": 0.003, + "loss": 4.2943, + "step": 1289 + }, + { + "epoch": 0.0129, + "grad_norm": 0.5997996926307678, + "learning_rate": 0.003, + "loss": 4.3178, + "step": 1290 + }, + { + "epoch": 0.01291, + "grad_norm": 0.6353629231452942, + "learning_rate": 0.003, + "loss": 4.2994, + "step": 1291 + }, + { + "epoch": 0.01292, + "grad_norm": 0.7124042510986328, + "learning_rate": 0.003, + "loss": 4.3335, + "step": 1292 + }, + { + "epoch": 0.01293, + "grad_norm": 0.8257373571395874, + "learning_rate": 0.003, + "loss": 4.3265, + "step": 1293 + }, + { + "epoch": 0.01294, + "grad_norm": 0.7939269542694092, + "learning_rate": 0.003, + "loss": 4.2974, + "step": 1294 + }, + { + "epoch": 0.01295, + "grad_norm": 0.703215479850769, + "learning_rate": 0.003, + "loss": 4.3191, + "step": 1295 + }, + { + "epoch": 0.01296, + "grad_norm": 0.851357102394104, + "learning_rate": 0.003, + "loss": 4.2949, + "step": 1296 + }, + { + "epoch": 0.01297, + "grad_norm": 0.8115617036819458, + "learning_rate": 0.003, + "loss": 4.309, + "step": 1297 + }, + { + "epoch": 0.01298, + "grad_norm": 0.8384767174720764, + "learning_rate": 0.003, + "loss": 4.2888, + "step": 1298 + }, + { + "epoch": 0.01299, + "grad_norm": 0.8158665895462036, + "learning_rate": 0.003, + "loss": 4.305, + "step": 1299 + }, + { + "epoch": 0.013, + "grad_norm": 0.7459951639175415, + "learning_rate": 0.003, + "loss": 4.29, + "step": 1300 + }, + { + "epoch": 0.01301, + "grad_norm": 0.7418009042739868, + "learning_rate": 0.003, + "loss": 4.3054, + "step": 1301 + }, + { + "epoch": 0.01302, + "grad_norm": 0.8234617710113525, + "learning_rate": 0.003, + "loss": 4.358, + "step": 1302 + }, + { + "epoch": 0.01303, + "grad_norm": 0.7982010841369629, + "learning_rate": 0.003, + "loss": 4.3068, + "step": 1303 + }, + { + "epoch": 0.01304, + "grad_norm": 0.812175452709198, + "learning_rate": 0.003, + "loss": 4.3282, + "step": 1304 + }, + { + "epoch": 0.01305, + "grad_norm": 0.8789085745811462, + "learning_rate": 0.003, + "loss": 4.322, + "step": 1305 + }, + { + "epoch": 0.01306, + "grad_norm": 0.828743577003479, + "learning_rate": 0.003, + "loss": 4.3122, + "step": 1306 + }, + { + "epoch": 0.01307, + "grad_norm": 0.7142960429191589, + "learning_rate": 0.003, + "loss": 4.3298, + "step": 1307 + }, + { + "epoch": 0.01308, + "grad_norm": 0.7313523888587952, + "learning_rate": 0.003, + "loss": 4.3053, + "step": 1308 + }, + { + "epoch": 0.01309, + "grad_norm": 0.7293263077735901, + "learning_rate": 0.003, + "loss": 4.3473, + "step": 1309 + }, + { + "epoch": 0.0131, + "grad_norm": 0.7924617528915405, + "learning_rate": 0.003, + "loss": 4.2932, + "step": 1310 + }, + { + "epoch": 0.01311, + "grad_norm": 0.8199212551116943, + "learning_rate": 0.003, + "loss": 4.3417, + "step": 1311 + }, + { + "epoch": 0.01312, + "grad_norm": 0.8725687861442566, + "learning_rate": 0.003, + "loss": 4.3076, + "step": 1312 + }, + { + "epoch": 0.01313, + "grad_norm": 0.8202559351921082, + "learning_rate": 0.003, + "loss": 4.2971, + "step": 1313 + }, + { + "epoch": 0.01314, + "grad_norm": 0.7738103866577148, + "learning_rate": 0.003, + "loss": 4.2944, + "step": 1314 + }, + { + "epoch": 0.01315, + "grad_norm": 0.6587685346603394, + "learning_rate": 0.003, + "loss": 4.3124, + "step": 1315 + }, + { + "epoch": 0.01316, + "grad_norm": 0.7233019471168518, + "learning_rate": 0.003, + "loss": 4.3037, + "step": 1316 + }, + { + "epoch": 0.01317, + "grad_norm": 0.6440667510032654, + "learning_rate": 0.003, + "loss": 4.3256, + "step": 1317 + }, + { + "epoch": 0.01318, + "grad_norm": 0.5893415212631226, + "learning_rate": 0.003, + "loss": 4.301, + "step": 1318 + }, + { + "epoch": 0.01319, + "grad_norm": 0.7347000241279602, + "learning_rate": 0.003, + "loss": 4.3133, + "step": 1319 + }, + { + "epoch": 0.0132, + "grad_norm": 0.8620561957359314, + "learning_rate": 0.003, + "loss": 4.3153, + "step": 1320 + }, + { + "epoch": 0.01321, + "grad_norm": 0.9148700833320618, + "learning_rate": 0.003, + "loss": 4.2941, + "step": 1321 + }, + { + "epoch": 0.01322, + "grad_norm": 0.8122279047966003, + "learning_rate": 0.003, + "loss": 4.3273, + "step": 1322 + }, + { + "epoch": 0.01323, + "grad_norm": 0.7629367113113403, + "learning_rate": 0.003, + "loss": 4.3161, + "step": 1323 + }, + { + "epoch": 0.01324, + "grad_norm": 0.6652523875236511, + "learning_rate": 0.003, + "loss": 4.2981, + "step": 1324 + }, + { + "epoch": 0.01325, + "grad_norm": 0.6290608048439026, + "learning_rate": 0.003, + "loss": 4.3001, + "step": 1325 + }, + { + "epoch": 0.01326, + "grad_norm": 0.6334658861160278, + "learning_rate": 0.003, + "loss": 4.2848, + "step": 1326 + }, + { + "epoch": 0.01327, + "grad_norm": 0.5682603716850281, + "learning_rate": 0.003, + "loss": 4.3018, + "step": 1327 + }, + { + "epoch": 0.01328, + "grad_norm": 0.4706239700317383, + "learning_rate": 0.003, + "loss": 4.2884, + "step": 1328 + }, + { + "epoch": 0.01329, + "grad_norm": 0.46228259801864624, + "learning_rate": 0.003, + "loss": 4.2518, + "step": 1329 + }, + { + "epoch": 0.0133, + "grad_norm": 0.4335189163684845, + "learning_rate": 0.003, + "loss": 4.293, + "step": 1330 + }, + { + "epoch": 0.01331, + "grad_norm": 0.383666068315506, + "learning_rate": 0.003, + "loss": 4.2772, + "step": 1331 + }, + { + "epoch": 0.01332, + "grad_norm": 0.39924779534339905, + "learning_rate": 0.003, + "loss": 4.2955, + "step": 1332 + }, + { + "epoch": 0.01333, + "grad_norm": 0.4323978126049042, + "learning_rate": 0.003, + "loss": 4.2692, + "step": 1333 + }, + { + "epoch": 0.01334, + "grad_norm": 0.43190306425094604, + "learning_rate": 0.003, + "loss": 4.2574, + "step": 1334 + }, + { + "epoch": 0.01335, + "grad_norm": 0.49212101101875305, + "learning_rate": 0.003, + "loss": 4.2803, + "step": 1335 + }, + { + "epoch": 0.01336, + "grad_norm": 0.6307611465454102, + "learning_rate": 0.003, + "loss": 4.2833, + "step": 1336 + }, + { + "epoch": 0.01337, + "grad_norm": 0.9523991942405701, + "learning_rate": 0.003, + "loss": 4.31, + "step": 1337 + }, + { + "epoch": 0.01338, + "grad_norm": 1.0491547584533691, + "learning_rate": 0.003, + "loss": 4.3099, + "step": 1338 + }, + { + "epoch": 0.01339, + "grad_norm": 0.7694852948188782, + "learning_rate": 0.003, + "loss": 4.307, + "step": 1339 + }, + { + "epoch": 0.0134, + "grad_norm": 0.7887849807739258, + "learning_rate": 0.003, + "loss": 4.3001, + "step": 1340 + }, + { + "epoch": 0.01341, + "grad_norm": 0.7728468179702759, + "learning_rate": 0.003, + "loss": 4.3115, + "step": 1341 + }, + { + "epoch": 0.01342, + "grad_norm": 0.7008172273635864, + "learning_rate": 0.003, + "loss": 4.2807, + "step": 1342 + }, + { + "epoch": 0.01343, + "grad_norm": 0.7046615481376648, + "learning_rate": 0.003, + "loss": 4.2868, + "step": 1343 + }, + { + "epoch": 0.01344, + "grad_norm": 0.6420544385910034, + "learning_rate": 0.003, + "loss": 4.2829, + "step": 1344 + }, + { + "epoch": 0.01345, + "grad_norm": 0.5914574861526489, + "learning_rate": 0.003, + "loss": 4.3212, + "step": 1345 + }, + { + "epoch": 0.01346, + "grad_norm": 0.5600994229316711, + "learning_rate": 0.003, + "loss": 4.2911, + "step": 1346 + }, + { + "epoch": 0.01347, + "grad_norm": 0.46645763516426086, + "learning_rate": 0.003, + "loss": 4.2774, + "step": 1347 + }, + { + "epoch": 0.01348, + "grad_norm": 0.49815306067466736, + "learning_rate": 0.003, + "loss": 4.2778, + "step": 1348 + }, + { + "epoch": 0.01349, + "grad_norm": 0.525874674320221, + "learning_rate": 0.003, + "loss": 4.2987, + "step": 1349 + }, + { + "epoch": 0.0135, + "grad_norm": 0.487930566072464, + "learning_rate": 0.003, + "loss": 4.2878, + "step": 1350 + }, + { + "epoch": 0.01351, + "grad_norm": 0.5213081240653992, + "learning_rate": 0.003, + "loss": 4.3174, + "step": 1351 + }, + { + "epoch": 0.01352, + "grad_norm": 0.6742727160453796, + "learning_rate": 0.003, + "loss": 4.2961, + "step": 1352 + }, + { + "epoch": 0.01353, + "grad_norm": 0.7189781069755554, + "learning_rate": 0.003, + "loss": 4.2871, + "step": 1353 + }, + { + "epoch": 0.01354, + "grad_norm": 0.59047532081604, + "learning_rate": 0.003, + "loss": 4.2676, + "step": 1354 + }, + { + "epoch": 0.01355, + "grad_norm": 0.5686421394348145, + "learning_rate": 0.003, + "loss": 4.2819, + "step": 1355 + }, + { + "epoch": 0.01356, + "grad_norm": 0.5455272793769836, + "learning_rate": 0.003, + "loss": 4.2746, + "step": 1356 + }, + { + "epoch": 0.01357, + "grad_norm": 0.6083900332450867, + "learning_rate": 0.003, + "loss": 4.2628, + "step": 1357 + }, + { + "epoch": 0.01358, + "grad_norm": 0.5685535073280334, + "learning_rate": 0.003, + "loss": 4.258, + "step": 1358 + }, + { + "epoch": 0.01359, + "grad_norm": 0.5254443883895874, + "learning_rate": 0.003, + "loss": 4.2971, + "step": 1359 + }, + { + "epoch": 0.0136, + "grad_norm": 0.6204063892364502, + "learning_rate": 0.003, + "loss": 4.2647, + "step": 1360 + }, + { + "epoch": 0.01361, + "grad_norm": 0.7256515026092529, + "learning_rate": 0.003, + "loss": 4.2884, + "step": 1361 + }, + { + "epoch": 0.01362, + "grad_norm": 0.8139166831970215, + "learning_rate": 0.003, + "loss": 4.3042, + "step": 1362 + }, + { + "epoch": 0.01363, + "grad_norm": 0.7577769756317139, + "learning_rate": 0.003, + "loss": 4.3089, + "step": 1363 + }, + { + "epoch": 0.01364, + "grad_norm": 0.6858768463134766, + "learning_rate": 0.003, + "loss": 4.2972, + "step": 1364 + }, + { + "epoch": 0.01365, + "grad_norm": 0.7445408701896667, + "learning_rate": 0.003, + "loss": 4.315, + "step": 1365 + }, + { + "epoch": 0.01366, + "grad_norm": 0.9198596477508545, + "learning_rate": 0.003, + "loss": 4.2795, + "step": 1366 + }, + { + "epoch": 0.01367, + "grad_norm": 0.8477755188941956, + "learning_rate": 0.003, + "loss": 4.3279, + "step": 1367 + }, + { + "epoch": 0.01368, + "grad_norm": 0.8082318305969238, + "learning_rate": 0.003, + "loss": 4.302, + "step": 1368 + }, + { + "epoch": 0.01369, + "grad_norm": 0.8816357851028442, + "learning_rate": 0.003, + "loss": 4.3, + "step": 1369 + }, + { + "epoch": 0.0137, + "grad_norm": 0.8340180516242981, + "learning_rate": 0.003, + "loss": 4.326, + "step": 1370 + }, + { + "epoch": 0.01371, + "grad_norm": 0.7977674007415771, + "learning_rate": 0.003, + "loss": 4.3179, + "step": 1371 + }, + { + "epoch": 0.01372, + "grad_norm": 0.7062758803367615, + "learning_rate": 0.003, + "loss": 4.2633, + "step": 1372 + }, + { + "epoch": 0.01373, + "grad_norm": 0.6788503527641296, + "learning_rate": 0.003, + "loss": 4.3, + "step": 1373 + }, + { + "epoch": 0.01374, + "grad_norm": 0.6201223731040955, + "learning_rate": 0.003, + "loss": 4.3053, + "step": 1374 + }, + { + "epoch": 0.01375, + "grad_norm": 0.6361509561538696, + "learning_rate": 0.003, + "loss": 4.264, + "step": 1375 + }, + { + "epoch": 0.01376, + "grad_norm": 0.5589747428894043, + "learning_rate": 0.003, + "loss": 4.2871, + "step": 1376 + }, + { + "epoch": 0.01377, + "grad_norm": 0.6119049787521362, + "learning_rate": 0.003, + "loss": 4.2875, + "step": 1377 + }, + { + "epoch": 0.01378, + "grad_norm": 0.5476299524307251, + "learning_rate": 0.003, + "loss": 4.3133, + "step": 1378 + }, + { + "epoch": 0.01379, + "grad_norm": 0.5394819974899292, + "learning_rate": 0.003, + "loss": 4.2841, + "step": 1379 + }, + { + "epoch": 0.0138, + "grad_norm": 0.49164652824401855, + "learning_rate": 0.003, + "loss": 4.2791, + "step": 1380 + }, + { + "epoch": 0.01381, + "grad_norm": 0.4505774974822998, + "learning_rate": 0.003, + "loss": 4.2845, + "step": 1381 + }, + { + "epoch": 0.01382, + "grad_norm": 0.4613495469093323, + "learning_rate": 0.003, + "loss": 4.279, + "step": 1382 + }, + { + "epoch": 0.01383, + "grad_norm": 0.6032357215881348, + "learning_rate": 0.003, + "loss": 4.2785, + "step": 1383 + }, + { + "epoch": 0.01384, + "grad_norm": 0.8419365882873535, + "learning_rate": 0.003, + "loss": 4.3193, + "step": 1384 + }, + { + "epoch": 0.01385, + "grad_norm": 1.1751699447631836, + "learning_rate": 0.003, + "loss": 4.3253, + "step": 1385 + }, + { + "epoch": 0.01386, + "grad_norm": 0.6834710240364075, + "learning_rate": 0.003, + "loss": 4.2903, + "step": 1386 + }, + { + "epoch": 0.01387, + "grad_norm": 0.5957779288291931, + "learning_rate": 0.003, + "loss": 4.2761, + "step": 1387 + }, + { + "epoch": 0.01388, + "grad_norm": 0.6788092851638794, + "learning_rate": 0.003, + "loss": 4.3125, + "step": 1388 + }, + { + "epoch": 0.01389, + "grad_norm": 0.6753062605857849, + "learning_rate": 0.003, + "loss": 4.2893, + "step": 1389 + }, + { + "epoch": 0.0139, + "grad_norm": 0.6624464392662048, + "learning_rate": 0.003, + "loss": 4.2919, + "step": 1390 + }, + { + "epoch": 0.01391, + "grad_norm": 0.5935611724853516, + "learning_rate": 0.003, + "loss": 4.2895, + "step": 1391 + }, + { + "epoch": 0.01392, + "grad_norm": 0.5433003902435303, + "learning_rate": 0.003, + "loss": 4.2973, + "step": 1392 + }, + { + "epoch": 0.01393, + "grad_norm": 0.5145422220230103, + "learning_rate": 0.003, + "loss": 4.2666, + "step": 1393 + }, + { + "epoch": 0.01394, + "grad_norm": 0.49970656633377075, + "learning_rate": 0.003, + "loss": 4.2526, + "step": 1394 + }, + { + "epoch": 0.01395, + "grad_norm": 0.5094105005264282, + "learning_rate": 0.003, + "loss": 4.283, + "step": 1395 + }, + { + "epoch": 0.01396, + "grad_norm": 0.5121984481811523, + "learning_rate": 0.003, + "loss": 4.274, + "step": 1396 + }, + { + "epoch": 0.01397, + "grad_norm": 0.5195156931877136, + "learning_rate": 0.003, + "loss": 4.286, + "step": 1397 + }, + { + "epoch": 0.01398, + "grad_norm": 0.47007647156715393, + "learning_rate": 0.003, + "loss": 4.2651, + "step": 1398 + }, + { + "epoch": 0.01399, + "grad_norm": 0.4805718660354614, + "learning_rate": 0.003, + "loss": 4.287, + "step": 1399 + }, + { + "epoch": 0.014, + "grad_norm": 0.6213575005531311, + "learning_rate": 0.003, + "loss": 4.2485, + "step": 1400 + }, + { + "epoch": 0.01401, + "grad_norm": 0.7338332533836365, + "learning_rate": 0.003, + "loss": 4.2762, + "step": 1401 + }, + { + "epoch": 0.01402, + "grad_norm": 0.7789960503578186, + "learning_rate": 0.003, + "loss": 4.2883, + "step": 1402 + }, + { + "epoch": 0.01403, + "grad_norm": 0.7047394514083862, + "learning_rate": 0.003, + "loss": 4.2622, + "step": 1403 + }, + { + "epoch": 0.01404, + "grad_norm": 0.5580564737319946, + "learning_rate": 0.003, + "loss": 4.279, + "step": 1404 + }, + { + "epoch": 0.01405, + "grad_norm": 0.6150562763214111, + "learning_rate": 0.003, + "loss": 4.2978, + "step": 1405 + }, + { + "epoch": 0.01406, + "grad_norm": 0.6412274241447449, + "learning_rate": 0.003, + "loss": 4.2817, + "step": 1406 + }, + { + "epoch": 0.01407, + "grad_norm": 0.6951134204864502, + "learning_rate": 0.003, + "loss": 4.2975, + "step": 1407 + }, + { + "epoch": 0.01408, + "grad_norm": 0.7907094955444336, + "learning_rate": 0.003, + "loss": 4.2629, + "step": 1408 + }, + { + "epoch": 0.01409, + "grad_norm": 0.6941542625427246, + "learning_rate": 0.003, + "loss": 4.2885, + "step": 1409 + }, + { + "epoch": 0.0141, + "grad_norm": 0.6482591032981873, + "learning_rate": 0.003, + "loss": 4.2693, + "step": 1410 + }, + { + "epoch": 0.01411, + "grad_norm": 0.6562476754188538, + "learning_rate": 0.003, + "loss": 4.2437, + "step": 1411 + }, + { + "epoch": 0.01412, + "grad_norm": 0.7608663439750671, + "learning_rate": 0.003, + "loss": 4.2874, + "step": 1412 + }, + { + "epoch": 0.01413, + "grad_norm": 0.763931930065155, + "learning_rate": 0.003, + "loss": 4.2842, + "step": 1413 + }, + { + "epoch": 0.01414, + "grad_norm": 0.7897149324417114, + "learning_rate": 0.003, + "loss": 4.2662, + "step": 1414 + }, + { + "epoch": 0.01415, + "grad_norm": 0.7817095518112183, + "learning_rate": 0.003, + "loss": 4.2634, + "step": 1415 + }, + { + "epoch": 0.01416, + "grad_norm": 0.7488892674446106, + "learning_rate": 0.003, + "loss": 4.2852, + "step": 1416 + }, + { + "epoch": 0.01417, + "grad_norm": 0.8105233907699585, + "learning_rate": 0.003, + "loss": 4.2802, + "step": 1417 + }, + { + "epoch": 0.01418, + "grad_norm": 0.7555150389671326, + "learning_rate": 0.003, + "loss": 4.2974, + "step": 1418 + }, + { + "epoch": 0.01419, + "grad_norm": 0.7873964309692383, + "learning_rate": 0.003, + "loss": 4.2928, + "step": 1419 + }, + { + "epoch": 0.0142, + "grad_norm": 0.942128598690033, + "learning_rate": 0.003, + "loss": 4.3049, + "step": 1420 + }, + { + "epoch": 0.01421, + "grad_norm": 1.1309813261032104, + "learning_rate": 0.003, + "loss": 4.2967, + "step": 1421 + }, + { + "epoch": 0.01422, + "grad_norm": 0.8632763028144836, + "learning_rate": 0.003, + "loss": 4.2868, + "step": 1422 + }, + { + "epoch": 0.01423, + "grad_norm": 0.8987135887145996, + "learning_rate": 0.003, + "loss": 4.2941, + "step": 1423 + }, + { + "epoch": 0.01424, + "grad_norm": 0.7247830629348755, + "learning_rate": 0.003, + "loss": 4.3046, + "step": 1424 + }, + { + "epoch": 0.01425, + "grad_norm": 0.6278427839279175, + "learning_rate": 0.003, + "loss": 4.3054, + "step": 1425 + }, + { + "epoch": 0.01426, + "grad_norm": 0.6600518226623535, + "learning_rate": 0.003, + "loss": 4.2819, + "step": 1426 + }, + { + "epoch": 0.01427, + "grad_norm": 0.744967520236969, + "learning_rate": 0.003, + "loss": 4.2813, + "step": 1427 + }, + { + "epoch": 0.01428, + "grad_norm": 0.6531183123588562, + "learning_rate": 0.003, + "loss": 4.2709, + "step": 1428 + }, + { + "epoch": 0.01429, + "grad_norm": 0.6728975176811218, + "learning_rate": 0.003, + "loss": 4.2816, + "step": 1429 + }, + { + "epoch": 0.0143, + "grad_norm": 0.6716436147689819, + "learning_rate": 0.003, + "loss": 4.2945, + "step": 1430 + }, + { + "epoch": 0.01431, + "grad_norm": 0.674777090549469, + "learning_rate": 0.003, + "loss": 4.2804, + "step": 1431 + }, + { + "epoch": 0.01432, + "grad_norm": 0.6107741594314575, + "learning_rate": 0.003, + "loss": 4.2694, + "step": 1432 + }, + { + "epoch": 0.01433, + "grad_norm": 0.5782887935638428, + "learning_rate": 0.003, + "loss": 4.2651, + "step": 1433 + }, + { + "epoch": 0.01434, + "grad_norm": 0.5728036761283875, + "learning_rate": 0.003, + "loss": 4.2497, + "step": 1434 + }, + { + "epoch": 0.01435, + "grad_norm": 0.5406824350357056, + "learning_rate": 0.003, + "loss": 4.2562, + "step": 1435 + }, + { + "epoch": 0.01436, + "grad_norm": 0.5029864311218262, + "learning_rate": 0.003, + "loss": 4.2703, + "step": 1436 + }, + { + "epoch": 0.01437, + "grad_norm": 0.5657121539115906, + "learning_rate": 0.003, + "loss": 4.2852, + "step": 1437 + }, + { + "epoch": 0.01438, + "grad_norm": 0.5845190286636353, + "learning_rate": 0.003, + "loss": 4.2793, + "step": 1438 + }, + { + "epoch": 0.01439, + "grad_norm": 0.591201901435852, + "learning_rate": 0.003, + "loss": 4.2448, + "step": 1439 + }, + { + "epoch": 0.0144, + "grad_norm": 0.8200764656066895, + "learning_rate": 0.003, + "loss": 4.2641, + "step": 1440 + }, + { + "epoch": 0.01441, + "grad_norm": 0.888174831867218, + "learning_rate": 0.003, + "loss": 4.2921, + "step": 1441 + }, + { + "epoch": 0.01442, + "grad_norm": 0.79758220911026, + "learning_rate": 0.003, + "loss": 4.2638, + "step": 1442 + }, + { + "epoch": 0.01443, + "grad_norm": 0.745823085308075, + "learning_rate": 0.003, + "loss": 4.3231, + "step": 1443 + }, + { + "epoch": 0.01444, + "grad_norm": 0.8156546354293823, + "learning_rate": 0.003, + "loss": 4.2923, + "step": 1444 + }, + { + "epoch": 0.01445, + "grad_norm": 0.7982717752456665, + "learning_rate": 0.003, + "loss": 4.2557, + "step": 1445 + }, + { + "epoch": 0.01446, + "grad_norm": 0.6928796172142029, + "learning_rate": 0.003, + "loss": 4.264, + "step": 1446 + }, + { + "epoch": 0.01447, + "grad_norm": 0.6646214723587036, + "learning_rate": 0.003, + "loss": 4.2746, + "step": 1447 + }, + { + "epoch": 0.01448, + "grad_norm": 0.6782436966896057, + "learning_rate": 0.003, + "loss": 4.2822, + "step": 1448 + }, + { + "epoch": 0.01449, + "grad_norm": 0.6758451461791992, + "learning_rate": 0.003, + "loss": 4.2758, + "step": 1449 + }, + { + "epoch": 0.0145, + "grad_norm": 0.6399548649787903, + "learning_rate": 0.003, + "loss": 4.291, + "step": 1450 + }, + { + "epoch": 0.01451, + "grad_norm": 0.6663489937782288, + "learning_rate": 0.003, + "loss": 4.2675, + "step": 1451 + }, + { + "epoch": 0.01452, + "grad_norm": 0.676326334476471, + "learning_rate": 0.003, + "loss": 4.2687, + "step": 1452 + }, + { + "epoch": 0.01453, + "grad_norm": 0.5554569959640503, + "learning_rate": 0.003, + "loss": 4.2765, + "step": 1453 + }, + { + "epoch": 0.01454, + "grad_norm": 0.5639092922210693, + "learning_rate": 0.003, + "loss": 4.2445, + "step": 1454 + }, + { + "epoch": 0.01455, + "grad_norm": 0.49674221873283386, + "learning_rate": 0.003, + "loss": 4.2722, + "step": 1455 + }, + { + "epoch": 0.01456, + "grad_norm": 0.5553675293922424, + "learning_rate": 0.003, + "loss": 4.2521, + "step": 1456 + }, + { + "epoch": 0.01457, + "grad_norm": 0.6361852288246155, + "learning_rate": 0.003, + "loss": 4.2449, + "step": 1457 + }, + { + "epoch": 0.01458, + "grad_norm": 0.7916824817657471, + "learning_rate": 0.003, + "loss": 4.2458, + "step": 1458 + }, + { + "epoch": 0.01459, + "grad_norm": 0.8565419316291809, + "learning_rate": 0.003, + "loss": 4.2834, + "step": 1459 + }, + { + "epoch": 0.0146, + "grad_norm": 0.7326977252960205, + "learning_rate": 0.003, + "loss": 4.28, + "step": 1460 + }, + { + "epoch": 0.01461, + "grad_norm": 0.7401906251907349, + "learning_rate": 0.003, + "loss": 4.2599, + "step": 1461 + }, + { + "epoch": 0.01462, + "grad_norm": 0.7613895535469055, + "learning_rate": 0.003, + "loss": 4.2676, + "step": 1462 + }, + { + "epoch": 0.01463, + "grad_norm": 0.642987072467804, + "learning_rate": 0.003, + "loss": 4.2603, + "step": 1463 + }, + { + "epoch": 0.01464, + "grad_norm": 0.705771803855896, + "learning_rate": 0.003, + "loss": 4.2361, + "step": 1464 + }, + { + "epoch": 0.01465, + "grad_norm": 0.6884810328483582, + "learning_rate": 0.003, + "loss": 4.284, + "step": 1465 + }, + { + "epoch": 0.01466, + "grad_norm": 0.5892930626869202, + "learning_rate": 0.003, + "loss": 4.2489, + "step": 1466 + }, + { + "epoch": 0.01467, + "grad_norm": 0.6196565628051758, + "learning_rate": 0.003, + "loss": 4.2848, + "step": 1467 + }, + { + "epoch": 0.01468, + "grad_norm": 0.7562909126281738, + "learning_rate": 0.003, + "loss": 4.2639, + "step": 1468 + }, + { + "epoch": 0.01469, + "grad_norm": 0.7464540004730225, + "learning_rate": 0.003, + "loss": 4.2693, + "step": 1469 + }, + { + "epoch": 0.0147, + "grad_norm": 0.6821353435516357, + "learning_rate": 0.003, + "loss": 4.2569, + "step": 1470 + }, + { + "epoch": 0.01471, + "grad_norm": 0.5921043753623962, + "learning_rate": 0.003, + "loss": 4.2689, + "step": 1471 + }, + { + "epoch": 0.01472, + "grad_norm": 0.531998336315155, + "learning_rate": 0.003, + "loss": 4.252, + "step": 1472 + }, + { + "epoch": 0.01473, + "grad_norm": 0.5447686910629272, + "learning_rate": 0.003, + "loss": 4.2658, + "step": 1473 + }, + { + "epoch": 0.01474, + "grad_norm": 0.6622011661529541, + "learning_rate": 0.003, + "loss": 4.2515, + "step": 1474 + }, + { + "epoch": 0.01475, + "grad_norm": 0.7866929173469543, + "learning_rate": 0.003, + "loss": 4.2783, + "step": 1475 + }, + { + "epoch": 0.01476, + "grad_norm": 0.965038537979126, + "learning_rate": 0.003, + "loss": 4.2726, + "step": 1476 + }, + { + "epoch": 0.01477, + "grad_norm": 0.9310076236724854, + "learning_rate": 0.003, + "loss": 4.2926, + "step": 1477 + }, + { + "epoch": 0.01478, + "grad_norm": 0.7643823027610779, + "learning_rate": 0.003, + "loss": 4.2629, + "step": 1478 + }, + { + "epoch": 0.01479, + "grad_norm": 0.867812991142273, + "learning_rate": 0.003, + "loss": 4.3023, + "step": 1479 + }, + { + "epoch": 0.0148, + "grad_norm": 0.7325505018234253, + "learning_rate": 0.003, + "loss": 4.2689, + "step": 1480 + }, + { + "epoch": 0.01481, + "grad_norm": 0.673205554485321, + "learning_rate": 0.003, + "loss": 4.2668, + "step": 1481 + }, + { + "epoch": 0.01482, + "grad_norm": 0.7312605381011963, + "learning_rate": 0.003, + "loss": 4.2525, + "step": 1482 + }, + { + "epoch": 0.01483, + "grad_norm": 0.7356554269790649, + "learning_rate": 0.003, + "loss": 4.256, + "step": 1483 + }, + { + "epoch": 0.01484, + "grad_norm": 0.6896143555641174, + "learning_rate": 0.003, + "loss": 4.3028, + "step": 1484 + }, + { + "epoch": 0.01485, + "grad_norm": 0.6053803563117981, + "learning_rate": 0.003, + "loss": 4.2617, + "step": 1485 + }, + { + "epoch": 0.01486, + "grad_norm": 0.5422018766403198, + "learning_rate": 0.003, + "loss": 4.2576, + "step": 1486 + }, + { + "epoch": 0.01487, + "grad_norm": 0.584571897983551, + "learning_rate": 0.003, + "loss": 4.2611, + "step": 1487 + }, + { + "epoch": 0.01488, + "grad_norm": 0.58668452501297, + "learning_rate": 0.003, + "loss": 4.2703, + "step": 1488 + }, + { + "epoch": 0.01489, + "grad_norm": 0.5606329441070557, + "learning_rate": 0.003, + "loss": 4.2644, + "step": 1489 + }, + { + "epoch": 0.0149, + "grad_norm": 0.5718882083892822, + "learning_rate": 0.003, + "loss": 4.2562, + "step": 1490 + }, + { + "epoch": 0.01491, + "grad_norm": 0.609704852104187, + "learning_rate": 0.003, + "loss": 4.2714, + "step": 1491 + }, + { + "epoch": 0.01492, + "grad_norm": 0.6093578338623047, + "learning_rate": 0.003, + "loss": 4.2341, + "step": 1492 + }, + { + "epoch": 0.01493, + "grad_norm": 0.6492394208908081, + "learning_rate": 0.003, + "loss": 4.2659, + "step": 1493 + }, + { + "epoch": 0.01494, + "grad_norm": 0.7214344143867493, + "learning_rate": 0.003, + "loss": 4.2478, + "step": 1494 + }, + { + "epoch": 0.01495, + "grad_norm": 0.6858140826225281, + "learning_rate": 0.003, + "loss": 4.2587, + "step": 1495 + }, + { + "epoch": 0.01496, + "grad_norm": 0.6935936808586121, + "learning_rate": 0.003, + "loss": 4.2367, + "step": 1496 + }, + { + "epoch": 0.01497, + "grad_norm": 0.6599563360214233, + "learning_rate": 0.003, + "loss": 4.2117, + "step": 1497 + }, + { + "epoch": 0.01498, + "grad_norm": 0.6300835609436035, + "learning_rate": 0.003, + "loss": 4.2634, + "step": 1498 + }, + { + "epoch": 0.01499, + "grad_norm": 0.721996009349823, + "learning_rate": 0.003, + "loss": 4.2933, + "step": 1499 + }, + { + "epoch": 0.015, + "grad_norm": 0.7603088021278381, + "learning_rate": 0.003, + "loss": 4.2446, + "step": 1500 + }, + { + "epoch": 0.01501, + "grad_norm": 0.7853468060493469, + "learning_rate": 0.003, + "loss": 4.2524, + "step": 1501 + }, + { + "epoch": 0.01502, + "grad_norm": 0.6559277772903442, + "learning_rate": 0.003, + "loss": 4.2862, + "step": 1502 + }, + { + "epoch": 0.01503, + "grad_norm": 0.6914763450622559, + "learning_rate": 0.003, + "loss": 4.2669, + "step": 1503 + }, + { + "epoch": 0.01504, + "grad_norm": 0.8051386475563049, + "learning_rate": 0.003, + "loss": 4.2746, + "step": 1504 + }, + { + "epoch": 0.01505, + "grad_norm": 0.7562403082847595, + "learning_rate": 0.003, + "loss": 4.2564, + "step": 1505 + }, + { + "epoch": 0.01506, + "grad_norm": 0.6991645693778992, + "learning_rate": 0.003, + "loss": 4.2715, + "step": 1506 + }, + { + "epoch": 0.01507, + "grad_norm": 0.6562188267707825, + "learning_rate": 0.003, + "loss": 4.2831, + "step": 1507 + }, + { + "epoch": 0.01508, + "grad_norm": 0.6328547596931458, + "learning_rate": 0.003, + "loss": 4.2611, + "step": 1508 + }, + { + "epoch": 0.01509, + "grad_norm": 0.6792221665382385, + "learning_rate": 0.003, + "loss": 4.2777, + "step": 1509 + }, + { + "epoch": 0.0151, + "grad_norm": 0.7432823777198792, + "learning_rate": 0.003, + "loss": 4.2532, + "step": 1510 + }, + { + "epoch": 0.01511, + "grad_norm": 0.8595705628395081, + "learning_rate": 0.003, + "loss": 4.2519, + "step": 1511 + }, + { + "epoch": 0.01512, + "grad_norm": 0.9392295479774475, + "learning_rate": 0.003, + "loss": 4.2664, + "step": 1512 + }, + { + "epoch": 0.01513, + "grad_norm": 0.706947386264801, + "learning_rate": 0.003, + "loss": 4.2625, + "step": 1513 + }, + { + "epoch": 0.01514, + "grad_norm": 0.6648169159889221, + "learning_rate": 0.003, + "loss": 4.2697, + "step": 1514 + }, + { + "epoch": 0.01515, + "grad_norm": 0.7984362244606018, + "learning_rate": 0.003, + "loss": 4.2642, + "step": 1515 + }, + { + "epoch": 0.01516, + "grad_norm": 0.8379718065261841, + "learning_rate": 0.003, + "loss": 4.2755, + "step": 1516 + }, + { + "epoch": 0.01517, + "grad_norm": 0.7840726375579834, + "learning_rate": 0.003, + "loss": 4.2252, + "step": 1517 + }, + { + "epoch": 0.01518, + "grad_norm": 0.7350525856018066, + "learning_rate": 0.003, + "loss": 4.2609, + "step": 1518 + }, + { + "epoch": 0.01519, + "grad_norm": 0.637050449848175, + "learning_rate": 0.003, + "loss": 4.2905, + "step": 1519 + }, + { + "epoch": 0.0152, + "grad_norm": 0.6365798711776733, + "learning_rate": 0.003, + "loss": 4.2441, + "step": 1520 + }, + { + "epoch": 0.01521, + "grad_norm": 0.5890719890594482, + "learning_rate": 0.003, + "loss": 4.2552, + "step": 1521 + }, + { + "epoch": 0.01522, + "grad_norm": 0.5176184773445129, + "learning_rate": 0.003, + "loss": 4.2486, + "step": 1522 + }, + { + "epoch": 0.01523, + "grad_norm": 0.5967531800270081, + "learning_rate": 0.003, + "loss": 4.2464, + "step": 1523 + }, + { + "epoch": 0.01524, + "grad_norm": 0.57130366563797, + "learning_rate": 0.003, + "loss": 4.2759, + "step": 1524 + }, + { + "epoch": 0.01525, + "grad_norm": 0.68913733959198, + "learning_rate": 0.003, + "loss": 4.2789, + "step": 1525 + }, + { + "epoch": 0.01526, + "grad_norm": 0.6771203279495239, + "learning_rate": 0.003, + "loss": 4.2725, + "step": 1526 + }, + { + "epoch": 0.01527, + "grad_norm": 0.7048394083976746, + "learning_rate": 0.003, + "loss": 4.2371, + "step": 1527 + }, + { + "epoch": 0.01528, + "grad_norm": 0.7917721271514893, + "learning_rate": 0.003, + "loss": 4.2607, + "step": 1528 + }, + { + "epoch": 0.01529, + "grad_norm": 0.7461357116699219, + "learning_rate": 0.003, + "loss": 4.2575, + "step": 1529 + }, + { + "epoch": 0.0153, + "grad_norm": 0.6422540545463562, + "learning_rate": 0.003, + "loss": 4.2576, + "step": 1530 + }, + { + "epoch": 0.01531, + "grad_norm": 0.6448599696159363, + "learning_rate": 0.003, + "loss": 4.2541, + "step": 1531 + }, + { + "epoch": 0.01532, + "grad_norm": 0.556128203868866, + "learning_rate": 0.003, + "loss": 4.2286, + "step": 1532 + }, + { + "epoch": 0.01533, + "grad_norm": 0.511016845703125, + "learning_rate": 0.003, + "loss": 4.2532, + "step": 1533 + }, + { + "epoch": 0.01534, + "grad_norm": 0.4958668351173401, + "learning_rate": 0.003, + "loss": 4.2504, + "step": 1534 + }, + { + "epoch": 0.01535, + "grad_norm": 0.472829669713974, + "learning_rate": 0.003, + "loss": 4.2611, + "step": 1535 + }, + { + "epoch": 0.01536, + "grad_norm": 0.4443574547767639, + "learning_rate": 0.003, + "loss": 4.2397, + "step": 1536 + }, + { + "epoch": 0.01537, + "grad_norm": 0.5031164884567261, + "learning_rate": 0.003, + "loss": 4.2348, + "step": 1537 + }, + { + "epoch": 0.01538, + "grad_norm": 0.6832772493362427, + "learning_rate": 0.003, + "loss": 4.2195, + "step": 1538 + }, + { + "epoch": 0.01539, + "grad_norm": 0.875408947467804, + "learning_rate": 0.003, + "loss": 4.2638, + "step": 1539 + }, + { + "epoch": 0.0154, + "grad_norm": 0.9963869452476501, + "learning_rate": 0.003, + "loss": 4.2726, + "step": 1540 + }, + { + "epoch": 0.01541, + "grad_norm": 0.8254619836807251, + "learning_rate": 0.003, + "loss": 4.249, + "step": 1541 + }, + { + "epoch": 0.01542, + "grad_norm": 0.8195037245750427, + "learning_rate": 0.003, + "loss": 4.2801, + "step": 1542 + }, + { + "epoch": 0.01543, + "grad_norm": 0.9007482528686523, + "learning_rate": 0.003, + "loss": 4.268, + "step": 1543 + }, + { + "epoch": 0.01544, + "grad_norm": 0.8777545690536499, + "learning_rate": 0.003, + "loss": 4.2622, + "step": 1544 + }, + { + "epoch": 0.01545, + "grad_norm": 0.7702150344848633, + "learning_rate": 0.003, + "loss": 4.3234, + "step": 1545 + }, + { + "epoch": 0.01546, + "grad_norm": 0.903275191783905, + "learning_rate": 0.003, + "loss": 4.2965, + "step": 1546 + }, + { + "epoch": 0.01547, + "grad_norm": 1.0503301620483398, + "learning_rate": 0.003, + "loss": 4.307, + "step": 1547 + }, + { + "epoch": 0.01548, + "grad_norm": 0.868294894695282, + "learning_rate": 0.003, + "loss": 4.2647, + "step": 1548 + }, + { + "epoch": 0.01549, + "grad_norm": 0.8206138610839844, + "learning_rate": 0.003, + "loss": 4.2515, + "step": 1549 + }, + { + "epoch": 0.0155, + "grad_norm": 0.9134683012962341, + "learning_rate": 0.003, + "loss": 4.2824, + "step": 1550 + }, + { + "epoch": 0.01551, + "grad_norm": 0.9439716935157776, + "learning_rate": 0.003, + "loss": 4.2675, + "step": 1551 + }, + { + "epoch": 0.01552, + "grad_norm": 0.7804630994796753, + "learning_rate": 0.003, + "loss": 4.28, + "step": 1552 + }, + { + "epoch": 0.01553, + "grad_norm": 0.6432939171791077, + "learning_rate": 0.003, + "loss": 4.2993, + "step": 1553 + }, + { + "epoch": 0.01554, + "grad_norm": 0.6834948062896729, + "learning_rate": 0.003, + "loss": 4.2822, + "step": 1554 + }, + { + "epoch": 0.01555, + "grad_norm": 0.7107034921646118, + "learning_rate": 0.003, + "loss": 4.2553, + "step": 1555 + }, + { + "epoch": 0.01556, + "grad_norm": 0.6976543664932251, + "learning_rate": 0.003, + "loss": 4.2901, + "step": 1556 + }, + { + "epoch": 0.01557, + "grad_norm": 0.6529300808906555, + "learning_rate": 0.003, + "loss": 4.2623, + "step": 1557 + }, + { + "epoch": 0.01558, + "grad_norm": 0.6456514000892639, + "learning_rate": 0.003, + "loss": 4.2503, + "step": 1558 + }, + { + "epoch": 0.01559, + "grad_norm": 0.5664896368980408, + "learning_rate": 0.003, + "loss": 4.2838, + "step": 1559 + }, + { + "epoch": 0.0156, + "grad_norm": 0.5338277816772461, + "learning_rate": 0.003, + "loss": 4.2719, + "step": 1560 + }, + { + "epoch": 0.01561, + "grad_norm": 0.45519402623176575, + "learning_rate": 0.003, + "loss": 4.2494, + "step": 1561 + }, + { + "epoch": 0.01562, + "grad_norm": 0.4456147849559784, + "learning_rate": 0.003, + "loss": 4.2411, + "step": 1562 + }, + { + "epoch": 0.01563, + "grad_norm": 0.3998737335205078, + "learning_rate": 0.003, + "loss": 4.2455, + "step": 1563 + }, + { + "epoch": 0.01564, + "grad_norm": 0.4089747965335846, + "learning_rate": 0.003, + "loss": 4.2731, + "step": 1564 + }, + { + "epoch": 0.01565, + "grad_norm": 0.41704586148262024, + "learning_rate": 0.003, + "loss": 4.2464, + "step": 1565 + }, + { + "epoch": 0.01566, + "grad_norm": 0.4549218416213989, + "learning_rate": 0.003, + "loss": 4.265, + "step": 1566 + }, + { + "epoch": 0.01567, + "grad_norm": 0.45544859766960144, + "learning_rate": 0.003, + "loss": 4.2568, + "step": 1567 + }, + { + "epoch": 0.01568, + "grad_norm": 0.496439129114151, + "learning_rate": 0.003, + "loss": 4.2394, + "step": 1568 + }, + { + "epoch": 0.01569, + "grad_norm": 0.5869415998458862, + "learning_rate": 0.003, + "loss": 4.2472, + "step": 1569 + }, + { + "epoch": 0.0157, + "grad_norm": 0.7246452569961548, + "learning_rate": 0.003, + "loss": 4.2683, + "step": 1570 + }, + { + "epoch": 0.01571, + "grad_norm": 0.807225227355957, + "learning_rate": 0.003, + "loss": 4.268, + "step": 1571 + }, + { + "epoch": 0.01572, + "grad_norm": 0.7089520692825317, + "learning_rate": 0.003, + "loss": 4.2658, + "step": 1572 + }, + { + "epoch": 0.01573, + "grad_norm": 0.7833078503608704, + "learning_rate": 0.003, + "loss": 4.2544, + "step": 1573 + }, + { + "epoch": 0.01574, + "grad_norm": 0.8009607791900635, + "learning_rate": 0.003, + "loss": 4.2723, + "step": 1574 + }, + { + "epoch": 0.01575, + "grad_norm": 0.7152323126792908, + "learning_rate": 0.003, + "loss": 4.2303, + "step": 1575 + }, + { + "epoch": 0.01576, + "grad_norm": 0.8425666093826294, + "learning_rate": 0.003, + "loss": 4.258, + "step": 1576 + }, + { + "epoch": 0.01577, + "grad_norm": 0.9029209613800049, + "learning_rate": 0.003, + "loss": 4.2227, + "step": 1577 + }, + { + "epoch": 0.01578, + "grad_norm": 0.8221943378448486, + "learning_rate": 0.003, + "loss": 4.2532, + "step": 1578 + }, + { + "epoch": 0.01579, + "grad_norm": 0.6562032699584961, + "learning_rate": 0.003, + "loss": 4.2422, + "step": 1579 + }, + { + "epoch": 0.0158, + "grad_norm": 0.6161283254623413, + "learning_rate": 0.003, + "loss": 4.2534, + "step": 1580 + }, + { + "epoch": 0.01581, + "grad_norm": 0.5998492240905762, + "learning_rate": 0.003, + "loss": 4.2595, + "step": 1581 + }, + { + "epoch": 0.01582, + "grad_norm": 0.6605640053749084, + "learning_rate": 0.003, + "loss": 4.2418, + "step": 1582 + }, + { + "epoch": 0.01583, + "grad_norm": 0.6276175379753113, + "learning_rate": 0.003, + "loss": 4.2293, + "step": 1583 + }, + { + "epoch": 0.01584, + "grad_norm": 0.5631775856018066, + "learning_rate": 0.003, + "loss": 4.2575, + "step": 1584 + }, + { + "epoch": 0.01585, + "grad_norm": 0.5975695252418518, + "learning_rate": 0.003, + "loss": 4.2521, + "step": 1585 + }, + { + "epoch": 0.01586, + "grad_norm": 0.7194235324859619, + "learning_rate": 0.003, + "loss": 4.229, + "step": 1586 + }, + { + "epoch": 0.01587, + "grad_norm": 0.7284054160118103, + "learning_rate": 0.003, + "loss": 4.246, + "step": 1587 + }, + { + "epoch": 0.01588, + "grad_norm": 0.6996057629585266, + "learning_rate": 0.003, + "loss": 4.2728, + "step": 1588 + }, + { + "epoch": 0.01589, + "grad_norm": 0.5657826662063599, + "learning_rate": 0.003, + "loss": 4.2616, + "step": 1589 + }, + { + "epoch": 0.0159, + "grad_norm": 0.6453202366828918, + "learning_rate": 0.003, + "loss": 4.2584, + "step": 1590 + }, + { + "epoch": 0.01591, + "grad_norm": 0.5882779955863953, + "learning_rate": 0.003, + "loss": 4.2627, + "step": 1591 + }, + { + "epoch": 0.01592, + "grad_norm": 0.5240422487258911, + "learning_rate": 0.003, + "loss": 4.2485, + "step": 1592 + }, + { + "epoch": 0.01593, + "grad_norm": 0.5395859479904175, + "learning_rate": 0.003, + "loss": 4.2507, + "step": 1593 + }, + { + "epoch": 0.01594, + "grad_norm": 0.5181586742401123, + "learning_rate": 0.003, + "loss": 4.2388, + "step": 1594 + }, + { + "epoch": 0.01595, + "grad_norm": 0.5144811272621155, + "learning_rate": 0.003, + "loss": 4.2661, + "step": 1595 + }, + { + "epoch": 0.01596, + "grad_norm": 0.5639116764068604, + "learning_rate": 0.003, + "loss": 4.2583, + "step": 1596 + }, + { + "epoch": 0.01597, + "grad_norm": 0.616133987903595, + "learning_rate": 0.003, + "loss": 4.2207, + "step": 1597 + }, + { + "epoch": 0.01598, + "grad_norm": 0.7612146735191345, + "learning_rate": 0.003, + "loss": 4.2233, + "step": 1598 + }, + { + "epoch": 0.01599, + "grad_norm": 0.8201577663421631, + "learning_rate": 0.003, + "loss": 4.2494, + "step": 1599 + }, + { + "epoch": 0.016, + "grad_norm": 0.7031198143959045, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 1600 + }, + { + "epoch": 0.01601, + "grad_norm": 0.7635224461555481, + "learning_rate": 0.003, + "loss": 4.2634, + "step": 1601 + }, + { + "epoch": 0.01602, + "grad_norm": 0.8013474941253662, + "learning_rate": 0.003, + "loss": 4.2394, + "step": 1602 + }, + { + "epoch": 0.01603, + "grad_norm": 0.7904717326164246, + "learning_rate": 0.003, + "loss": 4.2591, + "step": 1603 + }, + { + "epoch": 0.01604, + "grad_norm": 0.6602920293807983, + "learning_rate": 0.003, + "loss": 4.2474, + "step": 1604 + }, + { + "epoch": 0.01605, + "grad_norm": 0.7017102241516113, + "learning_rate": 0.003, + "loss": 4.2323, + "step": 1605 + }, + { + "epoch": 0.01606, + "grad_norm": 0.9051344394683838, + "learning_rate": 0.003, + "loss": 4.2411, + "step": 1606 + }, + { + "epoch": 0.01607, + "grad_norm": 0.8984688520431519, + "learning_rate": 0.003, + "loss": 4.2807, + "step": 1607 + }, + { + "epoch": 0.01608, + "grad_norm": 0.7925190329551697, + "learning_rate": 0.003, + "loss": 4.2482, + "step": 1608 + }, + { + "epoch": 0.01609, + "grad_norm": 0.8031415939331055, + "learning_rate": 0.003, + "loss": 4.2542, + "step": 1609 + }, + { + "epoch": 0.0161, + "grad_norm": 0.8899487257003784, + "learning_rate": 0.003, + "loss": 4.2282, + "step": 1610 + }, + { + "epoch": 0.01611, + "grad_norm": 0.7361058592796326, + "learning_rate": 0.003, + "loss": 4.2408, + "step": 1611 + }, + { + "epoch": 0.01612, + "grad_norm": 0.6385438442230225, + "learning_rate": 0.003, + "loss": 4.2281, + "step": 1612 + }, + { + "epoch": 0.01613, + "grad_norm": 0.5635811686515808, + "learning_rate": 0.003, + "loss": 4.2693, + "step": 1613 + }, + { + "epoch": 0.01614, + "grad_norm": 0.49793457984924316, + "learning_rate": 0.003, + "loss": 4.2898, + "step": 1614 + }, + { + "epoch": 0.01615, + "grad_norm": 0.5850211381912231, + "learning_rate": 0.003, + "loss": 4.2593, + "step": 1615 + }, + { + "epoch": 0.01616, + "grad_norm": 0.557797908782959, + "learning_rate": 0.003, + "loss": 4.2378, + "step": 1616 + }, + { + "epoch": 0.01617, + "grad_norm": 0.5186588764190674, + "learning_rate": 0.003, + "loss": 4.226, + "step": 1617 + }, + { + "epoch": 0.01618, + "grad_norm": 0.5103740692138672, + "learning_rate": 0.003, + "loss": 4.2319, + "step": 1618 + }, + { + "epoch": 0.01619, + "grad_norm": 0.4262300729751587, + "learning_rate": 0.003, + "loss": 4.2167, + "step": 1619 + }, + { + "epoch": 0.0162, + "grad_norm": 0.43414008617401123, + "learning_rate": 0.003, + "loss": 4.2243, + "step": 1620 + }, + { + "epoch": 0.01621, + "grad_norm": 0.47554972767829895, + "learning_rate": 0.003, + "loss": 4.2198, + "step": 1621 + }, + { + "epoch": 0.01622, + "grad_norm": 0.49989646673202515, + "learning_rate": 0.003, + "loss": 4.2531, + "step": 1622 + }, + { + "epoch": 0.01623, + "grad_norm": 0.514892578125, + "learning_rate": 0.003, + "loss": 4.2306, + "step": 1623 + }, + { + "epoch": 0.01624, + "grad_norm": 0.606673538684845, + "learning_rate": 0.003, + "loss": 4.2463, + "step": 1624 + }, + { + "epoch": 0.01625, + "grad_norm": 0.7891315817832947, + "learning_rate": 0.003, + "loss": 4.2309, + "step": 1625 + }, + { + "epoch": 0.01626, + "grad_norm": 0.9178404808044434, + "learning_rate": 0.003, + "loss": 4.2593, + "step": 1626 + }, + { + "epoch": 0.01627, + "grad_norm": 1.0693062543869019, + "learning_rate": 0.003, + "loss": 4.2387, + "step": 1627 + }, + { + "epoch": 0.01628, + "grad_norm": 0.7644784450531006, + "learning_rate": 0.003, + "loss": 4.2438, + "step": 1628 + }, + { + "epoch": 0.01629, + "grad_norm": 0.7061607241630554, + "learning_rate": 0.003, + "loss": 4.2389, + "step": 1629 + }, + { + "epoch": 0.0163, + "grad_norm": 0.8324339985847473, + "learning_rate": 0.003, + "loss": 4.2364, + "step": 1630 + }, + { + "epoch": 0.01631, + "grad_norm": 0.7869639992713928, + "learning_rate": 0.003, + "loss": 4.2544, + "step": 1631 + }, + { + "epoch": 0.01632, + "grad_norm": 0.7570949792861938, + "learning_rate": 0.003, + "loss": 4.264, + "step": 1632 + }, + { + "epoch": 0.01633, + "grad_norm": 0.8038742542266846, + "learning_rate": 0.003, + "loss": 4.2404, + "step": 1633 + }, + { + "epoch": 0.01634, + "grad_norm": 0.7716854810714722, + "learning_rate": 0.003, + "loss": 4.2566, + "step": 1634 + }, + { + "epoch": 0.01635, + "grad_norm": 0.6350013613700867, + "learning_rate": 0.003, + "loss": 4.2343, + "step": 1635 + }, + { + "epoch": 0.01636, + "grad_norm": 0.5491396188735962, + "learning_rate": 0.003, + "loss": 4.247, + "step": 1636 + }, + { + "epoch": 0.01637, + "grad_norm": 0.5923768281936646, + "learning_rate": 0.003, + "loss": 4.2317, + "step": 1637 + }, + { + "epoch": 0.01638, + "grad_norm": 0.5720195174217224, + "learning_rate": 0.003, + "loss": 4.2361, + "step": 1638 + }, + { + "epoch": 0.01639, + "grad_norm": 0.5404289364814758, + "learning_rate": 0.003, + "loss": 4.2596, + "step": 1639 + }, + { + "epoch": 0.0164, + "grad_norm": 0.5601854920387268, + "learning_rate": 0.003, + "loss": 4.2389, + "step": 1640 + }, + { + "epoch": 0.01641, + "grad_norm": 0.5230163335800171, + "learning_rate": 0.003, + "loss": 4.212, + "step": 1641 + }, + { + "epoch": 0.01642, + "grad_norm": 0.5209521651268005, + "learning_rate": 0.003, + "loss": 4.2547, + "step": 1642 + }, + { + "epoch": 0.01643, + "grad_norm": 0.6072290539741516, + "learning_rate": 0.003, + "loss": 4.2421, + "step": 1643 + }, + { + "epoch": 0.01644, + "grad_norm": 0.6505742073059082, + "learning_rate": 0.003, + "loss": 4.2612, + "step": 1644 + }, + { + "epoch": 0.01645, + "grad_norm": 0.6426966786384583, + "learning_rate": 0.003, + "loss": 4.2068, + "step": 1645 + }, + { + "epoch": 0.01646, + "grad_norm": 0.6992736458778381, + "learning_rate": 0.003, + "loss": 4.2429, + "step": 1646 + }, + { + "epoch": 0.01647, + "grad_norm": 0.6382941603660583, + "learning_rate": 0.003, + "loss": 4.2378, + "step": 1647 + }, + { + "epoch": 0.01648, + "grad_norm": 0.6025996804237366, + "learning_rate": 0.003, + "loss": 4.2157, + "step": 1648 + }, + { + "epoch": 0.01649, + "grad_norm": 0.5261057615280151, + "learning_rate": 0.003, + "loss": 4.2469, + "step": 1649 + }, + { + "epoch": 0.0165, + "grad_norm": 0.6138892769813538, + "learning_rate": 0.003, + "loss": 4.2268, + "step": 1650 + }, + { + "epoch": 0.01651, + "grad_norm": 0.5947723984718323, + "learning_rate": 0.003, + "loss": 4.2358, + "step": 1651 + }, + { + "epoch": 0.01652, + "grad_norm": 0.6178387403488159, + "learning_rate": 0.003, + "loss": 4.1929, + "step": 1652 + }, + { + "epoch": 0.01653, + "grad_norm": 0.6459247469902039, + "learning_rate": 0.003, + "loss": 4.2379, + "step": 1653 + }, + { + "epoch": 0.01654, + "grad_norm": 0.6552532911300659, + "learning_rate": 0.003, + "loss": 4.2241, + "step": 1654 + }, + { + "epoch": 0.01655, + "grad_norm": 0.780878484249115, + "learning_rate": 0.003, + "loss": 4.2406, + "step": 1655 + }, + { + "epoch": 0.01656, + "grad_norm": 0.9652771949768066, + "learning_rate": 0.003, + "loss": 4.2288, + "step": 1656 + }, + { + "epoch": 0.01657, + "grad_norm": 1.1051970720291138, + "learning_rate": 0.003, + "loss": 4.2586, + "step": 1657 + }, + { + "epoch": 0.01658, + "grad_norm": 0.7399364709854126, + "learning_rate": 0.003, + "loss": 4.2318, + "step": 1658 + }, + { + "epoch": 0.01659, + "grad_norm": 0.6876490116119385, + "learning_rate": 0.003, + "loss": 4.2377, + "step": 1659 + }, + { + "epoch": 0.0166, + "grad_norm": 0.8514830470085144, + "learning_rate": 0.003, + "loss": 4.2247, + "step": 1660 + }, + { + "epoch": 0.01661, + "grad_norm": 0.7782738208770752, + "learning_rate": 0.003, + "loss": 4.2268, + "step": 1661 + }, + { + "epoch": 0.01662, + "grad_norm": 0.6898196339607239, + "learning_rate": 0.003, + "loss": 4.2305, + "step": 1662 + }, + { + "epoch": 0.01663, + "grad_norm": 0.6155083179473877, + "learning_rate": 0.003, + "loss": 4.2454, + "step": 1663 + }, + { + "epoch": 0.01664, + "grad_norm": 0.5973721742630005, + "learning_rate": 0.003, + "loss": 4.2626, + "step": 1664 + }, + { + "epoch": 0.01665, + "grad_norm": 0.6198952794075012, + "learning_rate": 0.003, + "loss": 4.2273, + "step": 1665 + }, + { + "epoch": 0.01666, + "grad_norm": 0.5744661092758179, + "learning_rate": 0.003, + "loss": 4.2297, + "step": 1666 + }, + { + "epoch": 0.01667, + "grad_norm": 0.5517554879188538, + "learning_rate": 0.003, + "loss": 4.241, + "step": 1667 + }, + { + "epoch": 0.01668, + "grad_norm": 0.6101126670837402, + "learning_rate": 0.003, + "loss": 4.2451, + "step": 1668 + }, + { + "epoch": 0.01669, + "grad_norm": 0.6228256821632385, + "learning_rate": 0.003, + "loss": 4.2315, + "step": 1669 + }, + { + "epoch": 0.0167, + "grad_norm": 0.5298703908920288, + "learning_rate": 0.003, + "loss": 4.2179, + "step": 1670 + }, + { + "epoch": 0.01671, + "grad_norm": 0.5599346160888672, + "learning_rate": 0.003, + "loss": 4.2066, + "step": 1671 + }, + { + "epoch": 0.01672, + "grad_norm": 0.6326868534088135, + "learning_rate": 0.003, + "loss": 4.2204, + "step": 1672 + }, + { + "epoch": 0.01673, + "grad_norm": 0.566566526889801, + "learning_rate": 0.003, + "loss": 4.2135, + "step": 1673 + }, + { + "epoch": 0.01674, + "grad_norm": 0.5534862875938416, + "learning_rate": 0.003, + "loss": 4.2339, + "step": 1674 + }, + { + "epoch": 0.01675, + "grad_norm": 0.6360026597976685, + "learning_rate": 0.003, + "loss": 4.2196, + "step": 1675 + }, + { + "epoch": 0.01676, + "grad_norm": 0.754280686378479, + "learning_rate": 0.003, + "loss": 4.2342, + "step": 1676 + }, + { + "epoch": 0.01677, + "grad_norm": 0.9305175542831421, + "learning_rate": 0.003, + "loss": 4.2304, + "step": 1677 + }, + { + "epoch": 0.01678, + "grad_norm": 0.9641466736793518, + "learning_rate": 0.003, + "loss": 4.2519, + "step": 1678 + }, + { + "epoch": 0.01679, + "grad_norm": 0.9234570264816284, + "learning_rate": 0.003, + "loss": 4.2348, + "step": 1679 + }, + { + "epoch": 0.0168, + "grad_norm": 0.8640843629837036, + "learning_rate": 0.003, + "loss": 4.2518, + "step": 1680 + }, + { + "epoch": 0.01681, + "grad_norm": 0.802136242389679, + "learning_rate": 0.003, + "loss": 4.2519, + "step": 1681 + }, + { + "epoch": 0.01682, + "grad_norm": 0.8513818383216858, + "learning_rate": 0.003, + "loss": 4.2466, + "step": 1682 + }, + { + "epoch": 0.01683, + "grad_norm": 0.8904326558113098, + "learning_rate": 0.003, + "loss": 4.265, + "step": 1683 + }, + { + "epoch": 0.01684, + "grad_norm": 0.8687060475349426, + "learning_rate": 0.003, + "loss": 4.2429, + "step": 1684 + }, + { + "epoch": 0.01685, + "grad_norm": 0.8002411127090454, + "learning_rate": 0.003, + "loss": 4.2528, + "step": 1685 + }, + { + "epoch": 0.01686, + "grad_norm": 0.8148955702781677, + "learning_rate": 0.003, + "loss": 4.2527, + "step": 1686 + }, + { + "epoch": 0.01687, + "grad_norm": 0.8790806531906128, + "learning_rate": 0.003, + "loss": 4.251, + "step": 1687 + }, + { + "epoch": 0.01688, + "grad_norm": 0.9963980317115784, + "learning_rate": 0.003, + "loss": 4.2442, + "step": 1688 + }, + { + "epoch": 0.01689, + "grad_norm": 0.9348352551460266, + "learning_rate": 0.003, + "loss": 4.2605, + "step": 1689 + }, + { + "epoch": 0.0169, + "grad_norm": 0.9107605814933777, + "learning_rate": 0.003, + "loss": 4.2483, + "step": 1690 + }, + { + "epoch": 0.01691, + "grad_norm": 0.8572477698326111, + "learning_rate": 0.003, + "loss": 4.2603, + "step": 1691 + }, + { + "epoch": 0.01692, + "grad_norm": 0.8438946008682251, + "learning_rate": 0.003, + "loss": 4.2653, + "step": 1692 + }, + { + "epoch": 0.01693, + "grad_norm": 0.880574107170105, + "learning_rate": 0.003, + "loss": 4.2867, + "step": 1693 + }, + { + "epoch": 0.01694, + "grad_norm": 0.8747463822364807, + "learning_rate": 0.003, + "loss": 4.2402, + "step": 1694 + }, + { + "epoch": 0.01695, + "grad_norm": 0.9774526953697205, + "learning_rate": 0.003, + "loss": 4.2739, + "step": 1695 + }, + { + "epoch": 0.01696, + "grad_norm": 1.0337826013565063, + "learning_rate": 0.003, + "loss": 4.2798, + "step": 1696 + }, + { + "epoch": 0.01697, + "grad_norm": 0.852648138999939, + "learning_rate": 0.003, + "loss": 4.2511, + "step": 1697 + }, + { + "epoch": 0.01698, + "grad_norm": 0.7840266227722168, + "learning_rate": 0.003, + "loss": 4.2388, + "step": 1698 + }, + { + "epoch": 0.01699, + "grad_norm": 0.7520343065261841, + "learning_rate": 0.003, + "loss": 4.2503, + "step": 1699 + }, + { + "epoch": 0.017, + "grad_norm": 0.686873197555542, + "learning_rate": 0.003, + "loss": 4.2629, + "step": 1700 + }, + { + "epoch": 0.01701, + "grad_norm": 0.6256718039512634, + "learning_rate": 0.003, + "loss": 4.2424, + "step": 1701 + }, + { + "epoch": 0.01702, + "grad_norm": 0.5732917785644531, + "learning_rate": 0.003, + "loss": 4.2618, + "step": 1702 + }, + { + "epoch": 0.01703, + "grad_norm": 0.5769745707511902, + "learning_rate": 0.003, + "loss": 4.2478, + "step": 1703 + }, + { + "epoch": 0.01704, + "grad_norm": 0.5204162001609802, + "learning_rate": 0.003, + "loss": 4.23, + "step": 1704 + }, + { + "epoch": 0.01705, + "grad_norm": 0.4713651239871979, + "learning_rate": 0.003, + "loss": 4.2307, + "step": 1705 + }, + { + "epoch": 0.01706, + "grad_norm": 0.3806135058403015, + "learning_rate": 0.003, + "loss": 4.2216, + "step": 1706 + }, + { + "epoch": 0.01707, + "grad_norm": 0.4176587164402008, + "learning_rate": 0.003, + "loss": 4.2444, + "step": 1707 + }, + { + "epoch": 0.01708, + "grad_norm": 0.36861827969551086, + "learning_rate": 0.003, + "loss": 4.2182, + "step": 1708 + }, + { + "epoch": 0.01709, + "grad_norm": 0.35236138105392456, + "learning_rate": 0.003, + "loss": 4.239, + "step": 1709 + }, + { + "epoch": 0.0171, + "grad_norm": 0.311535120010376, + "learning_rate": 0.003, + "loss": 4.1837, + "step": 1710 + }, + { + "epoch": 0.01711, + "grad_norm": 0.3165230453014374, + "learning_rate": 0.003, + "loss": 4.2135, + "step": 1711 + }, + { + "epoch": 0.01712, + "grad_norm": 0.3468313217163086, + "learning_rate": 0.003, + "loss": 4.2236, + "step": 1712 + }, + { + "epoch": 0.01713, + "grad_norm": 0.36829516291618347, + "learning_rate": 0.003, + "loss": 4.1963, + "step": 1713 + }, + { + "epoch": 0.01714, + "grad_norm": 0.37751680612564087, + "learning_rate": 0.003, + "loss": 4.2329, + "step": 1714 + }, + { + "epoch": 0.01715, + "grad_norm": 0.40578410029411316, + "learning_rate": 0.003, + "loss": 4.2399, + "step": 1715 + }, + { + "epoch": 0.01716, + "grad_norm": 0.46360400319099426, + "learning_rate": 0.003, + "loss": 4.218, + "step": 1716 + }, + { + "epoch": 0.01717, + "grad_norm": 0.628455638885498, + "learning_rate": 0.003, + "loss": 4.2545, + "step": 1717 + }, + { + "epoch": 0.01718, + "grad_norm": 0.8051602244377136, + "learning_rate": 0.003, + "loss": 4.2358, + "step": 1718 + }, + { + "epoch": 0.01719, + "grad_norm": 0.8738026022911072, + "learning_rate": 0.003, + "loss": 4.2571, + "step": 1719 + }, + { + "epoch": 0.0172, + "grad_norm": 0.7727829217910767, + "learning_rate": 0.003, + "loss": 4.2155, + "step": 1720 + }, + { + "epoch": 0.01721, + "grad_norm": 0.7172210812568665, + "learning_rate": 0.003, + "loss": 4.2297, + "step": 1721 + }, + { + "epoch": 0.01722, + "grad_norm": 0.7522033452987671, + "learning_rate": 0.003, + "loss": 4.2262, + "step": 1722 + }, + { + "epoch": 0.01723, + "grad_norm": 0.6683304905891418, + "learning_rate": 0.003, + "loss": 4.2285, + "step": 1723 + }, + { + "epoch": 0.01724, + "grad_norm": 0.5735771656036377, + "learning_rate": 0.003, + "loss": 4.2286, + "step": 1724 + }, + { + "epoch": 0.01725, + "grad_norm": 0.6008380651473999, + "learning_rate": 0.003, + "loss": 4.2128, + "step": 1725 + }, + { + "epoch": 0.01726, + "grad_norm": 0.57844078540802, + "learning_rate": 0.003, + "loss": 4.2055, + "step": 1726 + }, + { + "epoch": 0.01727, + "grad_norm": 0.5139830112457275, + "learning_rate": 0.003, + "loss": 4.2082, + "step": 1727 + }, + { + "epoch": 0.01728, + "grad_norm": 0.4524785876274109, + "learning_rate": 0.003, + "loss": 4.2098, + "step": 1728 + }, + { + "epoch": 0.01729, + "grad_norm": 0.40762364864349365, + "learning_rate": 0.003, + "loss": 4.2149, + "step": 1729 + }, + { + "epoch": 0.0173, + "grad_norm": 0.42627114057540894, + "learning_rate": 0.003, + "loss": 4.1889, + "step": 1730 + }, + { + "epoch": 0.01731, + "grad_norm": 0.41734832525253296, + "learning_rate": 0.003, + "loss": 4.208, + "step": 1731 + }, + { + "epoch": 0.01732, + "grad_norm": 0.46209970116615295, + "learning_rate": 0.003, + "loss": 4.2163, + "step": 1732 + }, + { + "epoch": 0.01733, + "grad_norm": 0.5026668906211853, + "learning_rate": 0.003, + "loss": 4.1988, + "step": 1733 + }, + { + "epoch": 0.01734, + "grad_norm": 0.4639485478401184, + "learning_rate": 0.003, + "loss": 4.2076, + "step": 1734 + }, + { + "epoch": 0.01735, + "grad_norm": 0.5392582416534424, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 1735 + }, + { + "epoch": 0.01736, + "grad_norm": 0.6344774961471558, + "learning_rate": 0.003, + "loss": 4.186, + "step": 1736 + }, + { + "epoch": 0.01737, + "grad_norm": 0.6952452063560486, + "learning_rate": 0.003, + "loss": 4.2203, + "step": 1737 + }, + { + "epoch": 0.01738, + "grad_norm": 0.7636409997940063, + "learning_rate": 0.003, + "loss": 4.2103, + "step": 1738 + }, + { + "epoch": 0.01739, + "grad_norm": 0.8100059628486633, + "learning_rate": 0.003, + "loss": 4.2313, + "step": 1739 + }, + { + "epoch": 0.0174, + "grad_norm": 0.7574781775474548, + "learning_rate": 0.003, + "loss": 4.2287, + "step": 1740 + }, + { + "epoch": 0.01741, + "grad_norm": 0.8603562116622925, + "learning_rate": 0.003, + "loss": 4.2408, + "step": 1741 + }, + { + "epoch": 0.01742, + "grad_norm": 1.0610615015029907, + "learning_rate": 0.003, + "loss": 4.2403, + "step": 1742 + }, + { + "epoch": 0.01743, + "grad_norm": 0.9020041823387146, + "learning_rate": 0.003, + "loss": 4.2526, + "step": 1743 + }, + { + "epoch": 0.01744, + "grad_norm": 0.7726603746414185, + "learning_rate": 0.003, + "loss": 4.2302, + "step": 1744 + }, + { + "epoch": 0.01745, + "grad_norm": 0.842292070388794, + "learning_rate": 0.003, + "loss": 4.1998, + "step": 1745 + }, + { + "epoch": 0.01746, + "grad_norm": 0.8620842099189758, + "learning_rate": 0.003, + "loss": 4.2285, + "step": 1746 + }, + { + "epoch": 0.01747, + "grad_norm": 0.8984124660491943, + "learning_rate": 0.003, + "loss": 4.2309, + "step": 1747 + }, + { + "epoch": 0.01748, + "grad_norm": 0.7890809178352356, + "learning_rate": 0.003, + "loss": 4.2184, + "step": 1748 + }, + { + "epoch": 0.01749, + "grad_norm": 0.9051846861839294, + "learning_rate": 0.003, + "loss": 4.2199, + "step": 1749 + }, + { + "epoch": 0.0175, + "grad_norm": 0.8282995820045471, + "learning_rate": 0.003, + "loss": 4.2356, + "step": 1750 + }, + { + "epoch": 0.01751, + "grad_norm": 0.7147263884544373, + "learning_rate": 0.003, + "loss": 4.2187, + "step": 1751 + }, + { + "epoch": 0.01752, + "grad_norm": 0.8123989701271057, + "learning_rate": 0.003, + "loss": 4.269, + "step": 1752 + }, + { + "epoch": 0.01753, + "grad_norm": 0.8011901378631592, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 1753 + }, + { + "epoch": 0.01754, + "grad_norm": 0.6684347987174988, + "learning_rate": 0.003, + "loss": 4.2455, + "step": 1754 + }, + { + "epoch": 0.01755, + "grad_norm": 0.5351077318191528, + "learning_rate": 0.003, + "loss": 4.1816, + "step": 1755 + }, + { + "epoch": 0.01756, + "grad_norm": 0.5229023098945618, + "learning_rate": 0.003, + "loss": 4.2561, + "step": 1756 + }, + { + "epoch": 0.01757, + "grad_norm": 0.6237030029296875, + "learning_rate": 0.003, + "loss": 4.2403, + "step": 1757 + }, + { + "epoch": 0.01758, + "grad_norm": 0.5928184986114502, + "learning_rate": 0.003, + "loss": 4.2396, + "step": 1758 + }, + { + "epoch": 0.01759, + "grad_norm": 0.5645509362220764, + "learning_rate": 0.003, + "loss": 4.2107, + "step": 1759 + }, + { + "epoch": 0.0176, + "grad_norm": 0.6238541603088379, + "learning_rate": 0.003, + "loss": 4.2329, + "step": 1760 + }, + { + "epoch": 0.01761, + "grad_norm": 0.8237802982330322, + "learning_rate": 0.003, + "loss": 4.2266, + "step": 1761 + }, + { + "epoch": 0.01762, + "grad_norm": 1.0283610820770264, + "learning_rate": 0.003, + "loss": 4.2546, + "step": 1762 + }, + { + "epoch": 0.01763, + "grad_norm": 0.9172603487968445, + "learning_rate": 0.003, + "loss": 4.2466, + "step": 1763 + }, + { + "epoch": 0.01764, + "grad_norm": 0.7957273721694946, + "learning_rate": 0.003, + "loss": 4.2112, + "step": 1764 + }, + { + "epoch": 0.01765, + "grad_norm": 0.7934739589691162, + "learning_rate": 0.003, + "loss": 4.2362, + "step": 1765 + }, + { + "epoch": 0.01766, + "grad_norm": 0.8055094480514526, + "learning_rate": 0.003, + "loss": 4.2499, + "step": 1766 + }, + { + "epoch": 0.01767, + "grad_norm": 0.7152431011199951, + "learning_rate": 0.003, + "loss": 4.21, + "step": 1767 + }, + { + "epoch": 0.01768, + "grad_norm": 0.6750852465629578, + "learning_rate": 0.003, + "loss": 4.2147, + "step": 1768 + }, + { + "epoch": 0.01769, + "grad_norm": 0.62043297290802, + "learning_rate": 0.003, + "loss": 4.2126, + "step": 1769 + }, + { + "epoch": 0.0177, + "grad_norm": 0.7558618783950806, + "learning_rate": 0.003, + "loss": 4.2018, + "step": 1770 + }, + { + "epoch": 0.01771, + "grad_norm": 0.7306205034255981, + "learning_rate": 0.003, + "loss": 4.235, + "step": 1771 + }, + { + "epoch": 0.01772, + "grad_norm": 0.5451275706291199, + "learning_rate": 0.003, + "loss": 4.1989, + "step": 1772 + }, + { + "epoch": 0.01773, + "grad_norm": 0.5318989157676697, + "learning_rate": 0.003, + "loss": 4.1925, + "step": 1773 + }, + { + "epoch": 0.01774, + "grad_norm": 0.5252489447593689, + "learning_rate": 0.003, + "loss": 4.2149, + "step": 1774 + }, + { + "epoch": 0.01775, + "grad_norm": 0.4652414321899414, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 1775 + }, + { + "epoch": 0.01776, + "grad_norm": 0.4579496681690216, + "learning_rate": 0.003, + "loss": 4.1905, + "step": 1776 + }, + { + "epoch": 0.01777, + "grad_norm": 0.46492844820022583, + "learning_rate": 0.003, + "loss": 4.2099, + "step": 1777 + }, + { + "epoch": 0.01778, + "grad_norm": 0.46276724338531494, + "learning_rate": 0.003, + "loss": 4.2021, + "step": 1778 + }, + { + "epoch": 0.01779, + "grad_norm": 0.6184820532798767, + "learning_rate": 0.003, + "loss": 4.2227, + "step": 1779 + }, + { + "epoch": 0.0178, + "grad_norm": 0.7079681158065796, + "learning_rate": 0.003, + "loss": 4.1963, + "step": 1780 + }, + { + "epoch": 0.01781, + "grad_norm": 0.6819452047348022, + "learning_rate": 0.003, + "loss": 4.2105, + "step": 1781 + }, + { + "epoch": 0.01782, + "grad_norm": 0.6064656376838684, + "learning_rate": 0.003, + "loss": 4.2325, + "step": 1782 + }, + { + "epoch": 0.01783, + "grad_norm": 0.5331481695175171, + "learning_rate": 0.003, + "loss": 4.2083, + "step": 1783 + }, + { + "epoch": 0.01784, + "grad_norm": 0.5265206098556519, + "learning_rate": 0.003, + "loss": 4.2093, + "step": 1784 + }, + { + "epoch": 0.01785, + "grad_norm": 0.5389939546585083, + "learning_rate": 0.003, + "loss": 4.1711, + "step": 1785 + }, + { + "epoch": 0.01786, + "grad_norm": 0.7006595134735107, + "learning_rate": 0.003, + "loss": 4.2034, + "step": 1786 + }, + { + "epoch": 0.01787, + "grad_norm": 0.7858084440231323, + "learning_rate": 0.003, + "loss": 4.2233, + "step": 1787 + }, + { + "epoch": 0.01788, + "grad_norm": 0.7923018336296082, + "learning_rate": 0.003, + "loss": 4.228, + "step": 1788 + }, + { + "epoch": 0.01789, + "grad_norm": 0.593018651008606, + "learning_rate": 0.003, + "loss": 4.1934, + "step": 1789 + }, + { + "epoch": 0.0179, + "grad_norm": 0.6494895815849304, + "learning_rate": 0.003, + "loss": 4.2138, + "step": 1790 + }, + { + "epoch": 0.01791, + "grad_norm": 0.7846472859382629, + "learning_rate": 0.003, + "loss": 4.2301, + "step": 1791 + }, + { + "epoch": 0.01792, + "grad_norm": 0.8576614260673523, + "learning_rate": 0.003, + "loss": 4.239, + "step": 1792 + }, + { + "epoch": 0.01793, + "grad_norm": 0.8833402991294861, + "learning_rate": 0.003, + "loss": 4.2068, + "step": 1793 + }, + { + "epoch": 0.01794, + "grad_norm": 1.0114383697509766, + "learning_rate": 0.003, + "loss": 4.2273, + "step": 1794 + }, + { + "epoch": 0.01795, + "grad_norm": 0.879726231098175, + "learning_rate": 0.003, + "loss": 4.2331, + "step": 1795 + }, + { + "epoch": 0.01796, + "grad_norm": 0.6577330827713013, + "learning_rate": 0.003, + "loss": 4.2394, + "step": 1796 + }, + { + "epoch": 0.01797, + "grad_norm": 0.7508055567741394, + "learning_rate": 0.003, + "loss": 4.2031, + "step": 1797 + }, + { + "epoch": 0.01798, + "grad_norm": 0.7287778854370117, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 1798 + }, + { + "epoch": 0.01799, + "grad_norm": 0.7758158445358276, + "learning_rate": 0.003, + "loss": 4.2381, + "step": 1799 + }, + { + "epoch": 0.018, + "grad_norm": 0.652176558971405, + "learning_rate": 0.003, + "loss": 4.2204, + "step": 1800 + }, + { + "epoch": 0.01801, + "grad_norm": 0.6343629360198975, + "learning_rate": 0.003, + "loss": 4.2139, + "step": 1801 + }, + { + "epoch": 0.01802, + "grad_norm": 0.6776105165481567, + "learning_rate": 0.003, + "loss": 4.2317, + "step": 1802 + }, + { + "epoch": 0.01803, + "grad_norm": 0.7177024483680725, + "learning_rate": 0.003, + "loss": 4.2423, + "step": 1803 + }, + { + "epoch": 0.01804, + "grad_norm": 0.8185440301895142, + "learning_rate": 0.003, + "loss": 4.2159, + "step": 1804 + }, + { + "epoch": 0.01805, + "grad_norm": 0.8301059603691101, + "learning_rate": 0.003, + "loss": 4.2394, + "step": 1805 + }, + { + "epoch": 0.01806, + "grad_norm": 0.7407498359680176, + "learning_rate": 0.003, + "loss": 4.191, + "step": 1806 + }, + { + "epoch": 0.01807, + "grad_norm": 0.7113837003707886, + "learning_rate": 0.003, + "loss": 4.2277, + "step": 1807 + }, + { + "epoch": 0.01808, + "grad_norm": 0.5918108224868774, + "learning_rate": 0.003, + "loss": 4.2149, + "step": 1808 + }, + { + "epoch": 0.01809, + "grad_norm": 0.5566163659095764, + "learning_rate": 0.003, + "loss": 4.2318, + "step": 1809 + }, + { + "epoch": 0.0181, + "grad_norm": 0.4694720506668091, + "learning_rate": 0.003, + "loss": 4.224, + "step": 1810 + }, + { + "epoch": 0.01811, + "grad_norm": 0.45558327436447144, + "learning_rate": 0.003, + "loss": 4.2076, + "step": 1811 + }, + { + "epoch": 0.01812, + "grad_norm": 0.4792705476284027, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 1812 + }, + { + "epoch": 0.01813, + "grad_norm": 0.43152645230293274, + "learning_rate": 0.003, + "loss": 4.2329, + "step": 1813 + }, + { + "epoch": 0.01814, + "grad_norm": 0.4722409248352051, + "learning_rate": 0.003, + "loss": 4.2001, + "step": 1814 + }, + { + "epoch": 0.01815, + "grad_norm": 0.516764760017395, + "learning_rate": 0.003, + "loss": 4.1882, + "step": 1815 + }, + { + "epoch": 0.01816, + "grad_norm": 0.6370428800582886, + "learning_rate": 0.003, + "loss": 4.2183, + "step": 1816 + }, + { + "epoch": 0.01817, + "grad_norm": 0.7442272305488586, + "learning_rate": 0.003, + "loss": 4.2141, + "step": 1817 + }, + { + "epoch": 0.01818, + "grad_norm": 0.7647615671157837, + "learning_rate": 0.003, + "loss": 4.2145, + "step": 1818 + }, + { + "epoch": 0.01819, + "grad_norm": 0.683918833732605, + "learning_rate": 0.003, + "loss": 4.1944, + "step": 1819 + }, + { + "epoch": 0.0182, + "grad_norm": 0.6988905072212219, + "learning_rate": 0.003, + "loss": 4.2155, + "step": 1820 + }, + { + "epoch": 0.01821, + "grad_norm": 0.632123589515686, + "learning_rate": 0.003, + "loss": 4.1997, + "step": 1821 + }, + { + "epoch": 0.01822, + "grad_norm": 0.5962942838668823, + "learning_rate": 0.003, + "loss": 4.2102, + "step": 1822 + }, + { + "epoch": 0.01823, + "grad_norm": 0.5034074783325195, + "learning_rate": 0.003, + "loss": 4.2224, + "step": 1823 + }, + { + "epoch": 0.01824, + "grad_norm": 0.4946107268333435, + "learning_rate": 0.003, + "loss": 4.2025, + "step": 1824 + }, + { + "epoch": 0.01825, + "grad_norm": 0.5331458449363708, + "learning_rate": 0.003, + "loss": 4.2091, + "step": 1825 + }, + { + "epoch": 0.01826, + "grad_norm": 0.4815158247947693, + "learning_rate": 0.003, + "loss": 4.2164, + "step": 1826 + }, + { + "epoch": 0.01827, + "grad_norm": 0.5388825535774231, + "learning_rate": 0.003, + "loss": 4.1869, + "step": 1827 + }, + { + "epoch": 0.01828, + "grad_norm": 0.6578684449195862, + "learning_rate": 0.003, + "loss": 4.2037, + "step": 1828 + }, + { + "epoch": 0.01829, + "grad_norm": 0.7422965168952942, + "learning_rate": 0.003, + "loss": 4.2232, + "step": 1829 + }, + { + "epoch": 0.0183, + "grad_norm": 0.7160323262214661, + "learning_rate": 0.003, + "loss": 4.2196, + "step": 1830 + }, + { + "epoch": 0.01831, + "grad_norm": 0.7727608680725098, + "learning_rate": 0.003, + "loss": 4.2097, + "step": 1831 + }, + { + "epoch": 0.01832, + "grad_norm": 0.9163352847099304, + "learning_rate": 0.003, + "loss": 4.2196, + "step": 1832 + }, + { + "epoch": 0.01833, + "grad_norm": 0.7811264395713806, + "learning_rate": 0.003, + "loss": 4.2469, + "step": 1833 + }, + { + "epoch": 0.01834, + "grad_norm": 0.7502668499946594, + "learning_rate": 0.003, + "loss": 4.214, + "step": 1834 + }, + { + "epoch": 0.01835, + "grad_norm": 0.7814200520515442, + "learning_rate": 0.003, + "loss": 4.2163, + "step": 1835 + }, + { + "epoch": 0.01836, + "grad_norm": 0.7723317742347717, + "learning_rate": 0.003, + "loss": 4.2127, + "step": 1836 + }, + { + "epoch": 0.01837, + "grad_norm": 0.7659316062927246, + "learning_rate": 0.003, + "loss": 4.2181, + "step": 1837 + }, + { + "epoch": 0.01838, + "grad_norm": 0.789987325668335, + "learning_rate": 0.003, + "loss": 4.2248, + "step": 1838 + }, + { + "epoch": 0.01839, + "grad_norm": 0.8428659439086914, + "learning_rate": 0.003, + "loss": 4.2435, + "step": 1839 + }, + { + "epoch": 0.0184, + "grad_norm": 0.9034631252288818, + "learning_rate": 0.003, + "loss": 4.2468, + "step": 1840 + }, + { + "epoch": 0.01841, + "grad_norm": 0.8657104969024658, + "learning_rate": 0.003, + "loss": 4.1845, + "step": 1841 + }, + { + "epoch": 0.01842, + "grad_norm": 0.8507790565490723, + "learning_rate": 0.003, + "loss": 4.2207, + "step": 1842 + }, + { + "epoch": 0.01843, + "grad_norm": 0.942220151424408, + "learning_rate": 0.003, + "loss": 4.2062, + "step": 1843 + }, + { + "epoch": 0.01844, + "grad_norm": 1.0908007621765137, + "learning_rate": 0.003, + "loss": 4.2172, + "step": 1844 + }, + { + "epoch": 0.01845, + "grad_norm": 1.0858932733535767, + "learning_rate": 0.003, + "loss": 4.2239, + "step": 1845 + }, + { + "epoch": 0.01846, + "grad_norm": 0.8852041959762573, + "learning_rate": 0.003, + "loss": 4.2518, + "step": 1846 + }, + { + "epoch": 0.01847, + "grad_norm": 0.818362295627594, + "learning_rate": 0.003, + "loss": 4.243, + "step": 1847 + }, + { + "epoch": 0.01848, + "grad_norm": 0.8533189296722412, + "learning_rate": 0.003, + "loss": 4.2588, + "step": 1848 + }, + { + "epoch": 0.01849, + "grad_norm": 0.9039400219917297, + "learning_rate": 0.003, + "loss": 4.2557, + "step": 1849 + }, + { + "epoch": 0.0185, + "grad_norm": 0.769622802734375, + "learning_rate": 0.003, + "loss": 4.2424, + "step": 1850 + }, + { + "epoch": 0.01851, + "grad_norm": 0.6524956226348877, + "learning_rate": 0.003, + "loss": 4.2169, + "step": 1851 + }, + { + "epoch": 0.01852, + "grad_norm": 0.6401670575141907, + "learning_rate": 0.003, + "loss": 4.2199, + "step": 1852 + }, + { + "epoch": 0.01853, + "grad_norm": 0.593191921710968, + "learning_rate": 0.003, + "loss": 4.2181, + "step": 1853 + }, + { + "epoch": 0.01854, + "grad_norm": 0.47404181957244873, + "learning_rate": 0.003, + "loss": 4.2422, + "step": 1854 + }, + { + "epoch": 0.01855, + "grad_norm": 0.3873991072177887, + "learning_rate": 0.003, + "loss": 4.2245, + "step": 1855 + }, + { + "epoch": 0.01856, + "grad_norm": 0.3870847821235657, + "learning_rate": 0.003, + "loss": 4.2074, + "step": 1856 + }, + { + "epoch": 0.01857, + "grad_norm": 0.41644003987312317, + "learning_rate": 0.003, + "loss": 4.19, + "step": 1857 + }, + { + "epoch": 0.01858, + "grad_norm": 0.380154550075531, + "learning_rate": 0.003, + "loss": 4.196, + "step": 1858 + }, + { + "epoch": 0.01859, + "grad_norm": 0.39841675758361816, + "learning_rate": 0.003, + "loss": 4.1921, + "step": 1859 + }, + { + "epoch": 0.0186, + "grad_norm": 0.3521161377429962, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 1860 + }, + { + "epoch": 0.01861, + "grad_norm": 0.41872426867485046, + "learning_rate": 0.003, + "loss": 4.2174, + "step": 1861 + }, + { + "epoch": 0.01862, + "grad_norm": 0.47171056270599365, + "learning_rate": 0.003, + "loss": 4.1831, + "step": 1862 + }, + { + "epoch": 0.01863, + "grad_norm": 0.5940234661102295, + "learning_rate": 0.003, + "loss": 4.1933, + "step": 1863 + }, + { + "epoch": 0.01864, + "grad_norm": 0.659092366695404, + "learning_rate": 0.003, + "loss": 4.2305, + "step": 1864 + }, + { + "epoch": 0.01865, + "grad_norm": 0.6307891607284546, + "learning_rate": 0.003, + "loss": 4.2044, + "step": 1865 + }, + { + "epoch": 0.01866, + "grad_norm": 0.46219319105148315, + "learning_rate": 0.003, + "loss": 4.1993, + "step": 1866 + }, + { + "epoch": 0.01867, + "grad_norm": 0.5417385697364807, + "learning_rate": 0.003, + "loss": 4.1771, + "step": 1867 + }, + { + "epoch": 0.01868, + "grad_norm": 0.6495358943939209, + "learning_rate": 0.003, + "loss": 4.197, + "step": 1868 + }, + { + "epoch": 0.01869, + "grad_norm": 0.6564030647277832, + "learning_rate": 0.003, + "loss": 4.2127, + "step": 1869 + }, + { + "epoch": 0.0187, + "grad_norm": 0.6381605267524719, + "learning_rate": 0.003, + "loss": 4.2029, + "step": 1870 + }, + { + "epoch": 0.01871, + "grad_norm": 0.6345733404159546, + "learning_rate": 0.003, + "loss": 4.1787, + "step": 1871 + }, + { + "epoch": 0.01872, + "grad_norm": 0.6442919969558716, + "learning_rate": 0.003, + "loss": 4.2256, + "step": 1872 + }, + { + "epoch": 0.01873, + "grad_norm": 0.6332488656044006, + "learning_rate": 0.003, + "loss": 4.1849, + "step": 1873 + }, + { + "epoch": 0.01874, + "grad_norm": 0.6798062324523926, + "learning_rate": 0.003, + "loss": 4.191, + "step": 1874 + }, + { + "epoch": 0.01875, + "grad_norm": 0.6715080738067627, + "learning_rate": 0.003, + "loss": 4.2097, + "step": 1875 + }, + { + "epoch": 0.01876, + "grad_norm": 0.7180529832839966, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 1876 + }, + { + "epoch": 0.01877, + "grad_norm": 0.663026750087738, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 1877 + }, + { + "epoch": 0.01878, + "grad_norm": 0.7030567526817322, + "learning_rate": 0.003, + "loss": 4.2041, + "step": 1878 + }, + { + "epoch": 0.01879, + "grad_norm": 0.5851782560348511, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 1879 + }, + { + "epoch": 0.0188, + "grad_norm": 0.6897128820419312, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 1880 + }, + { + "epoch": 0.01881, + "grad_norm": 0.7664402723312378, + "learning_rate": 0.003, + "loss": 4.1954, + "step": 1881 + }, + { + "epoch": 0.01882, + "grad_norm": 0.9874089360237122, + "learning_rate": 0.003, + "loss": 4.2309, + "step": 1882 + }, + { + "epoch": 0.01883, + "grad_norm": 1.0450137853622437, + "learning_rate": 0.003, + "loss": 4.2229, + "step": 1883 + }, + { + "epoch": 0.01884, + "grad_norm": 0.8142443895339966, + "learning_rate": 0.003, + "loss": 4.2104, + "step": 1884 + }, + { + "epoch": 0.01885, + "grad_norm": 0.6901335120201111, + "learning_rate": 0.003, + "loss": 4.1818, + "step": 1885 + }, + { + "epoch": 0.01886, + "grad_norm": 0.7235998511314392, + "learning_rate": 0.003, + "loss": 4.224, + "step": 1886 + }, + { + "epoch": 0.01887, + "grad_norm": 0.7312121987342834, + "learning_rate": 0.003, + "loss": 4.2148, + "step": 1887 + }, + { + "epoch": 0.01888, + "grad_norm": 0.772678017616272, + "learning_rate": 0.003, + "loss": 4.2048, + "step": 1888 + }, + { + "epoch": 0.01889, + "grad_norm": 0.7756654024124146, + "learning_rate": 0.003, + "loss": 4.2176, + "step": 1889 + }, + { + "epoch": 0.0189, + "grad_norm": 0.7287212610244751, + "learning_rate": 0.003, + "loss": 4.1971, + "step": 1890 + }, + { + "epoch": 0.01891, + "grad_norm": 0.7076724767684937, + "learning_rate": 0.003, + "loss": 4.2462, + "step": 1891 + }, + { + "epoch": 0.01892, + "grad_norm": 0.6671635508537292, + "learning_rate": 0.003, + "loss": 4.215, + "step": 1892 + }, + { + "epoch": 0.01893, + "grad_norm": 0.6953762769699097, + "learning_rate": 0.003, + "loss": 4.201, + "step": 1893 + }, + { + "epoch": 0.01894, + "grad_norm": 0.5940777063369751, + "learning_rate": 0.003, + "loss": 4.2018, + "step": 1894 + }, + { + "epoch": 0.01895, + "grad_norm": 0.5938246250152588, + "learning_rate": 0.003, + "loss": 4.1879, + "step": 1895 + }, + { + "epoch": 0.01896, + "grad_norm": 0.6224337816238403, + "learning_rate": 0.003, + "loss": 4.2109, + "step": 1896 + }, + { + "epoch": 0.01897, + "grad_norm": 0.5562435388565063, + "learning_rate": 0.003, + "loss": 4.2071, + "step": 1897 + }, + { + "epoch": 0.01898, + "grad_norm": 0.5165086984634399, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 1898 + }, + { + "epoch": 0.01899, + "grad_norm": 0.46164852380752563, + "learning_rate": 0.003, + "loss": 4.1881, + "step": 1899 + }, + { + "epoch": 0.019, + "grad_norm": 0.42392870783805847, + "learning_rate": 0.003, + "loss": 4.179, + "step": 1900 + }, + { + "epoch": 0.01901, + "grad_norm": 0.41818806529045105, + "learning_rate": 0.003, + "loss": 4.2264, + "step": 1901 + }, + { + "epoch": 0.01902, + "grad_norm": 0.5019712448120117, + "learning_rate": 0.003, + "loss": 4.2029, + "step": 1902 + }, + { + "epoch": 0.01903, + "grad_norm": 0.6581493020057678, + "learning_rate": 0.003, + "loss": 4.1638, + "step": 1903 + }, + { + "epoch": 0.01904, + "grad_norm": 0.9066566228866577, + "learning_rate": 0.003, + "loss": 4.2291, + "step": 1904 + }, + { + "epoch": 0.01905, + "grad_norm": 0.9273806214332581, + "learning_rate": 0.003, + "loss": 4.2144, + "step": 1905 + }, + { + "epoch": 0.01906, + "grad_norm": 0.6842753291130066, + "learning_rate": 0.003, + "loss": 4.2198, + "step": 1906 + }, + { + "epoch": 0.01907, + "grad_norm": 0.7864841818809509, + "learning_rate": 0.003, + "loss": 4.217, + "step": 1907 + }, + { + "epoch": 0.01908, + "grad_norm": 0.9169145822525024, + "learning_rate": 0.003, + "loss": 4.2438, + "step": 1908 + }, + { + "epoch": 0.01909, + "grad_norm": 0.7539937496185303, + "learning_rate": 0.003, + "loss": 4.1979, + "step": 1909 + }, + { + "epoch": 0.0191, + "grad_norm": 1.0344467163085938, + "learning_rate": 0.003, + "loss": 4.2215, + "step": 1910 + }, + { + "epoch": 0.01911, + "grad_norm": 0.8631539940834045, + "learning_rate": 0.003, + "loss": 4.1974, + "step": 1911 + }, + { + "epoch": 0.01912, + "grad_norm": 0.691514790058136, + "learning_rate": 0.003, + "loss": 4.1992, + "step": 1912 + }, + { + "epoch": 0.01913, + "grad_norm": 0.6569962501525879, + "learning_rate": 0.003, + "loss": 4.2167, + "step": 1913 + }, + { + "epoch": 0.01914, + "grad_norm": 0.7453441619873047, + "learning_rate": 0.003, + "loss": 4.2353, + "step": 1914 + }, + { + "epoch": 0.01915, + "grad_norm": 0.6612529754638672, + "learning_rate": 0.003, + "loss": 4.2154, + "step": 1915 + }, + { + "epoch": 0.01916, + "grad_norm": 0.5441082119941711, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 1916 + }, + { + "epoch": 0.01917, + "grad_norm": 0.5467451214790344, + "learning_rate": 0.003, + "loss": 4.2117, + "step": 1917 + }, + { + "epoch": 0.01918, + "grad_norm": 0.6295916438102722, + "learning_rate": 0.003, + "loss": 4.1987, + "step": 1918 + }, + { + "epoch": 0.01919, + "grad_norm": 0.6276860237121582, + "learning_rate": 0.003, + "loss": 4.2016, + "step": 1919 + }, + { + "epoch": 0.0192, + "grad_norm": 0.6265702247619629, + "learning_rate": 0.003, + "loss": 4.2023, + "step": 1920 + }, + { + "epoch": 0.01921, + "grad_norm": 0.6072244644165039, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 1921 + }, + { + "epoch": 0.01922, + "grad_norm": 0.577560305595398, + "learning_rate": 0.003, + "loss": 4.2062, + "step": 1922 + }, + { + "epoch": 0.01923, + "grad_norm": 0.528586208820343, + "learning_rate": 0.003, + "loss": 4.2098, + "step": 1923 + }, + { + "epoch": 0.01924, + "grad_norm": 0.437764436006546, + "learning_rate": 0.003, + "loss": 4.2065, + "step": 1924 + }, + { + "epoch": 0.01925, + "grad_norm": 0.3857106566429138, + "learning_rate": 0.003, + "loss": 4.1935, + "step": 1925 + }, + { + "epoch": 0.01926, + "grad_norm": 0.4072262942790985, + "learning_rate": 0.003, + "loss": 4.2001, + "step": 1926 + }, + { + "epoch": 0.01927, + "grad_norm": 0.5231996774673462, + "learning_rate": 0.003, + "loss": 4.1922, + "step": 1927 + }, + { + "epoch": 0.01928, + "grad_norm": 0.6032196283340454, + "learning_rate": 0.003, + "loss": 4.1716, + "step": 1928 + }, + { + "epoch": 0.01929, + "grad_norm": 0.7323647737503052, + "learning_rate": 0.003, + "loss": 4.1937, + "step": 1929 + }, + { + "epoch": 0.0193, + "grad_norm": 0.7953292727470398, + "learning_rate": 0.003, + "loss": 4.1919, + "step": 1930 + }, + { + "epoch": 0.01931, + "grad_norm": 0.6991308927536011, + "learning_rate": 0.003, + "loss": 4.1873, + "step": 1931 + }, + { + "epoch": 0.01932, + "grad_norm": 0.6141433715820312, + "learning_rate": 0.003, + "loss": 4.1945, + "step": 1932 + }, + { + "epoch": 0.01933, + "grad_norm": 0.6607575416564941, + "learning_rate": 0.003, + "loss": 4.1875, + "step": 1933 + }, + { + "epoch": 0.01934, + "grad_norm": 0.665708065032959, + "learning_rate": 0.003, + "loss": 4.1913, + "step": 1934 + }, + { + "epoch": 0.01935, + "grad_norm": 0.6778123378753662, + "learning_rate": 0.003, + "loss": 4.2028, + "step": 1935 + }, + { + "epoch": 0.01936, + "grad_norm": 0.660306453704834, + "learning_rate": 0.003, + "loss": 4.188, + "step": 1936 + }, + { + "epoch": 0.01937, + "grad_norm": 0.6556859612464905, + "learning_rate": 0.003, + "loss": 4.201, + "step": 1937 + }, + { + "epoch": 0.01938, + "grad_norm": 0.6403252482414246, + "learning_rate": 0.003, + "loss": 4.1998, + "step": 1938 + }, + { + "epoch": 0.01939, + "grad_norm": 0.6520726084709167, + "learning_rate": 0.003, + "loss": 4.2214, + "step": 1939 + }, + { + "epoch": 0.0194, + "grad_norm": 0.6492891907691956, + "learning_rate": 0.003, + "loss": 4.1979, + "step": 1940 + }, + { + "epoch": 0.01941, + "grad_norm": 0.6015598773956299, + "learning_rate": 0.003, + "loss": 4.18, + "step": 1941 + }, + { + "epoch": 0.01942, + "grad_norm": 0.5949015021324158, + "learning_rate": 0.003, + "loss": 4.2027, + "step": 1942 + }, + { + "epoch": 0.01943, + "grad_norm": 0.5752372145652771, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 1943 + }, + { + "epoch": 0.01944, + "grad_norm": 0.5773752331733704, + "learning_rate": 0.003, + "loss": 4.2158, + "step": 1944 + }, + { + "epoch": 0.01945, + "grad_norm": 0.6772210597991943, + "learning_rate": 0.003, + "loss": 4.1897, + "step": 1945 + }, + { + "epoch": 0.01946, + "grad_norm": 0.817459225654602, + "learning_rate": 0.003, + "loss": 4.1898, + "step": 1946 + }, + { + "epoch": 0.01947, + "grad_norm": 0.9603843092918396, + "learning_rate": 0.003, + "loss": 4.2053, + "step": 1947 + }, + { + "epoch": 0.01948, + "grad_norm": 0.9629886746406555, + "learning_rate": 0.003, + "loss": 4.2304, + "step": 1948 + }, + { + "epoch": 0.01949, + "grad_norm": 1.0253958702087402, + "learning_rate": 0.003, + "loss": 4.2301, + "step": 1949 + }, + { + "epoch": 0.0195, + "grad_norm": 1.0213903188705444, + "learning_rate": 0.003, + "loss": 4.2249, + "step": 1950 + }, + { + "epoch": 0.01951, + "grad_norm": 0.7116432189941406, + "learning_rate": 0.003, + "loss": 4.2054, + "step": 1951 + }, + { + "epoch": 0.01952, + "grad_norm": 0.6784964203834534, + "learning_rate": 0.003, + "loss": 4.2146, + "step": 1952 + }, + { + "epoch": 0.01953, + "grad_norm": 0.755315899848938, + "learning_rate": 0.003, + "loss": 4.2206, + "step": 1953 + }, + { + "epoch": 0.01954, + "grad_norm": 0.8843002319335938, + "learning_rate": 0.003, + "loss": 4.2306, + "step": 1954 + }, + { + "epoch": 0.01955, + "grad_norm": 0.9285549521446228, + "learning_rate": 0.003, + "loss": 4.2258, + "step": 1955 + }, + { + "epoch": 0.01956, + "grad_norm": 0.7131898999214172, + "learning_rate": 0.003, + "loss": 4.2135, + "step": 1956 + }, + { + "epoch": 0.01957, + "grad_norm": 0.7296762466430664, + "learning_rate": 0.003, + "loss": 4.1819, + "step": 1957 + }, + { + "epoch": 0.01958, + "grad_norm": 0.76239013671875, + "learning_rate": 0.003, + "loss": 4.1787, + "step": 1958 + }, + { + "epoch": 0.01959, + "grad_norm": 0.7252810001373291, + "learning_rate": 0.003, + "loss": 4.2032, + "step": 1959 + }, + { + "epoch": 0.0196, + "grad_norm": 0.7949764132499695, + "learning_rate": 0.003, + "loss": 4.1992, + "step": 1960 + }, + { + "epoch": 0.01961, + "grad_norm": 0.7223983407020569, + "learning_rate": 0.003, + "loss": 4.197, + "step": 1961 + }, + { + "epoch": 0.01962, + "grad_norm": 0.683202862739563, + "learning_rate": 0.003, + "loss": 4.1965, + "step": 1962 + }, + { + "epoch": 0.01963, + "grad_norm": 0.5978757739067078, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 1963 + }, + { + "epoch": 0.01964, + "grad_norm": 0.6187034249305725, + "learning_rate": 0.003, + "loss": 4.2107, + "step": 1964 + }, + { + "epoch": 0.01965, + "grad_norm": 0.5564098358154297, + "learning_rate": 0.003, + "loss": 4.1988, + "step": 1965 + }, + { + "epoch": 0.01966, + "grad_norm": 0.6056374311447144, + "learning_rate": 0.003, + "loss": 4.2015, + "step": 1966 + }, + { + "epoch": 0.01967, + "grad_norm": 0.5678548812866211, + "learning_rate": 0.003, + "loss": 4.174, + "step": 1967 + }, + { + "epoch": 0.01968, + "grad_norm": 0.5833330750465393, + "learning_rate": 0.003, + "loss": 4.1938, + "step": 1968 + }, + { + "epoch": 0.01969, + "grad_norm": 0.681247889995575, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 1969 + }, + { + "epoch": 0.0197, + "grad_norm": 0.6020233035087585, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 1970 + }, + { + "epoch": 0.01971, + "grad_norm": 0.586289644241333, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 1971 + }, + { + "epoch": 0.01972, + "grad_norm": 0.6138716340065002, + "learning_rate": 0.003, + "loss": 4.1822, + "step": 1972 + }, + { + "epoch": 0.01973, + "grad_norm": 0.6565843820571899, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 1973 + }, + { + "epoch": 0.01974, + "grad_norm": 0.7141625881195068, + "learning_rate": 0.003, + "loss": 4.2062, + "step": 1974 + }, + { + "epoch": 0.01975, + "grad_norm": 0.9425281882286072, + "learning_rate": 0.003, + "loss": 4.2065, + "step": 1975 + }, + { + "epoch": 0.01976, + "grad_norm": 1.02041757106781, + "learning_rate": 0.003, + "loss": 4.2127, + "step": 1976 + }, + { + "epoch": 0.01977, + "grad_norm": 0.7708873748779297, + "learning_rate": 0.003, + "loss": 4.23, + "step": 1977 + }, + { + "epoch": 0.01978, + "grad_norm": 0.6835176348686218, + "learning_rate": 0.003, + "loss": 4.2285, + "step": 1978 + }, + { + "epoch": 0.01979, + "grad_norm": 0.6171081066131592, + "learning_rate": 0.003, + "loss": 4.2144, + "step": 1979 + }, + { + "epoch": 0.0198, + "grad_norm": 0.5269233584403992, + "learning_rate": 0.003, + "loss": 4.1914, + "step": 1980 + }, + { + "epoch": 0.01981, + "grad_norm": 0.5263850092887878, + "learning_rate": 0.003, + "loss": 4.175, + "step": 1981 + }, + { + "epoch": 0.01982, + "grad_norm": 0.50020831823349, + "learning_rate": 0.003, + "loss": 4.1986, + "step": 1982 + }, + { + "epoch": 0.01983, + "grad_norm": 0.5846083760261536, + "learning_rate": 0.003, + "loss": 4.194, + "step": 1983 + }, + { + "epoch": 0.01984, + "grad_norm": 0.6941750645637512, + "learning_rate": 0.003, + "loss": 4.2073, + "step": 1984 + }, + { + "epoch": 0.01985, + "grad_norm": 0.6278181672096252, + "learning_rate": 0.003, + "loss": 4.1955, + "step": 1985 + }, + { + "epoch": 0.01986, + "grad_norm": 0.5164998769760132, + "learning_rate": 0.003, + "loss": 4.2189, + "step": 1986 + }, + { + "epoch": 0.01987, + "grad_norm": 0.6220480799674988, + "learning_rate": 0.003, + "loss": 4.1777, + "step": 1987 + }, + { + "epoch": 0.01988, + "grad_norm": 0.668171226978302, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 1988 + }, + { + "epoch": 0.01989, + "grad_norm": 0.7721487879753113, + "learning_rate": 0.003, + "loss": 4.1845, + "step": 1989 + }, + { + "epoch": 0.0199, + "grad_norm": 0.9032546281814575, + "learning_rate": 0.003, + "loss": 4.2066, + "step": 1990 + }, + { + "epoch": 0.01991, + "grad_norm": 0.9958043098449707, + "learning_rate": 0.003, + "loss": 4.1939, + "step": 1991 + }, + { + "epoch": 0.01992, + "grad_norm": 0.9264521598815918, + "learning_rate": 0.003, + "loss": 4.2269, + "step": 1992 + }, + { + "epoch": 0.01993, + "grad_norm": 0.8177193999290466, + "learning_rate": 0.003, + "loss": 4.2101, + "step": 1993 + }, + { + "epoch": 0.01994, + "grad_norm": 0.6321484446525574, + "learning_rate": 0.003, + "loss": 4.2233, + "step": 1994 + }, + { + "epoch": 0.01995, + "grad_norm": 0.6378874182701111, + "learning_rate": 0.003, + "loss": 4.216, + "step": 1995 + }, + { + "epoch": 0.01996, + "grad_norm": 0.5592439770698547, + "learning_rate": 0.003, + "loss": 4.1789, + "step": 1996 + }, + { + "epoch": 0.01997, + "grad_norm": 0.5063359141349792, + "learning_rate": 0.003, + "loss": 4.2089, + "step": 1997 + }, + { + "epoch": 0.01998, + "grad_norm": 0.498551607131958, + "learning_rate": 0.003, + "loss": 4.1917, + "step": 1998 + }, + { + "epoch": 0.01999, + "grad_norm": 0.5395573377609253, + "learning_rate": 0.003, + "loss": 4.1836, + "step": 1999 + }, + { + "epoch": 0.02, + "grad_norm": 0.5799160003662109, + "learning_rate": 0.003, + "loss": 4.1718, + "step": 2000 + }, + { + "epoch": 0.02001, + "grad_norm": 0.6742599606513977, + "learning_rate": 0.003, + "loss": 4.2094, + "step": 2001 + }, + { + "epoch": 0.02002, + "grad_norm": 0.616278350353241, + "learning_rate": 0.003, + "loss": 4.2036, + "step": 2002 + }, + { + "epoch": 0.02003, + "grad_norm": 0.3921935260295868, + "learning_rate": 0.003, + "loss": 4.1952, + "step": 2003 + }, + { + "epoch": 0.02004, + "grad_norm": 0.3965532183647156, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 2004 + }, + { + "epoch": 0.02005, + "grad_norm": 0.4591590166091919, + "learning_rate": 0.003, + "loss": 4.1874, + "step": 2005 + }, + { + "epoch": 0.02006, + "grad_norm": 0.43843862414360046, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 2006 + }, + { + "epoch": 0.02007, + "grad_norm": 0.5198654532432556, + "learning_rate": 0.003, + "loss": 4.2, + "step": 2007 + }, + { + "epoch": 0.02008, + "grad_norm": 0.5515919327735901, + "learning_rate": 0.003, + "loss": 4.1918, + "step": 2008 + }, + { + "epoch": 0.02009, + "grad_norm": 0.5351531505584717, + "learning_rate": 0.003, + "loss": 4.1889, + "step": 2009 + }, + { + "epoch": 0.0201, + "grad_norm": 0.5307760834693909, + "learning_rate": 0.003, + "loss": 4.1996, + "step": 2010 + }, + { + "epoch": 0.02011, + "grad_norm": 0.46869173645973206, + "learning_rate": 0.003, + "loss": 4.183, + "step": 2011 + }, + { + "epoch": 0.02012, + "grad_norm": 0.45753300189971924, + "learning_rate": 0.003, + "loss": 4.1822, + "step": 2012 + }, + { + "epoch": 0.02013, + "grad_norm": 0.5114355087280273, + "learning_rate": 0.003, + "loss": 4.1902, + "step": 2013 + }, + { + "epoch": 0.02014, + "grad_norm": 0.62235426902771, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 2014 + }, + { + "epoch": 0.02015, + "grad_norm": 0.8762567043304443, + "learning_rate": 0.003, + "loss": 4.1818, + "step": 2015 + }, + { + "epoch": 0.02016, + "grad_norm": 1.1758772134780884, + "learning_rate": 0.003, + "loss": 4.2184, + "step": 2016 + }, + { + "epoch": 0.02017, + "grad_norm": 0.8140804171562195, + "learning_rate": 0.003, + "loss": 4.196, + "step": 2017 + }, + { + "epoch": 0.02018, + "grad_norm": 0.8363609910011292, + "learning_rate": 0.003, + "loss": 4.201, + "step": 2018 + }, + { + "epoch": 0.02019, + "grad_norm": 0.8395445346832275, + "learning_rate": 0.003, + "loss": 4.2077, + "step": 2019 + }, + { + "epoch": 0.0202, + "grad_norm": 0.7397228479385376, + "learning_rate": 0.003, + "loss": 4.2085, + "step": 2020 + }, + { + "epoch": 0.02021, + "grad_norm": 0.8446124792098999, + "learning_rate": 0.003, + "loss": 4.2309, + "step": 2021 + }, + { + "epoch": 0.02022, + "grad_norm": 0.790037989616394, + "learning_rate": 0.003, + "loss": 4.1867, + "step": 2022 + }, + { + "epoch": 0.02023, + "grad_norm": 0.8900567889213562, + "learning_rate": 0.003, + "loss": 4.1644, + "step": 2023 + }, + { + "epoch": 0.02024, + "grad_norm": 0.8839752078056335, + "learning_rate": 0.003, + "loss": 4.1963, + "step": 2024 + }, + { + "epoch": 0.02025, + "grad_norm": 0.9088333249092102, + "learning_rate": 0.003, + "loss": 4.2446, + "step": 2025 + }, + { + "epoch": 0.02026, + "grad_norm": 0.8618097901344299, + "learning_rate": 0.003, + "loss": 4.226, + "step": 2026 + }, + { + "epoch": 0.02027, + "grad_norm": 0.6961638927459717, + "learning_rate": 0.003, + "loss": 4.2132, + "step": 2027 + }, + { + "epoch": 0.02028, + "grad_norm": 0.7510376572608948, + "learning_rate": 0.003, + "loss": 4.2002, + "step": 2028 + }, + { + "epoch": 0.02029, + "grad_norm": 0.6982995271682739, + "learning_rate": 0.003, + "loss": 4.2032, + "step": 2029 + }, + { + "epoch": 0.0203, + "grad_norm": 0.8095003366470337, + "learning_rate": 0.003, + "loss": 4.1988, + "step": 2030 + }, + { + "epoch": 0.02031, + "grad_norm": 0.8760609030723572, + "learning_rate": 0.003, + "loss": 4.2121, + "step": 2031 + }, + { + "epoch": 0.02032, + "grad_norm": 0.8105490207672119, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 2032 + }, + { + "epoch": 0.02033, + "grad_norm": 0.7400681376457214, + "learning_rate": 0.003, + "loss": 4.2221, + "step": 2033 + }, + { + "epoch": 0.02034, + "grad_norm": 0.8544787168502808, + "learning_rate": 0.003, + "loss": 4.2195, + "step": 2034 + }, + { + "epoch": 0.02035, + "grad_norm": 0.8666634559631348, + "learning_rate": 0.003, + "loss": 4.2081, + "step": 2035 + }, + { + "epoch": 0.02036, + "grad_norm": 0.7516500949859619, + "learning_rate": 0.003, + "loss": 4.2028, + "step": 2036 + }, + { + "epoch": 0.02037, + "grad_norm": 0.6464591026306152, + "learning_rate": 0.003, + "loss": 4.2133, + "step": 2037 + }, + { + "epoch": 0.02038, + "grad_norm": 0.5785326957702637, + "learning_rate": 0.003, + "loss": 4.1922, + "step": 2038 + }, + { + "epoch": 0.02039, + "grad_norm": 0.5690309405326843, + "learning_rate": 0.003, + "loss": 4.2006, + "step": 2039 + }, + { + "epoch": 0.0204, + "grad_norm": 0.47884753346443176, + "learning_rate": 0.003, + "loss": 4.2009, + "step": 2040 + }, + { + "epoch": 0.02041, + "grad_norm": 0.43236783146858215, + "learning_rate": 0.003, + "loss": 4.1612, + "step": 2041 + }, + { + "epoch": 0.02042, + "grad_norm": 0.3718264102935791, + "learning_rate": 0.003, + "loss": 4.16, + "step": 2042 + }, + { + "epoch": 0.02043, + "grad_norm": 0.38901960849761963, + "learning_rate": 0.003, + "loss": 4.2028, + "step": 2043 + }, + { + "epoch": 0.02044, + "grad_norm": 0.391589492559433, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 2044 + }, + { + "epoch": 0.02045, + "grad_norm": 0.35344189405441284, + "learning_rate": 0.003, + "loss": 4.1881, + "step": 2045 + }, + { + "epoch": 0.02046, + "grad_norm": 0.36751314997673035, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 2046 + }, + { + "epoch": 0.02047, + "grad_norm": 0.3489588499069214, + "learning_rate": 0.003, + "loss": 4.1999, + "step": 2047 + }, + { + "epoch": 0.02048, + "grad_norm": 0.46102413535118103, + "learning_rate": 0.003, + "loss": 4.1826, + "step": 2048 + }, + { + "epoch": 0.02049, + "grad_norm": 0.5971166491508484, + "learning_rate": 0.003, + "loss": 4.2081, + "step": 2049 + }, + { + "epoch": 0.0205, + "grad_norm": 0.8196846842765808, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 2050 + }, + { + "epoch": 0.02051, + "grad_norm": 1.046412467956543, + "learning_rate": 0.003, + "loss": 4.179, + "step": 2051 + }, + { + "epoch": 0.02052, + "grad_norm": 0.8385469317436218, + "learning_rate": 0.003, + "loss": 4.1822, + "step": 2052 + }, + { + "epoch": 0.02053, + "grad_norm": 0.7001754641532898, + "learning_rate": 0.003, + "loss": 4.1758, + "step": 2053 + }, + { + "epoch": 0.02054, + "grad_norm": 0.7512269020080566, + "learning_rate": 0.003, + "loss": 4.2184, + "step": 2054 + }, + { + "epoch": 0.02055, + "grad_norm": 0.7145476937294006, + "learning_rate": 0.003, + "loss": 4.1785, + "step": 2055 + }, + { + "epoch": 0.02056, + "grad_norm": 0.5185825824737549, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 2056 + }, + { + "epoch": 0.02057, + "grad_norm": 0.7188176512718201, + "learning_rate": 0.003, + "loss": 4.2106, + "step": 2057 + }, + { + "epoch": 0.02058, + "grad_norm": 0.6570911407470703, + "learning_rate": 0.003, + "loss": 4.1847, + "step": 2058 + }, + { + "epoch": 0.02059, + "grad_norm": 0.6945111155509949, + "learning_rate": 0.003, + "loss": 4.1692, + "step": 2059 + }, + { + "epoch": 0.0206, + "grad_norm": 0.6718025207519531, + "learning_rate": 0.003, + "loss": 4.2144, + "step": 2060 + }, + { + "epoch": 0.02061, + "grad_norm": 0.7464396953582764, + "learning_rate": 0.003, + "loss": 4.1868, + "step": 2061 + }, + { + "epoch": 0.02062, + "grad_norm": 0.8595008850097656, + "learning_rate": 0.003, + "loss": 4.1982, + "step": 2062 + }, + { + "epoch": 0.02063, + "grad_norm": 0.7188435196876526, + "learning_rate": 0.003, + "loss": 4.186, + "step": 2063 + }, + { + "epoch": 0.02064, + "grad_norm": 0.7377521991729736, + "learning_rate": 0.003, + "loss": 4.1964, + "step": 2064 + }, + { + "epoch": 0.02065, + "grad_norm": 0.8290489315986633, + "learning_rate": 0.003, + "loss": 4.1817, + "step": 2065 + }, + { + "epoch": 0.02066, + "grad_norm": 0.8223605155944824, + "learning_rate": 0.003, + "loss": 4.1986, + "step": 2066 + }, + { + "epoch": 0.02067, + "grad_norm": 0.6963474154472351, + "learning_rate": 0.003, + "loss": 4.1927, + "step": 2067 + }, + { + "epoch": 0.02068, + "grad_norm": 0.6809067130088806, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 2068 + }, + { + "epoch": 0.02069, + "grad_norm": 0.6437063813209534, + "learning_rate": 0.003, + "loss": 4.1953, + "step": 2069 + }, + { + "epoch": 0.0207, + "grad_norm": 0.6656236052513123, + "learning_rate": 0.003, + "loss": 4.1692, + "step": 2070 + }, + { + "epoch": 0.02071, + "grad_norm": 0.5361942052841187, + "learning_rate": 0.003, + "loss": 4.1939, + "step": 2071 + }, + { + "epoch": 0.02072, + "grad_norm": 0.5347724556922913, + "learning_rate": 0.003, + "loss": 4.199, + "step": 2072 + }, + { + "epoch": 0.02073, + "grad_norm": 0.5662330985069275, + "learning_rate": 0.003, + "loss": 4.1792, + "step": 2073 + }, + { + "epoch": 0.02074, + "grad_norm": 0.7545875310897827, + "learning_rate": 0.003, + "loss": 4.1916, + "step": 2074 + }, + { + "epoch": 0.02075, + "grad_norm": 0.7776836156845093, + "learning_rate": 0.003, + "loss": 4.2211, + "step": 2075 + }, + { + "epoch": 0.02076, + "grad_norm": 0.665153443813324, + "learning_rate": 0.003, + "loss": 4.2109, + "step": 2076 + }, + { + "epoch": 0.02077, + "grad_norm": 0.6429176926612854, + "learning_rate": 0.003, + "loss": 4.1758, + "step": 2077 + }, + { + "epoch": 0.02078, + "grad_norm": 0.6930047273635864, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 2078 + }, + { + "epoch": 0.02079, + "grad_norm": 0.885644793510437, + "learning_rate": 0.003, + "loss": 4.1694, + "step": 2079 + }, + { + "epoch": 0.0208, + "grad_norm": 0.8631641864776611, + "learning_rate": 0.003, + "loss": 4.1873, + "step": 2080 + }, + { + "epoch": 0.02081, + "grad_norm": 0.7464285492897034, + "learning_rate": 0.003, + "loss": 4.2136, + "step": 2081 + }, + { + "epoch": 0.02082, + "grad_norm": 0.6033406257629395, + "learning_rate": 0.003, + "loss": 4.2032, + "step": 2082 + }, + { + "epoch": 0.02083, + "grad_norm": 0.5763504505157471, + "learning_rate": 0.003, + "loss": 4.1974, + "step": 2083 + }, + { + "epoch": 0.02084, + "grad_norm": 0.5885117650032043, + "learning_rate": 0.003, + "loss": 4.1894, + "step": 2084 + }, + { + "epoch": 0.02085, + "grad_norm": 0.6575955152511597, + "learning_rate": 0.003, + "loss": 4.1685, + "step": 2085 + }, + { + "epoch": 0.02086, + "grad_norm": 0.7422707080841064, + "learning_rate": 0.003, + "loss": 4.1768, + "step": 2086 + }, + { + "epoch": 0.02087, + "grad_norm": 0.8073996901512146, + "learning_rate": 0.003, + "loss": 4.183, + "step": 2087 + }, + { + "epoch": 0.02088, + "grad_norm": 0.7600259184837341, + "learning_rate": 0.003, + "loss": 4.2126, + "step": 2088 + }, + { + "epoch": 0.02089, + "grad_norm": 0.7548388242721558, + "learning_rate": 0.003, + "loss": 4.1805, + "step": 2089 + }, + { + "epoch": 0.0209, + "grad_norm": 0.6929354667663574, + "learning_rate": 0.003, + "loss": 4.1946, + "step": 2090 + }, + { + "epoch": 0.02091, + "grad_norm": 0.6573588252067566, + "learning_rate": 0.003, + "loss": 4.1839, + "step": 2091 + }, + { + "epoch": 0.02092, + "grad_norm": 0.5934063792228699, + "learning_rate": 0.003, + "loss": 4.1787, + "step": 2092 + }, + { + "epoch": 0.02093, + "grad_norm": 0.5749050974845886, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 2093 + }, + { + "epoch": 0.02094, + "grad_norm": 0.6522166728973389, + "learning_rate": 0.003, + "loss": 4.1551, + "step": 2094 + }, + { + "epoch": 0.02095, + "grad_norm": 0.6819972991943359, + "learning_rate": 0.003, + "loss": 4.1987, + "step": 2095 + }, + { + "epoch": 0.02096, + "grad_norm": 0.8537513613700867, + "learning_rate": 0.003, + "loss": 4.1954, + "step": 2096 + }, + { + "epoch": 0.02097, + "grad_norm": 0.8873555660247803, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 2097 + }, + { + "epoch": 0.02098, + "grad_norm": 0.8346237540245056, + "learning_rate": 0.003, + "loss": 4.1983, + "step": 2098 + }, + { + "epoch": 0.02099, + "grad_norm": 0.7050203084945679, + "learning_rate": 0.003, + "loss": 4.2048, + "step": 2099 + }, + { + "epoch": 0.021, + "grad_norm": 0.7459020614624023, + "learning_rate": 0.003, + "loss": 4.219, + "step": 2100 + }, + { + "epoch": 0.02101, + "grad_norm": 0.6101992726325989, + "learning_rate": 0.003, + "loss": 4.205, + "step": 2101 + }, + { + "epoch": 0.02102, + "grad_norm": 0.5766392350196838, + "learning_rate": 0.003, + "loss": 4.208, + "step": 2102 + }, + { + "epoch": 0.02103, + "grad_norm": 0.4940197765827179, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 2103 + }, + { + "epoch": 0.02104, + "grad_norm": 0.5101218819618225, + "learning_rate": 0.003, + "loss": 4.1715, + "step": 2104 + }, + { + "epoch": 0.02105, + "grad_norm": 0.4358375072479248, + "learning_rate": 0.003, + "loss": 4.1703, + "step": 2105 + }, + { + "epoch": 0.02106, + "grad_norm": 0.4337053596973419, + "learning_rate": 0.003, + "loss": 4.1631, + "step": 2106 + }, + { + "epoch": 0.02107, + "grad_norm": 0.42299988865852356, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 2107 + }, + { + "epoch": 0.02108, + "grad_norm": 0.408107191324234, + "learning_rate": 0.003, + "loss": 4.1546, + "step": 2108 + }, + { + "epoch": 0.02109, + "grad_norm": 0.36309558153152466, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 2109 + }, + { + "epoch": 0.0211, + "grad_norm": 0.44200608134269714, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 2110 + }, + { + "epoch": 0.02111, + "grad_norm": 0.5945785641670227, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 2111 + }, + { + "epoch": 0.02112, + "grad_norm": 0.8337014317512512, + "learning_rate": 0.003, + "loss": 4.1799, + "step": 2112 + }, + { + "epoch": 0.02113, + "grad_norm": 0.9096565246582031, + "learning_rate": 0.003, + "loss": 4.1882, + "step": 2113 + }, + { + "epoch": 0.02114, + "grad_norm": 0.8218700289726257, + "learning_rate": 0.003, + "loss": 4.1728, + "step": 2114 + }, + { + "epoch": 0.02115, + "grad_norm": 0.7097974419593811, + "learning_rate": 0.003, + "loss": 4.1612, + "step": 2115 + }, + { + "epoch": 0.02116, + "grad_norm": 0.7614986300468445, + "learning_rate": 0.003, + "loss": 4.1951, + "step": 2116 + }, + { + "epoch": 0.02117, + "grad_norm": 0.7237206697463989, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 2117 + }, + { + "epoch": 0.02118, + "grad_norm": 0.6130908727645874, + "learning_rate": 0.003, + "loss": 4.1776, + "step": 2118 + }, + { + "epoch": 0.02119, + "grad_norm": 0.5855314135551453, + "learning_rate": 0.003, + "loss": 4.1889, + "step": 2119 + }, + { + "epoch": 0.0212, + "grad_norm": 0.6109283566474915, + "learning_rate": 0.003, + "loss": 4.1619, + "step": 2120 + }, + { + "epoch": 0.02121, + "grad_norm": 0.610881507396698, + "learning_rate": 0.003, + "loss": 4.158, + "step": 2121 + }, + { + "epoch": 0.02122, + "grad_norm": 0.5126116275787354, + "learning_rate": 0.003, + "loss": 4.1642, + "step": 2122 + }, + { + "epoch": 0.02123, + "grad_norm": 0.50606369972229, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 2123 + }, + { + "epoch": 0.02124, + "grad_norm": 0.5823047161102295, + "learning_rate": 0.003, + "loss": 4.1677, + "step": 2124 + }, + { + "epoch": 0.02125, + "grad_norm": 0.6061305403709412, + "learning_rate": 0.003, + "loss": 4.1848, + "step": 2125 + }, + { + "epoch": 0.02126, + "grad_norm": 0.6525632739067078, + "learning_rate": 0.003, + "loss": 4.1807, + "step": 2126 + }, + { + "epoch": 0.02127, + "grad_norm": 0.6677929759025574, + "learning_rate": 0.003, + "loss": 4.1921, + "step": 2127 + }, + { + "epoch": 0.02128, + "grad_norm": 0.6721459031105042, + "learning_rate": 0.003, + "loss": 4.1793, + "step": 2128 + }, + { + "epoch": 0.02129, + "grad_norm": 0.5789133310317993, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 2129 + }, + { + "epoch": 0.0213, + "grad_norm": 0.48718854784965515, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 2130 + }, + { + "epoch": 0.02131, + "grad_norm": 0.5584967136383057, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 2131 + }, + { + "epoch": 0.02132, + "grad_norm": 0.6503726840019226, + "learning_rate": 0.003, + "loss": 4.1942, + "step": 2132 + }, + { + "epoch": 0.02133, + "grad_norm": 0.8396037817001343, + "learning_rate": 0.003, + "loss": 4.1927, + "step": 2133 + }, + { + "epoch": 0.02134, + "grad_norm": 0.9249106645584106, + "learning_rate": 0.003, + "loss": 4.2046, + "step": 2134 + }, + { + "epoch": 0.02135, + "grad_norm": 0.9640594124794006, + "learning_rate": 0.003, + "loss": 4.1994, + "step": 2135 + }, + { + "epoch": 0.02136, + "grad_norm": 0.8782548904418945, + "learning_rate": 0.003, + "loss": 4.177, + "step": 2136 + }, + { + "epoch": 0.02137, + "grad_norm": 0.8113679885864258, + "learning_rate": 0.003, + "loss": 4.2115, + "step": 2137 + }, + { + "epoch": 0.02138, + "grad_norm": 0.8208100199699402, + "learning_rate": 0.003, + "loss": 4.2211, + "step": 2138 + }, + { + "epoch": 0.02139, + "grad_norm": 0.8544081449508667, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 2139 + }, + { + "epoch": 0.0214, + "grad_norm": 0.9318257570266724, + "learning_rate": 0.003, + "loss": 4.2125, + "step": 2140 + }, + { + "epoch": 0.02141, + "grad_norm": 0.9511024951934814, + "learning_rate": 0.003, + "loss": 4.1925, + "step": 2141 + }, + { + "epoch": 0.02142, + "grad_norm": 0.8391996026039124, + "learning_rate": 0.003, + "loss": 4.1851, + "step": 2142 + }, + { + "epoch": 0.02143, + "grad_norm": 0.8300689458847046, + "learning_rate": 0.003, + "loss": 4.2047, + "step": 2143 + }, + { + "epoch": 0.02144, + "grad_norm": 0.7621930837631226, + "learning_rate": 0.003, + "loss": 4.2055, + "step": 2144 + }, + { + "epoch": 0.02145, + "grad_norm": 0.7281203269958496, + "learning_rate": 0.003, + "loss": 4.1893, + "step": 2145 + }, + { + "epoch": 0.02146, + "grad_norm": 0.6785754561424255, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 2146 + }, + { + "epoch": 0.02147, + "grad_norm": 0.6476329565048218, + "learning_rate": 0.003, + "loss": 4.2144, + "step": 2147 + }, + { + "epoch": 0.02148, + "grad_norm": 0.5711135268211365, + "learning_rate": 0.003, + "loss": 4.2235, + "step": 2148 + }, + { + "epoch": 0.02149, + "grad_norm": 0.5608777403831482, + "learning_rate": 0.003, + "loss": 4.2064, + "step": 2149 + }, + { + "epoch": 0.0215, + "grad_norm": 0.5449571013450623, + "learning_rate": 0.003, + "loss": 4.1918, + "step": 2150 + }, + { + "epoch": 0.02151, + "grad_norm": 0.5975837707519531, + "learning_rate": 0.003, + "loss": 4.1981, + "step": 2151 + }, + { + "epoch": 0.02152, + "grad_norm": 0.6778717041015625, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 2152 + }, + { + "epoch": 0.02153, + "grad_norm": 0.6393144726753235, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 2153 + }, + { + "epoch": 0.02154, + "grad_norm": 0.6100119352340698, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 2154 + }, + { + "epoch": 0.02155, + "grad_norm": 0.5750409364700317, + "learning_rate": 0.003, + "loss": 4.2008, + "step": 2155 + }, + { + "epoch": 0.02156, + "grad_norm": 0.6067488789558411, + "learning_rate": 0.003, + "loss": 4.1875, + "step": 2156 + }, + { + "epoch": 0.02157, + "grad_norm": 0.6373497247695923, + "learning_rate": 0.003, + "loss": 4.1553, + "step": 2157 + }, + { + "epoch": 0.02158, + "grad_norm": 0.6033819317817688, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 2158 + }, + { + "epoch": 0.02159, + "grad_norm": 0.5368500351905823, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 2159 + }, + { + "epoch": 0.0216, + "grad_norm": 0.5477511882781982, + "learning_rate": 0.003, + "loss": 4.1876, + "step": 2160 + }, + { + "epoch": 0.02161, + "grad_norm": 0.6277697086334229, + "learning_rate": 0.003, + "loss": 4.1773, + "step": 2161 + }, + { + "epoch": 0.02162, + "grad_norm": 0.6784037351608276, + "learning_rate": 0.003, + "loss": 4.1846, + "step": 2162 + }, + { + "epoch": 0.02163, + "grad_norm": 0.6868694424629211, + "learning_rate": 0.003, + "loss": 4.1997, + "step": 2163 + }, + { + "epoch": 0.02164, + "grad_norm": 0.7093749642372131, + "learning_rate": 0.003, + "loss": 4.1944, + "step": 2164 + }, + { + "epoch": 0.02165, + "grad_norm": 0.7334204912185669, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 2165 + }, + { + "epoch": 0.02166, + "grad_norm": 0.8240237236022949, + "learning_rate": 0.003, + "loss": 4.2062, + "step": 2166 + }, + { + "epoch": 0.02167, + "grad_norm": 0.8271484375, + "learning_rate": 0.003, + "loss": 4.1857, + "step": 2167 + }, + { + "epoch": 0.02168, + "grad_norm": 0.9684041142463684, + "learning_rate": 0.003, + "loss": 4.1878, + "step": 2168 + }, + { + "epoch": 0.02169, + "grad_norm": 0.9603534936904907, + "learning_rate": 0.003, + "loss": 4.1956, + "step": 2169 + }, + { + "epoch": 0.0217, + "grad_norm": 0.9623264074325562, + "learning_rate": 0.003, + "loss": 4.2029, + "step": 2170 + }, + { + "epoch": 0.02171, + "grad_norm": 0.8587528467178345, + "learning_rate": 0.003, + "loss": 4.2163, + "step": 2171 + }, + { + "epoch": 0.02172, + "grad_norm": 0.7502791285514832, + "learning_rate": 0.003, + "loss": 4.1819, + "step": 2172 + }, + { + "epoch": 0.02173, + "grad_norm": 0.6895846724510193, + "learning_rate": 0.003, + "loss": 4.1826, + "step": 2173 + }, + { + "epoch": 0.02174, + "grad_norm": 0.6692303419113159, + "learning_rate": 0.003, + "loss": 4.1793, + "step": 2174 + }, + { + "epoch": 0.02175, + "grad_norm": 0.6576927900314331, + "learning_rate": 0.003, + "loss": 4.2023, + "step": 2175 + }, + { + "epoch": 0.02176, + "grad_norm": 0.6110131740570068, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 2176 + }, + { + "epoch": 0.02177, + "grad_norm": 0.5583458542823792, + "learning_rate": 0.003, + "loss": 4.1855, + "step": 2177 + }, + { + "epoch": 0.02178, + "grad_norm": 0.5579090118408203, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 2178 + }, + { + "epoch": 0.02179, + "grad_norm": 0.54999840259552, + "learning_rate": 0.003, + "loss": 4.1765, + "step": 2179 + }, + { + "epoch": 0.0218, + "grad_norm": 0.5893214344978333, + "learning_rate": 0.003, + "loss": 4.193, + "step": 2180 + }, + { + "epoch": 0.02181, + "grad_norm": 0.5285120606422424, + "learning_rate": 0.003, + "loss": 4.19, + "step": 2181 + }, + { + "epoch": 0.02182, + "grad_norm": 0.537371814250946, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 2182 + }, + { + "epoch": 0.02183, + "grad_norm": 0.7419714331626892, + "learning_rate": 0.003, + "loss": 4.1927, + "step": 2183 + }, + { + "epoch": 0.02184, + "grad_norm": 0.8745968341827393, + "learning_rate": 0.003, + "loss": 4.1753, + "step": 2184 + }, + { + "epoch": 0.02185, + "grad_norm": 0.8310950398445129, + "learning_rate": 0.003, + "loss": 4.1708, + "step": 2185 + }, + { + "epoch": 0.02186, + "grad_norm": 0.6509782075881958, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 2186 + }, + { + "epoch": 0.02187, + "grad_norm": 0.6584124565124512, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 2187 + }, + { + "epoch": 0.02188, + "grad_norm": 0.6546313762664795, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 2188 + }, + { + "epoch": 0.02189, + "grad_norm": 0.6025269031524658, + "learning_rate": 0.003, + "loss": 4.1873, + "step": 2189 + }, + { + "epoch": 0.0219, + "grad_norm": 0.5339919924736023, + "learning_rate": 0.003, + "loss": 4.1761, + "step": 2190 + }, + { + "epoch": 0.02191, + "grad_norm": 0.5982394814491272, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 2191 + }, + { + "epoch": 0.02192, + "grad_norm": 0.6466907262802124, + "learning_rate": 0.003, + "loss": 4.1981, + "step": 2192 + }, + { + "epoch": 0.02193, + "grad_norm": 0.6886336207389832, + "learning_rate": 0.003, + "loss": 4.1664, + "step": 2193 + }, + { + "epoch": 0.02194, + "grad_norm": 0.6955612897872925, + "learning_rate": 0.003, + "loss": 4.2136, + "step": 2194 + }, + { + "epoch": 0.02195, + "grad_norm": 0.634350061416626, + "learning_rate": 0.003, + "loss": 4.1789, + "step": 2195 + }, + { + "epoch": 0.02196, + "grad_norm": 0.6464403867721558, + "learning_rate": 0.003, + "loss": 4.1877, + "step": 2196 + }, + { + "epoch": 0.02197, + "grad_norm": 0.7355444431304932, + "learning_rate": 0.003, + "loss": 4.1843, + "step": 2197 + }, + { + "epoch": 0.02198, + "grad_norm": 0.6861805319786072, + "learning_rate": 0.003, + "loss": 4.1732, + "step": 2198 + }, + { + "epoch": 0.02199, + "grad_norm": 0.5390181541442871, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 2199 + }, + { + "epoch": 0.022, + "grad_norm": 0.5449665188789368, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 2200 + }, + { + "epoch": 0.02201, + "grad_norm": 0.5550421476364136, + "learning_rate": 0.003, + "loss": 4.1819, + "step": 2201 + }, + { + "epoch": 0.02202, + "grad_norm": 0.7112802863121033, + "learning_rate": 0.003, + "loss": 4.1711, + "step": 2202 + }, + { + "epoch": 0.02203, + "grad_norm": 0.806663990020752, + "learning_rate": 0.003, + "loss": 4.2082, + "step": 2203 + }, + { + "epoch": 0.02204, + "grad_norm": 0.7858209013938904, + "learning_rate": 0.003, + "loss": 4.1998, + "step": 2204 + }, + { + "epoch": 0.02205, + "grad_norm": 0.782494306564331, + "learning_rate": 0.003, + "loss": 4.2034, + "step": 2205 + }, + { + "epoch": 0.02206, + "grad_norm": 0.6905097365379333, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 2206 + }, + { + "epoch": 0.02207, + "grad_norm": 0.7511074542999268, + "learning_rate": 0.003, + "loss": 4.1796, + "step": 2207 + }, + { + "epoch": 0.02208, + "grad_norm": 0.801884651184082, + "learning_rate": 0.003, + "loss": 4.1654, + "step": 2208 + }, + { + "epoch": 0.02209, + "grad_norm": 0.7684618234634399, + "learning_rate": 0.003, + "loss": 4.1791, + "step": 2209 + }, + { + "epoch": 0.0221, + "grad_norm": 0.7751032114028931, + "learning_rate": 0.003, + "loss": 4.1653, + "step": 2210 + }, + { + "epoch": 0.02211, + "grad_norm": 0.7961976528167725, + "learning_rate": 0.003, + "loss": 4.1777, + "step": 2211 + }, + { + "epoch": 0.02212, + "grad_norm": 0.694131076335907, + "learning_rate": 0.003, + "loss": 4.174, + "step": 2212 + }, + { + "epoch": 0.02213, + "grad_norm": 0.731791079044342, + "learning_rate": 0.003, + "loss": 4.1834, + "step": 2213 + }, + { + "epoch": 0.02214, + "grad_norm": 0.5831606388092041, + "learning_rate": 0.003, + "loss": 4.171, + "step": 2214 + }, + { + "epoch": 0.02215, + "grad_norm": 0.6400359869003296, + "learning_rate": 0.003, + "loss": 4.1778, + "step": 2215 + }, + { + "epoch": 0.02216, + "grad_norm": 0.6402711868286133, + "learning_rate": 0.003, + "loss": 4.156, + "step": 2216 + }, + { + "epoch": 0.02217, + "grad_norm": 0.5657291412353516, + "learning_rate": 0.003, + "loss": 4.153, + "step": 2217 + }, + { + "epoch": 0.02218, + "grad_norm": 0.5510069727897644, + "learning_rate": 0.003, + "loss": 4.1828, + "step": 2218 + }, + { + "epoch": 0.02219, + "grad_norm": 0.567196786403656, + "learning_rate": 0.003, + "loss": 4.173, + "step": 2219 + }, + { + "epoch": 0.0222, + "grad_norm": 0.5382395386695862, + "learning_rate": 0.003, + "loss": 4.1691, + "step": 2220 + }, + { + "epoch": 0.02221, + "grad_norm": 0.5838103890419006, + "learning_rate": 0.003, + "loss": 4.1595, + "step": 2221 + }, + { + "epoch": 0.02222, + "grad_norm": 0.6669290661811829, + "learning_rate": 0.003, + "loss": 4.1608, + "step": 2222 + }, + { + "epoch": 0.02223, + "grad_norm": 0.8421687483787537, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 2223 + }, + { + "epoch": 0.02224, + "grad_norm": 1.0844712257385254, + "learning_rate": 0.003, + "loss": 4.1789, + "step": 2224 + }, + { + "epoch": 0.02225, + "grad_norm": 0.7447740435600281, + "learning_rate": 0.003, + "loss": 4.1706, + "step": 2225 + }, + { + "epoch": 0.02226, + "grad_norm": 0.6103411912918091, + "learning_rate": 0.003, + "loss": 4.184, + "step": 2226 + }, + { + "epoch": 0.02227, + "grad_norm": 0.6984069347381592, + "learning_rate": 0.003, + "loss": 4.1806, + "step": 2227 + }, + { + "epoch": 0.02228, + "grad_norm": 0.6870824694633484, + "learning_rate": 0.003, + "loss": 4.1674, + "step": 2228 + }, + { + "epoch": 0.02229, + "grad_norm": 0.6159563064575195, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 2229 + }, + { + "epoch": 0.0223, + "grad_norm": 0.6422266364097595, + "learning_rate": 0.003, + "loss": 4.1847, + "step": 2230 + }, + { + "epoch": 0.02231, + "grad_norm": 0.6304430961608887, + "learning_rate": 0.003, + "loss": 4.1891, + "step": 2231 + }, + { + "epoch": 0.02232, + "grad_norm": 0.6334310173988342, + "learning_rate": 0.003, + "loss": 4.1933, + "step": 2232 + }, + { + "epoch": 0.02233, + "grad_norm": 0.729522705078125, + "learning_rate": 0.003, + "loss": 4.1662, + "step": 2233 + }, + { + "epoch": 0.02234, + "grad_norm": 0.839550256729126, + "learning_rate": 0.003, + "loss": 4.1911, + "step": 2234 + }, + { + "epoch": 0.02235, + "grad_norm": 0.8429378271102905, + "learning_rate": 0.003, + "loss": 4.2154, + "step": 2235 + }, + { + "epoch": 0.02236, + "grad_norm": 0.7225827574729919, + "learning_rate": 0.003, + "loss": 4.1659, + "step": 2236 + }, + { + "epoch": 0.02237, + "grad_norm": 0.6989867091178894, + "learning_rate": 0.003, + "loss": 4.1975, + "step": 2237 + }, + { + "epoch": 0.02238, + "grad_norm": 0.772613525390625, + "learning_rate": 0.003, + "loss": 4.1837, + "step": 2238 + }, + { + "epoch": 0.02239, + "grad_norm": 0.7962609529495239, + "learning_rate": 0.003, + "loss": 4.1688, + "step": 2239 + }, + { + "epoch": 0.0224, + "grad_norm": 0.7535995244979858, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 2240 + }, + { + "epoch": 0.02241, + "grad_norm": 0.8866783976554871, + "learning_rate": 0.003, + "loss": 4.1902, + "step": 2241 + }, + { + "epoch": 0.02242, + "grad_norm": 1.0102195739746094, + "learning_rate": 0.003, + "loss": 4.1875, + "step": 2242 + }, + { + "epoch": 0.02243, + "grad_norm": 1.013528823852539, + "learning_rate": 0.003, + "loss": 4.2245, + "step": 2243 + }, + { + "epoch": 0.02244, + "grad_norm": 0.8773812055587769, + "learning_rate": 0.003, + "loss": 4.1919, + "step": 2244 + }, + { + "epoch": 0.02245, + "grad_norm": 0.8694789409637451, + "learning_rate": 0.003, + "loss": 4.182, + "step": 2245 + }, + { + "epoch": 0.02246, + "grad_norm": 0.8814519047737122, + "learning_rate": 0.003, + "loss": 4.2131, + "step": 2246 + }, + { + "epoch": 0.02247, + "grad_norm": 0.7133143544197083, + "learning_rate": 0.003, + "loss": 4.1799, + "step": 2247 + }, + { + "epoch": 0.02248, + "grad_norm": 0.6402725577354431, + "learning_rate": 0.003, + "loss": 4.1839, + "step": 2248 + }, + { + "epoch": 0.02249, + "grad_norm": 0.5918011665344238, + "learning_rate": 0.003, + "loss": 4.1724, + "step": 2249 + }, + { + "epoch": 0.0225, + "grad_norm": 0.5062255859375, + "learning_rate": 0.003, + "loss": 4.1824, + "step": 2250 + }, + { + "epoch": 0.02251, + "grad_norm": 0.48150739073753357, + "learning_rate": 0.003, + "loss": 4.2037, + "step": 2251 + }, + { + "epoch": 0.02252, + "grad_norm": 0.4211646616458893, + "learning_rate": 0.003, + "loss": 4.1917, + "step": 2252 + }, + { + "epoch": 0.02253, + "grad_norm": 0.4155518114566803, + "learning_rate": 0.003, + "loss": 4.1719, + "step": 2253 + }, + { + "epoch": 0.02254, + "grad_norm": 0.3938300311565399, + "learning_rate": 0.003, + "loss": 4.1747, + "step": 2254 + }, + { + "epoch": 0.02255, + "grad_norm": 0.3917562961578369, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 2255 + }, + { + "epoch": 0.02256, + "grad_norm": 0.34900254011154175, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 2256 + }, + { + "epoch": 0.02257, + "grad_norm": 0.46977952122688293, + "learning_rate": 0.003, + "loss": 4.1641, + "step": 2257 + }, + { + "epoch": 0.02258, + "grad_norm": 0.6035822629928589, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 2258 + }, + { + "epoch": 0.02259, + "grad_norm": 0.8085506558418274, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 2259 + }, + { + "epoch": 0.0226, + "grad_norm": 1.0156970024108887, + "learning_rate": 0.003, + "loss": 4.2107, + "step": 2260 + }, + { + "epoch": 0.02261, + "grad_norm": 0.8568210601806641, + "learning_rate": 0.003, + "loss": 4.1861, + "step": 2261 + }, + { + "epoch": 0.02262, + "grad_norm": 0.7774007320404053, + "learning_rate": 0.003, + "loss": 4.1822, + "step": 2262 + }, + { + "epoch": 0.02263, + "grad_norm": 0.9870522022247314, + "learning_rate": 0.003, + "loss": 4.2046, + "step": 2263 + }, + { + "epoch": 0.02264, + "grad_norm": 0.7220656871795654, + "learning_rate": 0.003, + "loss": 4.1915, + "step": 2264 + }, + { + "epoch": 0.02265, + "grad_norm": 0.5900349020957947, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 2265 + }, + { + "epoch": 0.02266, + "grad_norm": 0.6756983995437622, + "learning_rate": 0.003, + "loss": 4.1642, + "step": 2266 + }, + { + "epoch": 0.02267, + "grad_norm": 0.6196969747543335, + "learning_rate": 0.003, + "loss": 4.1867, + "step": 2267 + }, + { + "epoch": 0.02268, + "grad_norm": 0.4943087697029114, + "learning_rate": 0.003, + "loss": 4.1589, + "step": 2268 + }, + { + "epoch": 0.02269, + "grad_norm": 0.45949792861938477, + "learning_rate": 0.003, + "loss": 4.1791, + "step": 2269 + }, + { + "epoch": 0.0227, + "grad_norm": 0.5057196617126465, + "learning_rate": 0.003, + "loss": 4.176, + "step": 2270 + }, + { + "epoch": 0.02271, + "grad_norm": 0.587051510810852, + "learning_rate": 0.003, + "loss": 4.1796, + "step": 2271 + }, + { + "epoch": 0.02272, + "grad_norm": 0.6179952621459961, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 2272 + }, + { + "epoch": 0.02273, + "grad_norm": 0.549956738948822, + "learning_rate": 0.003, + "loss": 4.1786, + "step": 2273 + }, + { + "epoch": 0.02274, + "grad_norm": 0.5753885507583618, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 2274 + }, + { + "epoch": 0.02275, + "grad_norm": 0.5730156898498535, + "learning_rate": 0.003, + "loss": 4.1582, + "step": 2275 + }, + { + "epoch": 0.02276, + "grad_norm": 0.633211076259613, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 2276 + }, + { + "epoch": 0.02277, + "grad_norm": 0.6606276631355286, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 2277 + }, + { + "epoch": 0.02278, + "grad_norm": 0.6212934255599976, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 2278 + }, + { + "epoch": 0.02279, + "grad_norm": 0.5762414932250977, + "learning_rate": 0.003, + "loss": 4.1803, + "step": 2279 + }, + { + "epoch": 0.0228, + "grad_norm": 0.6208809018135071, + "learning_rate": 0.003, + "loss": 4.1721, + "step": 2280 + }, + { + "epoch": 0.02281, + "grad_norm": 0.5790427327156067, + "learning_rate": 0.003, + "loss": 4.1659, + "step": 2281 + }, + { + "epoch": 0.02282, + "grad_norm": 0.5124678015708923, + "learning_rate": 0.003, + "loss": 4.1737, + "step": 2282 + }, + { + "epoch": 0.02283, + "grad_norm": 0.5517794489860535, + "learning_rate": 0.003, + "loss": 4.1895, + "step": 2283 + }, + { + "epoch": 0.02284, + "grad_norm": 0.5746618509292603, + "learning_rate": 0.003, + "loss": 4.1861, + "step": 2284 + }, + { + "epoch": 0.02285, + "grad_norm": 0.6934999823570251, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 2285 + }, + { + "epoch": 0.02286, + "grad_norm": 0.8188515305519104, + "learning_rate": 0.003, + "loss": 4.1994, + "step": 2286 + }, + { + "epoch": 0.02287, + "grad_norm": 0.7512263655662537, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 2287 + }, + { + "epoch": 0.02288, + "grad_norm": 0.7637566328048706, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 2288 + }, + { + "epoch": 0.02289, + "grad_norm": 0.700842559337616, + "learning_rate": 0.003, + "loss": 4.1993, + "step": 2289 + }, + { + "epoch": 0.0229, + "grad_norm": 0.6298280358314514, + "learning_rate": 0.003, + "loss": 4.1719, + "step": 2290 + }, + { + "epoch": 0.02291, + "grad_norm": 0.6435158848762512, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 2291 + }, + { + "epoch": 0.02292, + "grad_norm": 0.5874335169792175, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 2292 + }, + { + "epoch": 0.02293, + "grad_norm": 0.6026025414466858, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 2293 + }, + { + "epoch": 0.02294, + "grad_norm": 0.7460681200027466, + "learning_rate": 0.003, + "loss": 4.1566, + "step": 2294 + }, + { + "epoch": 0.02295, + "grad_norm": 0.7140766978263855, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 2295 + }, + { + "epoch": 0.02296, + "grad_norm": 0.6355901956558228, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 2296 + }, + { + "epoch": 0.02297, + "grad_norm": 0.673784077167511, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 2297 + }, + { + "epoch": 0.02298, + "grad_norm": 0.6826474666595459, + "learning_rate": 0.003, + "loss": 4.1671, + "step": 2298 + }, + { + "epoch": 0.02299, + "grad_norm": 0.6509355306625366, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 2299 + }, + { + "epoch": 0.023, + "grad_norm": 0.6510788202285767, + "learning_rate": 0.003, + "loss": 4.1909, + "step": 2300 + }, + { + "epoch": 0.02301, + "grad_norm": 0.6393023133277893, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 2301 + }, + { + "epoch": 0.02302, + "grad_norm": 0.7208176255226135, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 2302 + }, + { + "epoch": 0.02303, + "grad_norm": 0.8771877884864807, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 2303 + }, + { + "epoch": 0.02304, + "grad_norm": 1.1240605115890503, + "learning_rate": 0.003, + "loss": 4.2129, + "step": 2304 + }, + { + "epoch": 0.02305, + "grad_norm": 1.0586861371994019, + "learning_rate": 0.003, + "loss": 4.1816, + "step": 2305 + }, + { + "epoch": 0.02306, + "grad_norm": 0.7870573401451111, + "learning_rate": 0.003, + "loss": 4.1572, + "step": 2306 + }, + { + "epoch": 0.02307, + "grad_norm": 0.7953494191169739, + "learning_rate": 0.003, + "loss": 4.1864, + "step": 2307 + }, + { + "epoch": 0.02308, + "grad_norm": 0.800268292427063, + "learning_rate": 0.003, + "loss": 4.1782, + "step": 2308 + }, + { + "epoch": 0.02309, + "grad_norm": 0.7769814133644104, + "learning_rate": 0.003, + "loss": 4.2068, + "step": 2309 + }, + { + "epoch": 0.0231, + "grad_norm": 0.8404530882835388, + "learning_rate": 0.003, + "loss": 4.1894, + "step": 2310 + }, + { + "epoch": 0.02311, + "grad_norm": 0.8363377451896667, + "learning_rate": 0.003, + "loss": 4.1952, + "step": 2311 + }, + { + "epoch": 0.02312, + "grad_norm": 0.770296037197113, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 2312 + }, + { + "epoch": 0.02313, + "grad_norm": 0.7528126835823059, + "learning_rate": 0.003, + "loss": 4.1622, + "step": 2313 + }, + { + "epoch": 0.02314, + "grad_norm": 0.74263995885849, + "learning_rate": 0.003, + "loss": 4.1779, + "step": 2314 + }, + { + "epoch": 0.02315, + "grad_norm": 0.6081147789955139, + "learning_rate": 0.003, + "loss": 4.1766, + "step": 2315 + }, + { + "epoch": 0.02316, + "grad_norm": 0.5721566677093506, + "learning_rate": 0.003, + "loss": 4.1882, + "step": 2316 + }, + { + "epoch": 0.02317, + "grad_norm": 0.5732864141464233, + "learning_rate": 0.003, + "loss": 4.1684, + "step": 2317 + }, + { + "epoch": 0.02318, + "grad_norm": 0.5256015062332153, + "learning_rate": 0.003, + "loss": 4.1551, + "step": 2318 + }, + { + "epoch": 0.02319, + "grad_norm": 0.49379223585128784, + "learning_rate": 0.003, + "loss": 4.1772, + "step": 2319 + }, + { + "epoch": 0.0232, + "grad_norm": 0.5656861662864685, + "learning_rate": 0.003, + "loss": 4.1857, + "step": 2320 + }, + { + "epoch": 0.02321, + "grad_norm": 0.5173032879829407, + "learning_rate": 0.003, + "loss": 4.1628, + "step": 2321 + }, + { + "epoch": 0.02322, + "grad_norm": 0.4566805362701416, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 2322 + }, + { + "epoch": 0.02323, + "grad_norm": 0.42646524310112, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 2323 + }, + { + "epoch": 0.02324, + "grad_norm": 0.46551966667175293, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 2324 + }, + { + "epoch": 0.02325, + "grad_norm": 0.4420718252658844, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 2325 + }, + { + "epoch": 0.02326, + "grad_norm": 0.42437517642974854, + "learning_rate": 0.003, + "loss": 4.1773, + "step": 2326 + }, + { + "epoch": 0.02327, + "grad_norm": 0.457309365272522, + "learning_rate": 0.003, + "loss": 4.164, + "step": 2327 + }, + { + "epoch": 0.02328, + "grad_norm": 0.515424370765686, + "learning_rate": 0.003, + "loss": 4.1652, + "step": 2328 + }, + { + "epoch": 0.02329, + "grad_norm": 0.5420231223106384, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 2329 + }, + { + "epoch": 0.0233, + "grad_norm": 0.5254723429679871, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 2330 + }, + { + "epoch": 0.02331, + "grad_norm": 0.47864091396331787, + "learning_rate": 0.003, + "loss": 4.1851, + "step": 2331 + }, + { + "epoch": 0.02332, + "grad_norm": 0.6108183264732361, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 2332 + }, + { + "epoch": 0.02333, + "grad_norm": 0.7649471759796143, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 2333 + }, + { + "epoch": 0.02334, + "grad_norm": 0.9433549642562866, + "learning_rate": 0.003, + "loss": 4.1411, + "step": 2334 + }, + { + "epoch": 0.02335, + "grad_norm": 1.05660080909729, + "learning_rate": 0.003, + "loss": 4.1889, + "step": 2335 + }, + { + "epoch": 0.02336, + "grad_norm": 0.7674508690834045, + "learning_rate": 0.003, + "loss": 4.1752, + "step": 2336 + }, + { + "epoch": 0.02337, + "grad_norm": 0.6039113402366638, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 2337 + }, + { + "epoch": 0.02338, + "grad_norm": 0.678247332572937, + "learning_rate": 0.003, + "loss": 4.1635, + "step": 2338 + }, + { + "epoch": 0.02339, + "grad_norm": 0.6683263182640076, + "learning_rate": 0.003, + "loss": 4.1573, + "step": 2339 + }, + { + "epoch": 0.0234, + "grad_norm": 0.6628241539001465, + "learning_rate": 0.003, + "loss": 4.1711, + "step": 2340 + }, + { + "epoch": 0.02341, + "grad_norm": 0.6385108232498169, + "learning_rate": 0.003, + "loss": 4.145, + "step": 2341 + }, + { + "epoch": 0.02342, + "grad_norm": 0.7277436852455139, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 2342 + }, + { + "epoch": 0.02343, + "grad_norm": 0.7746118903160095, + "learning_rate": 0.003, + "loss": 4.1654, + "step": 2343 + }, + { + "epoch": 0.02344, + "grad_norm": 0.8185151815414429, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 2344 + }, + { + "epoch": 0.02345, + "grad_norm": 0.8307933211326599, + "learning_rate": 0.003, + "loss": 4.163, + "step": 2345 + }, + { + "epoch": 0.02346, + "grad_norm": 0.7920862436294556, + "learning_rate": 0.003, + "loss": 4.1608, + "step": 2346 + }, + { + "epoch": 0.02347, + "grad_norm": 0.7341541051864624, + "learning_rate": 0.003, + "loss": 4.1955, + "step": 2347 + }, + { + "epoch": 0.02348, + "grad_norm": 0.7185869216918945, + "learning_rate": 0.003, + "loss": 4.1581, + "step": 2348 + }, + { + "epoch": 0.02349, + "grad_norm": 0.7363864183425903, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 2349 + }, + { + "epoch": 0.0235, + "grad_norm": 0.8115167021751404, + "learning_rate": 0.003, + "loss": 4.1777, + "step": 2350 + }, + { + "epoch": 0.02351, + "grad_norm": 0.8712372779846191, + "learning_rate": 0.003, + "loss": 4.1958, + "step": 2351 + }, + { + "epoch": 0.02352, + "grad_norm": 0.7720403075218201, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 2352 + }, + { + "epoch": 0.02353, + "grad_norm": 0.7881413102149963, + "learning_rate": 0.003, + "loss": 4.1809, + "step": 2353 + }, + { + "epoch": 0.02354, + "grad_norm": 0.9158604741096497, + "learning_rate": 0.003, + "loss": 4.1959, + "step": 2354 + }, + { + "epoch": 0.02355, + "grad_norm": 0.8635379076004028, + "learning_rate": 0.003, + "loss": 4.1914, + "step": 2355 + }, + { + "epoch": 0.02356, + "grad_norm": 0.9214644432067871, + "learning_rate": 0.003, + "loss": 4.182, + "step": 2356 + }, + { + "epoch": 0.02357, + "grad_norm": 1.004699945449829, + "learning_rate": 0.003, + "loss": 4.1873, + "step": 2357 + }, + { + "epoch": 0.02358, + "grad_norm": 0.8610268235206604, + "learning_rate": 0.003, + "loss": 4.1999, + "step": 2358 + }, + { + "epoch": 0.02359, + "grad_norm": 0.7828254699707031, + "learning_rate": 0.003, + "loss": 4.1779, + "step": 2359 + }, + { + "epoch": 0.0236, + "grad_norm": 0.6673213839530945, + "learning_rate": 0.003, + "loss": 4.2078, + "step": 2360 + }, + { + "epoch": 0.02361, + "grad_norm": 0.6459137201309204, + "learning_rate": 0.003, + "loss": 4.1883, + "step": 2361 + }, + { + "epoch": 0.02362, + "grad_norm": 0.5431698560714722, + "learning_rate": 0.003, + "loss": 4.1762, + "step": 2362 + }, + { + "epoch": 0.02363, + "grad_norm": 0.551616907119751, + "learning_rate": 0.003, + "loss": 4.151, + "step": 2363 + }, + { + "epoch": 0.02364, + "grad_norm": 0.5682998299598694, + "learning_rate": 0.003, + "loss": 4.1714, + "step": 2364 + }, + { + "epoch": 0.02365, + "grad_norm": 0.651782751083374, + "learning_rate": 0.003, + "loss": 4.1916, + "step": 2365 + }, + { + "epoch": 0.02366, + "grad_norm": 0.653306782245636, + "learning_rate": 0.003, + "loss": 4.164, + "step": 2366 + }, + { + "epoch": 0.02367, + "grad_norm": 0.6368426084518433, + "learning_rate": 0.003, + "loss": 4.1718, + "step": 2367 + }, + { + "epoch": 0.02368, + "grad_norm": 0.713473379611969, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 2368 + }, + { + "epoch": 0.02369, + "grad_norm": 0.8085301518440247, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 2369 + }, + { + "epoch": 0.0237, + "grad_norm": 0.7926542162895203, + "learning_rate": 0.003, + "loss": 4.1888, + "step": 2370 + }, + { + "epoch": 0.02371, + "grad_norm": 0.828406572341919, + "learning_rate": 0.003, + "loss": 4.1795, + "step": 2371 + }, + { + "epoch": 0.02372, + "grad_norm": 0.7606370449066162, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 2372 + }, + { + "epoch": 0.02373, + "grad_norm": 0.6590718626976013, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 2373 + }, + { + "epoch": 0.02374, + "grad_norm": 0.5918402075767517, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 2374 + }, + { + "epoch": 0.02375, + "grad_norm": 0.6583318710327148, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 2375 + }, + { + "epoch": 0.02376, + "grad_norm": 0.6908389329910278, + "learning_rate": 0.003, + "loss": 4.1806, + "step": 2376 + }, + { + "epoch": 0.02377, + "grad_norm": 0.6131519675254822, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 2377 + }, + { + "epoch": 0.02378, + "grad_norm": 0.4756273031234741, + "learning_rate": 0.003, + "loss": 4.1538, + "step": 2378 + }, + { + "epoch": 0.02379, + "grad_norm": 0.5182189345359802, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 2379 + }, + { + "epoch": 0.0238, + "grad_norm": 0.4939170181751251, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 2380 + }, + { + "epoch": 0.02381, + "grad_norm": 0.4250797629356384, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 2381 + }, + { + "epoch": 0.02382, + "grad_norm": 0.44852757453918457, + "learning_rate": 0.003, + "loss": 4.1818, + "step": 2382 + }, + { + "epoch": 0.02383, + "grad_norm": 0.4787205457687378, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 2383 + }, + { + "epoch": 0.02384, + "grad_norm": 0.5154638290405273, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 2384 + }, + { + "epoch": 0.02385, + "grad_norm": 0.5779361128807068, + "learning_rate": 0.003, + "loss": 4.1594, + "step": 2385 + }, + { + "epoch": 0.02386, + "grad_norm": 0.6726837158203125, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 2386 + }, + { + "epoch": 0.02387, + "grad_norm": 0.6705544590950012, + "learning_rate": 0.003, + "loss": 4.138, + "step": 2387 + }, + { + "epoch": 0.02388, + "grad_norm": 0.520875096321106, + "learning_rate": 0.003, + "loss": 4.1869, + "step": 2388 + }, + { + "epoch": 0.02389, + "grad_norm": 0.4918033182621002, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 2389 + }, + { + "epoch": 0.0239, + "grad_norm": 0.5085783004760742, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 2390 + }, + { + "epoch": 0.02391, + "grad_norm": 0.5585421323776245, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 2391 + }, + { + "epoch": 0.02392, + "grad_norm": 0.6892051100730896, + "learning_rate": 0.003, + "loss": 4.173, + "step": 2392 + }, + { + "epoch": 0.02393, + "grad_norm": 0.7382843494415283, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 2393 + }, + { + "epoch": 0.02394, + "grad_norm": 0.6831554770469666, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 2394 + }, + { + "epoch": 0.02395, + "grad_norm": 0.5855069756507874, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 2395 + }, + { + "epoch": 0.02396, + "grad_norm": 0.618294894695282, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 2396 + }, + { + "epoch": 0.02397, + "grad_norm": 0.6775050163269043, + "learning_rate": 0.003, + "loss": 4.1604, + "step": 2397 + }, + { + "epoch": 0.02398, + "grad_norm": 0.6511969566345215, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 2398 + }, + { + "epoch": 0.02399, + "grad_norm": 0.538463294506073, + "learning_rate": 0.003, + "loss": 4.142, + "step": 2399 + }, + { + "epoch": 0.024, + "grad_norm": 0.6128776669502258, + "learning_rate": 0.003, + "loss": 4.1581, + "step": 2400 + }, + { + "epoch": 0.02401, + "grad_norm": 0.7419440150260925, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 2401 + }, + { + "epoch": 0.02402, + "grad_norm": 0.8263295888900757, + "learning_rate": 0.003, + "loss": 4.1752, + "step": 2402 + }, + { + "epoch": 0.02403, + "grad_norm": 0.7187276482582092, + "learning_rate": 0.003, + "loss": 4.143, + "step": 2403 + }, + { + "epoch": 0.02404, + "grad_norm": 0.6155825257301331, + "learning_rate": 0.003, + "loss": 4.1775, + "step": 2404 + }, + { + "epoch": 0.02405, + "grad_norm": 0.584798276424408, + "learning_rate": 0.003, + "loss": 4.1805, + "step": 2405 + }, + { + "epoch": 0.02406, + "grad_norm": 0.7314527034759521, + "learning_rate": 0.003, + "loss": 4.169, + "step": 2406 + }, + { + "epoch": 0.02407, + "grad_norm": 0.9434296488761902, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 2407 + }, + { + "epoch": 0.02408, + "grad_norm": 1.003302812576294, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 2408 + }, + { + "epoch": 0.02409, + "grad_norm": 0.9447243809700012, + "learning_rate": 0.003, + "loss": 4.1773, + "step": 2409 + }, + { + "epoch": 0.0241, + "grad_norm": 0.7472692131996155, + "learning_rate": 0.003, + "loss": 4.1837, + "step": 2410 + }, + { + "epoch": 0.02411, + "grad_norm": 0.8666222095489502, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 2411 + }, + { + "epoch": 0.02412, + "grad_norm": 0.9741396307945251, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 2412 + }, + { + "epoch": 0.02413, + "grad_norm": 1.0521008968353271, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 2413 + }, + { + "epoch": 0.02414, + "grad_norm": 1.098198652267456, + "learning_rate": 0.003, + "loss": 4.198, + "step": 2414 + }, + { + "epoch": 0.02415, + "grad_norm": 1.0509668588638306, + "learning_rate": 0.003, + "loss": 4.1933, + "step": 2415 + }, + { + "epoch": 0.02416, + "grad_norm": 0.9923770427703857, + "learning_rate": 0.003, + "loss": 4.1791, + "step": 2416 + }, + { + "epoch": 0.02417, + "grad_norm": 0.9507772922515869, + "learning_rate": 0.003, + "loss": 4.2103, + "step": 2417 + }, + { + "epoch": 0.02418, + "grad_norm": 0.7804825305938721, + "learning_rate": 0.003, + "loss": 4.1914, + "step": 2418 + }, + { + "epoch": 0.02419, + "grad_norm": 0.6353334188461304, + "learning_rate": 0.003, + "loss": 4.1871, + "step": 2419 + }, + { + "epoch": 0.0242, + "grad_norm": 0.7049343585968018, + "learning_rate": 0.003, + "loss": 4.1989, + "step": 2420 + }, + { + "epoch": 0.02421, + "grad_norm": 0.8157212734222412, + "learning_rate": 0.003, + "loss": 4.1685, + "step": 2421 + }, + { + "epoch": 0.02422, + "grad_norm": 0.9302619099617004, + "learning_rate": 0.003, + "loss": 4.2265, + "step": 2422 + }, + { + "epoch": 0.02423, + "grad_norm": 1.0272700786590576, + "learning_rate": 0.003, + "loss": 4.2247, + "step": 2423 + }, + { + "epoch": 0.02424, + "grad_norm": 1.0545587539672852, + "learning_rate": 0.003, + "loss": 4.2182, + "step": 2424 + }, + { + "epoch": 0.02425, + "grad_norm": 0.8540631532669067, + "learning_rate": 0.003, + "loss": 4.2022, + "step": 2425 + }, + { + "epoch": 0.02426, + "grad_norm": 0.7790244817733765, + "learning_rate": 0.003, + "loss": 4.2053, + "step": 2426 + }, + { + "epoch": 0.02427, + "grad_norm": 0.6194785237312317, + "learning_rate": 0.003, + "loss": 4.2245, + "step": 2427 + }, + { + "epoch": 0.02428, + "grad_norm": 0.723381519317627, + "learning_rate": 0.003, + "loss": 4.1967, + "step": 2428 + }, + { + "epoch": 0.02429, + "grad_norm": 0.8042016625404358, + "learning_rate": 0.003, + "loss": 4.1857, + "step": 2429 + }, + { + "epoch": 0.0243, + "grad_norm": 0.8838525414466858, + "learning_rate": 0.003, + "loss": 4.211, + "step": 2430 + }, + { + "epoch": 0.02431, + "grad_norm": 0.9250835180282593, + "learning_rate": 0.003, + "loss": 4.1912, + "step": 2431 + }, + { + "epoch": 0.02432, + "grad_norm": 0.7146485447883606, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 2432 + }, + { + "epoch": 0.02433, + "grad_norm": 0.689758837223053, + "learning_rate": 0.003, + "loss": 4.2175, + "step": 2433 + }, + { + "epoch": 0.02434, + "grad_norm": 0.7109194993972778, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 2434 + }, + { + "epoch": 0.02435, + "grad_norm": 0.6944730281829834, + "learning_rate": 0.003, + "loss": 4.1786, + "step": 2435 + }, + { + "epoch": 0.02436, + "grad_norm": 0.5854509472846985, + "learning_rate": 0.003, + "loss": 4.1804, + "step": 2436 + }, + { + "epoch": 0.02437, + "grad_norm": 0.5410053730010986, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 2437 + }, + { + "epoch": 0.02438, + "grad_norm": 0.4632667303085327, + "learning_rate": 0.003, + "loss": 4.1763, + "step": 2438 + }, + { + "epoch": 0.02439, + "grad_norm": 0.47085174918174744, + "learning_rate": 0.003, + "loss": 4.143, + "step": 2439 + }, + { + "epoch": 0.0244, + "grad_norm": 0.4175461232662201, + "learning_rate": 0.003, + "loss": 4.1908, + "step": 2440 + }, + { + "epoch": 0.02441, + "grad_norm": 0.4487816095352173, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 2441 + }, + { + "epoch": 0.02442, + "grad_norm": 0.3787577450275421, + "learning_rate": 0.003, + "loss": 4.1845, + "step": 2442 + }, + { + "epoch": 0.02443, + "grad_norm": 0.4035150706768036, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 2443 + }, + { + "epoch": 0.02444, + "grad_norm": 0.399797648191452, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 2444 + }, + { + "epoch": 0.02445, + "grad_norm": 0.4260754883289337, + "learning_rate": 0.003, + "loss": 4.1538, + "step": 2445 + }, + { + "epoch": 0.02446, + "grad_norm": 0.5048103928565979, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 2446 + }, + { + "epoch": 0.02447, + "grad_norm": 0.6289964914321899, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 2447 + }, + { + "epoch": 0.02448, + "grad_norm": 0.8102890253067017, + "learning_rate": 0.003, + "loss": 4.1719, + "step": 2448 + }, + { + "epoch": 0.02449, + "grad_norm": 0.8461275100708008, + "learning_rate": 0.003, + "loss": 4.1705, + "step": 2449 + }, + { + "epoch": 0.0245, + "grad_norm": 0.7120311260223389, + "learning_rate": 0.003, + "loss": 4.1692, + "step": 2450 + }, + { + "epoch": 0.02451, + "grad_norm": 0.4928838908672333, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 2451 + }, + { + "epoch": 0.02452, + "grad_norm": 0.517375648021698, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 2452 + }, + { + "epoch": 0.02453, + "grad_norm": 0.5950134992599487, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 2453 + }, + { + "epoch": 0.02454, + "grad_norm": 0.6498395800590515, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 2454 + }, + { + "epoch": 0.02455, + "grad_norm": 0.666580319404602, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 2455 + }, + { + "epoch": 0.02456, + "grad_norm": 0.6841810941696167, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 2456 + }, + { + "epoch": 0.02457, + "grad_norm": 0.6185785531997681, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 2457 + }, + { + "epoch": 0.02458, + "grad_norm": 0.6619728207588196, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 2458 + }, + { + "epoch": 0.02459, + "grad_norm": 0.721472978591919, + "learning_rate": 0.003, + "loss": 4.1746, + "step": 2459 + }, + { + "epoch": 0.0246, + "grad_norm": 0.7174915075302124, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 2460 + }, + { + "epoch": 0.02461, + "grad_norm": 0.7339833378791809, + "learning_rate": 0.003, + "loss": 4.1527, + "step": 2461 + }, + { + "epoch": 0.02462, + "grad_norm": 0.738292932510376, + "learning_rate": 0.003, + "loss": 4.1817, + "step": 2462 + }, + { + "epoch": 0.02463, + "grad_norm": 0.6757227778434753, + "learning_rate": 0.003, + "loss": 4.1783, + "step": 2463 + }, + { + "epoch": 0.02464, + "grad_norm": 0.5629785656929016, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 2464 + }, + { + "epoch": 0.02465, + "grad_norm": 0.619929850101471, + "learning_rate": 0.003, + "loss": 4.1723, + "step": 2465 + }, + { + "epoch": 0.02466, + "grad_norm": 0.5323424339294434, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 2466 + }, + { + "epoch": 0.02467, + "grad_norm": 0.5933912396430969, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 2467 + }, + { + "epoch": 0.02468, + "grad_norm": 0.5487152338027954, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 2468 + }, + { + "epoch": 0.02469, + "grad_norm": 0.5824567675590515, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 2469 + }, + { + "epoch": 0.0247, + "grad_norm": 0.674126148223877, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 2470 + }, + { + "epoch": 0.02471, + "grad_norm": 0.6640561819076538, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 2471 + }, + { + "epoch": 0.02472, + "grad_norm": 0.6775668859481812, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 2472 + }, + { + "epoch": 0.02473, + "grad_norm": 0.668128252029419, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 2473 + }, + { + "epoch": 0.02474, + "grad_norm": 0.6847355961799622, + "learning_rate": 0.003, + "loss": 4.1625, + "step": 2474 + }, + { + "epoch": 0.02475, + "grad_norm": 0.8420220017433167, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 2475 + }, + { + "epoch": 0.02476, + "grad_norm": 0.8548656702041626, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 2476 + }, + { + "epoch": 0.02477, + "grad_norm": 0.6734539866447449, + "learning_rate": 0.003, + "loss": 4.1705, + "step": 2477 + }, + { + "epoch": 0.02478, + "grad_norm": 0.6893376111984253, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 2478 + }, + { + "epoch": 0.02479, + "grad_norm": 0.5793731808662415, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 2479 + }, + { + "epoch": 0.0248, + "grad_norm": 0.5962329506874084, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 2480 + }, + { + "epoch": 0.02481, + "grad_norm": 0.5665598511695862, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 2481 + }, + { + "epoch": 0.02482, + "grad_norm": 0.6422098875045776, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 2482 + }, + { + "epoch": 0.02483, + "grad_norm": 0.6842206716537476, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 2483 + }, + { + "epoch": 0.02484, + "grad_norm": 0.6730716824531555, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 2484 + }, + { + "epoch": 0.02485, + "grad_norm": 0.7416301369667053, + "learning_rate": 0.003, + "loss": 4.1761, + "step": 2485 + }, + { + "epoch": 0.02486, + "grad_norm": 0.8094339370727539, + "learning_rate": 0.003, + "loss": 4.2033, + "step": 2486 + }, + { + "epoch": 0.02487, + "grad_norm": 0.7401948571205139, + "learning_rate": 0.003, + "loss": 4.1575, + "step": 2487 + }, + { + "epoch": 0.02488, + "grad_norm": 0.7254106402397156, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 2488 + }, + { + "epoch": 0.02489, + "grad_norm": 0.7380350232124329, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 2489 + }, + { + "epoch": 0.0249, + "grad_norm": 0.8402035236358643, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 2490 + }, + { + "epoch": 0.02491, + "grad_norm": 0.9696373343467712, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 2491 + }, + { + "epoch": 0.02492, + "grad_norm": 0.9300618171691895, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 2492 + }, + { + "epoch": 0.02493, + "grad_norm": 0.7866774201393127, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 2493 + }, + { + "epoch": 0.02494, + "grad_norm": 0.7883460521697998, + "learning_rate": 0.003, + "loss": 4.1631, + "step": 2494 + }, + { + "epoch": 0.02495, + "grad_norm": 0.7333797216415405, + "learning_rate": 0.003, + "loss": 4.1705, + "step": 2495 + }, + { + "epoch": 0.02496, + "grad_norm": 0.7359468936920166, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 2496 + }, + { + "epoch": 0.02497, + "grad_norm": 0.7477230429649353, + "learning_rate": 0.003, + "loss": 4.1701, + "step": 2497 + }, + { + "epoch": 0.02498, + "grad_norm": 0.6632980704307556, + "learning_rate": 0.003, + "loss": 4.1549, + "step": 2498 + }, + { + "epoch": 0.02499, + "grad_norm": 0.5524125099182129, + "learning_rate": 0.003, + "loss": 4.1622, + "step": 2499 + }, + { + "epoch": 0.025, + "grad_norm": 0.5341687202453613, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 2500 + }, + { + "epoch": 0.02501, + "grad_norm": 0.5111349821090698, + "learning_rate": 0.003, + "loss": 4.1766, + "step": 2501 + }, + { + "epoch": 0.02502, + "grad_norm": 0.5449102520942688, + "learning_rate": 0.003, + "loss": 4.159, + "step": 2502 + }, + { + "epoch": 0.02503, + "grad_norm": 0.5349955558776855, + "learning_rate": 0.003, + "loss": 4.1642, + "step": 2503 + }, + { + "epoch": 0.02504, + "grad_norm": 0.48626482486724854, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 2504 + }, + { + "epoch": 0.02505, + "grad_norm": 0.527682363986969, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 2505 + }, + { + "epoch": 0.02506, + "grad_norm": 0.6103768348693848, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 2506 + }, + { + "epoch": 0.02507, + "grad_norm": 0.6964853405952454, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 2507 + }, + { + "epoch": 0.02508, + "grad_norm": 0.8648868203163147, + "learning_rate": 0.003, + "loss": 4.1494, + "step": 2508 + }, + { + "epoch": 0.02509, + "grad_norm": 0.9269753694534302, + "learning_rate": 0.003, + "loss": 4.1639, + "step": 2509 + }, + { + "epoch": 0.0251, + "grad_norm": 1.022930383682251, + "learning_rate": 0.003, + "loss": 4.1832, + "step": 2510 + }, + { + "epoch": 0.02511, + "grad_norm": 0.7745629549026489, + "learning_rate": 0.003, + "loss": 4.1837, + "step": 2511 + }, + { + "epoch": 0.02512, + "grad_norm": 0.6751075983047485, + "learning_rate": 0.003, + "loss": 4.1635, + "step": 2512 + }, + { + "epoch": 0.02513, + "grad_norm": 0.7629876732826233, + "learning_rate": 0.003, + "loss": 4.1626, + "step": 2513 + }, + { + "epoch": 0.02514, + "grad_norm": 0.7923519015312195, + "learning_rate": 0.003, + "loss": 4.1521, + "step": 2514 + }, + { + "epoch": 0.02515, + "grad_norm": 0.7363923192024231, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 2515 + }, + { + "epoch": 0.02516, + "grad_norm": 0.6523436903953552, + "learning_rate": 0.003, + "loss": 4.1588, + "step": 2516 + }, + { + "epoch": 0.02517, + "grad_norm": 0.669247031211853, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 2517 + }, + { + "epoch": 0.02518, + "grad_norm": 0.5932192802429199, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 2518 + }, + { + "epoch": 0.02519, + "grad_norm": 0.5920613408088684, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 2519 + }, + { + "epoch": 0.0252, + "grad_norm": 0.5889076590538025, + "learning_rate": 0.003, + "loss": 4.1566, + "step": 2520 + }, + { + "epoch": 0.02521, + "grad_norm": 0.650814414024353, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 2521 + }, + { + "epoch": 0.02522, + "grad_norm": 0.5603437423706055, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 2522 + }, + { + "epoch": 0.02523, + "grad_norm": 0.4928959906101227, + "learning_rate": 0.003, + "loss": 4.146, + "step": 2523 + }, + { + "epoch": 0.02524, + "grad_norm": 0.4817637801170349, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 2524 + }, + { + "epoch": 0.02525, + "grad_norm": 0.4819410741329193, + "learning_rate": 0.003, + "loss": 4.169, + "step": 2525 + }, + { + "epoch": 0.02526, + "grad_norm": 0.5578210353851318, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 2526 + }, + { + "epoch": 0.02527, + "grad_norm": 0.7309604287147522, + "learning_rate": 0.003, + "loss": 4.157, + "step": 2527 + }, + { + "epoch": 0.02528, + "grad_norm": 0.9689713716506958, + "learning_rate": 0.003, + "loss": 4.1882, + "step": 2528 + }, + { + "epoch": 0.02529, + "grad_norm": 1.002740502357483, + "learning_rate": 0.003, + "loss": 4.1697, + "step": 2529 + }, + { + "epoch": 0.0253, + "grad_norm": 0.845522940158844, + "learning_rate": 0.003, + "loss": 4.1791, + "step": 2530 + }, + { + "epoch": 0.02531, + "grad_norm": 0.8186164498329163, + "learning_rate": 0.003, + "loss": 4.181, + "step": 2531 + }, + { + "epoch": 0.02532, + "grad_norm": 0.7776713371276855, + "learning_rate": 0.003, + "loss": 4.15, + "step": 2532 + }, + { + "epoch": 0.02533, + "grad_norm": 0.6073544025421143, + "learning_rate": 0.003, + "loss": 4.1843, + "step": 2533 + }, + { + "epoch": 0.02534, + "grad_norm": 0.7059157490730286, + "learning_rate": 0.003, + "loss": 4.1674, + "step": 2534 + }, + { + "epoch": 0.02535, + "grad_norm": 0.6908336281776428, + "learning_rate": 0.003, + "loss": 4.1601, + "step": 2535 + }, + { + "epoch": 0.02536, + "grad_norm": 0.6330790519714355, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 2536 + }, + { + "epoch": 0.02537, + "grad_norm": 0.5009950995445251, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 2537 + }, + { + "epoch": 0.02538, + "grad_norm": 0.5122132897377014, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 2538 + }, + { + "epoch": 0.02539, + "grad_norm": 0.5557868480682373, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 2539 + }, + { + "epoch": 0.0254, + "grad_norm": 0.5659569501876831, + "learning_rate": 0.003, + "loss": 4.1601, + "step": 2540 + }, + { + "epoch": 0.02541, + "grad_norm": 0.4916652739048004, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 2541 + }, + { + "epoch": 0.02542, + "grad_norm": 0.5319804549217224, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 2542 + }, + { + "epoch": 0.02543, + "grad_norm": 0.4849517047405243, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 2543 + }, + { + "epoch": 0.02544, + "grad_norm": 0.46546679735183716, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 2544 + }, + { + "epoch": 0.02545, + "grad_norm": 0.4903736710548401, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 2545 + }, + { + "epoch": 0.02546, + "grad_norm": 0.5320796370506287, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 2546 + }, + { + "epoch": 0.02547, + "grad_norm": 0.6876251101493835, + "learning_rate": 0.003, + "loss": 4.1736, + "step": 2547 + }, + { + "epoch": 0.02548, + "grad_norm": 0.8988205194473267, + "learning_rate": 0.003, + "loss": 4.1808, + "step": 2548 + }, + { + "epoch": 0.02549, + "grad_norm": 0.9427040815353394, + "learning_rate": 0.003, + "loss": 4.1752, + "step": 2549 + }, + { + "epoch": 0.0255, + "grad_norm": 0.8955228924751282, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 2550 + }, + { + "epoch": 0.02551, + "grad_norm": 0.6933260560035706, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 2551 + }, + { + "epoch": 0.02552, + "grad_norm": 0.6283506751060486, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 2552 + }, + { + "epoch": 0.02553, + "grad_norm": 0.6629008054733276, + "learning_rate": 0.003, + "loss": 4.1984, + "step": 2553 + }, + { + "epoch": 0.02554, + "grad_norm": 0.6356085538864136, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 2554 + }, + { + "epoch": 0.02555, + "grad_norm": 0.6418399214744568, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 2555 + }, + { + "epoch": 0.02556, + "grad_norm": 0.689191997051239, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 2556 + }, + { + "epoch": 0.02557, + "grad_norm": 0.6599099636077881, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 2557 + }, + { + "epoch": 0.02558, + "grad_norm": 0.6565999984741211, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 2558 + }, + { + "epoch": 0.02559, + "grad_norm": 0.6738297939300537, + "learning_rate": 0.003, + "loss": 4.1674, + "step": 2559 + }, + { + "epoch": 0.0256, + "grad_norm": 0.5509478449821472, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 2560 + }, + { + "epoch": 0.02561, + "grad_norm": 0.5239382982254028, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 2561 + }, + { + "epoch": 0.02562, + "grad_norm": 0.5102459788322449, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 2562 + }, + { + "epoch": 0.02563, + "grad_norm": 0.5850344896316528, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 2563 + }, + { + "epoch": 0.02564, + "grad_norm": 0.6901075839996338, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 2564 + }, + { + "epoch": 0.02565, + "grad_norm": 0.8146087527275085, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 2565 + }, + { + "epoch": 0.02566, + "grad_norm": 0.8885420560836792, + "learning_rate": 0.003, + "loss": 4.1712, + "step": 2566 + }, + { + "epoch": 0.02567, + "grad_norm": 0.9885812997817993, + "learning_rate": 0.003, + "loss": 4.1808, + "step": 2567 + }, + { + "epoch": 0.02568, + "grad_norm": 1.0153335332870483, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 2568 + }, + { + "epoch": 0.02569, + "grad_norm": 0.9595790505409241, + "learning_rate": 0.003, + "loss": 4.1596, + "step": 2569 + }, + { + "epoch": 0.0257, + "grad_norm": 1.0054306983947754, + "learning_rate": 0.003, + "loss": 4.1689, + "step": 2570 + }, + { + "epoch": 0.02571, + "grad_norm": 0.997405469417572, + "learning_rate": 0.003, + "loss": 4.1908, + "step": 2571 + }, + { + "epoch": 0.02572, + "grad_norm": 0.9328084588050842, + "learning_rate": 0.003, + "loss": 4.1687, + "step": 2572 + }, + { + "epoch": 0.02573, + "grad_norm": 0.8243227005004883, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 2573 + }, + { + "epoch": 0.02574, + "grad_norm": 0.7659904360771179, + "learning_rate": 0.003, + "loss": 4.1984, + "step": 2574 + }, + { + "epoch": 0.02575, + "grad_norm": 0.7300676107406616, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 2575 + }, + { + "epoch": 0.02576, + "grad_norm": 0.6190217733383179, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 2576 + }, + { + "epoch": 0.02577, + "grad_norm": 0.6222746968269348, + "learning_rate": 0.003, + "loss": 4.138, + "step": 2577 + }, + { + "epoch": 0.02578, + "grad_norm": 0.698691725730896, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 2578 + }, + { + "epoch": 0.02579, + "grad_norm": 0.6583962440490723, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 2579 + }, + { + "epoch": 0.0258, + "grad_norm": 0.6283578872680664, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 2580 + }, + { + "epoch": 0.02581, + "grad_norm": 0.48666515946388245, + "learning_rate": 0.003, + "loss": 4.1808, + "step": 2581 + }, + { + "epoch": 0.02582, + "grad_norm": 0.5155734419822693, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 2582 + }, + { + "epoch": 0.02583, + "grad_norm": 0.46086037158966064, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 2583 + }, + { + "epoch": 0.02584, + "grad_norm": 0.4467635154724121, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 2584 + }, + { + "epoch": 0.02585, + "grad_norm": 0.4732924997806549, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 2585 + }, + { + "epoch": 0.02586, + "grad_norm": 0.40596628189086914, + "learning_rate": 0.003, + "loss": 4.159, + "step": 2586 + }, + { + "epoch": 0.02587, + "grad_norm": 0.41176727414131165, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 2587 + }, + { + "epoch": 0.02588, + "grad_norm": 0.4998745620250702, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 2588 + }, + { + "epoch": 0.02589, + "grad_norm": 0.6378880143165588, + "learning_rate": 0.003, + "loss": 4.1526, + "step": 2589 + }, + { + "epoch": 0.0259, + "grad_norm": 0.9064441919326782, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 2590 + }, + { + "epoch": 0.02591, + "grad_norm": 1.16274094581604, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 2591 + }, + { + "epoch": 0.02592, + "grad_norm": 0.8324390053749084, + "learning_rate": 0.003, + "loss": 4.1564, + "step": 2592 + }, + { + "epoch": 0.02593, + "grad_norm": 0.8290135860443115, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 2593 + }, + { + "epoch": 0.02594, + "grad_norm": 0.8181279897689819, + "learning_rate": 0.003, + "loss": 4.1811, + "step": 2594 + }, + { + "epoch": 0.02595, + "grad_norm": 0.7736456394195557, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 2595 + }, + { + "epoch": 0.02596, + "grad_norm": 0.7409265041351318, + "learning_rate": 0.003, + "loss": 4.1736, + "step": 2596 + }, + { + "epoch": 0.02597, + "grad_norm": 0.7836101651191711, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 2597 + }, + { + "epoch": 0.02598, + "grad_norm": 0.6943440437316895, + "learning_rate": 0.003, + "loss": 4.1827, + "step": 2598 + }, + { + "epoch": 0.02599, + "grad_norm": 0.6729668378829956, + "learning_rate": 0.003, + "loss": 4.1732, + "step": 2599 + }, + { + "epoch": 0.026, + "grad_norm": 0.7131978869438171, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 2600 + }, + { + "epoch": 0.02601, + "grad_norm": 0.7535547018051147, + "learning_rate": 0.003, + "loss": 4.1785, + "step": 2601 + }, + { + "epoch": 0.02602, + "grad_norm": 0.8210467100143433, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 2602 + }, + { + "epoch": 0.02603, + "grad_norm": 0.7161123752593994, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 2603 + }, + { + "epoch": 0.02604, + "grad_norm": 0.6834149956703186, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 2604 + }, + { + "epoch": 0.02605, + "grad_norm": 0.5697470307350159, + "learning_rate": 0.003, + "loss": 4.1503, + "step": 2605 + }, + { + "epoch": 0.02606, + "grad_norm": 0.5395636558532715, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 2606 + }, + { + "epoch": 0.02607, + "grad_norm": 0.47388994693756104, + "learning_rate": 0.003, + "loss": 4.157, + "step": 2607 + }, + { + "epoch": 0.02608, + "grad_norm": 0.435537189245224, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 2608 + }, + { + "epoch": 0.02609, + "grad_norm": 0.4369211792945862, + "learning_rate": 0.003, + "loss": 4.1788, + "step": 2609 + }, + { + "epoch": 0.0261, + "grad_norm": 0.46976083517074585, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 2610 + }, + { + "epoch": 0.02611, + "grad_norm": 0.5262687802314758, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 2611 + }, + { + "epoch": 0.02612, + "grad_norm": 0.7051580548286438, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 2612 + }, + { + "epoch": 0.02613, + "grad_norm": 0.8811540603637695, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 2613 + }, + { + "epoch": 0.02614, + "grad_norm": 0.8296162486076355, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 2614 + }, + { + "epoch": 0.02615, + "grad_norm": 0.5731111764907837, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 2615 + }, + { + "epoch": 0.02616, + "grad_norm": 0.6465540528297424, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 2616 + }, + { + "epoch": 0.02617, + "grad_norm": 0.7469308972358704, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 2617 + }, + { + "epoch": 0.02618, + "grad_norm": 0.7493590712547302, + "learning_rate": 0.003, + "loss": 4.1589, + "step": 2618 + }, + { + "epoch": 0.02619, + "grad_norm": 0.6884176731109619, + "learning_rate": 0.003, + "loss": 4.1704, + "step": 2619 + }, + { + "epoch": 0.0262, + "grad_norm": 0.6231204271316528, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 2620 + }, + { + "epoch": 0.02621, + "grad_norm": 0.5638678669929504, + "learning_rate": 0.003, + "loss": 4.1548, + "step": 2621 + }, + { + "epoch": 0.02622, + "grad_norm": 0.6251258254051208, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 2622 + }, + { + "epoch": 0.02623, + "grad_norm": 0.6526143550872803, + "learning_rate": 0.003, + "loss": 4.1604, + "step": 2623 + }, + { + "epoch": 0.02624, + "grad_norm": 0.7207937836647034, + "learning_rate": 0.003, + "loss": 4.1973, + "step": 2624 + }, + { + "epoch": 0.02625, + "grad_norm": 0.7291254997253418, + "learning_rate": 0.003, + "loss": 4.1951, + "step": 2625 + }, + { + "epoch": 0.02626, + "grad_norm": 0.6461899280548096, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 2626 + }, + { + "epoch": 0.02627, + "grad_norm": 0.7499544620513916, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 2627 + }, + { + "epoch": 0.02628, + "grad_norm": 0.7836819887161255, + "learning_rate": 0.003, + "loss": 4.1795, + "step": 2628 + }, + { + "epoch": 0.02629, + "grad_norm": 0.7647614479064941, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 2629 + }, + { + "epoch": 0.0263, + "grad_norm": 0.6021679639816284, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 2630 + }, + { + "epoch": 0.02631, + "grad_norm": 0.5774012804031372, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 2631 + }, + { + "epoch": 0.02632, + "grad_norm": 0.5787044167518616, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 2632 + }, + { + "epoch": 0.02633, + "grad_norm": 0.6322444677352905, + "learning_rate": 0.003, + "loss": 4.14, + "step": 2633 + }, + { + "epoch": 0.02634, + "grad_norm": 0.6454491019248962, + "learning_rate": 0.003, + "loss": 4.161, + "step": 2634 + }, + { + "epoch": 0.02635, + "grad_norm": 0.6432276368141174, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 2635 + }, + { + "epoch": 0.02636, + "grad_norm": 0.7024194002151489, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 2636 + }, + { + "epoch": 0.02637, + "grad_norm": 0.8076905012130737, + "learning_rate": 0.003, + "loss": 4.1774, + "step": 2637 + }, + { + "epoch": 0.02638, + "grad_norm": 0.7787435054779053, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 2638 + }, + { + "epoch": 0.02639, + "grad_norm": 0.6584726572036743, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 2639 + }, + { + "epoch": 0.0264, + "grad_norm": 0.6226917505264282, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 2640 + }, + { + "epoch": 0.02641, + "grad_norm": 0.5553937554359436, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 2641 + }, + { + "epoch": 0.02642, + "grad_norm": 0.5754583477973938, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 2642 + }, + { + "epoch": 0.02643, + "grad_norm": 0.6242693662643433, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 2643 + }, + { + "epoch": 0.02644, + "grad_norm": 0.6401628255844116, + "learning_rate": 0.003, + "loss": 4.1486, + "step": 2644 + }, + { + "epoch": 0.02645, + "grad_norm": 0.6735085844993591, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 2645 + }, + { + "epoch": 0.02646, + "grad_norm": 0.7502104640007019, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 2646 + }, + { + "epoch": 0.02647, + "grad_norm": 0.9201660752296448, + "learning_rate": 0.003, + "loss": 4.1609, + "step": 2647 + }, + { + "epoch": 0.02648, + "grad_norm": 0.8611843585968018, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 2648 + }, + { + "epoch": 0.02649, + "grad_norm": 0.6259444355964661, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 2649 + }, + { + "epoch": 0.0265, + "grad_norm": 0.6097885966300964, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 2650 + }, + { + "epoch": 0.02651, + "grad_norm": 0.696152925491333, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 2651 + }, + { + "epoch": 0.02652, + "grad_norm": 0.6412234902381897, + "learning_rate": 0.003, + "loss": 4.1679, + "step": 2652 + }, + { + "epoch": 0.02653, + "grad_norm": 0.6418108940124512, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 2653 + }, + { + "epoch": 0.02654, + "grad_norm": 0.5534683465957642, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 2654 + }, + { + "epoch": 0.02655, + "grad_norm": 0.5514299273490906, + "learning_rate": 0.003, + "loss": 4.1394, + "step": 2655 + }, + { + "epoch": 0.02656, + "grad_norm": 0.48489049077033997, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 2656 + }, + { + "epoch": 0.02657, + "grad_norm": 0.5331873297691345, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 2657 + }, + { + "epoch": 0.02658, + "grad_norm": 0.5732285380363464, + "learning_rate": 0.003, + "loss": 4.149, + "step": 2658 + }, + { + "epoch": 0.02659, + "grad_norm": 0.6610122323036194, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 2659 + }, + { + "epoch": 0.0266, + "grad_norm": 0.7037291526794434, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 2660 + }, + { + "epoch": 0.02661, + "grad_norm": 0.8676835894584656, + "learning_rate": 0.003, + "loss": 4.161, + "step": 2661 + }, + { + "epoch": 0.02662, + "grad_norm": 0.9771322011947632, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 2662 + }, + { + "epoch": 0.02663, + "grad_norm": 1.1501520872116089, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 2663 + }, + { + "epoch": 0.02664, + "grad_norm": 0.8324311971664429, + "learning_rate": 0.003, + "loss": 4.1441, + "step": 2664 + }, + { + "epoch": 0.02665, + "grad_norm": 0.852374255657196, + "learning_rate": 0.003, + "loss": 4.1931, + "step": 2665 + }, + { + "epoch": 0.02666, + "grad_norm": 1.0820196866989136, + "learning_rate": 0.003, + "loss": 4.198, + "step": 2666 + }, + { + "epoch": 0.02667, + "grad_norm": 0.7490083575248718, + "learning_rate": 0.003, + "loss": 4.1632, + "step": 2667 + }, + { + "epoch": 0.02668, + "grad_norm": 0.679391622543335, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 2668 + }, + { + "epoch": 0.02669, + "grad_norm": 0.6063788533210754, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 2669 + }, + { + "epoch": 0.0267, + "grad_norm": 0.6576395034790039, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 2670 + }, + { + "epoch": 0.02671, + "grad_norm": 0.8710551261901855, + "learning_rate": 0.003, + "loss": 4.1782, + "step": 2671 + }, + { + "epoch": 0.02672, + "grad_norm": 1.043360710144043, + "learning_rate": 0.003, + "loss": 4.179, + "step": 2672 + }, + { + "epoch": 0.02673, + "grad_norm": 0.8524636030197144, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 2673 + }, + { + "epoch": 0.02674, + "grad_norm": 0.7167258858680725, + "learning_rate": 0.003, + "loss": 4.135, + "step": 2674 + }, + { + "epoch": 0.02675, + "grad_norm": 0.8229854106903076, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 2675 + }, + { + "epoch": 0.02676, + "grad_norm": 0.8725418448448181, + "learning_rate": 0.003, + "loss": 4.1608, + "step": 2676 + }, + { + "epoch": 0.02677, + "grad_norm": 0.8751718997955322, + "learning_rate": 0.003, + "loss": 4.189, + "step": 2677 + }, + { + "epoch": 0.02678, + "grad_norm": 0.9639006853103638, + "learning_rate": 0.003, + "loss": 4.1897, + "step": 2678 + }, + { + "epoch": 0.02679, + "grad_norm": 0.9028424024581909, + "learning_rate": 0.003, + "loss": 4.1596, + "step": 2679 + }, + { + "epoch": 0.0268, + "grad_norm": 0.8419076204299927, + "learning_rate": 0.003, + "loss": 4.1719, + "step": 2680 + }, + { + "epoch": 0.02681, + "grad_norm": 0.8327796459197998, + "learning_rate": 0.003, + "loss": 4.1966, + "step": 2681 + }, + { + "epoch": 0.02682, + "grad_norm": 0.6412435173988342, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 2682 + }, + { + "epoch": 0.02683, + "grad_norm": 0.5319331288337708, + "learning_rate": 0.003, + "loss": 4.1829, + "step": 2683 + }, + { + "epoch": 0.02684, + "grad_norm": 0.5018413662910461, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 2684 + }, + { + "epoch": 0.02685, + "grad_norm": 0.4933565557003021, + "learning_rate": 0.003, + "loss": 4.1592, + "step": 2685 + }, + { + "epoch": 0.02686, + "grad_norm": 0.5294405221939087, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 2686 + }, + { + "epoch": 0.02687, + "grad_norm": 0.6580759882926941, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 2687 + }, + { + "epoch": 0.02688, + "grad_norm": 0.7823148369789124, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 2688 + }, + { + "epoch": 0.02689, + "grad_norm": 0.8333227038383484, + "learning_rate": 0.003, + "loss": 4.1677, + "step": 2689 + }, + { + "epoch": 0.0269, + "grad_norm": 0.6743582487106323, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 2690 + }, + { + "epoch": 0.02691, + "grad_norm": 0.5275107622146606, + "learning_rate": 0.003, + "loss": 4.154, + "step": 2691 + }, + { + "epoch": 0.02692, + "grad_norm": 0.585838258266449, + "learning_rate": 0.003, + "loss": 4.1589, + "step": 2692 + }, + { + "epoch": 0.02693, + "grad_norm": 0.6433229446411133, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 2693 + }, + { + "epoch": 0.02694, + "grad_norm": 0.5260990858078003, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 2694 + }, + { + "epoch": 0.02695, + "grad_norm": 0.49656516313552856, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 2695 + }, + { + "epoch": 0.02696, + "grad_norm": 0.46913599967956543, + "learning_rate": 0.003, + "loss": 4.1441, + "step": 2696 + }, + { + "epoch": 0.02697, + "grad_norm": 0.4466596841812134, + "learning_rate": 0.003, + "loss": 4.1534, + "step": 2697 + }, + { + "epoch": 0.02698, + "grad_norm": 0.4291961193084717, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 2698 + }, + { + "epoch": 0.02699, + "grad_norm": 0.43313321471214294, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 2699 + }, + { + "epoch": 0.027, + "grad_norm": 0.4425189793109894, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 2700 + }, + { + "epoch": 0.02701, + "grad_norm": 0.41129767894744873, + "learning_rate": 0.003, + "loss": 4.112, + "step": 2701 + }, + { + "epoch": 0.02702, + "grad_norm": 0.3028920888900757, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 2702 + }, + { + "epoch": 0.02703, + "grad_norm": 0.36113715171813965, + "learning_rate": 0.003, + "loss": 4.1575, + "step": 2703 + }, + { + "epoch": 0.02704, + "grad_norm": 0.42614156007766724, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 2704 + }, + { + "epoch": 0.02705, + "grad_norm": 0.4877123534679413, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 2705 + }, + { + "epoch": 0.02706, + "grad_norm": 0.6430928707122803, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 2706 + }, + { + "epoch": 0.02707, + "grad_norm": 0.7538681626319885, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 2707 + }, + { + "epoch": 0.02708, + "grad_norm": 0.9270827770233154, + "learning_rate": 0.003, + "loss": 4.1612, + "step": 2708 + }, + { + "epoch": 0.02709, + "grad_norm": 1.0371726751327515, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 2709 + }, + { + "epoch": 0.0271, + "grad_norm": 0.7252341508865356, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 2710 + }, + { + "epoch": 0.02711, + "grad_norm": 0.6913469433784485, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 2711 + }, + { + "epoch": 0.02712, + "grad_norm": 0.7029024362564087, + "learning_rate": 0.003, + "loss": 4.142, + "step": 2712 + }, + { + "epoch": 0.02713, + "grad_norm": 0.6771293878555298, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 2713 + }, + { + "epoch": 0.02714, + "grad_norm": 0.6258983016014099, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 2714 + }, + { + "epoch": 0.02715, + "grad_norm": 0.7235097289085388, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 2715 + }, + { + "epoch": 0.02716, + "grad_norm": 0.8248934745788574, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 2716 + }, + { + "epoch": 0.02717, + "grad_norm": 0.8002756237983704, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 2717 + }, + { + "epoch": 0.02718, + "grad_norm": 0.7360194325447083, + "learning_rate": 0.003, + "loss": 4.1647, + "step": 2718 + }, + { + "epoch": 0.02719, + "grad_norm": 0.8694934844970703, + "learning_rate": 0.003, + "loss": 4.145, + "step": 2719 + }, + { + "epoch": 0.0272, + "grad_norm": 0.8528920412063599, + "learning_rate": 0.003, + "loss": 4.137, + "step": 2720 + }, + { + "epoch": 0.02721, + "grad_norm": 1.0115561485290527, + "learning_rate": 0.003, + "loss": 4.1594, + "step": 2721 + }, + { + "epoch": 0.02722, + "grad_norm": 1.050325632095337, + "learning_rate": 0.003, + "loss": 4.1686, + "step": 2722 + }, + { + "epoch": 0.02723, + "grad_norm": 0.8528364896774292, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 2723 + }, + { + "epoch": 0.02724, + "grad_norm": 0.8148915767669678, + "learning_rate": 0.003, + "loss": 4.1665, + "step": 2724 + }, + { + "epoch": 0.02725, + "grad_norm": 0.707343339920044, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 2725 + }, + { + "epoch": 0.02726, + "grad_norm": 0.7041299939155579, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 2726 + }, + { + "epoch": 0.02727, + "grad_norm": 0.5910518765449524, + "learning_rate": 0.003, + "loss": 4.1551, + "step": 2727 + }, + { + "epoch": 0.02728, + "grad_norm": 0.6046881079673767, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 2728 + }, + { + "epoch": 0.02729, + "grad_norm": 0.5683781504631042, + "learning_rate": 0.003, + "loss": 4.1716, + "step": 2729 + }, + { + "epoch": 0.0273, + "grad_norm": 0.5593625903129578, + "learning_rate": 0.003, + "loss": 4.143, + "step": 2730 + }, + { + "epoch": 0.02731, + "grad_norm": 0.5543289184570312, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 2731 + }, + { + "epoch": 0.02732, + "grad_norm": 0.5714956521987915, + "learning_rate": 0.003, + "loss": 4.135, + "step": 2732 + }, + { + "epoch": 0.02733, + "grad_norm": 0.5906261205673218, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 2733 + }, + { + "epoch": 0.02734, + "grad_norm": 0.7874712944030762, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 2734 + }, + { + "epoch": 0.02735, + "grad_norm": 0.9163006544113159, + "learning_rate": 0.003, + "loss": 4.174, + "step": 2735 + }, + { + "epoch": 0.02736, + "grad_norm": 0.8792824745178223, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 2736 + }, + { + "epoch": 0.02737, + "grad_norm": 0.7901833057403564, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 2737 + }, + { + "epoch": 0.02738, + "grad_norm": 0.8548569083213806, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 2738 + }, + { + "epoch": 0.02739, + "grad_norm": 0.7134705185890198, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 2739 + }, + { + "epoch": 0.0274, + "grad_norm": 0.6430677175521851, + "learning_rate": 0.003, + "loss": 4.139, + "step": 2740 + }, + { + "epoch": 0.02741, + "grad_norm": 0.5635836720466614, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 2741 + }, + { + "epoch": 0.02742, + "grad_norm": 0.5489112138748169, + "learning_rate": 0.003, + "loss": 4.1367, + "step": 2742 + }, + { + "epoch": 0.02743, + "grad_norm": 0.5366271734237671, + "learning_rate": 0.003, + "loss": 4.1572, + "step": 2743 + }, + { + "epoch": 0.02744, + "grad_norm": 0.489491730928421, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 2744 + }, + { + "epoch": 0.02745, + "grad_norm": 0.47551363706588745, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 2745 + }, + { + "epoch": 0.02746, + "grad_norm": 0.39822816848754883, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 2746 + }, + { + "epoch": 0.02747, + "grad_norm": 0.3974153399467468, + "learning_rate": 0.003, + "loss": 4.1538, + "step": 2747 + }, + { + "epoch": 0.02748, + "grad_norm": 0.47189366817474365, + "learning_rate": 0.003, + "loss": 4.107, + "step": 2748 + }, + { + "epoch": 0.02749, + "grad_norm": 0.5153442025184631, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 2749 + }, + { + "epoch": 0.0275, + "grad_norm": 0.565412700176239, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 2750 + }, + { + "epoch": 0.02751, + "grad_norm": 0.6354933977127075, + "learning_rate": 0.003, + "loss": 4.1638, + "step": 2751 + }, + { + "epoch": 0.02752, + "grad_norm": 0.6897895932197571, + "learning_rate": 0.003, + "loss": 4.1494, + "step": 2752 + }, + { + "epoch": 0.02753, + "grad_norm": 0.6377696394920349, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 2753 + }, + { + "epoch": 0.02754, + "grad_norm": 0.5976959466934204, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 2754 + }, + { + "epoch": 0.02755, + "grad_norm": 0.702274739742279, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 2755 + }, + { + "epoch": 0.02756, + "grad_norm": 0.7118279337882996, + "learning_rate": 0.003, + "loss": 4.1439, + "step": 2756 + }, + { + "epoch": 0.02757, + "grad_norm": 0.660001277923584, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 2757 + }, + { + "epoch": 0.02758, + "grad_norm": 0.7159073948860168, + "learning_rate": 0.003, + "loss": 4.106, + "step": 2758 + }, + { + "epoch": 0.02759, + "grad_norm": 0.8859509825706482, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 2759 + }, + { + "epoch": 0.0276, + "grad_norm": 0.9749924540519714, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 2760 + }, + { + "epoch": 0.02761, + "grad_norm": 0.9062779545783997, + "learning_rate": 0.003, + "loss": 4.1742, + "step": 2761 + }, + { + "epoch": 0.02762, + "grad_norm": 0.8129538893699646, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 2762 + }, + { + "epoch": 0.02763, + "grad_norm": 0.7007825970649719, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 2763 + }, + { + "epoch": 0.02764, + "grad_norm": 0.8174701929092407, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 2764 + }, + { + "epoch": 0.02765, + "grad_norm": 0.8394188284873962, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 2765 + }, + { + "epoch": 0.02766, + "grad_norm": 0.8616805672645569, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 2766 + }, + { + "epoch": 0.02767, + "grad_norm": 0.8441250920295715, + "learning_rate": 0.003, + "loss": 4.1869, + "step": 2767 + }, + { + "epoch": 0.02768, + "grad_norm": 0.7080286741256714, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 2768 + }, + { + "epoch": 0.02769, + "grad_norm": 0.708198070526123, + "learning_rate": 0.003, + "loss": 4.174, + "step": 2769 + }, + { + "epoch": 0.0277, + "grad_norm": 0.7906506061553955, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 2770 + }, + { + "epoch": 0.02771, + "grad_norm": 0.7239770889282227, + "learning_rate": 0.003, + "loss": 4.1314, + "step": 2771 + }, + { + "epoch": 0.02772, + "grad_norm": 0.5857943296432495, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 2772 + }, + { + "epoch": 0.02773, + "grad_norm": 0.5903287529945374, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 2773 + }, + { + "epoch": 0.02774, + "grad_norm": 0.6580007672309875, + "learning_rate": 0.003, + "loss": 4.176, + "step": 2774 + }, + { + "epoch": 0.02775, + "grad_norm": 0.6529462933540344, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 2775 + }, + { + "epoch": 0.02776, + "grad_norm": 0.7036746144294739, + "learning_rate": 0.003, + "loss": 4.138, + "step": 2776 + }, + { + "epoch": 0.02777, + "grad_norm": 0.7649575471878052, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 2777 + }, + { + "epoch": 0.02778, + "grad_norm": 0.6718697547912598, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 2778 + }, + { + "epoch": 0.02779, + "grad_norm": 0.587509036064148, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 2779 + }, + { + "epoch": 0.0278, + "grad_norm": 0.5817381143569946, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 2780 + }, + { + "epoch": 0.02781, + "grad_norm": 0.5770044326782227, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 2781 + }, + { + "epoch": 0.02782, + "grad_norm": 0.687736451625824, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 2782 + }, + { + "epoch": 0.02783, + "grad_norm": 0.6473103165626526, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 2783 + }, + { + "epoch": 0.02784, + "grad_norm": 0.5239824056625366, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 2784 + }, + { + "epoch": 0.02785, + "grad_norm": 0.4588249921798706, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 2785 + }, + { + "epoch": 0.02786, + "grad_norm": 0.4880278408527374, + "learning_rate": 0.003, + "loss": 4.1394, + "step": 2786 + }, + { + "epoch": 0.02787, + "grad_norm": 0.502045750617981, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 2787 + }, + { + "epoch": 0.02788, + "grad_norm": 0.6067155003547668, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 2788 + }, + { + "epoch": 0.02789, + "grad_norm": 0.7814184427261353, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 2789 + }, + { + "epoch": 0.0279, + "grad_norm": 1.1954095363616943, + "learning_rate": 0.003, + "loss": 4.132, + "step": 2790 + }, + { + "epoch": 0.02791, + "grad_norm": 0.7128247022628784, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 2791 + }, + { + "epoch": 0.02792, + "grad_norm": 0.6396022439002991, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 2792 + }, + { + "epoch": 0.02793, + "grad_norm": 0.8452949523925781, + "learning_rate": 0.003, + "loss": 4.1513, + "step": 2793 + }, + { + "epoch": 0.02794, + "grad_norm": 0.7837446928024292, + "learning_rate": 0.003, + "loss": 4.1464, + "step": 2794 + }, + { + "epoch": 0.02795, + "grad_norm": 0.7633419036865234, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 2795 + }, + { + "epoch": 0.02796, + "grad_norm": 0.7906386852264404, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 2796 + }, + { + "epoch": 0.02797, + "grad_norm": 0.7294182777404785, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 2797 + }, + { + "epoch": 0.02798, + "grad_norm": 0.7420005798339844, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 2798 + }, + { + "epoch": 0.02799, + "grad_norm": 0.722224235534668, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 2799 + }, + { + "epoch": 0.028, + "grad_norm": 0.677977442741394, + "learning_rate": 0.003, + "loss": 4.1596, + "step": 2800 + }, + { + "epoch": 0.02801, + "grad_norm": 0.6493065357208252, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 2801 + }, + { + "epoch": 0.02802, + "grad_norm": 0.6755994558334351, + "learning_rate": 0.003, + "loss": 4.1363, + "step": 2802 + }, + { + "epoch": 0.02803, + "grad_norm": 0.6203683614730835, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 2803 + }, + { + "epoch": 0.02804, + "grad_norm": 0.5352902412414551, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 2804 + }, + { + "epoch": 0.02805, + "grad_norm": 0.4998801350593567, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 2805 + }, + { + "epoch": 0.02806, + "grad_norm": 0.43145909905433655, + "learning_rate": 0.003, + "loss": 4.13, + "step": 2806 + }, + { + "epoch": 0.02807, + "grad_norm": 0.4628302752971649, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 2807 + }, + { + "epoch": 0.02808, + "grad_norm": 0.4608153998851776, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 2808 + }, + { + "epoch": 0.02809, + "grad_norm": 0.47452715039253235, + "learning_rate": 0.003, + "loss": 4.1551, + "step": 2809 + }, + { + "epoch": 0.0281, + "grad_norm": 0.5173685550689697, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 2810 + }, + { + "epoch": 0.02811, + "grad_norm": 0.5137503147125244, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 2811 + }, + { + "epoch": 0.02812, + "grad_norm": 0.5507237315177917, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 2812 + }, + { + "epoch": 0.02813, + "grad_norm": 0.667034387588501, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 2813 + }, + { + "epoch": 0.02814, + "grad_norm": 0.8353255987167358, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 2814 + }, + { + "epoch": 0.02815, + "grad_norm": 0.8815486431121826, + "learning_rate": 0.003, + "loss": 4.1736, + "step": 2815 + }, + { + "epoch": 0.02816, + "grad_norm": 0.7531535029411316, + "learning_rate": 0.003, + "loss": 4.1477, + "step": 2816 + }, + { + "epoch": 0.02817, + "grad_norm": 0.6786168217658997, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 2817 + }, + { + "epoch": 0.02818, + "grad_norm": 0.5562347173690796, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 2818 + }, + { + "epoch": 0.02819, + "grad_norm": 0.6471225619316101, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 2819 + }, + { + "epoch": 0.0282, + "grad_norm": 0.7505398392677307, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 2820 + }, + { + "epoch": 0.02821, + "grad_norm": 0.845917820930481, + "learning_rate": 0.003, + "loss": 4.1632, + "step": 2821 + }, + { + "epoch": 0.02822, + "grad_norm": 0.8887820839881897, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 2822 + }, + { + "epoch": 0.02823, + "grad_norm": 0.953726589679718, + "learning_rate": 0.003, + "loss": 4.1686, + "step": 2823 + }, + { + "epoch": 0.02824, + "grad_norm": 0.9745505452156067, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 2824 + }, + { + "epoch": 0.02825, + "grad_norm": 0.9194844961166382, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 2825 + }, + { + "epoch": 0.02826, + "grad_norm": 0.9426382184028625, + "learning_rate": 0.003, + "loss": 4.1439, + "step": 2826 + }, + { + "epoch": 0.02827, + "grad_norm": 0.8329664468765259, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 2827 + }, + { + "epoch": 0.02828, + "grad_norm": 0.7950755953788757, + "learning_rate": 0.003, + "loss": 4.1685, + "step": 2828 + }, + { + "epoch": 0.02829, + "grad_norm": 0.7973052263259888, + "learning_rate": 0.003, + "loss": 4.1943, + "step": 2829 + }, + { + "epoch": 0.0283, + "grad_norm": 0.7938761115074158, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 2830 + }, + { + "epoch": 0.02831, + "grad_norm": 0.900173008441925, + "learning_rate": 0.003, + "loss": 4.1604, + "step": 2831 + }, + { + "epoch": 0.02832, + "grad_norm": 0.9947631359100342, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 2832 + }, + { + "epoch": 0.02833, + "grad_norm": 0.8765170574188232, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 2833 + }, + { + "epoch": 0.02834, + "grad_norm": 0.8053457140922546, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 2834 + }, + { + "epoch": 0.02835, + "grad_norm": 0.728701651096344, + "learning_rate": 0.003, + "loss": 4.147, + "step": 2835 + }, + { + "epoch": 0.02836, + "grad_norm": 0.6180861592292786, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 2836 + }, + { + "epoch": 0.02837, + "grad_norm": 0.5929837226867676, + "learning_rate": 0.003, + "loss": 4.1342, + "step": 2837 + }, + { + "epoch": 0.02838, + "grad_norm": 0.5861077308654785, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 2838 + }, + { + "epoch": 0.02839, + "grad_norm": 0.5293443202972412, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 2839 + }, + { + "epoch": 0.0284, + "grad_norm": 0.5219327211380005, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 2840 + }, + { + "epoch": 0.02841, + "grad_norm": 0.494223028421402, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 2841 + }, + { + "epoch": 0.02842, + "grad_norm": 0.5202710628509521, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 2842 + }, + { + "epoch": 0.02843, + "grad_norm": 0.6077046990394592, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 2843 + }, + { + "epoch": 0.02844, + "grad_norm": 0.6248916983604431, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 2844 + }, + { + "epoch": 0.02845, + "grad_norm": 0.5447239875793457, + "learning_rate": 0.003, + "loss": 4.124, + "step": 2845 + }, + { + "epoch": 0.02846, + "grad_norm": 0.5102211236953735, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 2846 + }, + { + "epoch": 0.02847, + "grad_norm": 0.47447827458381653, + "learning_rate": 0.003, + "loss": 4.171, + "step": 2847 + }, + { + "epoch": 0.02848, + "grad_norm": 0.5006018877029419, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 2848 + }, + { + "epoch": 0.02849, + "grad_norm": 0.6120002865791321, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 2849 + }, + { + "epoch": 0.0285, + "grad_norm": 0.6846702098846436, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 2850 + }, + { + "epoch": 0.02851, + "grad_norm": 0.7738217115402222, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 2851 + }, + { + "epoch": 0.02852, + "grad_norm": 0.8055019974708557, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 2852 + }, + { + "epoch": 0.02853, + "grad_norm": 0.8890880942344666, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 2853 + }, + { + "epoch": 0.02854, + "grad_norm": 0.8885093927383423, + "learning_rate": 0.003, + "loss": 4.1639, + "step": 2854 + }, + { + "epoch": 0.02855, + "grad_norm": 0.8279180526733398, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 2855 + }, + { + "epoch": 0.02856, + "grad_norm": 0.7752195596694946, + "learning_rate": 0.003, + "loss": 4.1614, + "step": 2856 + }, + { + "epoch": 0.02857, + "grad_norm": 0.8694232702255249, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 2857 + }, + { + "epoch": 0.02858, + "grad_norm": 0.9445026516914368, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 2858 + }, + { + "epoch": 0.02859, + "grad_norm": 0.8522076606750488, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 2859 + }, + { + "epoch": 0.0286, + "grad_norm": 0.8931873440742493, + "learning_rate": 0.003, + "loss": 4.1628, + "step": 2860 + }, + { + "epoch": 0.02861, + "grad_norm": 0.8646109700202942, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 2861 + }, + { + "epoch": 0.02862, + "grad_norm": 0.7844035625457764, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 2862 + }, + { + "epoch": 0.02863, + "grad_norm": 0.7147074341773987, + "learning_rate": 0.003, + "loss": 4.169, + "step": 2863 + }, + { + "epoch": 0.02864, + "grad_norm": 0.7001998424530029, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 2864 + }, + { + "epoch": 0.02865, + "grad_norm": 0.78515625, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 2865 + }, + { + "epoch": 0.02866, + "grad_norm": 0.6479020714759827, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 2866 + }, + { + "epoch": 0.02867, + "grad_norm": 0.720973014831543, + "learning_rate": 0.003, + "loss": 4.1431, + "step": 2867 + }, + { + "epoch": 0.02868, + "grad_norm": 0.6694818735122681, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 2868 + }, + { + "epoch": 0.02869, + "grad_norm": 0.5998570919036865, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 2869 + }, + { + "epoch": 0.0287, + "grad_norm": 0.5683268308639526, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 2870 + }, + { + "epoch": 0.02871, + "grad_norm": 0.5723297595977783, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 2871 + }, + { + "epoch": 0.02872, + "grad_norm": 0.5388701558113098, + "learning_rate": 0.003, + "loss": 4.1708, + "step": 2872 + }, + { + "epoch": 0.02873, + "grad_norm": 0.46427324414253235, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 2873 + }, + { + "epoch": 0.02874, + "grad_norm": 0.4973464012145996, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 2874 + }, + { + "epoch": 0.02875, + "grad_norm": 0.5175668001174927, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 2875 + }, + { + "epoch": 0.02876, + "grad_norm": 0.5334407091140747, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 2876 + }, + { + "epoch": 0.02877, + "grad_norm": 0.5334872007369995, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 2877 + }, + { + "epoch": 0.02878, + "grad_norm": 0.5646567344665527, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 2878 + }, + { + "epoch": 0.02879, + "grad_norm": 0.560286819934845, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 2879 + }, + { + "epoch": 0.0288, + "grad_norm": 0.6549745202064514, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 2880 + }, + { + "epoch": 0.02881, + "grad_norm": 0.8416467905044556, + "learning_rate": 0.003, + "loss": 4.1365, + "step": 2881 + }, + { + "epoch": 0.02882, + "grad_norm": 0.7693246603012085, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 2882 + }, + { + "epoch": 0.02883, + "grad_norm": 0.6049852967262268, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 2883 + }, + { + "epoch": 0.02884, + "grad_norm": 0.561359703540802, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 2884 + }, + { + "epoch": 0.02885, + "grad_norm": 0.630006730556488, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 2885 + }, + { + "epoch": 0.02886, + "grad_norm": 0.6453554630279541, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 2886 + }, + { + "epoch": 0.02887, + "grad_norm": 0.5455800890922546, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 2887 + }, + { + "epoch": 0.02888, + "grad_norm": 0.4928850531578064, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 2888 + }, + { + "epoch": 0.02889, + "grad_norm": 0.4772005081176758, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 2889 + }, + { + "epoch": 0.0289, + "grad_norm": 0.6117534041404724, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 2890 + }, + { + "epoch": 0.02891, + "grad_norm": 0.8205873370170593, + "learning_rate": 0.003, + "loss": 4.1647, + "step": 2891 + }, + { + "epoch": 0.02892, + "grad_norm": 0.8640782833099365, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 2892 + }, + { + "epoch": 0.02893, + "grad_norm": 0.8006044626235962, + "learning_rate": 0.003, + "loss": 4.13, + "step": 2893 + }, + { + "epoch": 0.02894, + "grad_norm": 0.8116544485092163, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 2894 + }, + { + "epoch": 0.02895, + "grad_norm": 0.8600641489028931, + "learning_rate": 0.003, + "loss": 4.1864, + "step": 2895 + }, + { + "epoch": 0.02896, + "grad_norm": 0.6702893376350403, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 2896 + }, + { + "epoch": 0.02897, + "grad_norm": 0.6657308340072632, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 2897 + }, + { + "epoch": 0.02898, + "grad_norm": 0.662657618522644, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 2898 + }, + { + "epoch": 0.02899, + "grad_norm": 0.7254528999328613, + "learning_rate": 0.003, + "loss": 4.148, + "step": 2899 + }, + { + "epoch": 0.029, + "grad_norm": 0.7490177750587463, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 2900 + }, + { + "epoch": 0.02901, + "grad_norm": 0.7398015856742859, + "learning_rate": 0.003, + "loss": 4.1513, + "step": 2901 + }, + { + "epoch": 0.02902, + "grad_norm": 0.8604675531387329, + "learning_rate": 0.003, + "loss": 4.1725, + "step": 2902 + }, + { + "epoch": 0.02903, + "grad_norm": 0.9788482785224915, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 2903 + }, + { + "epoch": 0.02904, + "grad_norm": 0.9012083411216736, + "learning_rate": 0.003, + "loss": 4.1596, + "step": 2904 + }, + { + "epoch": 0.02905, + "grad_norm": 0.6638503074645996, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 2905 + }, + { + "epoch": 0.02906, + "grad_norm": 0.640064001083374, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 2906 + }, + { + "epoch": 0.02907, + "grad_norm": 0.6573886871337891, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 2907 + }, + { + "epoch": 0.02908, + "grad_norm": 0.6549690961837769, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 2908 + }, + { + "epoch": 0.02909, + "grad_norm": 0.6671884059906006, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 2909 + }, + { + "epoch": 0.0291, + "grad_norm": 0.6334506273269653, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 2910 + }, + { + "epoch": 0.02911, + "grad_norm": 0.6618163585662842, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 2911 + }, + { + "epoch": 0.02912, + "grad_norm": 0.7299465537071228, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 2912 + }, + { + "epoch": 0.02913, + "grad_norm": 0.7096904516220093, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 2913 + }, + { + "epoch": 0.02914, + "grad_norm": 0.7228156924247742, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 2914 + }, + { + "epoch": 0.02915, + "grad_norm": 0.6965159177780151, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 2915 + }, + { + "epoch": 0.02916, + "grad_norm": 0.5864951014518738, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 2916 + }, + { + "epoch": 0.02917, + "grad_norm": 0.682627260684967, + "learning_rate": 0.003, + "loss": 4.1647, + "step": 2917 + }, + { + "epoch": 0.02918, + "grad_norm": 0.7035491466522217, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 2918 + }, + { + "epoch": 0.02919, + "grad_norm": 0.6258479952812195, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 2919 + }, + { + "epoch": 0.0292, + "grad_norm": 0.7004368305206299, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 2920 + }, + { + "epoch": 0.02921, + "grad_norm": 0.7259970903396606, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 2921 + }, + { + "epoch": 0.02922, + "grad_norm": 0.8203588724136353, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 2922 + }, + { + "epoch": 0.02923, + "grad_norm": 0.7598622441291809, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 2923 + }, + { + "epoch": 0.02924, + "grad_norm": 0.713235080242157, + "learning_rate": 0.003, + "loss": 4.1518, + "step": 2924 + }, + { + "epoch": 0.02925, + "grad_norm": 0.6576932668685913, + "learning_rate": 0.003, + "loss": 4.1548, + "step": 2925 + }, + { + "epoch": 0.02926, + "grad_norm": 0.5912261009216309, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 2926 + }, + { + "epoch": 0.02927, + "grad_norm": 0.6221820712089539, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 2927 + }, + { + "epoch": 0.02928, + "grad_norm": 0.5222737193107605, + "learning_rate": 0.003, + "loss": 4.1557, + "step": 2928 + }, + { + "epoch": 0.02929, + "grad_norm": 0.448812872171402, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 2929 + }, + { + "epoch": 0.0293, + "grad_norm": 0.4934072196483612, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 2930 + }, + { + "epoch": 0.02931, + "grad_norm": 0.6587796211242676, + "learning_rate": 0.003, + "loss": 4.136, + "step": 2931 + }, + { + "epoch": 0.02932, + "grad_norm": 0.8658855557441711, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 2932 + }, + { + "epoch": 0.02933, + "grad_norm": 1.1174193620681763, + "learning_rate": 0.003, + "loss": 4.1685, + "step": 2933 + }, + { + "epoch": 0.02934, + "grad_norm": 0.7016493678092957, + "learning_rate": 0.003, + "loss": 4.145, + "step": 2934 + }, + { + "epoch": 0.02935, + "grad_norm": 0.543066680431366, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 2935 + }, + { + "epoch": 0.02936, + "grad_norm": 0.6172811388969421, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 2936 + }, + { + "epoch": 0.02937, + "grad_norm": 0.7183346152305603, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 2937 + }, + { + "epoch": 0.02938, + "grad_norm": 0.7085850834846497, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 2938 + }, + { + "epoch": 0.02939, + "grad_norm": 0.5843592882156372, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 2939 + }, + { + "epoch": 0.0294, + "grad_norm": 0.5836790800094604, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 2940 + }, + { + "epoch": 0.02941, + "grad_norm": 0.6397603154182434, + "learning_rate": 0.003, + "loss": 4.1546, + "step": 2941 + }, + { + "epoch": 0.02942, + "grad_norm": 0.6955589056015015, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 2942 + }, + { + "epoch": 0.02943, + "grad_norm": 0.754485011100769, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 2943 + }, + { + "epoch": 0.02944, + "grad_norm": 0.8499577045440674, + "learning_rate": 0.003, + "loss": 4.1566, + "step": 2944 + }, + { + "epoch": 0.02945, + "grad_norm": 0.8511131405830383, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 2945 + }, + { + "epoch": 0.02946, + "grad_norm": 0.7081684470176697, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 2946 + }, + { + "epoch": 0.02947, + "grad_norm": 0.6551468968391418, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 2947 + }, + { + "epoch": 0.02948, + "grad_norm": 0.6386260390281677, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 2948 + }, + { + "epoch": 0.02949, + "grad_norm": 0.5764749050140381, + "learning_rate": 0.003, + "loss": 4.113, + "step": 2949 + }, + { + "epoch": 0.0295, + "grad_norm": 0.5615102648735046, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 2950 + }, + { + "epoch": 0.02951, + "grad_norm": 0.5903642773628235, + "learning_rate": 0.003, + "loss": 4.088, + "step": 2951 + }, + { + "epoch": 0.02952, + "grad_norm": 0.6850778460502625, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 2952 + }, + { + "epoch": 0.02953, + "grad_norm": 0.6652215123176575, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 2953 + }, + { + "epoch": 0.02954, + "grad_norm": 0.6463000178337097, + "learning_rate": 0.003, + "loss": 4.1725, + "step": 2954 + }, + { + "epoch": 0.02955, + "grad_norm": 0.7008712291717529, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 2955 + }, + { + "epoch": 0.02956, + "grad_norm": 0.7221593260765076, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 2956 + }, + { + "epoch": 0.02957, + "grad_norm": 0.8538970351219177, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 2957 + }, + { + "epoch": 0.02958, + "grad_norm": 0.9920998811721802, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 2958 + }, + { + "epoch": 0.02959, + "grad_norm": 1.0068269968032837, + "learning_rate": 0.003, + "loss": 4.1654, + "step": 2959 + }, + { + "epoch": 0.0296, + "grad_norm": 0.9310505986213684, + "learning_rate": 0.003, + "loss": 4.1729, + "step": 2960 + }, + { + "epoch": 0.02961, + "grad_norm": 0.878132164478302, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 2961 + }, + { + "epoch": 0.02962, + "grad_norm": 0.953723132610321, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 2962 + }, + { + "epoch": 0.02963, + "grad_norm": 0.9499011039733887, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 2963 + }, + { + "epoch": 0.02964, + "grad_norm": 0.8707635402679443, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 2964 + }, + { + "epoch": 0.02965, + "grad_norm": 0.8199546337127686, + "learning_rate": 0.003, + "loss": 4.1628, + "step": 2965 + }, + { + "epoch": 0.02966, + "grad_norm": 0.7343395948410034, + "learning_rate": 0.003, + "loss": 4.1728, + "step": 2966 + }, + { + "epoch": 0.02967, + "grad_norm": 0.6565076112747192, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 2967 + }, + { + "epoch": 0.02968, + "grad_norm": 0.6714766025543213, + "learning_rate": 0.003, + "loss": 4.1518, + "step": 2968 + }, + { + "epoch": 0.02969, + "grad_norm": 0.6478135585784912, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 2969 + }, + { + "epoch": 0.0297, + "grad_norm": 0.5846438407897949, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 2970 + }, + { + "epoch": 0.02971, + "grad_norm": 0.5542798042297363, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 2971 + }, + { + "epoch": 0.02972, + "grad_norm": 0.5505024790763855, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 2972 + }, + { + "epoch": 0.02973, + "grad_norm": 0.5356234908103943, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 2973 + }, + { + "epoch": 0.02974, + "grad_norm": 0.6036863327026367, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 2974 + }, + { + "epoch": 0.02975, + "grad_norm": 0.48086002469062805, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 2975 + }, + { + "epoch": 0.02976, + "grad_norm": 0.43674689531326294, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 2976 + }, + { + "epoch": 0.02977, + "grad_norm": 0.4434605836868286, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 2977 + }, + { + "epoch": 0.02978, + "grad_norm": 0.5155782103538513, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 2978 + }, + { + "epoch": 0.02979, + "grad_norm": 0.6957648396492004, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 2979 + }, + { + "epoch": 0.0298, + "grad_norm": 0.8462650179862976, + "learning_rate": 0.003, + "loss": 4.139, + "step": 2980 + }, + { + "epoch": 0.02981, + "grad_norm": 0.7909703254699707, + "learning_rate": 0.003, + "loss": 4.146, + "step": 2981 + }, + { + "epoch": 0.02982, + "grad_norm": 0.5594244003295898, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 2982 + }, + { + "epoch": 0.02983, + "grad_norm": 0.5850778222084045, + "learning_rate": 0.003, + "loss": 4.1518, + "step": 2983 + }, + { + "epoch": 0.02984, + "grad_norm": 0.6843441128730774, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 2984 + }, + { + "epoch": 0.02985, + "grad_norm": 0.6832881569862366, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 2985 + }, + { + "epoch": 0.02986, + "grad_norm": 0.5421028733253479, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 2986 + }, + { + "epoch": 0.02987, + "grad_norm": 0.551857590675354, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 2987 + }, + { + "epoch": 0.02988, + "grad_norm": 0.6235337853431702, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 2988 + }, + { + "epoch": 0.02989, + "grad_norm": 0.7048717737197876, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 2989 + }, + { + "epoch": 0.0299, + "grad_norm": 0.7494181990623474, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 2990 + }, + { + "epoch": 0.02991, + "grad_norm": 0.7830327749252319, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 2991 + }, + { + "epoch": 0.02992, + "grad_norm": 0.8855887651443481, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 2992 + }, + { + "epoch": 0.02993, + "grad_norm": 0.888844907283783, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 2993 + }, + { + "epoch": 0.02994, + "grad_norm": 0.822543203830719, + "learning_rate": 0.003, + "loss": 4.1363, + "step": 2994 + }, + { + "epoch": 0.02995, + "grad_norm": 0.8102987408638, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 2995 + }, + { + "epoch": 0.02996, + "grad_norm": 0.8138471841812134, + "learning_rate": 0.003, + "loss": 4.1624, + "step": 2996 + }, + { + "epoch": 0.02997, + "grad_norm": 0.7990952134132385, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 2997 + }, + { + "epoch": 0.02998, + "grad_norm": 0.7419268488883972, + "learning_rate": 0.003, + "loss": 4.124, + "step": 2998 + }, + { + "epoch": 0.02999, + "grad_norm": 0.6637421250343323, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 2999 + }, + { + "epoch": 0.03, + "grad_norm": 0.5517862439155579, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 3000 + }, + { + "epoch": 0.03001, + "grad_norm": 0.5322889089584351, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 3001 + }, + { + "epoch": 0.03002, + "grad_norm": 0.5043152570724487, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 3002 + }, + { + "epoch": 0.03003, + "grad_norm": 0.5661661028862, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 3003 + }, + { + "epoch": 0.03004, + "grad_norm": 0.6002155542373657, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 3004 + }, + { + "epoch": 0.03005, + "grad_norm": 0.5543268918991089, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 3005 + }, + { + "epoch": 0.03006, + "grad_norm": 0.6193807721138, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 3006 + }, + { + "epoch": 0.03007, + "grad_norm": 0.7129995822906494, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 3007 + }, + { + "epoch": 0.03008, + "grad_norm": 0.7921411991119385, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 3008 + }, + { + "epoch": 0.03009, + "grad_norm": 0.6867934465408325, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 3009 + }, + { + "epoch": 0.0301, + "grad_norm": 0.630652666091919, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 3010 + }, + { + "epoch": 0.03011, + "grad_norm": 0.7961447834968567, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 3011 + }, + { + "epoch": 0.03012, + "grad_norm": 0.6278183460235596, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 3012 + }, + { + "epoch": 0.03013, + "grad_norm": 0.5435069799423218, + "learning_rate": 0.003, + "loss": 4.087, + "step": 3013 + }, + { + "epoch": 0.03014, + "grad_norm": 0.5863070487976074, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 3014 + }, + { + "epoch": 0.03015, + "grad_norm": 0.6343687176704407, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 3015 + }, + { + "epoch": 0.03016, + "grad_norm": 0.72650146484375, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 3016 + }, + { + "epoch": 0.03017, + "grad_norm": 0.719595193862915, + "learning_rate": 0.003, + "loss": 4.126, + "step": 3017 + }, + { + "epoch": 0.03018, + "grad_norm": 0.5911171436309814, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 3018 + }, + { + "epoch": 0.03019, + "grad_norm": 0.5423696637153625, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 3019 + }, + { + "epoch": 0.0302, + "grad_norm": 0.5470551252365112, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 3020 + }, + { + "epoch": 0.03021, + "grad_norm": 0.7124951481819153, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 3021 + }, + { + "epoch": 0.03022, + "grad_norm": 0.8745585680007935, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 3022 + }, + { + "epoch": 0.03023, + "grad_norm": 0.866553008556366, + "learning_rate": 0.003, + "loss": 4.1734, + "step": 3023 + }, + { + "epoch": 0.03024, + "grad_norm": 0.7500194907188416, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 3024 + }, + { + "epoch": 0.03025, + "grad_norm": 0.7470998764038086, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 3025 + }, + { + "epoch": 0.03026, + "grad_norm": 0.8869830369949341, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 3026 + }, + { + "epoch": 0.03027, + "grad_norm": 0.7815242409706116, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 3027 + }, + { + "epoch": 0.03028, + "grad_norm": 0.6851547956466675, + "learning_rate": 0.003, + "loss": 4.098, + "step": 3028 + }, + { + "epoch": 0.03029, + "grad_norm": 0.6831088066101074, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 3029 + }, + { + "epoch": 0.0303, + "grad_norm": 0.6287125945091248, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 3030 + }, + { + "epoch": 0.03031, + "grad_norm": 0.5370122194290161, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 3031 + }, + { + "epoch": 0.03032, + "grad_norm": 0.5675404071807861, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 3032 + }, + { + "epoch": 0.03033, + "grad_norm": 0.7573124766349792, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 3033 + }, + { + "epoch": 0.03034, + "grad_norm": 0.8983549475669861, + "learning_rate": 0.003, + "loss": 4.1466, + "step": 3034 + }, + { + "epoch": 0.03035, + "grad_norm": 1.0109939575195312, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 3035 + }, + { + "epoch": 0.03036, + "grad_norm": 0.9054379463195801, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 3036 + }, + { + "epoch": 0.03037, + "grad_norm": 0.9631518125534058, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 3037 + }, + { + "epoch": 0.03038, + "grad_norm": 0.9668822288513184, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 3038 + }, + { + "epoch": 0.03039, + "grad_norm": 0.9021666049957275, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 3039 + }, + { + "epoch": 0.0304, + "grad_norm": 0.8186821937561035, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 3040 + }, + { + "epoch": 0.03041, + "grad_norm": 0.7148540616035461, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 3041 + }, + { + "epoch": 0.03042, + "grad_norm": 0.7441778182983398, + "learning_rate": 0.003, + "loss": 4.18, + "step": 3042 + }, + { + "epoch": 0.03043, + "grad_norm": 0.7269161343574524, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 3043 + }, + { + "epoch": 0.03044, + "grad_norm": 0.7291993498802185, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 3044 + }, + { + "epoch": 0.03045, + "grad_norm": 0.8323465585708618, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 3045 + }, + { + "epoch": 0.03046, + "grad_norm": 0.9850740432739258, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 3046 + }, + { + "epoch": 0.03047, + "grad_norm": 1.0652803182601929, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 3047 + }, + { + "epoch": 0.03048, + "grad_norm": 0.8505216836929321, + "learning_rate": 0.003, + "loss": 4.145, + "step": 3048 + }, + { + "epoch": 0.03049, + "grad_norm": 0.7812585234642029, + "learning_rate": 0.003, + "loss": 4.13, + "step": 3049 + }, + { + "epoch": 0.0305, + "grad_norm": 0.8137039542198181, + "learning_rate": 0.003, + "loss": 4.148, + "step": 3050 + }, + { + "epoch": 0.03051, + "grad_norm": 0.8009527325630188, + "learning_rate": 0.003, + "loss": 4.157, + "step": 3051 + }, + { + "epoch": 0.03052, + "grad_norm": 0.6827675104141235, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 3052 + }, + { + "epoch": 0.03053, + "grad_norm": 0.5847889184951782, + "learning_rate": 0.003, + "loss": 4.1564, + "step": 3053 + }, + { + "epoch": 0.03054, + "grad_norm": 0.5359497666358948, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 3054 + }, + { + "epoch": 0.03055, + "grad_norm": 0.5127337574958801, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 3055 + }, + { + "epoch": 0.03056, + "grad_norm": 0.4672796130180359, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 3056 + }, + { + "epoch": 0.03057, + "grad_norm": 0.4908005893230438, + "learning_rate": 0.003, + "loss": 4.141, + "step": 3057 + }, + { + "epoch": 0.03058, + "grad_norm": 0.4413350522518158, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 3058 + }, + { + "epoch": 0.03059, + "grad_norm": 0.42543938755989075, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 3059 + }, + { + "epoch": 0.0306, + "grad_norm": 0.43946075439453125, + "learning_rate": 0.003, + "loss": 4.107, + "step": 3060 + }, + { + "epoch": 0.03061, + "grad_norm": 0.4698311388492584, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 3061 + }, + { + "epoch": 0.03062, + "grad_norm": 0.44352906942367554, + "learning_rate": 0.003, + "loss": 4.154, + "step": 3062 + }, + { + "epoch": 0.03063, + "grad_norm": 0.4036288857460022, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 3063 + }, + { + "epoch": 0.03064, + "grad_norm": 0.34206268191337585, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 3064 + }, + { + "epoch": 0.03065, + "grad_norm": 0.31049665808677673, + "learning_rate": 0.003, + "loss": 4.094, + "step": 3065 + }, + { + "epoch": 0.03066, + "grad_norm": 0.309965580701828, + "learning_rate": 0.003, + "loss": 4.137, + "step": 3066 + }, + { + "epoch": 0.03067, + "grad_norm": 0.31760501861572266, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 3067 + }, + { + "epoch": 0.03068, + "grad_norm": 0.34679949283599854, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 3068 + }, + { + "epoch": 0.03069, + "grad_norm": 0.491832435131073, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 3069 + }, + { + "epoch": 0.0307, + "grad_norm": 0.7506191730499268, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 3070 + }, + { + "epoch": 0.03071, + "grad_norm": 0.9722976684570312, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 3071 + }, + { + "epoch": 0.03072, + "grad_norm": 0.9531710147857666, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 3072 + }, + { + "epoch": 0.03073, + "grad_norm": 0.7240149974822998, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 3073 + }, + { + "epoch": 0.03074, + "grad_norm": 0.8381486535072327, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 3074 + }, + { + "epoch": 0.03075, + "grad_norm": 1.0160551071166992, + "learning_rate": 0.003, + "loss": 4.163, + "step": 3075 + }, + { + "epoch": 0.03076, + "grad_norm": 0.9568767547607422, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 3076 + }, + { + "epoch": 0.03077, + "grad_norm": 0.855299174785614, + "learning_rate": 0.003, + "loss": 4.152, + "step": 3077 + }, + { + "epoch": 0.03078, + "grad_norm": 0.9153781533241272, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 3078 + }, + { + "epoch": 0.03079, + "grad_norm": 0.8517905473709106, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 3079 + }, + { + "epoch": 0.0308, + "grad_norm": 0.7446357607841492, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 3080 + }, + { + "epoch": 0.03081, + "grad_norm": 0.8622949719429016, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 3081 + }, + { + "epoch": 0.03082, + "grad_norm": 0.923214316368103, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 3082 + }, + { + "epoch": 0.03083, + "grad_norm": 0.7624456882476807, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 3083 + }, + { + "epoch": 0.03084, + "grad_norm": 0.7506299018859863, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 3084 + }, + { + "epoch": 0.03085, + "grad_norm": 0.7482979893684387, + "learning_rate": 0.003, + "loss": 4.149, + "step": 3085 + }, + { + "epoch": 0.03086, + "grad_norm": 0.8742256164550781, + "learning_rate": 0.003, + "loss": 4.1735, + "step": 3086 + }, + { + "epoch": 0.03087, + "grad_norm": 0.8986655473709106, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 3087 + }, + { + "epoch": 0.03088, + "grad_norm": 0.6525835990905762, + "learning_rate": 0.003, + "loss": 4.1581, + "step": 3088 + }, + { + "epoch": 0.03089, + "grad_norm": 0.5235843062400818, + "learning_rate": 0.003, + "loss": 4.119, + "step": 3089 + }, + { + "epoch": 0.0309, + "grad_norm": 0.5896126627922058, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 3090 + }, + { + "epoch": 0.03091, + "grad_norm": 0.6240640878677368, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 3091 + }, + { + "epoch": 0.03092, + "grad_norm": 0.7089477777481079, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 3092 + }, + { + "epoch": 0.03093, + "grad_norm": 0.7679318785667419, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 3093 + }, + { + "epoch": 0.03094, + "grad_norm": 0.829313337802887, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 3094 + }, + { + "epoch": 0.03095, + "grad_norm": 0.858989953994751, + "learning_rate": 0.003, + "loss": 4.1537, + "step": 3095 + }, + { + "epoch": 0.03096, + "grad_norm": 0.8490680456161499, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 3096 + }, + { + "epoch": 0.03097, + "grad_norm": 0.7456005811691284, + "learning_rate": 0.003, + "loss": 4.1564, + "step": 3097 + }, + { + "epoch": 0.03098, + "grad_norm": 0.6156277656555176, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 3098 + }, + { + "epoch": 0.03099, + "grad_norm": 0.7040721774101257, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 3099 + }, + { + "epoch": 0.031, + "grad_norm": 0.7396160960197449, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 3100 + }, + { + "epoch": 0.03101, + "grad_norm": 0.6194980144500732, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 3101 + }, + { + "epoch": 0.03102, + "grad_norm": 0.5160166025161743, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 3102 + }, + { + "epoch": 0.03103, + "grad_norm": 0.4513045847415924, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 3103 + }, + { + "epoch": 0.03104, + "grad_norm": 0.46098387241363525, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 3104 + }, + { + "epoch": 0.03105, + "grad_norm": 0.46439290046691895, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 3105 + }, + { + "epoch": 0.03106, + "grad_norm": 0.48921629786491394, + "learning_rate": 0.003, + "loss": 4.112, + "step": 3106 + }, + { + "epoch": 0.03107, + "grad_norm": 0.4456900954246521, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 3107 + }, + { + "epoch": 0.03108, + "grad_norm": 0.3602052330970764, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 3108 + }, + { + "epoch": 0.03109, + "grad_norm": 0.37874293327331543, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 3109 + }, + { + "epoch": 0.0311, + "grad_norm": 0.34511056542396545, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 3110 + }, + { + "epoch": 0.03111, + "grad_norm": 0.31971776485443115, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 3111 + }, + { + "epoch": 0.03112, + "grad_norm": 0.38327884674072266, + "learning_rate": 0.003, + "loss": 4.123, + "step": 3112 + }, + { + "epoch": 0.03113, + "grad_norm": 0.4181249141693115, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 3113 + }, + { + "epoch": 0.03114, + "grad_norm": 0.5446284413337708, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 3114 + }, + { + "epoch": 0.03115, + "grad_norm": 0.7353031635284424, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 3115 + }, + { + "epoch": 0.03116, + "grad_norm": 0.9837905168533325, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 3116 + }, + { + "epoch": 0.03117, + "grad_norm": 1.0615694522857666, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 3117 + }, + { + "epoch": 0.03118, + "grad_norm": 0.7411212921142578, + "learning_rate": 0.003, + "loss": 4.122, + "step": 3118 + }, + { + "epoch": 0.03119, + "grad_norm": 0.6940107941627502, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 3119 + }, + { + "epoch": 0.0312, + "grad_norm": 0.6913151144981384, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 3120 + }, + { + "epoch": 0.03121, + "grad_norm": 0.6097826361656189, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 3121 + }, + { + "epoch": 0.03122, + "grad_norm": 0.6615216732025146, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 3122 + }, + { + "epoch": 0.03123, + "grad_norm": 0.799041211605072, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 3123 + }, + { + "epoch": 0.03124, + "grad_norm": 0.8251768350601196, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 3124 + }, + { + "epoch": 0.03125, + "grad_norm": 0.7816169857978821, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 3125 + }, + { + "epoch": 0.03126, + "grad_norm": 0.7771303057670593, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 3126 + }, + { + "epoch": 0.03127, + "grad_norm": 0.7820833921432495, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 3127 + }, + { + "epoch": 0.03128, + "grad_norm": 0.7861484885215759, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 3128 + }, + { + "epoch": 0.03129, + "grad_norm": 0.7761921286582947, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 3129 + }, + { + "epoch": 0.0313, + "grad_norm": 0.7497063875198364, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 3130 + }, + { + "epoch": 0.03131, + "grad_norm": 0.673405647277832, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 3131 + }, + { + "epoch": 0.03132, + "grad_norm": 0.6740285158157349, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 3132 + }, + { + "epoch": 0.03133, + "grad_norm": 0.7402722835540771, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 3133 + }, + { + "epoch": 0.03134, + "grad_norm": 0.8292396068572998, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 3134 + }, + { + "epoch": 0.03135, + "grad_norm": 0.8309870958328247, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 3135 + }, + { + "epoch": 0.03136, + "grad_norm": 0.9130949378013611, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 3136 + }, + { + "epoch": 0.03137, + "grad_norm": 0.8444306254386902, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 3137 + }, + { + "epoch": 0.03138, + "grad_norm": 0.8266807198524475, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 3138 + }, + { + "epoch": 0.03139, + "grad_norm": 0.9213491082191467, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 3139 + }, + { + "epoch": 0.0314, + "grad_norm": 0.8733784556388855, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 3140 + }, + { + "epoch": 0.03141, + "grad_norm": 0.848698079586029, + "learning_rate": 0.003, + "loss": 4.1763, + "step": 3141 + }, + { + "epoch": 0.03142, + "grad_norm": 0.8152028918266296, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 3142 + }, + { + "epoch": 0.03143, + "grad_norm": 0.690168023109436, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 3143 + }, + { + "epoch": 0.03144, + "grad_norm": 0.6327980160713196, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 3144 + }, + { + "epoch": 0.03145, + "grad_norm": 0.5650879144668579, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 3145 + }, + { + "epoch": 0.03146, + "grad_norm": 0.5615544319152832, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 3146 + }, + { + "epoch": 0.03147, + "grad_norm": 0.5901345014572144, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 3147 + }, + { + "epoch": 0.03148, + "grad_norm": 0.5519311428070068, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 3148 + }, + { + "epoch": 0.03149, + "grad_norm": 0.5862347483634949, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 3149 + }, + { + "epoch": 0.0315, + "grad_norm": 0.48593467473983765, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 3150 + }, + { + "epoch": 0.03151, + "grad_norm": 0.44726112484931946, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 3151 + }, + { + "epoch": 0.03152, + "grad_norm": 0.5725032687187195, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 3152 + }, + { + "epoch": 0.03153, + "grad_norm": 0.7204610109329224, + "learning_rate": 0.003, + "loss": 4.1445, + "step": 3153 + }, + { + "epoch": 0.03154, + "grad_norm": 0.7772231101989746, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 3154 + }, + { + "epoch": 0.03155, + "grad_norm": 0.7544244527816772, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 3155 + }, + { + "epoch": 0.03156, + "grad_norm": 0.6843472719192505, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 3156 + }, + { + "epoch": 0.03157, + "grad_norm": 0.5659379363059998, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 3157 + }, + { + "epoch": 0.03158, + "grad_norm": 0.560870885848999, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 3158 + }, + { + "epoch": 0.03159, + "grad_norm": 0.6483632922172546, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 3159 + }, + { + "epoch": 0.0316, + "grad_norm": 0.7656121253967285, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 3160 + }, + { + "epoch": 0.03161, + "grad_norm": 0.7832038998603821, + "learning_rate": 0.003, + "loss": 4.148, + "step": 3161 + }, + { + "epoch": 0.03162, + "grad_norm": 0.5725152492523193, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 3162 + }, + { + "epoch": 0.03163, + "grad_norm": 0.5066990256309509, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 3163 + }, + { + "epoch": 0.03164, + "grad_norm": 0.5056422352790833, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 3164 + }, + { + "epoch": 0.03165, + "grad_norm": 0.5810751914978027, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 3165 + }, + { + "epoch": 0.03166, + "grad_norm": 0.6706964373588562, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 3166 + }, + { + "epoch": 0.03167, + "grad_norm": 0.8062258958816528, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 3167 + }, + { + "epoch": 0.03168, + "grad_norm": 0.8774738311767578, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 3168 + }, + { + "epoch": 0.03169, + "grad_norm": 0.817283034324646, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 3169 + }, + { + "epoch": 0.0317, + "grad_norm": 0.8635371327400208, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 3170 + }, + { + "epoch": 0.03171, + "grad_norm": 0.7711028456687927, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 3171 + }, + { + "epoch": 0.03172, + "grad_norm": 0.754484236240387, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 3172 + }, + { + "epoch": 0.03173, + "grad_norm": 0.6678244471549988, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 3173 + }, + { + "epoch": 0.03174, + "grad_norm": 0.588426411151886, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 3174 + }, + { + "epoch": 0.03175, + "grad_norm": 0.5396115779876709, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 3175 + }, + { + "epoch": 0.03176, + "grad_norm": 0.5115644931793213, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 3176 + }, + { + "epoch": 0.03177, + "grad_norm": 0.5066182017326355, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 3177 + }, + { + "epoch": 0.03178, + "grad_norm": 0.6152738928794861, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 3178 + }, + { + "epoch": 0.03179, + "grad_norm": 0.7892536520957947, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 3179 + }, + { + "epoch": 0.0318, + "grad_norm": 0.9117100834846497, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 3180 + }, + { + "epoch": 0.03181, + "grad_norm": 0.8305923938751221, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 3181 + }, + { + "epoch": 0.03182, + "grad_norm": 0.7435977458953857, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 3182 + }, + { + "epoch": 0.03183, + "grad_norm": 0.9169098138809204, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 3183 + }, + { + "epoch": 0.03184, + "grad_norm": 1.1370512247085571, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 3184 + }, + { + "epoch": 0.03185, + "grad_norm": 0.7811375856399536, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 3185 + }, + { + "epoch": 0.03186, + "grad_norm": 0.6801838278770447, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 3186 + }, + { + "epoch": 0.03187, + "grad_norm": 0.7519158124923706, + "learning_rate": 0.003, + "loss": 4.103, + "step": 3187 + }, + { + "epoch": 0.03188, + "grad_norm": 0.7719635963439941, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 3188 + }, + { + "epoch": 0.03189, + "grad_norm": 0.7892667055130005, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 3189 + }, + { + "epoch": 0.0319, + "grad_norm": 0.7617167234420776, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 3190 + }, + { + "epoch": 0.03191, + "grad_norm": 0.729390561580658, + "learning_rate": 0.003, + "loss": 4.1513, + "step": 3191 + }, + { + "epoch": 0.03192, + "grad_norm": 0.6866948008537292, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 3192 + }, + { + "epoch": 0.03193, + "grad_norm": 0.8149189352989197, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 3193 + }, + { + "epoch": 0.03194, + "grad_norm": 0.7727823257446289, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 3194 + }, + { + "epoch": 0.03195, + "grad_norm": 0.6907305121421814, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 3195 + }, + { + "epoch": 0.03196, + "grad_norm": 0.6480019688606262, + "learning_rate": 0.003, + "loss": 4.129, + "step": 3196 + }, + { + "epoch": 0.03197, + "grad_norm": 0.6377413272857666, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 3197 + }, + { + "epoch": 0.03198, + "grad_norm": 0.5232990384101868, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 3198 + }, + { + "epoch": 0.03199, + "grad_norm": 0.5117045044898987, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 3199 + }, + { + "epoch": 0.032, + "grad_norm": 0.49118801951408386, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 3200 + }, + { + "epoch": 0.03201, + "grad_norm": 0.39169371128082275, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 3201 + }, + { + "epoch": 0.03202, + "grad_norm": 0.4503386616706848, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 3202 + }, + { + "epoch": 0.03203, + "grad_norm": 0.44623667001724243, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 3203 + }, + { + "epoch": 0.03204, + "grad_norm": 0.4461396634578705, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 3204 + }, + { + "epoch": 0.03205, + "grad_norm": 0.47087642550468445, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 3205 + }, + { + "epoch": 0.03206, + "grad_norm": 0.4962460696697235, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 3206 + }, + { + "epoch": 0.03207, + "grad_norm": 0.6064698100090027, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 3207 + }, + { + "epoch": 0.03208, + "grad_norm": 0.759096622467041, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 3208 + }, + { + "epoch": 0.03209, + "grad_norm": 0.7810276746749878, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 3209 + }, + { + "epoch": 0.0321, + "grad_norm": 0.7179496884346008, + "learning_rate": 0.003, + "loss": 4.1363, + "step": 3210 + }, + { + "epoch": 0.03211, + "grad_norm": 0.7148532867431641, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 3211 + }, + { + "epoch": 0.03212, + "grad_norm": 0.7214945554733276, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 3212 + }, + { + "epoch": 0.03213, + "grad_norm": 0.7809195518493652, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 3213 + }, + { + "epoch": 0.03214, + "grad_norm": 0.9291434288024902, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 3214 + }, + { + "epoch": 0.03215, + "grad_norm": 0.9452233910560608, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 3215 + }, + { + "epoch": 0.03216, + "grad_norm": 0.8493017554283142, + "learning_rate": 0.003, + "loss": 4.125, + "step": 3216 + }, + { + "epoch": 0.03217, + "grad_norm": 0.7973296046257019, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 3217 + }, + { + "epoch": 0.03218, + "grad_norm": 0.7930925488471985, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 3218 + }, + { + "epoch": 0.03219, + "grad_norm": 0.7060784101486206, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 3219 + }, + { + "epoch": 0.0322, + "grad_norm": 0.6504276394844055, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 3220 + }, + { + "epoch": 0.03221, + "grad_norm": 0.7263512015342712, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 3221 + }, + { + "epoch": 0.03222, + "grad_norm": 0.7812563180923462, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 3222 + }, + { + "epoch": 0.03223, + "grad_norm": 0.718917727470398, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 3223 + }, + { + "epoch": 0.03224, + "grad_norm": 0.6486620306968689, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 3224 + }, + { + "epoch": 0.03225, + "grad_norm": 0.633409857749939, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 3225 + }, + { + "epoch": 0.03226, + "grad_norm": 0.6219311356544495, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 3226 + }, + { + "epoch": 0.03227, + "grad_norm": 0.6718069911003113, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 3227 + }, + { + "epoch": 0.03228, + "grad_norm": 0.6600261926651001, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 3228 + }, + { + "epoch": 0.03229, + "grad_norm": 0.7300699353218079, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 3229 + }, + { + "epoch": 0.0323, + "grad_norm": 0.8185812830924988, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 3230 + }, + { + "epoch": 0.03231, + "grad_norm": 0.8448948860168457, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 3231 + }, + { + "epoch": 0.03232, + "grad_norm": 0.8037069439888, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 3232 + }, + { + "epoch": 0.03233, + "grad_norm": 0.6786398887634277, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 3233 + }, + { + "epoch": 0.03234, + "grad_norm": 0.6006938219070435, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 3234 + }, + { + "epoch": 0.03235, + "grad_norm": 0.651039183139801, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 3235 + }, + { + "epoch": 0.03236, + "grad_norm": 0.635384202003479, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 3236 + }, + { + "epoch": 0.03237, + "grad_norm": 0.6635240912437439, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 3237 + }, + { + "epoch": 0.03238, + "grad_norm": 0.6811038851737976, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 3238 + }, + { + "epoch": 0.03239, + "grad_norm": 0.7615686655044556, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 3239 + }, + { + "epoch": 0.0324, + "grad_norm": 0.8041428923606873, + "learning_rate": 0.003, + "loss": 4.151, + "step": 3240 + }, + { + "epoch": 0.03241, + "grad_norm": 0.8027405738830566, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 3241 + }, + { + "epoch": 0.03242, + "grad_norm": 0.9180432558059692, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 3242 + }, + { + "epoch": 0.03243, + "grad_norm": 0.9836052060127258, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 3243 + }, + { + "epoch": 0.03244, + "grad_norm": 0.9347205758094788, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 3244 + }, + { + "epoch": 0.03245, + "grad_norm": 1.1407207250595093, + "learning_rate": 0.003, + "loss": 4.1692, + "step": 3245 + }, + { + "epoch": 0.03246, + "grad_norm": 0.7863346934318542, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 3246 + }, + { + "epoch": 0.03247, + "grad_norm": 0.6830831170082092, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 3247 + }, + { + "epoch": 0.03248, + "grad_norm": 0.6007001996040344, + "learning_rate": 0.003, + "loss": 4.1605, + "step": 3248 + }, + { + "epoch": 0.03249, + "grad_norm": 0.5720266103744507, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 3249 + }, + { + "epoch": 0.0325, + "grad_norm": 0.5476762056350708, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 3250 + }, + { + "epoch": 0.03251, + "grad_norm": 0.5844308137893677, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 3251 + }, + { + "epoch": 0.03252, + "grad_norm": 0.6554009318351746, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 3252 + }, + { + "epoch": 0.03253, + "grad_norm": 0.7417848110198975, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 3253 + }, + { + "epoch": 0.03254, + "grad_norm": 0.7940112352371216, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 3254 + }, + { + "epoch": 0.03255, + "grad_norm": 0.7129653096199036, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 3255 + }, + { + "epoch": 0.03256, + "grad_norm": 0.5819328427314758, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 3256 + }, + { + "epoch": 0.03257, + "grad_norm": 0.4663737118244171, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 3257 + }, + { + "epoch": 0.03258, + "grad_norm": 0.423995703458786, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 3258 + }, + { + "epoch": 0.03259, + "grad_norm": 0.43535032868385315, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 3259 + }, + { + "epoch": 0.0326, + "grad_norm": 0.39455822110176086, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 3260 + }, + { + "epoch": 0.03261, + "grad_norm": 0.40695109963417053, + "learning_rate": 0.003, + "loss": 4.088, + "step": 3261 + }, + { + "epoch": 0.03262, + "grad_norm": 0.3774632513523102, + "learning_rate": 0.003, + "loss": 4.113, + "step": 3262 + }, + { + "epoch": 0.03263, + "grad_norm": 0.46376833319664, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 3263 + }, + { + "epoch": 0.03264, + "grad_norm": 0.5567134618759155, + "learning_rate": 0.003, + "loss": 4.136, + "step": 3264 + }, + { + "epoch": 0.03265, + "grad_norm": 0.6502901315689087, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 3265 + }, + { + "epoch": 0.03266, + "grad_norm": 0.8112798929214478, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 3266 + }, + { + "epoch": 0.03267, + "grad_norm": 0.9981924295425415, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 3267 + }, + { + "epoch": 0.03268, + "grad_norm": 0.9973412752151489, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 3268 + }, + { + "epoch": 0.03269, + "grad_norm": 0.8641186356544495, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 3269 + }, + { + "epoch": 0.0327, + "grad_norm": 0.7888898253440857, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 3270 + }, + { + "epoch": 0.03271, + "grad_norm": 0.8556457757949829, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 3271 + }, + { + "epoch": 0.03272, + "grad_norm": 0.8728947639465332, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 3272 + }, + { + "epoch": 0.03273, + "grad_norm": 0.8285180926322937, + "learning_rate": 0.003, + "loss": 4.133, + "step": 3273 + }, + { + "epoch": 0.03274, + "grad_norm": 0.6249201893806458, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 3274 + }, + { + "epoch": 0.03275, + "grad_norm": 0.5604618787765503, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 3275 + }, + { + "epoch": 0.03276, + "grad_norm": 0.6792795062065125, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 3276 + }, + { + "epoch": 0.03277, + "grad_norm": 0.7676330804824829, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 3277 + }, + { + "epoch": 0.03278, + "grad_norm": 0.8335491418838501, + "learning_rate": 0.003, + "loss": 4.138, + "step": 3278 + }, + { + "epoch": 0.03279, + "grad_norm": 0.7494176626205444, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 3279 + }, + { + "epoch": 0.0328, + "grad_norm": 0.5933589935302734, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 3280 + }, + { + "epoch": 0.03281, + "grad_norm": 0.49470627307891846, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 3281 + }, + { + "epoch": 0.03282, + "grad_norm": 0.5273627638816833, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 3282 + }, + { + "epoch": 0.03283, + "grad_norm": 0.5509713292121887, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 3283 + }, + { + "epoch": 0.03284, + "grad_norm": 0.544580340385437, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 3284 + }, + { + "epoch": 0.03285, + "grad_norm": 0.5744161009788513, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 3285 + }, + { + "epoch": 0.03286, + "grad_norm": 0.531225860118866, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 3286 + }, + { + "epoch": 0.03287, + "grad_norm": 0.49495846033096313, + "learning_rate": 0.003, + "loss": 4.13, + "step": 3287 + }, + { + "epoch": 0.03288, + "grad_norm": 0.5203666687011719, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 3288 + }, + { + "epoch": 0.03289, + "grad_norm": 0.5833369493484497, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 3289 + }, + { + "epoch": 0.0329, + "grad_norm": 0.6156966686248779, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 3290 + }, + { + "epoch": 0.03291, + "grad_norm": 0.7284314036369324, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 3291 + }, + { + "epoch": 0.03292, + "grad_norm": 0.8301785588264465, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 3292 + }, + { + "epoch": 0.03293, + "grad_norm": 0.8880192041397095, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 3293 + }, + { + "epoch": 0.03294, + "grad_norm": 0.7777594923973083, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 3294 + }, + { + "epoch": 0.03295, + "grad_norm": 0.7415675520896912, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 3295 + }, + { + "epoch": 0.03296, + "grad_norm": 0.5945268273353577, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 3296 + }, + { + "epoch": 0.03297, + "grad_norm": 0.6904322504997253, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 3297 + }, + { + "epoch": 0.03298, + "grad_norm": 0.8207030892372131, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 3298 + }, + { + "epoch": 0.03299, + "grad_norm": 1.0657931566238403, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 3299 + }, + { + "epoch": 0.033, + "grad_norm": 0.9290387034416199, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 3300 + }, + { + "epoch": 0.03301, + "grad_norm": 0.6823791861534119, + "learning_rate": 0.003, + "loss": 4.1464, + "step": 3301 + }, + { + "epoch": 0.03302, + "grad_norm": 0.5951451659202576, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 3302 + }, + { + "epoch": 0.03303, + "grad_norm": 0.5956482887268066, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 3303 + }, + { + "epoch": 0.03304, + "grad_norm": 0.5632296204566956, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 3304 + }, + { + "epoch": 0.03305, + "grad_norm": 0.5975555181503296, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 3305 + }, + { + "epoch": 0.03306, + "grad_norm": 0.5866029262542725, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 3306 + }, + { + "epoch": 0.03307, + "grad_norm": 0.586901843547821, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 3307 + }, + { + "epoch": 0.03308, + "grad_norm": 0.5834401249885559, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 3308 + }, + { + "epoch": 0.03309, + "grad_norm": 0.6751212477684021, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 3309 + }, + { + "epoch": 0.0331, + "grad_norm": 0.5830533504486084, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 3310 + }, + { + "epoch": 0.03311, + "grad_norm": 0.6124199628829956, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 3311 + }, + { + "epoch": 0.03312, + "grad_norm": 0.5900564193725586, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 3312 + }, + { + "epoch": 0.03313, + "grad_norm": 0.6442069411277771, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 3313 + }, + { + "epoch": 0.03314, + "grad_norm": 0.7764161229133606, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 3314 + }, + { + "epoch": 0.03315, + "grad_norm": 0.8021034002304077, + "learning_rate": 0.003, + "loss": 4.13, + "step": 3315 + }, + { + "epoch": 0.03316, + "grad_norm": 0.8197445869445801, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 3316 + }, + { + "epoch": 0.03317, + "grad_norm": 0.7437990307807922, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 3317 + }, + { + "epoch": 0.03318, + "grad_norm": 0.7505268454551697, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 3318 + }, + { + "epoch": 0.03319, + "grad_norm": 0.7677083611488342, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 3319 + }, + { + "epoch": 0.0332, + "grad_norm": 0.8221232891082764, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 3320 + }, + { + "epoch": 0.03321, + "grad_norm": 0.8141359090805054, + "learning_rate": 0.003, + "loss": 4.1433, + "step": 3321 + }, + { + "epoch": 0.03322, + "grad_norm": 0.7189428806304932, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 3322 + }, + { + "epoch": 0.03323, + "grad_norm": 0.6044638156890869, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 3323 + }, + { + "epoch": 0.03324, + "grad_norm": 0.5276358127593994, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 3324 + }, + { + "epoch": 0.03325, + "grad_norm": 0.5766726136207581, + "learning_rate": 0.003, + "loss": 4.134, + "step": 3325 + }, + { + "epoch": 0.03326, + "grad_norm": 0.6931579113006592, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 3326 + }, + { + "epoch": 0.03327, + "grad_norm": 0.8419184684753418, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 3327 + }, + { + "epoch": 0.03328, + "grad_norm": 0.82012540102005, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 3328 + }, + { + "epoch": 0.03329, + "grad_norm": 0.6669595837593079, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 3329 + }, + { + "epoch": 0.0333, + "grad_norm": 0.6745646595954895, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 3330 + }, + { + "epoch": 0.03331, + "grad_norm": 0.7668302655220032, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 3331 + }, + { + "epoch": 0.03332, + "grad_norm": 0.7806735634803772, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 3332 + }, + { + "epoch": 0.03333, + "grad_norm": 0.8616372346878052, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 3333 + }, + { + "epoch": 0.03334, + "grad_norm": 0.8143258690834045, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 3334 + }, + { + "epoch": 0.03335, + "grad_norm": 0.7264704704284668, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 3335 + }, + { + "epoch": 0.03336, + "grad_norm": 0.7028628587722778, + "learning_rate": 0.003, + "loss": 4.134, + "step": 3336 + }, + { + "epoch": 0.03337, + "grad_norm": 0.7172381281852722, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 3337 + }, + { + "epoch": 0.03338, + "grad_norm": 0.7452136278152466, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 3338 + }, + { + "epoch": 0.03339, + "grad_norm": 0.6151822209358215, + "learning_rate": 0.003, + "loss": 4.122, + "step": 3339 + }, + { + "epoch": 0.0334, + "grad_norm": 0.48978814482688904, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 3340 + }, + { + "epoch": 0.03341, + "grad_norm": 0.5616298317909241, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 3341 + }, + { + "epoch": 0.03342, + "grad_norm": 0.5989934802055359, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 3342 + }, + { + "epoch": 0.03343, + "grad_norm": 0.6791746616363525, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 3343 + }, + { + "epoch": 0.03344, + "grad_norm": 0.8041501641273499, + "learning_rate": 0.003, + "loss": 4.139, + "step": 3344 + }, + { + "epoch": 0.03345, + "grad_norm": 0.9025998711585999, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 3345 + }, + { + "epoch": 0.03346, + "grad_norm": 0.8547161221504211, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 3346 + }, + { + "epoch": 0.03347, + "grad_norm": 0.8230910301208496, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 3347 + }, + { + "epoch": 0.03348, + "grad_norm": 0.6764378547668457, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 3348 + }, + { + "epoch": 0.03349, + "grad_norm": 0.6756121516227722, + "learning_rate": 0.003, + "loss": 4.1537, + "step": 3349 + }, + { + "epoch": 0.0335, + "grad_norm": 0.90619957447052, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 3350 + }, + { + "epoch": 0.03351, + "grad_norm": 0.995084822177887, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 3351 + }, + { + "epoch": 0.03352, + "grad_norm": 0.9932165741920471, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 3352 + }, + { + "epoch": 0.03353, + "grad_norm": 0.7695693373680115, + "learning_rate": 0.003, + "loss": 4.144, + "step": 3353 + }, + { + "epoch": 0.03354, + "grad_norm": 0.757735013961792, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 3354 + }, + { + "epoch": 0.03355, + "grad_norm": 0.7958679795265198, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 3355 + }, + { + "epoch": 0.03356, + "grad_norm": 0.8041818141937256, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 3356 + }, + { + "epoch": 0.03357, + "grad_norm": 0.6908081769943237, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 3357 + }, + { + "epoch": 0.03358, + "grad_norm": 0.5655892491340637, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 3358 + }, + { + "epoch": 0.03359, + "grad_norm": 0.6353917717933655, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 3359 + }, + { + "epoch": 0.0336, + "grad_norm": 0.6901819705963135, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 3360 + }, + { + "epoch": 0.03361, + "grad_norm": 0.6260518431663513, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 3361 + }, + { + "epoch": 0.03362, + "grad_norm": 0.5238988399505615, + "learning_rate": 0.003, + "loss": 4.132, + "step": 3362 + }, + { + "epoch": 0.03363, + "grad_norm": 0.4695539176464081, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 3363 + }, + { + "epoch": 0.03364, + "grad_norm": 0.4180772602558136, + "learning_rate": 0.003, + "loss": 4.103, + "step": 3364 + }, + { + "epoch": 0.03365, + "grad_norm": 0.46931684017181396, + "learning_rate": 0.003, + "loss": 4.098, + "step": 3365 + }, + { + "epoch": 0.03366, + "grad_norm": 0.4340416491031647, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 3366 + }, + { + "epoch": 0.03367, + "grad_norm": 0.3824833929538727, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 3367 + }, + { + "epoch": 0.03368, + "grad_norm": 0.391565203666687, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 3368 + }, + { + "epoch": 0.03369, + "grad_norm": 0.3979616165161133, + "learning_rate": 0.003, + "loss": 4.093, + "step": 3369 + }, + { + "epoch": 0.0337, + "grad_norm": 0.41400259733200073, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 3370 + }, + { + "epoch": 0.03371, + "grad_norm": 0.4535299241542816, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 3371 + }, + { + "epoch": 0.03372, + "grad_norm": 0.4900164008140564, + "learning_rate": 0.003, + "loss": 4.091, + "step": 3372 + }, + { + "epoch": 0.03373, + "grad_norm": 0.48760801553726196, + "learning_rate": 0.003, + "loss": 4.124, + "step": 3373 + }, + { + "epoch": 0.03374, + "grad_norm": 0.42620474100112915, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 3374 + }, + { + "epoch": 0.03375, + "grad_norm": 0.539411187171936, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 3375 + }, + { + "epoch": 0.03376, + "grad_norm": 0.7004280686378479, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 3376 + }, + { + "epoch": 0.03377, + "grad_norm": 0.9512114524841309, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 3377 + }, + { + "epoch": 0.03378, + "grad_norm": 1.173850178718567, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 3378 + }, + { + "epoch": 0.03379, + "grad_norm": 0.8208590149879456, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 3379 + }, + { + "epoch": 0.0338, + "grad_norm": 0.9167894721031189, + "learning_rate": 0.003, + "loss": 4.138, + "step": 3380 + }, + { + "epoch": 0.03381, + "grad_norm": 0.8736410140991211, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 3381 + }, + { + "epoch": 0.03382, + "grad_norm": 0.9413759112358093, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 3382 + }, + { + "epoch": 0.03383, + "grad_norm": 0.9034369587898254, + "learning_rate": 0.003, + "loss": 4.102, + "step": 3383 + }, + { + "epoch": 0.03384, + "grad_norm": 0.7863962650299072, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 3384 + }, + { + "epoch": 0.03385, + "grad_norm": 0.6559106707572937, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 3385 + }, + { + "epoch": 0.03386, + "grad_norm": 0.5466395020484924, + "learning_rate": 0.003, + "loss": 4.116, + "step": 3386 + }, + { + "epoch": 0.03387, + "grad_norm": 0.5657241344451904, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 3387 + }, + { + "epoch": 0.03388, + "grad_norm": 0.6148836016654968, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 3388 + }, + { + "epoch": 0.03389, + "grad_norm": 0.6511346101760864, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 3389 + }, + { + "epoch": 0.0339, + "grad_norm": 0.7301871180534363, + "learning_rate": 0.003, + "loss": 4.14, + "step": 3390 + }, + { + "epoch": 0.03391, + "grad_norm": 0.7425968647003174, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 3391 + }, + { + "epoch": 0.03392, + "grad_norm": 0.8300238251686096, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 3392 + }, + { + "epoch": 0.03393, + "grad_norm": 0.9328773617744446, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 3393 + }, + { + "epoch": 0.03394, + "grad_norm": 0.8949936628341675, + "learning_rate": 0.003, + "loss": 4.145, + "step": 3394 + }, + { + "epoch": 0.03395, + "grad_norm": 0.9411712884902954, + "learning_rate": 0.003, + "loss": 4.1549, + "step": 3395 + }, + { + "epoch": 0.03396, + "grad_norm": 0.9026573896408081, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 3396 + }, + { + "epoch": 0.03397, + "grad_norm": 0.8574052453041077, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 3397 + }, + { + "epoch": 0.03398, + "grad_norm": 0.9310076236724854, + "learning_rate": 0.003, + "loss": 4.1626, + "step": 3398 + }, + { + "epoch": 0.03399, + "grad_norm": 0.9323184490203857, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 3399 + }, + { + "epoch": 0.034, + "grad_norm": 0.996589720249176, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 3400 + }, + { + "epoch": 0.03401, + "grad_norm": 0.8762171268463135, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 3401 + }, + { + "epoch": 0.03402, + "grad_norm": 0.7612597942352295, + "learning_rate": 0.003, + "loss": 4.162, + "step": 3402 + }, + { + "epoch": 0.03403, + "grad_norm": 0.7073224782943726, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 3403 + }, + { + "epoch": 0.03404, + "grad_norm": 0.8644006848335266, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 3404 + }, + { + "epoch": 0.03405, + "grad_norm": 0.9565165638923645, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 3405 + }, + { + "epoch": 0.03406, + "grad_norm": 0.9732779264450073, + "learning_rate": 0.003, + "loss": 4.1351, + "step": 3406 + }, + { + "epoch": 0.03407, + "grad_norm": 0.8747183084487915, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 3407 + }, + { + "epoch": 0.03408, + "grad_norm": 0.8171975612640381, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 3408 + }, + { + "epoch": 0.03409, + "grad_norm": 0.8472719788551331, + "learning_rate": 0.003, + "loss": 4.139, + "step": 3409 + }, + { + "epoch": 0.0341, + "grad_norm": 0.772420346736908, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 3410 + }, + { + "epoch": 0.03411, + "grad_norm": 0.7122489809989929, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 3411 + }, + { + "epoch": 0.03412, + "grad_norm": 0.7397792339324951, + "learning_rate": 0.003, + "loss": 4.1595, + "step": 3412 + }, + { + "epoch": 0.03413, + "grad_norm": 0.5002322793006897, + "learning_rate": 0.003, + "loss": 4.1603, + "step": 3413 + }, + { + "epoch": 0.03414, + "grad_norm": 0.533718466758728, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 3414 + }, + { + "epoch": 0.03415, + "grad_norm": 0.5009657144546509, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 3415 + }, + { + "epoch": 0.03416, + "grad_norm": 0.49424099922180176, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 3416 + }, + { + "epoch": 0.03417, + "grad_norm": 0.5071776509284973, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 3417 + }, + { + "epoch": 0.03418, + "grad_norm": 0.600383996963501, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 3418 + }, + { + "epoch": 0.03419, + "grad_norm": 0.5728192329406738, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 3419 + }, + { + "epoch": 0.0342, + "grad_norm": 0.5360076427459717, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 3420 + }, + { + "epoch": 0.03421, + "grad_norm": 0.5283966660499573, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 3421 + }, + { + "epoch": 0.03422, + "grad_norm": 0.5383856892585754, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 3422 + }, + { + "epoch": 0.03423, + "grad_norm": 0.6316571831703186, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 3423 + }, + { + "epoch": 0.03424, + "grad_norm": 0.7047085762023926, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 3424 + }, + { + "epoch": 0.03425, + "grad_norm": 0.7236632704734802, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 3425 + }, + { + "epoch": 0.03426, + "grad_norm": 0.6646005511283875, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 3426 + }, + { + "epoch": 0.03427, + "grad_norm": 0.5403919816017151, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 3427 + }, + { + "epoch": 0.03428, + "grad_norm": 0.5291489958763123, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 3428 + }, + { + "epoch": 0.03429, + "grad_norm": 0.5223709940910339, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 3429 + }, + { + "epoch": 0.0343, + "grad_norm": 0.5112771987915039, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 3430 + }, + { + "epoch": 0.03431, + "grad_norm": 0.4984135031700134, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 3431 + }, + { + "epoch": 0.03432, + "grad_norm": 0.6651791334152222, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 3432 + }, + { + "epoch": 0.03433, + "grad_norm": 0.8698033094406128, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 3433 + }, + { + "epoch": 0.03434, + "grad_norm": 0.9080641865730286, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 3434 + }, + { + "epoch": 0.03435, + "grad_norm": 0.6624755263328552, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 3435 + }, + { + "epoch": 0.03436, + "grad_norm": 0.5985552668571472, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 3436 + }, + { + "epoch": 0.03437, + "grad_norm": 0.74169921875, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 3437 + }, + { + "epoch": 0.03438, + "grad_norm": 0.6739313006401062, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 3438 + }, + { + "epoch": 0.03439, + "grad_norm": 0.6183229684829712, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 3439 + }, + { + "epoch": 0.0344, + "grad_norm": 0.6372612118721008, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 3440 + }, + { + "epoch": 0.03441, + "grad_norm": 0.5561286211013794, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 3441 + }, + { + "epoch": 0.03442, + "grad_norm": 0.5146679282188416, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 3442 + }, + { + "epoch": 0.03443, + "grad_norm": 0.5086862444877625, + "learning_rate": 0.003, + "loss": 4.108, + "step": 3443 + }, + { + "epoch": 0.03444, + "grad_norm": 0.5029706358909607, + "learning_rate": 0.003, + "loss": 4.117, + "step": 3444 + }, + { + "epoch": 0.03445, + "grad_norm": 0.5076051950454712, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 3445 + }, + { + "epoch": 0.03446, + "grad_norm": 0.5356073379516602, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 3446 + }, + { + "epoch": 0.03447, + "grad_norm": 0.5907533764839172, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 3447 + }, + { + "epoch": 0.03448, + "grad_norm": 0.6191877722740173, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 3448 + }, + { + "epoch": 0.03449, + "grad_norm": 0.6186185479164124, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 3449 + }, + { + "epoch": 0.0345, + "grad_norm": 0.6868272423744202, + "learning_rate": 0.003, + "loss": 4.119, + "step": 3450 + }, + { + "epoch": 0.03451, + "grad_norm": 0.5868905186653137, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 3451 + }, + { + "epoch": 0.03452, + "grad_norm": 0.5602003931999207, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 3452 + }, + { + "epoch": 0.03453, + "grad_norm": 0.6827725172042847, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 3453 + }, + { + "epoch": 0.03454, + "grad_norm": 0.6988645792007446, + "learning_rate": 0.003, + "loss": 4.084, + "step": 3454 + }, + { + "epoch": 0.03455, + "grad_norm": 0.7984029650688171, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 3455 + }, + { + "epoch": 0.03456, + "grad_norm": 0.8589499592781067, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 3456 + }, + { + "epoch": 0.03457, + "grad_norm": 0.8572336435317993, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 3457 + }, + { + "epoch": 0.03458, + "grad_norm": 0.7515364289283752, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 3458 + }, + { + "epoch": 0.03459, + "grad_norm": 0.7732160091400146, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 3459 + }, + { + "epoch": 0.0346, + "grad_norm": 1.0194605588912964, + "learning_rate": 0.003, + "loss": 4.1595, + "step": 3460 + }, + { + "epoch": 0.03461, + "grad_norm": 1.2320705652236938, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 3461 + }, + { + "epoch": 0.03462, + "grad_norm": 0.8935246467590332, + "learning_rate": 0.003, + "loss": 4.134, + "step": 3462 + }, + { + "epoch": 0.03463, + "grad_norm": 0.8400885462760925, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 3463 + }, + { + "epoch": 0.03464, + "grad_norm": 0.8417510390281677, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 3464 + }, + { + "epoch": 0.03465, + "grad_norm": 0.9221370220184326, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 3465 + }, + { + "epoch": 0.03466, + "grad_norm": 0.9339236617088318, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 3466 + }, + { + "epoch": 0.03467, + "grad_norm": 1.027878999710083, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 3467 + }, + { + "epoch": 0.03468, + "grad_norm": 0.9010937213897705, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 3468 + }, + { + "epoch": 0.03469, + "grad_norm": 1.0625158548355103, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 3469 + }, + { + "epoch": 0.0347, + "grad_norm": 0.9459049701690674, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 3470 + }, + { + "epoch": 0.03471, + "grad_norm": 0.8470619320869446, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 3471 + }, + { + "epoch": 0.03472, + "grad_norm": 0.7621662020683289, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 3472 + }, + { + "epoch": 0.03473, + "grad_norm": 0.7190437316894531, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 3473 + }, + { + "epoch": 0.03474, + "grad_norm": 0.6625324487686157, + "learning_rate": 0.003, + "loss": 4.1605, + "step": 3474 + }, + { + "epoch": 0.03475, + "grad_norm": 0.5887710452079773, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 3475 + }, + { + "epoch": 0.03476, + "grad_norm": 0.6497656106948853, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 3476 + }, + { + "epoch": 0.03477, + "grad_norm": 0.6939454078674316, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 3477 + }, + { + "epoch": 0.03478, + "grad_norm": 0.7538445591926575, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 3478 + }, + { + "epoch": 0.03479, + "grad_norm": 0.8461519479751587, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 3479 + }, + { + "epoch": 0.0348, + "grad_norm": 0.9156594276428223, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 3480 + }, + { + "epoch": 0.03481, + "grad_norm": 0.830324649810791, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 3481 + }, + { + "epoch": 0.03482, + "grad_norm": 0.6343877911567688, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 3482 + }, + { + "epoch": 0.03483, + "grad_norm": 0.5798615217208862, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 3483 + }, + { + "epoch": 0.03484, + "grad_norm": 0.6012097001075745, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 3484 + }, + { + "epoch": 0.03485, + "grad_norm": 0.534057080745697, + "learning_rate": 0.003, + "loss": 4.14, + "step": 3485 + }, + { + "epoch": 0.03486, + "grad_norm": 0.5091946721076965, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 3486 + }, + { + "epoch": 0.03487, + "grad_norm": 0.45456182956695557, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 3487 + }, + { + "epoch": 0.03488, + "grad_norm": 0.4439913034439087, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 3488 + }, + { + "epoch": 0.03489, + "grad_norm": 0.48145756125450134, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 3489 + }, + { + "epoch": 0.0349, + "grad_norm": 0.543171763420105, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 3490 + }, + { + "epoch": 0.03491, + "grad_norm": 0.4846112132072449, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 3491 + }, + { + "epoch": 0.03492, + "grad_norm": 0.4532843828201294, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 3492 + }, + { + "epoch": 0.03493, + "grad_norm": 0.452328622341156, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 3493 + }, + { + "epoch": 0.03494, + "grad_norm": 0.43499302864074707, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 3494 + }, + { + "epoch": 0.03495, + "grad_norm": 0.3597606122493744, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 3495 + }, + { + "epoch": 0.03496, + "grad_norm": 0.36562833189964294, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 3496 + }, + { + "epoch": 0.03497, + "grad_norm": 0.3867475986480713, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 3497 + }, + { + "epoch": 0.03498, + "grad_norm": 0.42188137769699097, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 3498 + }, + { + "epoch": 0.03499, + "grad_norm": 0.5299180746078491, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 3499 + }, + { + "epoch": 0.035, + "grad_norm": 0.6962635517120361, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 3500 + }, + { + "epoch": 0.03501, + "grad_norm": 0.9552702307701111, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 3501 + }, + { + "epoch": 0.03502, + "grad_norm": 1.1519496440887451, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 3502 + }, + { + "epoch": 0.03503, + "grad_norm": 0.6880185604095459, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 3503 + }, + { + "epoch": 0.03504, + "grad_norm": 0.5892576575279236, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 3504 + }, + { + "epoch": 0.03505, + "grad_norm": 0.6985999345779419, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 3505 + }, + { + "epoch": 0.03506, + "grad_norm": 0.7576223611831665, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 3506 + }, + { + "epoch": 0.03507, + "grad_norm": 0.6434160470962524, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 3507 + }, + { + "epoch": 0.03508, + "grad_norm": 0.6596179604530334, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 3508 + }, + { + "epoch": 0.03509, + "grad_norm": 0.6995351314544678, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 3509 + }, + { + "epoch": 0.0351, + "grad_norm": 0.6974444389343262, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 3510 + }, + { + "epoch": 0.03511, + "grad_norm": 0.6773884296417236, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 3511 + }, + { + "epoch": 0.03512, + "grad_norm": 0.6800361275672913, + "learning_rate": 0.003, + "loss": 4.101, + "step": 3512 + }, + { + "epoch": 0.03513, + "grad_norm": 0.6331777572631836, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 3513 + }, + { + "epoch": 0.03514, + "grad_norm": 0.6398311853408813, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 3514 + }, + { + "epoch": 0.03515, + "grad_norm": 0.652564525604248, + "learning_rate": 0.003, + "loss": 4.123, + "step": 3515 + }, + { + "epoch": 0.03516, + "grad_norm": 0.6278548836708069, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 3516 + }, + { + "epoch": 0.03517, + "grad_norm": 0.5660470128059387, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 3517 + }, + { + "epoch": 0.03518, + "grad_norm": 0.6680671572685242, + "learning_rate": 0.003, + "loss": 4.088, + "step": 3518 + }, + { + "epoch": 0.03519, + "grad_norm": 0.8297867178916931, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 3519 + }, + { + "epoch": 0.0352, + "grad_norm": 1.0089613199234009, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 3520 + }, + { + "epoch": 0.03521, + "grad_norm": 0.8898510336875916, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 3521 + }, + { + "epoch": 0.03522, + "grad_norm": 0.7937150001525879, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 3522 + }, + { + "epoch": 0.03523, + "grad_norm": 0.9151129126548767, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 3523 + }, + { + "epoch": 0.03524, + "grad_norm": 1.016127109527588, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 3524 + }, + { + "epoch": 0.03525, + "grad_norm": 0.8876500129699707, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 3525 + }, + { + "epoch": 0.03526, + "grad_norm": 0.9732938408851624, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 3526 + }, + { + "epoch": 0.03527, + "grad_norm": 0.9365341067314148, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 3527 + }, + { + "epoch": 0.03528, + "grad_norm": 0.7851080298423767, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 3528 + }, + { + "epoch": 0.03529, + "grad_norm": 0.7726654410362244, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 3529 + }, + { + "epoch": 0.0353, + "grad_norm": 0.8136371374130249, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 3530 + }, + { + "epoch": 0.03531, + "grad_norm": 0.9276587963104248, + "learning_rate": 0.003, + "loss": 4.146, + "step": 3531 + }, + { + "epoch": 0.03532, + "grad_norm": 1.0662339925765991, + "learning_rate": 0.003, + "loss": 4.1553, + "step": 3532 + }, + { + "epoch": 0.03533, + "grad_norm": 0.9835387468338013, + "learning_rate": 0.003, + "loss": 4.154, + "step": 3533 + }, + { + "epoch": 0.03534, + "grad_norm": 0.8469735383987427, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 3534 + }, + { + "epoch": 0.03535, + "grad_norm": 0.8011348247528076, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 3535 + }, + { + "epoch": 0.03536, + "grad_norm": 0.7366276383399963, + "learning_rate": 0.003, + "loss": 4.113, + "step": 3536 + }, + { + "epoch": 0.03537, + "grad_norm": 0.7694060802459717, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 3537 + }, + { + "epoch": 0.03538, + "grad_norm": 0.7417018413543701, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 3538 + }, + { + "epoch": 0.03539, + "grad_norm": 0.6806791424751282, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 3539 + }, + { + "epoch": 0.0354, + "grad_norm": 0.7284207940101624, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 3540 + }, + { + "epoch": 0.03541, + "grad_norm": 0.8043148517608643, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 3541 + }, + { + "epoch": 0.03542, + "grad_norm": 0.9069803357124329, + "learning_rate": 0.003, + "loss": 4.121, + "step": 3542 + }, + { + "epoch": 0.03543, + "grad_norm": 0.8361301422119141, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 3543 + }, + { + "epoch": 0.03544, + "grad_norm": 0.6755576133728027, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 3544 + }, + { + "epoch": 0.03545, + "grad_norm": 0.6478848457336426, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 3545 + }, + { + "epoch": 0.03546, + "grad_norm": 0.6325247287750244, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 3546 + }, + { + "epoch": 0.03547, + "grad_norm": 0.551747739315033, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 3547 + }, + { + "epoch": 0.03548, + "grad_norm": 0.550105094909668, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 3548 + }, + { + "epoch": 0.03549, + "grad_norm": 0.5281518697738647, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 3549 + }, + { + "epoch": 0.0355, + "grad_norm": 0.5446397066116333, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 3550 + }, + { + "epoch": 0.03551, + "grad_norm": 0.5065484046936035, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 3551 + }, + { + "epoch": 0.03552, + "grad_norm": 0.5231883525848389, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 3552 + }, + { + "epoch": 0.03553, + "grad_norm": 0.5496423840522766, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 3553 + }, + { + "epoch": 0.03554, + "grad_norm": 0.5717417597770691, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 3554 + }, + { + "epoch": 0.03555, + "grad_norm": 0.5578296184539795, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 3555 + }, + { + "epoch": 0.03556, + "grad_norm": 0.5329107642173767, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 3556 + }, + { + "epoch": 0.03557, + "grad_norm": 0.6033030152320862, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 3557 + }, + { + "epoch": 0.03558, + "grad_norm": 0.7457759380340576, + "learning_rate": 0.003, + "loss": 4.135, + "step": 3558 + }, + { + "epoch": 0.03559, + "grad_norm": 0.9078201055526733, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 3559 + }, + { + "epoch": 0.0356, + "grad_norm": 0.848138689994812, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 3560 + }, + { + "epoch": 0.03561, + "grad_norm": 0.6350159645080566, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 3561 + }, + { + "epoch": 0.03562, + "grad_norm": 0.6772452592849731, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 3562 + }, + { + "epoch": 0.03563, + "grad_norm": 0.7804591655731201, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 3563 + }, + { + "epoch": 0.03564, + "grad_norm": 0.725709080696106, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 3564 + }, + { + "epoch": 0.03565, + "grad_norm": 0.6160607933998108, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 3565 + }, + { + "epoch": 0.03566, + "grad_norm": 0.6539621353149414, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 3566 + }, + { + "epoch": 0.03567, + "grad_norm": 0.7242776155471802, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 3567 + }, + { + "epoch": 0.03568, + "grad_norm": 0.623746395111084, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 3568 + }, + { + "epoch": 0.03569, + "grad_norm": 0.570114016532898, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 3569 + }, + { + "epoch": 0.0357, + "grad_norm": 0.5575445294380188, + "learning_rate": 0.003, + "loss": 4.098, + "step": 3570 + }, + { + "epoch": 0.03571, + "grad_norm": 0.5092292428016663, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 3571 + }, + { + "epoch": 0.03572, + "grad_norm": 0.5417683720588684, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 3572 + }, + { + "epoch": 0.03573, + "grad_norm": 0.5097172856330872, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 3573 + }, + { + "epoch": 0.03574, + "grad_norm": 0.4699859917163849, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 3574 + }, + { + "epoch": 0.03575, + "grad_norm": 0.47701337933540344, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 3575 + }, + { + "epoch": 0.03576, + "grad_norm": 0.48855581879615784, + "learning_rate": 0.003, + "loss": 4.079, + "step": 3576 + }, + { + "epoch": 0.03577, + "grad_norm": 0.5532284379005432, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 3577 + }, + { + "epoch": 0.03578, + "grad_norm": 0.6470955610275269, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 3578 + }, + { + "epoch": 0.03579, + "grad_norm": 0.7264450788497925, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 3579 + }, + { + "epoch": 0.0358, + "grad_norm": 0.8104549646377563, + "learning_rate": 0.003, + "loss": 4.083, + "step": 3580 + }, + { + "epoch": 0.03581, + "grad_norm": 0.7132370471954346, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 3581 + }, + { + "epoch": 0.03582, + "grad_norm": 0.5359878540039062, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 3582 + }, + { + "epoch": 0.03583, + "grad_norm": 0.45527881383895874, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 3583 + }, + { + "epoch": 0.03584, + "grad_norm": 0.5485943555831909, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 3584 + }, + { + "epoch": 0.03585, + "grad_norm": 0.5606132745742798, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 3585 + }, + { + "epoch": 0.03586, + "grad_norm": 0.5852906703948975, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 3586 + }, + { + "epoch": 0.03587, + "grad_norm": 0.6464812159538269, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 3587 + }, + { + "epoch": 0.03588, + "grad_norm": 0.6532333493232727, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 3588 + }, + { + "epoch": 0.03589, + "grad_norm": 0.729999303817749, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 3589 + }, + { + "epoch": 0.0359, + "grad_norm": 0.716708779335022, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 3590 + }, + { + "epoch": 0.03591, + "grad_norm": 0.6319497227668762, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 3591 + }, + { + "epoch": 0.03592, + "grad_norm": 0.7355323433876038, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 3592 + }, + { + "epoch": 0.03593, + "grad_norm": 0.74644935131073, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 3593 + }, + { + "epoch": 0.03594, + "grad_norm": 0.7125880122184753, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 3594 + }, + { + "epoch": 0.03595, + "grad_norm": 0.6679682731628418, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 3595 + }, + { + "epoch": 0.03596, + "grad_norm": 0.6546993851661682, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 3596 + }, + { + "epoch": 0.03597, + "grad_norm": 0.6869426369667053, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 3597 + }, + { + "epoch": 0.03598, + "grad_norm": 0.6870181560516357, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 3598 + }, + { + "epoch": 0.03599, + "grad_norm": 0.6337956190109253, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 3599 + }, + { + "epoch": 0.036, + "grad_norm": 0.7691097855567932, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 3600 + }, + { + "epoch": 0.03601, + "grad_norm": 0.938689649105072, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 3601 + }, + { + "epoch": 0.03602, + "grad_norm": 1.0124480724334717, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 3602 + }, + { + "epoch": 0.03603, + "grad_norm": 1.0977243185043335, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 3603 + }, + { + "epoch": 0.03604, + "grad_norm": 0.7794393301010132, + "learning_rate": 0.003, + "loss": 4.131, + "step": 3604 + }, + { + "epoch": 0.03605, + "grad_norm": 0.5185660719871521, + "learning_rate": 0.003, + "loss": 4.114, + "step": 3605 + }, + { + "epoch": 0.03606, + "grad_norm": 0.5366716384887695, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 3606 + }, + { + "epoch": 0.03607, + "grad_norm": 0.5930801630020142, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 3607 + }, + { + "epoch": 0.03608, + "grad_norm": 0.5576140284538269, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 3608 + }, + { + "epoch": 0.03609, + "grad_norm": 0.5525704622268677, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 3609 + }, + { + "epoch": 0.0361, + "grad_norm": 0.5817636847496033, + "learning_rate": 0.003, + "loss": 4.126, + "step": 3610 + }, + { + "epoch": 0.03611, + "grad_norm": 0.5388094782829285, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 3611 + }, + { + "epoch": 0.03612, + "grad_norm": 0.5407794117927551, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 3612 + }, + { + "epoch": 0.03613, + "grad_norm": 0.7358308434486389, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 3613 + }, + { + "epoch": 0.03614, + "grad_norm": 0.7770156264305115, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 3614 + }, + { + "epoch": 0.03615, + "grad_norm": 0.7676093578338623, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 3615 + }, + { + "epoch": 0.03616, + "grad_norm": 0.8209259510040283, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 3616 + }, + { + "epoch": 0.03617, + "grad_norm": 0.7589568495750427, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 3617 + }, + { + "epoch": 0.03618, + "grad_norm": 0.7976088523864746, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 3618 + }, + { + "epoch": 0.03619, + "grad_norm": 0.7740630507469177, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 3619 + }, + { + "epoch": 0.0362, + "grad_norm": 0.8788385391235352, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 3620 + }, + { + "epoch": 0.03621, + "grad_norm": 1.0235563516616821, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 3621 + }, + { + "epoch": 0.03622, + "grad_norm": 1.101138949394226, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 3622 + }, + { + "epoch": 0.03623, + "grad_norm": 0.782410204410553, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 3623 + }, + { + "epoch": 0.03624, + "grad_norm": 0.5812339186668396, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 3624 + }, + { + "epoch": 0.03625, + "grad_norm": 0.679203987121582, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 3625 + }, + { + "epoch": 0.03626, + "grad_norm": 0.7614912986755371, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 3626 + }, + { + "epoch": 0.03627, + "grad_norm": 0.7893409729003906, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 3627 + }, + { + "epoch": 0.03628, + "grad_norm": 0.6757506132125854, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 3628 + }, + { + "epoch": 0.03629, + "grad_norm": 0.5744339227676392, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 3629 + }, + { + "epoch": 0.0363, + "grad_norm": 0.5694675445556641, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 3630 + }, + { + "epoch": 0.03631, + "grad_norm": 0.5684549808502197, + "learning_rate": 0.003, + "loss": 4.107, + "step": 3631 + }, + { + "epoch": 0.03632, + "grad_norm": 0.7418304085731506, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 3632 + }, + { + "epoch": 0.03633, + "grad_norm": 0.8812902569770813, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 3633 + }, + { + "epoch": 0.03634, + "grad_norm": 0.960989236831665, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 3634 + }, + { + "epoch": 0.03635, + "grad_norm": 0.8940312266349792, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 3635 + }, + { + "epoch": 0.03636, + "grad_norm": 0.7907798886299133, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 3636 + }, + { + "epoch": 0.03637, + "grad_norm": 0.7844222187995911, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 3637 + }, + { + "epoch": 0.03638, + "grad_norm": 0.8917554616928101, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 3638 + }, + { + "epoch": 0.03639, + "grad_norm": 0.8074480891227722, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 3639 + }, + { + "epoch": 0.0364, + "grad_norm": 0.9071139097213745, + "learning_rate": 0.003, + "loss": 4.149, + "step": 3640 + }, + { + "epoch": 0.03641, + "grad_norm": 0.749542236328125, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 3641 + }, + { + "epoch": 0.03642, + "grad_norm": 0.7375038862228394, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 3642 + }, + { + "epoch": 0.03643, + "grad_norm": 0.7622751593589783, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 3643 + }, + { + "epoch": 0.03644, + "grad_norm": 0.7410848140716553, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 3644 + }, + { + "epoch": 0.03645, + "grad_norm": 0.7992194890975952, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 3645 + }, + { + "epoch": 0.03646, + "grad_norm": 0.7577066421508789, + "learning_rate": 0.003, + "loss": 4.124, + "step": 3646 + }, + { + "epoch": 0.03647, + "grad_norm": 0.7054228782653809, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 3647 + }, + { + "epoch": 0.03648, + "grad_norm": 0.664077877998352, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 3648 + }, + { + "epoch": 0.03649, + "grad_norm": 0.5441340208053589, + "learning_rate": 0.003, + "loss": 4.126, + "step": 3649 + }, + { + "epoch": 0.0365, + "grad_norm": 0.5746814012527466, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 3650 + }, + { + "epoch": 0.03651, + "grad_norm": 0.6534304618835449, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 3651 + }, + { + "epoch": 0.03652, + "grad_norm": 0.6664396524429321, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 3652 + }, + { + "epoch": 0.03653, + "grad_norm": 0.6238181591033936, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 3653 + }, + { + "epoch": 0.03654, + "grad_norm": 0.5265892148017883, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 3654 + }, + { + "epoch": 0.03655, + "grad_norm": 0.4858669936656952, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 3655 + }, + { + "epoch": 0.03656, + "grad_norm": 0.41965577006340027, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 3656 + }, + { + "epoch": 0.03657, + "grad_norm": 0.4169410765171051, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 3657 + }, + { + "epoch": 0.03658, + "grad_norm": 0.4434378445148468, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 3658 + }, + { + "epoch": 0.03659, + "grad_norm": 0.4580453634262085, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 3659 + }, + { + "epoch": 0.0366, + "grad_norm": 0.4966508448123932, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 3660 + }, + { + "epoch": 0.03661, + "grad_norm": 0.6382727026939392, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 3661 + }, + { + "epoch": 0.03662, + "grad_norm": 0.8349222540855408, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 3662 + }, + { + "epoch": 0.03663, + "grad_norm": 0.931993842124939, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 3663 + }, + { + "epoch": 0.03664, + "grad_norm": 0.837236762046814, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 3664 + }, + { + "epoch": 0.03665, + "grad_norm": 0.6735215783119202, + "learning_rate": 0.003, + "loss": 4.12, + "step": 3665 + }, + { + "epoch": 0.03666, + "grad_norm": 0.7142363786697388, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 3666 + }, + { + "epoch": 0.03667, + "grad_norm": 0.7672339081764221, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 3667 + }, + { + "epoch": 0.03668, + "grad_norm": 0.660099446773529, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 3668 + }, + { + "epoch": 0.03669, + "grad_norm": 0.664168119430542, + "learning_rate": 0.003, + "loss": 4.08, + "step": 3669 + }, + { + "epoch": 0.0367, + "grad_norm": 0.5606299042701721, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 3670 + }, + { + "epoch": 0.03671, + "grad_norm": 0.5958951711654663, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 3671 + }, + { + "epoch": 0.03672, + "grad_norm": 0.7330699563026428, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 3672 + }, + { + "epoch": 0.03673, + "grad_norm": 0.7183411717414856, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 3673 + }, + { + "epoch": 0.03674, + "grad_norm": 0.6538311243057251, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 3674 + }, + { + "epoch": 0.03675, + "grad_norm": 0.6396141052246094, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 3675 + }, + { + "epoch": 0.03676, + "grad_norm": 0.5857633948326111, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 3676 + }, + { + "epoch": 0.03677, + "grad_norm": 0.5811484456062317, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 3677 + }, + { + "epoch": 0.03678, + "grad_norm": 0.5666897892951965, + "learning_rate": 0.003, + "loss": 4.121, + "step": 3678 + }, + { + "epoch": 0.03679, + "grad_norm": 0.6068264245986938, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 3679 + }, + { + "epoch": 0.0368, + "grad_norm": 0.6787545084953308, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 3680 + }, + { + "epoch": 0.03681, + "grad_norm": 0.8176649212837219, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 3681 + }, + { + "epoch": 0.03682, + "grad_norm": 1.0537775754928589, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 3682 + }, + { + "epoch": 0.03683, + "grad_norm": 0.9726674556732178, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 3683 + }, + { + "epoch": 0.03684, + "grad_norm": 0.7920905351638794, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 3684 + }, + { + "epoch": 0.03685, + "grad_norm": 0.6423739194869995, + "learning_rate": 0.003, + "loss": 4.094, + "step": 3685 + }, + { + "epoch": 0.03686, + "grad_norm": 0.6905859112739563, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 3686 + }, + { + "epoch": 0.03687, + "grad_norm": 0.7596766948699951, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 3687 + }, + { + "epoch": 0.03688, + "grad_norm": 0.7518109679222107, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 3688 + }, + { + "epoch": 0.03689, + "grad_norm": 0.764574408531189, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 3689 + }, + { + "epoch": 0.0369, + "grad_norm": 0.721488893032074, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 3690 + }, + { + "epoch": 0.03691, + "grad_norm": 0.7047094702720642, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 3691 + }, + { + "epoch": 0.03692, + "grad_norm": 0.8295153975486755, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 3692 + }, + { + "epoch": 0.03693, + "grad_norm": 0.899239182472229, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 3693 + }, + { + "epoch": 0.03694, + "grad_norm": 0.9156467914581299, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 3694 + }, + { + "epoch": 0.03695, + "grad_norm": 0.8504350781440735, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 3695 + }, + { + "epoch": 0.03696, + "grad_norm": 0.7041043639183044, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 3696 + }, + { + "epoch": 0.03697, + "grad_norm": 0.643920361995697, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 3697 + }, + { + "epoch": 0.03698, + "grad_norm": 0.7144352197647095, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 3698 + }, + { + "epoch": 0.03699, + "grad_norm": 0.7363175749778748, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 3699 + }, + { + "epoch": 0.037, + "grad_norm": 0.649755597114563, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 3700 + }, + { + "epoch": 0.03701, + "grad_norm": 0.5577970743179321, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 3701 + }, + { + "epoch": 0.03702, + "grad_norm": 0.5865651369094849, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 3702 + }, + { + "epoch": 0.03703, + "grad_norm": 0.6238481402397156, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 3703 + }, + { + "epoch": 0.03704, + "grad_norm": 0.630403995513916, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 3704 + }, + { + "epoch": 0.03705, + "grad_norm": 0.6488320231437683, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 3705 + }, + { + "epoch": 0.03706, + "grad_norm": 0.6384865045547485, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 3706 + }, + { + "epoch": 0.03707, + "grad_norm": 0.6757420897483826, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 3707 + }, + { + "epoch": 0.03708, + "grad_norm": 0.7628651857376099, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 3708 + }, + { + "epoch": 0.03709, + "grad_norm": 0.6610810160636902, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 3709 + }, + { + "epoch": 0.0371, + "grad_norm": 0.6568567156791687, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 3710 + }, + { + "epoch": 0.03711, + "grad_norm": 0.6123287677764893, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 3711 + }, + { + "epoch": 0.03712, + "grad_norm": 0.5023053288459778, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 3712 + }, + { + "epoch": 0.03713, + "grad_norm": 0.561567485332489, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 3713 + }, + { + "epoch": 0.03714, + "grad_norm": 0.6204001903533936, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 3714 + }, + { + "epoch": 0.03715, + "grad_norm": 0.7001908421516418, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 3715 + }, + { + "epoch": 0.03716, + "grad_norm": 0.7154534459114075, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 3716 + }, + { + "epoch": 0.03717, + "grad_norm": 0.8945140242576599, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 3717 + }, + { + "epoch": 0.03718, + "grad_norm": 0.9586088061332703, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 3718 + }, + { + "epoch": 0.03719, + "grad_norm": 0.8600691556930542, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 3719 + }, + { + "epoch": 0.0372, + "grad_norm": 0.7804197669029236, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 3720 + }, + { + "epoch": 0.03721, + "grad_norm": 0.9556355476379395, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 3721 + }, + { + "epoch": 0.03722, + "grad_norm": 0.9721794724464417, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 3722 + }, + { + "epoch": 0.03723, + "grad_norm": 0.9020527005195618, + "learning_rate": 0.003, + "loss": 4.1534, + "step": 3723 + }, + { + "epoch": 0.03724, + "grad_norm": 0.7264366745948792, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 3724 + }, + { + "epoch": 0.03725, + "grad_norm": 0.6015045046806335, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 3725 + }, + { + "epoch": 0.03726, + "grad_norm": 0.5826119780540466, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 3726 + }, + { + "epoch": 0.03727, + "grad_norm": 0.5803874731063843, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 3727 + }, + { + "epoch": 0.03728, + "grad_norm": 0.5483809113502502, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 3728 + }, + { + "epoch": 0.03729, + "grad_norm": 0.49769601225852966, + "learning_rate": 0.003, + "loss": 4.117, + "step": 3729 + }, + { + "epoch": 0.0373, + "grad_norm": 0.5014387965202332, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 3730 + }, + { + "epoch": 0.03731, + "grad_norm": 0.5072956681251526, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 3731 + }, + { + "epoch": 0.03732, + "grad_norm": 0.5933743119239807, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 3732 + }, + { + "epoch": 0.03733, + "grad_norm": 0.644372820854187, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 3733 + }, + { + "epoch": 0.03734, + "grad_norm": 0.6615225076675415, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 3734 + }, + { + "epoch": 0.03735, + "grad_norm": 0.7122288942337036, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 3735 + }, + { + "epoch": 0.03736, + "grad_norm": 0.7745519876480103, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 3736 + }, + { + "epoch": 0.03737, + "grad_norm": 0.7451056241989136, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 3737 + }, + { + "epoch": 0.03738, + "grad_norm": 0.6421622037887573, + "learning_rate": 0.003, + "loss": 4.127, + "step": 3738 + }, + { + "epoch": 0.03739, + "grad_norm": 0.6206536889076233, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 3739 + }, + { + "epoch": 0.0374, + "grad_norm": 0.6064566373825073, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 3740 + }, + { + "epoch": 0.03741, + "grad_norm": 0.726904034614563, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 3741 + }, + { + "epoch": 0.03742, + "grad_norm": 0.8068590760231018, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 3742 + }, + { + "epoch": 0.03743, + "grad_norm": 0.6390416622161865, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 3743 + }, + { + "epoch": 0.03744, + "grad_norm": 0.6150771975517273, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 3744 + }, + { + "epoch": 0.03745, + "grad_norm": 0.664985716342926, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 3745 + }, + { + "epoch": 0.03746, + "grad_norm": 0.7317749857902527, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 3746 + }, + { + "epoch": 0.03747, + "grad_norm": 0.7559656500816345, + "learning_rate": 0.003, + "loss": 4.119, + "step": 3747 + }, + { + "epoch": 0.03748, + "grad_norm": 0.6956848502159119, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 3748 + }, + { + "epoch": 0.03749, + "grad_norm": 0.8748852014541626, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 3749 + }, + { + "epoch": 0.0375, + "grad_norm": 0.8952407240867615, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 3750 + }, + { + "epoch": 0.03751, + "grad_norm": 0.9176827669143677, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 3751 + }, + { + "epoch": 0.03752, + "grad_norm": 1.0543770790100098, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 3752 + }, + { + "epoch": 0.03753, + "grad_norm": 0.9552967548370361, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 3753 + }, + { + "epoch": 0.03754, + "grad_norm": 0.7647460699081421, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 3754 + }, + { + "epoch": 0.03755, + "grad_norm": 0.8259355425834656, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 3755 + }, + { + "epoch": 0.03756, + "grad_norm": 0.8404808044433594, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 3756 + }, + { + "epoch": 0.03757, + "grad_norm": 0.7295672297477722, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 3757 + }, + { + "epoch": 0.03758, + "grad_norm": 0.6218863129615784, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 3758 + }, + { + "epoch": 0.03759, + "grad_norm": 0.5583391785621643, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 3759 + }, + { + "epoch": 0.0376, + "grad_norm": 0.6009096503257751, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 3760 + }, + { + "epoch": 0.03761, + "grad_norm": 0.7737796902656555, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 3761 + }, + { + "epoch": 0.03762, + "grad_norm": 0.8101951479911804, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 3762 + }, + { + "epoch": 0.03763, + "grad_norm": 0.6714347004890442, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 3763 + }, + { + "epoch": 0.03764, + "grad_norm": 0.4961341321468353, + "learning_rate": 0.003, + "loss": 4.125, + "step": 3764 + }, + { + "epoch": 0.03765, + "grad_norm": 0.5392913818359375, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 3765 + }, + { + "epoch": 0.03766, + "grad_norm": 0.6497456431388855, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 3766 + }, + { + "epoch": 0.03767, + "grad_norm": 0.6586916446685791, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 3767 + }, + { + "epoch": 0.03768, + "grad_norm": 0.6380057334899902, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 3768 + }, + { + "epoch": 0.03769, + "grad_norm": 0.6120327711105347, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 3769 + }, + { + "epoch": 0.0377, + "grad_norm": 0.5535104274749756, + "learning_rate": 0.003, + "loss": 4.074, + "step": 3770 + }, + { + "epoch": 0.03771, + "grad_norm": 0.5369603633880615, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 3771 + }, + { + "epoch": 0.03772, + "grad_norm": 0.5378307700157166, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 3772 + }, + { + "epoch": 0.03773, + "grad_norm": 0.5350049734115601, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 3773 + }, + { + "epoch": 0.03774, + "grad_norm": 0.48990651965141296, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 3774 + }, + { + "epoch": 0.03775, + "grad_norm": 0.4350149929523468, + "learning_rate": 0.003, + "loss": 4.112, + "step": 3775 + }, + { + "epoch": 0.03776, + "grad_norm": 0.455160915851593, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 3776 + }, + { + "epoch": 0.03777, + "grad_norm": 0.4597632586956024, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 3777 + }, + { + "epoch": 0.03778, + "grad_norm": 0.5477138757705688, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 3778 + }, + { + "epoch": 0.03779, + "grad_norm": 0.7227920293807983, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 3779 + }, + { + "epoch": 0.0378, + "grad_norm": 0.9108830690383911, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 3780 + }, + { + "epoch": 0.03781, + "grad_norm": 1.0170267820358276, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 3781 + }, + { + "epoch": 0.03782, + "grad_norm": 0.9500049948692322, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 3782 + }, + { + "epoch": 0.03783, + "grad_norm": 0.7215931415557861, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 3783 + }, + { + "epoch": 0.03784, + "grad_norm": 0.6211499571800232, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 3784 + }, + { + "epoch": 0.03785, + "grad_norm": 0.6421756148338318, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 3785 + }, + { + "epoch": 0.03786, + "grad_norm": 0.6241987943649292, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 3786 + }, + { + "epoch": 0.03787, + "grad_norm": 0.6087479591369629, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 3787 + }, + { + "epoch": 0.03788, + "grad_norm": 0.5939789414405823, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 3788 + }, + { + "epoch": 0.03789, + "grad_norm": 0.5391459465026855, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 3789 + }, + { + "epoch": 0.0379, + "grad_norm": 0.5861315131187439, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 3790 + }, + { + "epoch": 0.03791, + "grad_norm": 0.6246805191040039, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 3791 + }, + { + "epoch": 0.03792, + "grad_norm": 0.7353853583335876, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 3792 + }, + { + "epoch": 0.03793, + "grad_norm": 0.6765242218971252, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 3793 + }, + { + "epoch": 0.03794, + "grad_norm": 0.7689555883407593, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 3794 + }, + { + "epoch": 0.03795, + "grad_norm": 0.9137020707130432, + "learning_rate": 0.003, + "loss": 4.121, + "step": 3795 + }, + { + "epoch": 0.03796, + "grad_norm": 1.0011204481124878, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 3796 + }, + { + "epoch": 0.03797, + "grad_norm": 1.0785057544708252, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 3797 + }, + { + "epoch": 0.03798, + "grad_norm": 1.0564720630645752, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 3798 + }, + { + "epoch": 0.03799, + "grad_norm": 0.9513949155807495, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 3799 + }, + { + "epoch": 0.038, + "grad_norm": 0.7991365790367126, + "learning_rate": 0.003, + "loss": 4.136, + "step": 3800 + }, + { + "epoch": 0.03801, + "grad_norm": 0.942852795124054, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 3801 + }, + { + "epoch": 0.03802, + "grad_norm": 0.9442994594573975, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 3802 + }, + { + "epoch": 0.03803, + "grad_norm": 0.917535662651062, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 3803 + }, + { + "epoch": 0.03804, + "grad_norm": 1.0844844579696655, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 3804 + }, + { + "epoch": 0.03805, + "grad_norm": 0.8430477380752563, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 3805 + }, + { + "epoch": 0.03806, + "grad_norm": 0.8205291032791138, + "learning_rate": 0.003, + "loss": 4.1397, + "step": 3806 + }, + { + "epoch": 0.03807, + "grad_norm": 0.8324175477027893, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 3807 + }, + { + "epoch": 0.03808, + "grad_norm": 0.8151392936706543, + "learning_rate": 0.003, + "loss": 4.146, + "step": 3808 + }, + { + "epoch": 0.03809, + "grad_norm": 0.6972253918647766, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 3809 + }, + { + "epoch": 0.0381, + "grad_norm": 0.7419064044952393, + "learning_rate": 0.003, + "loss": 4.109, + "step": 3810 + }, + { + "epoch": 0.03811, + "grad_norm": 0.72950679063797, + "learning_rate": 0.003, + "loss": 4.1397, + "step": 3811 + }, + { + "epoch": 0.03812, + "grad_norm": 0.7456114888191223, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 3812 + }, + { + "epoch": 0.03813, + "grad_norm": 0.868884801864624, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 3813 + }, + { + "epoch": 0.03814, + "grad_norm": 1.0820525884628296, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 3814 + }, + { + "epoch": 0.03815, + "grad_norm": 1.0806629657745361, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 3815 + }, + { + "epoch": 0.03816, + "grad_norm": 0.820555567741394, + "learning_rate": 0.003, + "loss": 4.126, + "step": 3816 + }, + { + "epoch": 0.03817, + "grad_norm": 0.7203388810157776, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 3817 + }, + { + "epoch": 0.03818, + "grad_norm": 0.6214293241500854, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 3818 + }, + { + "epoch": 0.03819, + "grad_norm": 0.5392916798591614, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 3819 + }, + { + "epoch": 0.0382, + "grad_norm": 0.5882942080497742, + "learning_rate": 0.003, + "loss": 4.1451, + "step": 3820 + }, + { + "epoch": 0.03821, + "grad_norm": 0.6876145005226135, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 3821 + }, + { + "epoch": 0.03822, + "grad_norm": 0.6543509364128113, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 3822 + }, + { + "epoch": 0.03823, + "grad_norm": 0.6853277087211609, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 3823 + }, + { + "epoch": 0.03824, + "grad_norm": 0.6686453819274902, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 3824 + }, + { + "epoch": 0.03825, + "grad_norm": 0.5822768211364746, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 3825 + }, + { + "epoch": 0.03826, + "grad_norm": 0.5187621712684631, + "learning_rate": 0.003, + "loss": 4.101, + "step": 3826 + }, + { + "epoch": 0.03827, + "grad_norm": 0.48330262303352356, + "learning_rate": 0.003, + "loss": 4.125, + "step": 3827 + }, + { + "epoch": 0.03828, + "grad_norm": 0.525285542011261, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 3828 + }, + { + "epoch": 0.03829, + "grad_norm": 0.5435876846313477, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 3829 + }, + { + "epoch": 0.0383, + "grad_norm": 0.5318546295166016, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 3830 + }, + { + "epoch": 0.03831, + "grad_norm": 0.5064510703086853, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 3831 + }, + { + "epoch": 0.03832, + "grad_norm": 0.5135751962661743, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 3832 + }, + { + "epoch": 0.03833, + "grad_norm": 0.5150834321975708, + "learning_rate": 0.003, + "loss": 4.091, + "step": 3833 + }, + { + "epoch": 0.03834, + "grad_norm": 0.5075490474700928, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 3834 + }, + { + "epoch": 0.03835, + "grad_norm": 0.47964203357696533, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 3835 + }, + { + "epoch": 0.03836, + "grad_norm": 0.5092160105705261, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 3836 + }, + { + "epoch": 0.03837, + "grad_norm": 0.5252288579940796, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 3837 + }, + { + "epoch": 0.03838, + "grad_norm": 0.4719623029232025, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 3838 + }, + { + "epoch": 0.03839, + "grad_norm": 0.525349497795105, + "learning_rate": 0.003, + "loss": 4.09, + "step": 3839 + }, + { + "epoch": 0.0384, + "grad_norm": 0.5973276495933533, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 3840 + }, + { + "epoch": 0.03841, + "grad_norm": 0.7389470338821411, + "learning_rate": 0.003, + "loss": 4.101, + "step": 3841 + }, + { + "epoch": 0.03842, + "grad_norm": 0.8890230059623718, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 3842 + }, + { + "epoch": 0.03843, + "grad_norm": 0.9227127432823181, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 3843 + }, + { + "epoch": 0.03844, + "grad_norm": 0.8832306861877441, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 3844 + }, + { + "epoch": 0.03845, + "grad_norm": 0.711075484752655, + "learning_rate": 0.003, + "loss": 4.104, + "step": 3845 + }, + { + "epoch": 0.03846, + "grad_norm": 0.5776762962341309, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 3846 + }, + { + "epoch": 0.03847, + "grad_norm": 0.6540825963020325, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 3847 + }, + { + "epoch": 0.03848, + "grad_norm": 0.6692461967468262, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 3848 + }, + { + "epoch": 0.03849, + "grad_norm": 0.6830011010169983, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 3849 + }, + { + "epoch": 0.0385, + "grad_norm": 0.6109870076179504, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 3850 + }, + { + "epoch": 0.03851, + "grad_norm": 0.6136434078216553, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 3851 + }, + { + "epoch": 0.03852, + "grad_norm": 0.6604334712028503, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 3852 + }, + { + "epoch": 0.03853, + "grad_norm": 0.59455806016922, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 3853 + }, + { + "epoch": 0.03854, + "grad_norm": 0.575217604637146, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 3854 + }, + { + "epoch": 0.03855, + "grad_norm": 0.6192313432693481, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 3855 + }, + { + "epoch": 0.03856, + "grad_norm": 0.6025145649909973, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 3856 + }, + { + "epoch": 0.03857, + "grad_norm": 0.6573590636253357, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 3857 + }, + { + "epoch": 0.03858, + "grad_norm": 0.6135656833648682, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 3858 + }, + { + "epoch": 0.03859, + "grad_norm": 0.6176097989082336, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 3859 + }, + { + "epoch": 0.0386, + "grad_norm": 0.7532007098197937, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 3860 + }, + { + "epoch": 0.03861, + "grad_norm": 0.8719074130058289, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 3861 + }, + { + "epoch": 0.03862, + "grad_norm": 1.079362154006958, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 3862 + }, + { + "epoch": 0.03863, + "grad_norm": 0.9231005311012268, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 3863 + }, + { + "epoch": 0.03864, + "grad_norm": 0.776627242565155, + "learning_rate": 0.003, + "loss": 4.105, + "step": 3864 + }, + { + "epoch": 0.03865, + "grad_norm": 0.6355084180831909, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 3865 + }, + { + "epoch": 0.03866, + "grad_norm": 0.6550437211990356, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 3866 + }, + { + "epoch": 0.03867, + "grad_norm": 0.8415738940238953, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 3867 + }, + { + "epoch": 0.03868, + "grad_norm": 0.8901708126068115, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 3868 + }, + { + "epoch": 0.03869, + "grad_norm": 0.697814404964447, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 3869 + }, + { + "epoch": 0.0387, + "grad_norm": 0.6080979108810425, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 3870 + }, + { + "epoch": 0.03871, + "grad_norm": 0.6336243748664856, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 3871 + }, + { + "epoch": 0.03872, + "grad_norm": 0.6221011877059937, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 3872 + }, + { + "epoch": 0.03873, + "grad_norm": 0.6594142317771912, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 3873 + }, + { + "epoch": 0.03874, + "grad_norm": 0.5940245389938354, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 3874 + }, + { + "epoch": 0.03875, + "grad_norm": 0.47639554738998413, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 3875 + }, + { + "epoch": 0.03876, + "grad_norm": 0.5326244831085205, + "learning_rate": 0.003, + "loss": 4.123, + "step": 3876 + }, + { + "epoch": 0.03877, + "grad_norm": 0.5539395213127136, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 3877 + }, + { + "epoch": 0.03878, + "grad_norm": 0.604924201965332, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 3878 + }, + { + "epoch": 0.03879, + "grad_norm": 0.6767284870147705, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 3879 + }, + { + "epoch": 0.0388, + "grad_norm": 0.8324589133262634, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 3880 + }, + { + "epoch": 0.03881, + "grad_norm": 1.0423182249069214, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 3881 + }, + { + "epoch": 0.03882, + "grad_norm": 1.036028504371643, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 3882 + }, + { + "epoch": 0.03883, + "grad_norm": 0.8236280679702759, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 3883 + }, + { + "epoch": 0.03884, + "grad_norm": 0.6207178831100464, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 3884 + }, + { + "epoch": 0.03885, + "grad_norm": 0.6633720397949219, + "learning_rate": 0.003, + "loss": 4.122, + "step": 3885 + }, + { + "epoch": 0.03886, + "grad_norm": 0.6237534880638123, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 3886 + }, + { + "epoch": 0.03887, + "grad_norm": 0.49867168068885803, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 3887 + }, + { + "epoch": 0.03888, + "grad_norm": 0.4786747395992279, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 3888 + }, + { + "epoch": 0.03889, + "grad_norm": 0.5030584931373596, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 3889 + }, + { + "epoch": 0.0389, + "grad_norm": 0.6045381426811218, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 3890 + }, + { + "epoch": 0.03891, + "grad_norm": 0.6243774890899658, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 3891 + }, + { + "epoch": 0.03892, + "grad_norm": 0.7076771855354309, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 3892 + }, + { + "epoch": 0.03893, + "grad_norm": 0.6316627264022827, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 3893 + }, + { + "epoch": 0.03894, + "grad_norm": 0.6711505055427551, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 3894 + }, + { + "epoch": 0.03895, + "grad_norm": 0.7016406655311584, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 3895 + }, + { + "epoch": 0.03896, + "grad_norm": 0.6848783493041992, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 3896 + }, + { + "epoch": 0.03897, + "grad_norm": 0.701120913028717, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 3897 + }, + { + "epoch": 0.03898, + "grad_norm": 0.678917407989502, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 3898 + }, + { + "epoch": 0.03899, + "grad_norm": 0.7172825336456299, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 3899 + }, + { + "epoch": 0.039, + "grad_norm": 0.6747111082077026, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 3900 + }, + { + "epoch": 0.03901, + "grad_norm": 0.7526534795761108, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 3901 + }, + { + "epoch": 0.03902, + "grad_norm": 0.7362321615219116, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 3902 + }, + { + "epoch": 0.03903, + "grad_norm": 0.8664571642875671, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 3903 + }, + { + "epoch": 0.03904, + "grad_norm": 1.0233482122421265, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 3904 + }, + { + "epoch": 0.03905, + "grad_norm": 0.9382357001304626, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 3905 + }, + { + "epoch": 0.03906, + "grad_norm": 0.687033474445343, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 3906 + }, + { + "epoch": 0.03907, + "grad_norm": 0.6406064033508301, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 3907 + }, + { + "epoch": 0.03908, + "grad_norm": 0.6577279567718506, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 3908 + }, + { + "epoch": 0.03909, + "grad_norm": 0.6069698333740234, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 3909 + }, + { + "epoch": 0.0391, + "grad_norm": 0.625200629234314, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 3910 + }, + { + "epoch": 0.03911, + "grad_norm": 0.6532104015350342, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 3911 + }, + { + "epoch": 0.03912, + "grad_norm": 0.5893357992172241, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 3912 + }, + { + "epoch": 0.03913, + "grad_norm": 0.6330502033233643, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 3913 + }, + { + "epoch": 0.03914, + "grad_norm": 0.7185488343238831, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 3914 + }, + { + "epoch": 0.03915, + "grad_norm": 0.8594509363174438, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 3915 + }, + { + "epoch": 0.03916, + "grad_norm": 0.9027094841003418, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 3916 + }, + { + "epoch": 0.03917, + "grad_norm": 0.9196935892105103, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 3917 + }, + { + "epoch": 0.03918, + "grad_norm": 1.003941535949707, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 3918 + }, + { + "epoch": 0.03919, + "grad_norm": 1.0010123252868652, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 3919 + }, + { + "epoch": 0.0392, + "grad_norm": 0.9059627056121826, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 3920 + }, + { + "epoch": 0.03921, + "grad_norm": 0.95637047290802, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 3921 + }, + { + "epoch": 0.03922, + "grad_norm": 0.8252183198928833, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 3922 + }, + { + "epoch": 0.03923, + "grad_norm": 0.8936133980751038, + "learning_rate": 0.003, + "loss": 4.103, + "step": 3923 + }, + { + "epoch": 0.03924, + "grad_norm": 0.7836267352104187, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 3924 + }, + { + "epoch": 0.03925, + "grad_norm": 0.7859375476837158, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 3925 + }, + { + "epoch": 0.03926, + "grad_norm": 0.7374272346496582, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 3926 + }, + { + "epoch": 0.03927, + "grad_norm": 0.8471856117248535, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 3927 + }, + { + "epoch": 0.03928, + "grad_norm": 0.85481858253479, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 3928 + }, + { + "epoch": 0.03929, + "grad_norm": 0.907612144947052, + "learning_rate": 0.003, + "loss": 4.1445, + "step": 3929 + }, + { + "epoch": 0.0393, + "grad_norm": 0.7699798345565796, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 3930 + }, + { + "epoch": 0.03931, + "grad_norm": 0.6045697331428528, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 3931 + }, + { + "epoch": 0.03932, + "grad_norm": 0.6712193489074707, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 3932 + }, + { + "epoch": 0.03933, + "grad_norm": 0.8346224427223206, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 3933 + }, + { + "epoch": 0.03934, + "grad_norm": 0.8903274536132812, + "learning_rate": 0.003, + "loss": 4.12, + "step": 3934 + }, + { + "epoch": 0.03935, + "grad_norm": 0.7318519353866577, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 3935 + }, + { + "epoch": 0.03936, + "grad_norm": 0.6603662967681885, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 3936 + }, + { + "epoch": 0.03937, + "grad_norm": 0.6446847915649414, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 3937 + }, + { + "epoch": 0.03938, + "grad_norm": 0.6806734204292297, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 3938 + }, + { + "epoch": 0.03939, + "grad_norm": 0.5945215225219727, + "learning_rate": 0.003, + "loss": 4.088, + "step": 3939 + }, + { + "epoch": 0.0394, + "grad_norm": 0.5750917196273804, + "learning_rate": 0.003, + "loss": 4.116, + "step": 3940 + }, + { + "epoch": 0.03941, + "grad_norm": 0.5394409894943237, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 3941 + }, + { + "epoch": 0.03942, + "grad_norm": 0.49402347207069397, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 3942 + }, + { + "epoch": 0.03943, + "grad_norm": 0.39833274483680725, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 3943 + }, + { + "epoch": 0.03944, + "grad_norm": 0.4659424126148224, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 3944 + }, + { + "epoch": 0.03945, + "grad_norm": 0.4315028488636017, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 3945 + }, + { + "epoch": 0.03946, + "grad_norm": 0.40673503279685974, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 3946 + }, + { + "epoch": 0.03947, + "grad_norm": 0.38524332642555237, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 3947 + }, + { + "epoch": 0.03948, + "grad_norm": 0.33258089423179626, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 3948 + }, + { + "epoch": 0.03949, + "grad_norm": 0.38337442278862, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 3949 + }, + { + "epoch": 0.0395, + "grad_norm": 0.46301016211509705, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 3950 + }, + { + "epoch": 0.03951, + "grad_norm": 0.7521165013313293, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 3951 + }, + { + "epoch": 0.03952, + "grad_norm": 1.1936590671539307, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 3952 + }, + { + "epoch": 0.03953, + "grad_norm": 0.9014281034469604, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 3953 + }, + { + "epoch": 0.03954, + "grad_norm": 0.5299249887466431, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 3954 + }, + { + "epoch": 0.03955, + "grad_norm": 0.621684193611145, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 3955 + }, + { + "epoch": 0.03956, + "grad_norm": 0.7184932827949524, + "learning_rate": 0.003, + "loss": 4.101, + "step": 3956 + }, + { + "epoch": 0.03957, + "grad_norm": 0.6237415075302124, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 3957 + }, + { + "epoch": 0.03958, + "grad_norm": 0.5475152134895325, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 3958 + }, + { + "epoch": 0.03959, + "grad_norm": 0.5959770083427429, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 3959 + }, + { + "epoch": 0.0396, + "grad_norm": 0.6074774861335754, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 3960 + }, + { + "epoch": 0.03961, + "grad_norm": 0.6236320734024048, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 3961 + }, + { + "epoch": 0.03962, + "grad_norm": 0.6013439297676086, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 3962 + }, + { + "epoch": 0.03963, + "grad_norm": 0.6263412237167358, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 3963 + }, + { + "epoch": 0.03964, + "grad_norm": 0.656822919845581, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 3964 + }, + { + "epoch": 0.03965, + "grad_norm": 0.6711769700050354, + "learning_rate": 0.003, + "loss": 4.096, + "step": 3965 + }, + { + "epoch": 0.03966, + "grad_norm": 0.7558622360229492, + "learning_rate": 0.003, + "loss": 4.1394, + "step": 3966 + }, + { + "epoch": 0.03967, + "grad_norm": 0.8678721785545349, + "learning_rate": 0.003, + "loss": 4.101, + "step": 3967 + }, + { + "epoch": 0.03968, + "grad_norm": 0.9608453512191772, + "learning_rate": 0.003, + "loss": 4.1351, + "step": 3968 + }, + { + "epoch": 0.03969, + "grad_norm": 0.92333984375, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 3969 + }, + { + "epoch": 0.0397, + "grad_norm": 1.0910831689834595, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 3970 + }, + { + "epoch": 0.03971, + "grad_norm": 0.8948251008987427, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 3971 + }, + { + "epoch": 0.03972, + "grad_norm": 0.8917779922485352, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 3972 + }, + { + "epoch": 0.03973, + "grad_norm": 0.9352078437805176, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 3973 + }, + { + "epoch": 0.03974, + "grad_norm": 0.8667543530464172, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 3974 + }, + { + "epoch": 0.03975, + "grad_norm": 0.7579936385154724, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 3975 + }, + { + "epoch": 0.03976, + "grad_norm": 0.7740127444267273, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 3976 + }, + { + "epoch": 0.03977, + "grad_norm": 0.7652050256729126, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 3977 + }, + { + "epoch": 0.03978, + "grad_norm": 0.7869767546653748, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 3978 + }, + { + "epoch": 0.03979, + "grad_norm": 0.8766190409660339, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 3979 + }, + { + "epoch": 0.0398, + "grad_norm": 0.9912131428718567, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 3980 + }, + { + "epoch": 0.03981, + "grad_norm": 1.0468101501464844, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 3981 + }, + { + "epoch": 0.03982, + "grad_norm": 0.8397706747055054, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 3982 + }, + { + "epoch": 0.03983, + "grad_norm": 0.6968387365341187, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 3983 + }, + { + "epoch": 0.03984, + "grad_norm": 0.6742430925369263, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 3984 + }, + { + "epoch": 0.03985, + "grad_norm": 0.7048434019088745, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 3985 + }, + { + "epoch": 0.03986, + "grad_norm": 0.7287977337837219, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 3986 + }, + { + "epoch": 0.03987, + "grad_norm": 0.6307860016822815, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 3987 + }, + { + "epoch": 0.03988, + "grad_norm": 0.6006014943122864, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 3988 + }, + { + "epoch": 0.03989, + "grad_norm": 0.614997386932373, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 3989 + }, + { + "epoch": 0.0399, + "grad_norm": 0.6875625252723694, + "learning_rate": 0.003, + "loss": 4.082, + "step": 3990 + }, + { + "epoch": 0.03991, + "grad_norm": 0.7122326493263245, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 3991 + }, + { + "epoch": 0.03992, + "grad_norm": 0.8080908060073853, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 3992 + }, + { + "epoch": 0.03993, + "grad_norm": 0.7636401653289795, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 3993 + }, + { + "epoch": 0.03994, + "grad_norm": 0.7076770663261414, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 3994 + }, + { + "epoch": 0.03995, + "grad_norm": 0.5683047771453857, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 3995 + }, + { + "epoch": 0.03996, + "grad_norm": 0.4724688231945038, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 3996 + }, + { + "epoch": 0.03997, + "grad_norm": 0.5588562488555908, + "learning_rate": 0.003, + "loss": 4.116, + "step": 3997 + }, + { + "epoch": 0.03998, + "grad_norm": 0.5613643527030945, + "learning_rate": 0.003, + "loss": 4.117, + "step": 3998 + }, + { + "epoch": 0.03999, + "grad_norm": 0.5133464932441711, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 3999 + }, + { + "epoch": 0.04, + "grad_norm": 0.46894940733909607, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 4000 + }, + { + "epoch": 0.04001, + "grad_norm": 0.4599146544933319, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 4001 + }, + { + "epoch": 0.04002, + "grad_norm": 0.5162279605865479, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 4002 + }, + { + "epoch": 0.04003, + "grad_norm": 0.5921787023544312, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 4003 + }, + { + "epoch": 0.04004, + "grad_norm": 0.6687480807304382, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 4004 + }, + { + "epoch": 0.04005, + "grad_norm": 0.6211435794830322, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 4005 + }, + { + "epoch": 0.04006, + "grad_norm": 0.6189464330673218, + "learning_rate": 0.003, + "loss": 4.08, + "step": 4006 + }, + { + "epoch": 0.04007, + "grad_norm": 0.6097008585929871, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 4007 + }, + { + "epoch": 0.04008, + "grad_norm": 0.6154138445854187, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 4008 + }, + { + "epoch": 0.04009, + "grad_norm": 0.7050077319145203, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 4009 + }, + { + "epoch": 0.0401, + "grad_norm": 0.7118995189666748, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 4010 + }, + { + "epoch": 0.04011, + "grad_norm": 0.7219801545143127, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 4011 + }, + { + "epoch": 0.04012, + "grad_norm": 0.7222893834114075, + "learning_rate": 0.003, + "loss": 4.103, + "step": 4012 + }, + { + "epoch": 0.04013, + "grad_norm": 0.7958313822746277, + "learning_rate": 0.003, + "loss": 4.107, + "step": 4013 + }, + { + "epoch": 0.04014, + "grad_norm": 0.6716787815093994, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 4014 + }, + { + "epoch": 0.04015, + "grad_norm": 0.6670402884483337, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 4015 + }, + { + "epoch": 0.04016, + "grad_norm": 0.8088099956512451, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 4016 + }, + { + "epoch": 0.04017, + "grad_norm": 0.9719444513320923, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 4017 + }, + { + "epoch": 0.04018, + "grad_norm": 0.9042829871177673, + "learning_rate": 0.003, + "loss": 4.109, + "step": 4018 + }, + { + "epoch": 0.04019, + "grad_norm": 0.701447069644928, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 4019 + }, + { + "epoch": 0.0402, + "grad_norm": 0.6191059350967407, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 4020 + }, + { + "epoch": 0.04021, + "grad_norm": 0.6724329590797424, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 4021 + }, + { + "epoch": 0.04022, + "grad_norm": 0.6767911911010742, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 4022 + }, + { + "epoch": 0.04023, + "grad_norm": 0.6602100133895874, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 4023 + }, + { + "epoch": 0.04024, + "grad_norm": 0.6715804934501648, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 4024 + }, + { + "epoch": 0.04025, + "grad_norm": 0.6610874533653259, + "learning_rate": 0.003, + "loss": 4.087, + "step": 4025 + }, + { + "epoch": 0.04026, + "grad_norm": 0.7230667471885681, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 4026 + }, + { + "epoch": 0.04027, + "grad_norm": 0.6509302258491516, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 4027 + }, + { + "epoch": 0.04028, + "grad_norm": 0.7460821866989136, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 4028 + }, + { + "epoch": 0.04029, + "grad_norm": 0.760715901851654, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 4029 + }, + { + "epoch": 0.0403, + "grad_norm": 0.7702258825302124, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 4030 + }, + { + "epoch": 0.04031, + "grad_norm": 0.7244075536727905, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 4031 + }, + { + "epoch": 0.04032, + "grad_norm": 0.7782586812973022, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 4032 + }, + { + "epoch": 0.04033, + "grad_norm": 0.7341326475143433, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 4033 + }, + { + "epoch": 0.04034, + "grad_norm": 0.7792084813117981, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 4034 + }, + { + "epoch": 0.04035, + "grad_norm": 0.8536400198936462, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 4035 + }, + { + "epoch": 0.04036, + "grad_norm": 0.7702471017837524, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 4036 + }, + { + "epoch": 0.04037, + "grad_norm": 0.7094979882240295, + "learning_rate": 0.003, + "loss": 4.078, + "step": 4037 + }, + { + "epoch": 0.04038, + "grad_norm": 0.6657090187072754, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 4038 + }, + { + "epoch": 0.04039, + "grad_norm": 0.8088093400001526, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 4039 + }, + { + "epoch": 0.0404, + "grad_norm": 0.879915177822113, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 4040 + }, + { + "epoch": 0.04041, + "grad_norm": 0.9084352850914001, + "learning_rate": 0.003, + "loss": 4.098, + "step": 4041 + }, + { + "epoch": 0.04042, + "grad_norm": 0.7722291946411133, + "learning_rate": 0.003, + "loss": 4.133, + "step": 4042 + }, + { + "epoch": 0.04043, + "grad_norm": 0.71905916929245, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 4043 + }, + { + "epoch": 0.04044, + "grad_norm": 0.7158097624778748, + "learning_rate": 0.003, + "loss": 4.113, + "step": 4044 + }, + { + "epoch": 0.04045, + "grad_norm": 0.6895061731338501, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 4045 + }, + { + "epoch": 0.04046, + "grad_norm": 0.681125283241272, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 4046 + }, + { + "epoch": 0.04047, + "grad_norm": 0.7285493016242981, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 4047 + }, + { + "epoch": 0.04048, + "grad_norm": 0.7340379357337952, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 4048 + }, + { + "epoch": 0.04049, + "grad_norm": 0.7420255541801453, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 4049 + }, + { + "epoch": 0.0405, + "grad_norm": 0.7484999299049377, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 4050 + }, + { + "epoch": 0.04051, + "grad_norm": 0.6874270439147949, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 4051 + }, + { + "epoch": 0.04052, + "grad_norm": 0.6694976091384888, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 4052 + }, + { + "epoch": 0.04053, + "grad_norm": 0.6139923930168152, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 4053 + }, + { + "epoch": 0.04054, + "grad_norm": 0.5986779928207397, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 4054 + }, + { + "epoch": 0.04055, + "grad_norm": 0.5122130513191223, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 4055 + }, + { + "epoch": 0.04056, + "grad_norm": 0.5276870727539062, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 4056 + }, + { + "epoch": 0.04057, + "grad_norm": 0.5313277244567871, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 4057 + }, + { + "epoch": 0.04058, + "grad_norm": 0.7404656410217285, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 4058 + }, + { + "epoch": 0.04059, + "grad_norm": 0.905890166759491, + "learning_rate": 0.003, + "loss": 4.065, + "step": 4059 + }, + { + "epoch": 0.0406, + "grad_norm": 0.9699053168296814, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 4060 + }, + { + "epoch": 0.04061, + "grad_norm": 0.7840712070465088, + "learning_rate": 0.003, + "loss": 4.107, + "step": 4061 + }, + { + "epoch": 0.04062, + "grad_norm": 0.6911324262619019, + "learning_rate": 0.003, + "loss": 4.088, + "step": 4062 + }, + { + "epoch": 0.04063, + "grad_norm": 0.664695143699646, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 4063 + }, + { + "epoch": 0.04064, + "grad_norm": 0.7076813578605652, + "learning_rate": 0.003, + "loss": 4.108, + "step": 4064 + }, + { + "epoch": 0.04065, + "grad_norm": 0.6082136034965515, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 4065 + }, + { + "epoch": 0.04066, + "grad_norm": 0.6230340003967285, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 4066 + }, + { + "epoch": 0.04067, + "grad_norm": 0.5546422004699707, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 4067 + }, + { + "epoch": 0.04068, + "grad_norm": 0.5739226937294006, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 4068 + }, + { + "epoch": 0.04069, + "grad_norm": 0.5410630702972412, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 4069 + }, + { + "epoch": 0.0407, + "grad_norm": 0.5467224717140198, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 4070 + }, + { + "epoch": 0.04071, + "grad_norm": 0.6489109992980957, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 4071 + }, + { + "epoch": 0.04072, + "grad_norm": 0.6744621992111206, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 4072 + }, + { + "epoch": 0.04073, + "grad_norm": 0.60545814037323, + "learning_rate": 0.003, + "loss": 4.078, + "step": 4073 + }, + { + "epoch": 0.04074, + "grad_norm": 0.6391845941543579, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 4074 + }, + { + "epoch": 0.04075, + "grad_norm": 0.7024280428886414, + "learning_rate": 0.003, + "loss": 4.104, + "step": 4075 + }, + { + "epoch": 0.04076, + "grad_norm": 0.7279633283615112, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 4076 + }, + { + "epoch": 0.04077, + "grad_norm": 0.9424205422401428, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 4077 + }, + { + "epoch": 0.04078, + "grad_norm": 1.1304396390914917, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 4078 + }, + { + "epoch": 0.04079, + "grad_norm": 0.8722311854362488, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 4079 + }, + { + "epoch": 0.0408, + "grad_norm": 0.760799765586853, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 4080 + }, + { + "epoch": 0.04081, + "grad_norm": 0.7816235423088074, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 4081 + }, + { + "epoch": 0.04082, + "grad_norm": 0.7581563591957092, + "learning_rate": 0.003, + "loss": 4.113, + "step": 4082 + }, + { + "epoch": 0.04083, + "grad_norm": 0.7936528921127319, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 4083 + }, + { + "epoch": 0.04084, + "grad_norm": 0.772882342338562, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 4084 + }, + { + "epoch": 0.04085, + "grad_norm": 0.7227948904037476, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 4085 + }, + { + "epoch": 0.04086, + "grad_norm": 0.6548799276351929, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 4086 + }, + { + "epoch": 0.04087, + "grad_norm": 0.5630327463150024, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 4087 + }, + { + "epoch": 0.04088, + "grad_norm": 0.527557373046875, + "learning_rate": 0.003, + "loss": 4.084, + "step": 4088 + }, + { + "epoch": 0.04089, + "grad_norm": 0.46402889490127563, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 4089 + }, + { + "epoch": 0.0409, + "grad_norm": 0.44137027859687805, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 4090 + }, + { + "epoch": 0.04091, + "grad_norm": 0.4541914165019989, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 4091 + }, + { + "epoch": 0.04092, + "grad_norm": 0.4600992500782013, + "learning_rate": 0.003, + "loss": 4.065, + "step": 4092 + }, + { + "epoch": 0.04093, + "grad_norm": 0.4791017472743988, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 4093 + }, + { + "epoch": 0.04094, + "grad_norm": 0.527622640132904, + "learning_rate": 0.003, + "loss": 4.083, + "step": 4094 + }, + { + "epoch": 0.04095, + "grad_norm": 0.5713629126548767, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 4095 + }, + { + "epoch": 0.04096, + "grad_norm": 0.6901856064796448, + "learning_rate": 0.003, + "loss": 4.069, + "step": 4096 + }, + { + "epoch": 0.04097, + "grad_norm": 0.8031033873558044, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 4097 + }, + { + "epoch": 0.04098, + "grad_norm": 0.9514955878257751, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 4098 + }, + { + "epoch": 0.04099, + "grad_norm": 1.1810649633407593, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 4099 + }, + { + "epoch": 0.041, + "grad_norm": 0.7748354077339172, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 4100 + }, + { + "epoch": 0.04101, + "grad_norm": 0.6190698146820068, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 4101 + }, + { + "epoch": 0.04102, + "grad_norm": 0.7298320531845093, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 4102 + }, + { + "epoch": 0.04103, + "grad_norm": 0.8667885661125183, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 4103 + }, + { + "epoch": 0.04104, + "grad_norm": 0.8904606699943542, + "learning_rate": 0.003, + "loss": 4.109, + "step": 4104 + }, + { + "epoch": 0.04105, + "grad_norm": 0.8015432953834534, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 4105 + }, + { + "epoch": 0.04106, + "grad_norm": 0.7645514011383057, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 4106 + }, + { + "epoch": 0.04107, + "grad_norm": 0.7201929092407227, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 4107 + }, + { + "epoch": 0.04108, + "grad_norm": 0.696420431137085, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 4108 + }, + { + "epoch": 0.04109, + "grad_norm": 0.7951487302780151, + "learning_rate": 0.003, + "loss": 4.086, + "step": 4109 + }, + { + "epoch": 0.0411, + "grad_norm": 0.7593970894813538, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 4110 + }, + { + "epoch": 0.04111, + "grad_norm": 0.6601749658584595, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 4111 + }, + { + "epoch": 0.04112, + "grad_norm": 0.5873904228210449, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 4112 + }, + { + "epoch": 0.04113, + "grad_norm": 0.6075222492218018, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 4113 + }, + { + "epoch": 0.04114, + "grad_norm": 0.5487315654754639, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 4114 + }, + { + "epoch": 0.04115, + "grad_norm": 0.5993032455444336, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 4115 + }, + { + "epoch": 0.04116, + "grad_norm": 0.5999573469161987, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 4116 + }, + { + "epoch": 0.04117, + "grad_norm": 0.7142841815948486, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 4117 + }, + { + "epoch": 0.04118, + "grad_norm": 0.8221079707145691, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 4118 + }, + { + "epoch": 0.04119, + "grad_norm": 0.8343155384063721, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 4119 + }, + { + "epoch": 0.0412, + "grad_norm": 0.7892839312553406, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 4120 + }, + { + "epoch": 0.04121, + "grad_norm": 0.7228533625602722, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 4121 + }, + { + "epoch": 0.04122, + "grad_norm": 0.6834357976913452, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 4122 + }, + { + "epoch": 0.04123, + "grad_norm": 0.6968315243721008, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 4123 + }, + { + "epoch": 0.04124, + "grad_norm": 0.5444803833961487, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 4124 + }, + { + "epoch": 0.04125, + "grad_norm": 0.6157690286636353, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 4125 + }, + { + "epoch": 0.04126, + "grad_norm": 0.5248627066612244, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 4126 + }, + { + "epoch": 0.04127, + "grad_norm": 0.606207549571991, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 4127 + }, + { + "epoch": 0.04128, + "grad_norm": 0.7370710968971252, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 4128 + }, + { + "epoch": 0.04129, + "grad_norm": 0.8737143278121948, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 4129 + }, + { + "epoch": 0.0413, + "grad_norm": 0.8792223334312439, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 4130 + }, + { + "epoch": 0.04131, + "grad_norm": 0.6732345819473267, + "learning_rate": 0.003, + "loss": 4.107, + "step": 4131 + }, + { + "epoch": 0.04132, + "grad_norm": 0.6241254210472107, + "learning_rate": 0.003, + "loss": 4.069, + "step": 4132 + }, + { + "epoch": 0.04133, + "grad_norm": 0.6735711693763733, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 4133 + }, + { + "epoch": 0.04134, + "grad_norm": 0.6963285803794861, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 4134 + }, + { + "epoch": 0.04135, + "grad_norm": 0.8845453262329102, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 4135 + }, + { + "epoch": 0.04136, + "grad_norm": 0.9322511553764343, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 4136 + }, + { + "epoch": 0.04137, + "grad_norm": 1.0508934259414673, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 4137 + }, + { + "epoch": 0.04138, + "grad_norm": 1.0728346109390259, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 4138 + }, + { + "epoch": 0.04139, + "grad_norm": 0.9741144180297852, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 4139 + }, + { + "epoch": 0.0414, + "grad_norm": 0.8693930506706238, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 4140 + }, + { + "epoch": 0.04141, + "grad_norm": 1.100584626197815, + "learning_rate": 0.003, + "loss": 4.121, + "step": 4141 + }, + { + "epoch": 0.04142, + "grad_norm": 0.9084172248840332, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 4142 + }, + { + "epoch": 0.04143, + "grad_norm": 0.9111177921295166, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 4143 + }, + { + "epoch": 0.04144, + "grad_norm": 1.0545722246170044, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 4144 + }, + { + "epoch": 0.04145, + "grad_norm": 1.0651071071624756, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 4145 + }, + { + "epoch": 0.04146, + "grad_norm": 0.8738028407096863, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 4146 + }, + { + "epoch": 0.04147, + "grad_norm": 0.8165324926376343, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 4147 + }, + { + "epoch": 0.04148, + "grad_norm": 0.7123237252235413, + "learning_rate": 0.003, + "loss": 4.1, + "step": 4148 + }, + { + "epoch": 0.04149, + "grad_norm": 0.6672841310501099, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 4149 + }, + { + "epoch": 0.0415, + "grad_norm": 0.6620326042175293, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 4150 + }, + { + "epoch": 0.04151, + "grad_norm": 0.6698993444442749, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 4151 + }, + { + "epoch": 0.04152, + "grad_norm": 0.6350617408752441, + "learning_rate": 0.003, + "loss": 4.123, + "step": 4152 + }, + { + "epoch": 0.04153, + "grad_norm": 0.47288987040519714, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 4153 + }, + { + "epoch": 0.04154, + "grad_norm": 0.4522964656352997, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 4154 + }, + { + "epoch": 0.04155, + "grad_norm": 0.426760196685791, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 4155 + }, + { + "epoch": 0.04156, + "grad_norm": 0.43916746973991394, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 4156 + }, + { + "epoch": 0.04157, + "grad_norm": 0.47124361991882324, + "learning_rate": 0.003, + "loss": 4.121, + "step": 4157 + }, + { + "epoch": 0.04158, + "grad_norm": 0.5592663288116455, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 4158 + }, + { + "epoch": 0.04159, + "grad_norm": 0.747325599193573, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 4159 + }, + { + "epoch": 0.0416, + "grad_norm": 1.02601957321167, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 4160 + }, + { + "epoch": 0.04161, + "grad_norm": 1.0634429454803467, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 4161 + }, + { + "epoch": 0.04162, + "grad_norm": 0.5860819220542908, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 4162 + }, + { + "epoch": 0.04163, + "grad_norm": 0.6616426706314087, + "learning_rate": 0.003, + "loss": 4.105, + "step": 4163 + }, + { + "epoch": 0.04164, + "grad_norm": 0.6689631938934326, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 4164 + }, + { + "epoch": 0.04165, + "grad_norm": 0.5675843954086304, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 4165 + }, + { + "epoch": 0.04166, + "grad_norm": 0.4940520226955414, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 4166 + }, + { + "epoch": 0.04167, + "grad_norm": 0.5212199091911316, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 4167 + }, + { + "epoch": 0.04168, + "grad_norm": 0.52357017993927, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 4168 + }, + { + "epoch": 0.04169, + "grad_norm": 0.43375542759895325, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 4169 + }, + { + "epoch": 0.0417, + "grad_norm": 0.44883838295936584, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 4170 + }, + { + "epoch": 0.04171, + "grad_norm": 0.4587332606315613, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 4171 + }, + { + "epoch": 0.04172, + "grad_norm": 0.5260483026504517, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 4172 + }, + { + "epoch": 0.04173, + "grad_norm": 0.7267121076583862, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 4173 + }, + { + "epoch": 0.04174, + "grad_norm": 0.8692538738250732, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 4174 + }, + { + "epoch": 0.04175, + "grad_norm": 0.8145063519477844, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 4175 + }, + { + "epoch": 0.04176, + "grad_norm": 0.7225742340087891, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 4176 + }, + { + "epoch": 0.04177, + "grad_norm": 0.6365543007850647, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 4177 + }, + { + "epoch": 0.04178, + "grad_norm": 0.6694263815879822, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 4178 + }, + { + "epoch": 0.04179, + "grad_norm": 0.7857269644737244, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 4179 + }, + { + "epoch": 0.0418, + "grad_norm": 0.8134697675704956, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 4180 + }, + { + "epoch": 0.04181, + "grad_norm": 0.687165379524231, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 4181 + }, + { + "epoch": 0.04182, + "grad_norm": 0.6623665690422058, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 4182 + }, + { + "epoch": 0.04183, + "grad_norm": 0.659334659576416, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 4183 + }, + { + "epoch": 0.04184, + "grad_norm": 0.6392843723297119, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 4184 + }, + { + "epoch": 0.04185, + "grad_norm": 0.5640465617179871, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 4185 + }, + { + "epoch": 0.04186, + "grad_norm": 0.5710539221763611, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 4186 + }, + { + "epoch": 0.04187, + "grad_norm": 0.4746825695037842, + "learning_rate": 0.003, + "loss": 4.109, + "step": 4187 + }, + { + "epoch": 0.04188, + "grad_norm": 0.4887617826461792, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 4188 + }, + { + "epoch": 0.04189, + "grad_norm": 0.40521180629730225, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 4189 + }, + { + "epoch": 0.0419, + "grad_norm": 0.47049522399902344, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 4190 + }, + { + "epoch": 0.04191, + "grad_norm": 0.5449577569961548, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 4191 + }, + { + "epoch": 0.04192, + "grad_norm": 0.6885932087898254, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 4192 + }, + { + "epoch": 0.04193, + "grad_norm": 0.9400091171264648, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 4193 + }, + { + "epoch": 0.04194, + "grad_norm": 1.1772154569625854, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 4194 + }, + { + "epoch": 0.04195, + "grad_norm": 0.7265138030052185, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 4195 + }, + { + "epoch": 0.04196, + "grad_norm": 0.702653169631958, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 4196 + }, + { + "epoch": 0.04197, + "grad_norm": 0.8408281803131104, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 4197 + }, + { + "epoch": 0.04198, + "grad_norm": 0.7780887484550476, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 4198 + }, + { + "epoch": 0.04199, + "grad_norm": 0.7158482074737549, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 4199 + }, + { + "epoch": 0.042, + "grad_norm": 0.6765742301940918, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 4200 + }, + { + "epoch": 0.04201, + "grad_norm": 0.6093567609786987, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 4201 + }, + { + "epoch": 0.04202, + "grad_norm": 0.6598721146583557, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 4202 + }, + { + "epoch": 0.04203, + "grad_norm": 0.5697882175445557, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 4203 + }, + { + "epoch": 0.04204, + "grad_norm": 0.6565430760383606, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 4204 + }, + { + "epoch": 0.04205, + "grad_norm": 0.8136803507804871, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 4205 + }, + { + "epoch": 0.04206, + "grad_norm": 0.828262984752655, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 4206 + }, + { + "epoch": 0.04207, + "grad_norm": 0.7792331576347351, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 4207 + }, + { + "epoch": 0.04208, + "grad_norm": 0.9303550124168396, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 4208 + }, + { + "epoch": 0.04209, + "grad_norm": 0.8936018347740173, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 4209 + }, + { + "epoch": 0.0421, + "grad_norm": 1.0039626359939575, + "learning_rate": 0.003, + "loss": 4.081, + "step": 4210 + }, + { + "epoch": 0.04211, + "grad_norm": 0.9981667995452881, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 4211 + }, + { + "epoch": 0.04212, + "grad_norm": 0.8137405514717102, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 4212 + }, + { + "epoch": 0.04213, + "grad_norm": 0.8376567959785461, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 4213 + }, + { + "epoch": 0.04214, + "grad_norm": 0.7845513820648193, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 4214 + }, + { + "epoch": 0.04215, + "grad_norm": 0.7330971360206604, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 4215 + }, + { + "epoch": 0.04216, + "grad_norm": 0.7391840219497681, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 4216 + }, + { + "epoch": 0.04217, + "grad_norm": 0.8007631897926331, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 4217 + }, + { + "epoch": 0.04218, + "grad_norm": 0.8993299007415771, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 4218 + }, + { + "epoch": 0.04219, + "grad_norm": 0.9009442925453186, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 4219 + }, + { + "epoch": 0.0422, + "grad_norm": 1.0134148597717285, + "learning_rate": 0.003, + "loss": 4.1351, + "step": 4220 + }, + { + "epoch": 0.04221, + "grad_norm": 0.9685496687889099, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 4221 + }, + { + "epoch": 0.04222, + "grad_norm": 0.8909528255462646, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 4222 + }, + { + "epoch": 0.04223, + "grad_norm": 0.806991457939148, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 4223 + }, + { + "epoch": 0.04224, + "grad_norm": 0.7712184190750122, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 4224 + }, + { + "epoch": 0.04225, + "grad_norm": 0.6012980341911316, + "learning_rate": 0.003, + "loss": 4.119, + "step": 4225 + }, + { + "epoch": 0.04226, + "grad_norm": 0.6349509358406067, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 4226 + }, + { + "epoch": 0.04227, + "grad_norm": 0.6691918969154358, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 4227 + }, + { + "epoch": 0.04228, + "grad_norm": 0.622721254825592, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 4228 + }, + { + "epoch": 0.04229, + "grad_norm": 0.57576984167099, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 4229 + }, + { + "epoch": 0.0423, + "grad_norm": 0.6181168556213379, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 4230 + }, + { + "epoch": 0.04231, + "grad_norm": 0.6395006775856018, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 4231 + }, + { + "epoch": 0.04232, + "grad_norm": 0.5587146878242493, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 4232 + }, + { + "epoch": 0.04233, + "grad_norm": 0.4529360234737396, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 4233 + }, + { + "epoch": 0.04234, + "grad_norm": 0.4365110397338867, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 4234 + }, + { + "epoch": 0.04235, + "grad_norm": 0.3945704996585846, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 4235 + }, + { + "epoch": 0.04236, + "grad_norm": 0.4571150541305542, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 4236 + }, + { + "epoch": 0.04237, + "grad_norm": 0.4355223476886749, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 4237 + }, + { + "epoch": 0.04238, + "grad_norm": 0.47384676337242126, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 4238 + }, + { + "epoch": 0.04239, + "grad_norm": 0.5304744243621826, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 4239 + }, + { + "epoch": 0.0424, + "grad_norm": 0.5990294814109802, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 4240 + }, + { + "epoch": 0.04241, + "grad_norm": 0.6438462734222412, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 4241 + }, + { + "epoch": 0.04242, + "grad_norm": 0.6008359789848328, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 4242 + }, + { + "epoch": 0.04243, + "grad_norm": 0.653829038143158, + "learning_rate": 0.003, + "loss": 4.08, + "step": 4243 + }, + { + "epoch": 0.04244, + "grad_norm": 0.6691821813583374, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 4244 + }, + { + "epoch": 0.04245, + "grad_norm": 0.62837815284729, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 4245 + }, + { + "epoch": 0.04246, + "grad_norm": 0.6627157330513, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 4246 + }, + { + "epoch": 0.04247, + "grad_norm": 0.7871987819671631, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 4247 + }, + { + "epoch": 0.04248, + "grad_norm": 0.9460709095001221, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 4248 + }, + { + "epoch": 0.04249, + "grad_norm": 0.9994218945503235, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 4249 + }, + { + "epoch": 0.0425, + "grad_norm": 0.8460412621498108, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 4250 + }, + { + "epoch": 0.04251, + "grad_norm": 0.6389533281326294, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 4251 + }, + { + "epoch": 0.04252, + "grad_norm": 0.7343540191650391, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 4252 + }, + { + "epoch": 0.04253, + "grad_norm": 0.8084548115730286, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 4253 + }, + { + "epoch": 0.04254, + "grad_norm": 0.804446280002594, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 4254 + }, + { + "epoch": 0.04255, + "grad_norm": 0.6976709365844727, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 4255 + }, + { + "epoch": 0.04256, + "grad_norm": 0.6876462697982788, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 4256 + }, + { + "epoch": 0.04257, + "grad_norm": 0.6768009662628174, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 4257 + }, + { + "epoch": 0.04258, + "grad_norm": 0.735969603061676, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 4258 + }, + { + "epoch": 0.04259, + "grad_norm": 0.83452969789505, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 4259 + }, + { + "epoch": 0.0426, + "grad_norm": 0.7557306289672852, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 4260 + }, + { + "epoch": 0.04261, + "grad_norm": 0.7074769735336304, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 4261 + }, + { + "epoch": 0.04262, + "grad_norm": 0.6406824588775635, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 4262 + }, + { + "epoch": 0.04263, + "grad_norm": 0.6755427122116089, + "learning_rate": 0.003, + "loss": 4.104, + "step": 4263 + }, + { + "epoch": 0.04264, + "grad_norm": 0.7064899206161499, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 4264 + }, + { + "epoch": 0.04265, + "grad_norm": 0.673495888710022, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 4265 + }, + { + "epoch": 0.04266, + "grad_norm": 0.6447572708129883, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 4266 + }, + { + "epoch": 0.04267, + "grad_norm": 0.5904159545898438, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 4267 + }, + { + "epoch": 0.04268, + "grad_norm": 0.6268547177314758, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 4268 + }, + { + "epoch": 0.04269, + "grad_norm": 0.5750144124031067, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 4269 + }, + { + "epoch": 0.0427, + "grad_norm": 0.44790661334991455, + "learning_rate": 0.003, + "loss": 4.08, + "step": 4270 + }, + { + "epoch": 0.04271, + "grad_norm": 0.4415220022201538, + "learning_rate": 0.003, + "loss": 4.094, + "step": 4271 + }, + { + "epoch": 0.04272, + "grad_norm": 0.39944225549697876, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 4272 + }, + { + "epoch": 0.04273, + "grad_norm": 0.3969830870628357, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 4273 + }, + { + "epoch": 0.04274, + "grad_norm": 0.41689732670783997, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 4274 + }, + { + "epoch": 0.04275, + "grad_norm": 0.4759061932563782, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 4275 + }, + { + "epoch": 0.04276, + "grad_norm": 0.6165413856506348, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 4276 + }, + { + "epoch": 0.04277, + "grad_norm": 0.9280685782432556, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 4277 + }, + { + "epoch": 0.04278, + "grad_norm": 1.2253838777542114, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 4278 + }, + { + "epoch": 0.04279, + "grad_norm": 0.5876243710517883, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 4279 + }, + { + "epoch": 0.0428, + "grad_norm": 0.5460189580917358, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 4280 + }, + { + "epoch": 0.04281, + "grad_norm": 0.81770920753479, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 4281 + }, + { + "epoch": 0.04282, + "grad_norm": 0.7643664479255676, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 4282 + }, + { + "epoch": 0.04283, + "grad_norm": 0.7183843851089478, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 4283 + }, + { + "epoch": 0.04284, + "grad_norm": 0.6911137104034424, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 4284 + }, + { + "epoch": 0.04285, + "grad_norm": 0.6681384444236755, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 4285 + }, + { + "epoch": 0.04286, + "grad_norm": 0.7612639665603638, + "learning_rate": 0.003, + "loss": 4.117, + "step": 4286 + }, + { + "epoch": 0.04287, + "grad_norm": 0.8821431398391724, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 4287 + }, + { + "epoch": 0.04288, + "grad_norm": 0.8891422748565674, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 4288 + }, + { + "epoch": 0.04289, + "grad_norm": 0.8912457823753357, + "learning_rate": 0.003, + "loss": 4.101, + "step": 4289 + }, + { + "epoch": 0.0429, + "grad_norm": 0.9452528953552246, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 4290 + }, + { + "epoch": 0.04291, + "grad_norm": 0.9560781121253967, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 4291 + }, + { + "epoch": 0.04292, + "grad_norm": 0.9647819995880127, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 4292 + }, + { + "epoch": 0.04293, + "grad_norm": 1.0337492227554321, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 4293 + }, + { + "epoch": 0.04294, + "grad_norm": 0.9274333119392395, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 4294 + }, + { + "epoch": 0.04295, + "grad_norm": 0.9098928570747375, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 4295 + }, + { + "epoch": 0.04296, + "grad_norm": 0.9552505612373352, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 4296 + }, + { + "epoch": 0.04297, + "grad_norm": 0.9055096507072449, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 4297 + }, + { + "epoch": 0.04298, + "grad_norm": 0.9436898231506348, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 4298 + }, + { + "epoch": 0.04299, + "grad_norm": 0.9265126585960388, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 4299 + }, + { + "epoch": 0.043, + "grad_norm": 0.9316915273666382, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 4300 + }, + { + "epoch": 0.04301, + "grad_norm": 0.9089905023574829, + "learning_rate": 0.003, + "loss": 4.1575, + "step": 4301 + }, + { + "epoch": 0.04302, + "grad_norm": 0.9315720200538635, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 4302 + }, + { + "epoch": 0.04303, + "grad_norm": 0.9286166429519653, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 4303 + }, + { + "epoch": 0.04304, + "grad_norm": 0.91733717918396, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 4304 + }, + { + "epoch": 0.04305, + "grad_norm": 1.065048098564148, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 4305 + }, + { + "epoch": 0.04306, + "grad_norm": 0.9198899269104004, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 4306 + }, + { + "epoch": 0.04307, + "grad_norm": 0.729816198348999, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 4307 + }, + { + "epoch": 0.04308, + "grad_norm": 0.6812875866889954, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 4308 + }, + { + "epoch": 0.04309, + "grad_norm": 0.6616823077201843, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 4309 + }, + { + "epoch": 0.0431, + "grad_norm": 0.7262138724327087, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 4310 + }, + { + "epoch": 0.04311, + "grad_norm": 0.6826511025428772, + "learning_rate": 0.003, + "loss": 4.12, + "step": 4311 + }, + { + "epoch": 0.04312, + "grad_norm": 0.5878548622131348, + "learning_rate": 0.003, + "loss": 4.097, + "step": 4312 + }, + { + "epoch": 0.04313, + "grad_norm": 0.52091383934021, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 4313 + }, + { + "epoch": 0.04314, + "grad_norm": 0.5465619564056396, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 4314 + }, + { + "epoch": 0.04315, + "grad_norm": 0.5617761015892029, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 4315 + }, + { + "epoch": 0.04316, + "grad_norm": 0.5584519505500793, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 4316 + }, + { + "epoch": 0.04317, + "grad_norm": 0.591358482837677, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 4317 + }, + { + "epoch": 0.04318, + "grad_norm": 0.6004156470298767, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 4318 + }, + { + "epoch": 0.04319, + "grad_norm": 0.6281335353851318, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 4319 + }, + { + "epoch": 0.0432, + "grad_norm": 0.5984190702438354, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 4320 + }, + { + "epoch": 0.04321, + "grad_norm": 0.4503112733364105, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 4321 + }, + { + "epoch": 0.04322, + "grad_norm": 0.44634100794792175, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 4322 + }, + { + "epoch": 0.04323, + "grad_norm": 0.4618377089500427, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 4323 + }, + { + "epoch": 0.04324, + "grad_norm": 0.4412762224674225, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 4324 + }, + { + "epoch": 0.04325, + "grad_norm": 0.4403208792209625, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 4325 + }, + { + "epoch": 0.04326, + "grad_norm": 0.4678369164466858, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 4326 + }, + { + "epoch": 0.04327, + "grad_norm": 0.5307203531265259, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 4327 + }, + { + "epoch": 0.04328, + "grad_norm": 0.6740878820419312, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 4328 + }, + { + "epoch": 0.04329, + "grad_norm": 0.827231764793396, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 4329 + }, + { + "epoch": 0.0433, + "grad_norm": 0.9776306748390198, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 4330 + }, + { + "epoch": 0.04331, + "grad_norm": 0.7999715805053711, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 4331 + }, + { + "epoch": 0.04332, + "grad_norm": 0.6406177282333374, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 4332 + }, + { + "epoch": 0.04333, + "grad_norm": 0.7425047159194946, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 4333 + }, + { + "epoch": 0.04334, + "grad_norm": 0.7017237544059753, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 4334 + }, + { + "epoch": 0.04335, + "grad_norm": 0.6225997805595398, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 4335 + }, + { + "epoch": 0.04336, + "grad_norm": 0.5787286162376404, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 4336 + }, + { + "epoch": 0.04337, + "grad_norm": 0.4820355474948883, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 4337 + }, + { + "epoch": 0.04338, + "grad_norm": 0.454394668340683, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 4338 + }, + { + "epoch": 0.04339, + "grad_norm": 0.48518452048301697, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 4339 + }, + { + "epoch": 0.0434, + "grad_norm": 0.6182829737663269, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 4340 + }, + { + "epoch": 0.04341, + "grad_norm": 0.6660934090614319, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 4341 + }, + { + "epoch": 0.04342, + "grad_norm": 0.7125284671783447, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 4342 + }, + { + "epoch": 0.04343, + "grad_norm": 0.6455511450767517, + "learning_rate": 0.003, + "loss": 4.114, + "step": 4343 + }, + { + "epoch": 0.04344, + "grad_norm": 0.5411390066146851, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 4344 + }, + { + "epoch": 0.04345, + "grad_norm": 0.6171619296073914, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 4345 + }, + { + "epoch": 0.04346, + "grad_norm": 0.8696421980857849, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 4346 + }, + { + "epoch": 0.04347, + "grad_norm": 1.0493338108062744, + "learning_rate": 0.003, + "loss": 4.112, + "step": 4347 + }, + { + "epoch": 0.04348, + "grad_norm": 0.9583601951599121, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 4348 + }, + { + "epoch": 0.04349, + "grad_norm": 0.8417112827301025, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 4349 + }, + { + "epoch": 0.0435, + "grad_norm": 0.7813431620597839, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 4350 + }, + { + "epoch": 0.04351, + "grad_norm": 0.7515244483947754, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 4351 + }, + { + "epoch": 0.04352, + "grad_norm": 0.7575326561927795, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 4352 + }, + { + "epoch": 0.04353, + "grad_norm": 0.8079462647438049, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 4353 + }, + { + "epoch": 0.04354, + "grad_norm": 0.8209161162376404, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 4354 + }, + { + "epoch": 0.04355, + "grad_norm": 0.8110191226005554, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 4355 + }, + { + "epoch": 0.04356, + "grad_norm": 0.8562638759613037, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 4356 + }, + { + "epoch": 0.04357, + "grad_norm": 0.773973822593689, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 4357 + }, + { + "epoch": 0.04358, + "grad_norm": 0.8397206664085388, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 4358 + }, + { + "epoch": 0.04359, + "grad_norm": 0.7626861929893494, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 4359 + }, + { + "epoch": 0.0436, + "grad_norm": 0.80926513671875, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 4360 + }, + { + "epoch": 0.04361, + "grad_norm": 0.8279222846031189, + "learning_rate": 0.003, + "loss": 4.077, + "step": 4361 + }, + { + "epoch": 0.04362, + "grad_norm": 0.7730289697647095, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 4362 + }, + { + "epoch": 0.04363, + "grad_norm": 0.6915876865386963, + "learning_rate": 0.003, + "loss": 4.089, + "step": 4363 + }, + { + "epoch": 0.04364, + "grad_norm": 0.6894673705101013, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 4364 + }, + { + "epoch": 0.04365, + "grad_norm": 0.6546920537948608, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 4365 + }, + { + "epoch": 0.04366, + "grad_norm": 0.6718312501907349, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 4366 + }, + { + "epoch": 0.04367, + "grad_norm": 0.7026388049125671, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 4367 + }, + { + "epoch": 0.04368, + "grad_norm": 0.664130687713623, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 4368 + }, + { + "epoch": 0.04369, + "grad_norm": 0.4905566871166229, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 4369 + }, + { + "epoch": 0.0437, + "grad_norm": 0.5002229809761047, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 4370 + }, + { + "epoch": 0.04371, + "grad_norm": 0.5647168159484863, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 4371 + }, + { + "epoch": 0.04372, + "grad_norm": 0.5772929787635803, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 4372 + }, + { + "epoch": 0.04373, + "grad_norm": 0.5668372511863708, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 4373 + }, + { + "epoch": 0.04374, + "grad_norm": 0.5907851457595825, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 4374 + }, + { + "epoch": 0.04375, + "grad_norm": 0.5372269749641418, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 4375 + }, + { + "epoch": 0.04376, + "grad_norm": 0.5136074423789978, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 4376 + }, + { + "epoch": 0.04377, + "grad_norm": 0.5012854933738708, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 4377 + }, + { + "epoch": 0.04378, + "grad_norm": 0.5583826899528503, + "learning_rate": 0.003, + "loss": 4.1, + "step": 4378 + }, + { + "epoch": 0.04379, + "grad_norm": 0.609851062297821, + "learning_rate": 0.003, + "loss": 4.115, + "step": 4379 + }, + { + "epoch": 0.0438, + "grad_norm": 0.6667990684509277, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 4380 + }, + { + "epoch": 0.04381, + "grad_norm": 0.769900381565094, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 4381 + }, + { + "epoch": 0.04382, + "grad_norm": 0.918383777141571, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 4382 + }, + { + "epoch": 0.04383, + "grad_norm": 1.0884766578674316, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 4383 + }, + { + "epoch": 0.04384, + "grad_norm": 0.753715991973877, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 4384 + }, + { + "epoch": 0.04385, + "grad_norm": 0.5258980989456177, + "learning_rate": 0.003, + "loss": 4.112, + "step": 4385 + }, + { + "epoch": 0.04386, + "grad_norm": 0.7058178186416626, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 4386 + }, + { + "epoch": 0.04387, + "grad_norm": 0.865214467048645, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 4387 + }, + { + "epoch": 0.04388, + "grad_norm": 0.872906506061554, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 4388 + }, + { + "epoch": 0.04389, + "grad_norm": 0.785380482673645, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 4389 + }, + { + "epoch": 0.0439, + "grad_norm": 0.6976808309555054, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 4390 + }, + { + "epoch": 0.04391, + "grad_norm": 0.7179602384567261, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 4391 + }, + { + "epoch": 0.04392, + "grad_norm": 0.7050849199295044, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 4392 + }, + { + "epoch": 0.04393, + "grad_norm": 0.6494352221488953, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 4393 + }, + { + "epoch": 0.04394, + "grad_norm": 0.5354273915290833, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 4394 + }, + { + "epoch": 0.04395, + "grad_norm": 0.5617183446884155, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 4395 + }, + { + "epoch": 0.04396, + "grad_norm": 0.5766990780830383, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 4396 + }, + { + "epoch": 0.04397, + "grad_norm": 0.6702291369438171, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 4397 + }, + { + "epoch": 0.04398, + "grad_norm": 0.7613652944564819, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 4398 + }, + { + "epoch": 0.04399, + "grad_norm": 0.8499076962471008, + "learning_rate": 0.003, + "loss": 4.1, + "step": 4399 + }, + { + "epoch": 0.044, + "grad_norm": 1.0589154958724976, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 4400 + }, + { + "epoch": 0.04401, + "grad_norm": 0.9029095768928528, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 4401 + }, + { + "epoch": 0.04402, + "grad_norm": 0.7743659019470215, + "learning_rate": 0.003, + "loss": 4.118, + "step": 4402 + }, + { + "epoch": 0.04403, + "grad_norm": 0.6107021570205688, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 4403 + }, + { + "epoch": 0.04404, + "grad_norm": 0.578995406627655, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 4404 + }, + { + "epoch": 0.04405, + "grad_norm": 0.5411956906318665, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 4405 + }, + { + "epoch": 0.04406, + "grad_norm": 0.5497092008590698, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 4406 + }, + { + "epoch": 0.04407, + "grad_norm": 0.5126742124557495, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 4407 + }, + { + "epoch": 0.04408, + "grad_norm": 0.4807611107826233, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 4408 + }, + { + "epoch": 0.04409, + "grad_norm": 0.4981795847415924, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 4409 + }, + { + "epoch": 0.0441, + "grad_norm": 0.5225616097450256, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 4410 + }, + { + "epoch": 0.04411, + "grad_norm": 0.525507390499115, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 4411 + }, + { + "epoch": 0.04412, + "grad_norm": 0.551207423210144, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 4412 + }, + { + "epoch": 0.04413, + "grad_norm": 0.5674916505813599, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 4413 + }, + { + "epoch": 0.04414, + "grad_norm": 0.5803220868110657, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 4414 + }, + { + "epoch": 0.04415, + "grad_norm": 0.6080368161201477, + "learning_rate": 0.003, + "loss": 4.08, + "step": 4415 + }, + { + "epoch": 0.04416, + "grad_norm": 0.6551415920257568, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 4416 + }, + { + "epoch": 0.04417, + "grad_norm": 0.727580189704895, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 4417 + }, + { + "epoch": 0.04418, + "grad_norm": 0.869605302810669, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 4418 + }, + { + "epoch": 0.04419, + "grad_norm": 0.9329266548156738, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 4419 + }, + { + "epoch": 0.0442, + "grad_norm": 0.971718966960907, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 4420 + }, + { + "epoch": 0.04421, + "grad_norm": 0.8899560570716858, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 4421 + }, + { + "epoch": 0.04422, + "grad_norm": 0.8837299942970276, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 4422 + }, + { + "epoch": 0.04423, + "grad_norm": 0.8218461871147156, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 4423 + }, + { + "epoch": 0.04424, + "grad_norm": 0.834577202796936, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 4424 + }, + { + "epoch": 0.04425, + "grad_norm": 0.7609786987304688, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 4425 + }, + { + "epoch": 0.04426, + "grad_norm": 0.71391761302948, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 4426 + }, + { + "epoch": 0.04427, + "grad_norm": 0.6980590224266052, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 4427 + }, + { + "epoch": 0.04428, + "grad_norm": 0.6043670773506165, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 4428 + }, + { + "epoch": 0.04429, + "grad_norm": 0.6160145401954651, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 4429 + }, + { + "epoch": 0.0443, + "grad_norm": 0.6831215620040894, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 4430 + }, + { + "epoch": 0.04431, + "grad_norm": 0.8066039681434631, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 4431 + }, + { + "epoch": 0.04432, + "grad_norm": 0.8776717782020569, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 4432 + }, + { + "epoch": 0.04433, + "grad_norm": 0.7881397008895874, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 4433 + }, + { + "epoch": 0.04434, + "grad_norm": 0.6989492177963257, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 4434 + }, + { + "epoch": 0.04435, + "grad_norm": 0.6727581024169922, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 4435 + }, + { + "epoch": 0.04436, + "grad_norm": 0.6831278204917908, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 4436 + }, + { + "epoch": 0.04437, + "grad_norm": 0.7069624066352844, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 4437 + }, + { + "epoch": 0.04438, + "grad_norm": 0.8832647204399109, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 4438 + }, + { + "epoch": 0.04439, + "grad_norm": 1.118896245956421, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 4439 + }, + { + "epoch": 0.0444, + "grad_norm": 0.8600191473960876, + "learning_rate": 0.003, + "loss": 4.069, + "step": 4440 + }, + { + "epoch": 0.04441, + "grad_norm": 0.7522971630096436, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 4441 + }, + { + "epoch": 0.04442, + "grad_norm": 0.6988781690597534, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 4442 + }, + { + "epoch": 0.04443, + "grad_norm": 0.7806941270828247, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 4443 + }, + { + "epoch": 0.04444, + "grad_norm": 0.7053735256195068, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 4444 + }, + { + "epoch": 0.04445, + "grad_norm": 0.6177310943603516, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 4445 + }, + { + "epoch": 0.04446, + "grad_norm": 0.6128456592559814, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 4446 + }, + { + "epoch": 0.04447, + "grad_norm": 0.6953614950180054, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 4447 + }, + { + "epoch": 0.04448, + "grad_norm": 0.6975845098495483, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 4448 + }, + { + "epoch": 0.04449, + "grad_norm": 0.5980611443519592, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 4449 + }, + { + "epoch": 0.0445, + "grad_norm": 0.5174282789230347, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 4450 + }, + { + "epoch": 0.04451, + "grad_norm": 0.44481992721557617, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 4451 + }, + { + "epoch": 0.04452, + "grad_norm": 0.38826221227645874, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 4452 + }, + { + "epoch": 0.04453, + "grad_norm": 0.4051187038421631, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 4453 + }, + { + "epoch": 0.04454, + "grad_norm": 0.3912207782268524, + "learning_rate": 0.003, + "loss": 4.082, + "step": 4454 + }, + { + "epoch": 0.04455, + "grad_norm": 0.35977083444595337, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 4455 + }, + { + "epoch": 0.04456, + "grad_norm": 0.32243481278419495, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 4456 + }, + { + "epoch": 0.04457, + "grad_norm": 0.39922428131103516, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 4457 + }, + { + "epoch": 0.04458, + "grad_norm": 0.43713390827178955, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 4458 + }, + { + "epoch": 0.04459, + "grad_norm": 0.4921168088912964, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 4459 + }, + { + "epoch": 0.0446, + "grad_norm": 0.5851754546165466, + "learning_rate": 0.003, + "loss": 4.101, + "step": 4460 + }, + { + "epoch": 0.04461, + "grad_norm": 0.7543833255767822, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 4461 + }, + { + "epoch": 0.04462, + "grad_norm": 0.8701854348182678, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 4462 + }, + { + "epoch": 0.04463, + "grad_norm": 0.928347110748291, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 4463 + }, + { + "epoch": 0.04464, + "grad_norm": 0.923896312713623, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 4464 + }, + { + "epoch": 0.04465, + "grad_norm": 0.9406884908676147, + "learning_rate": 0.003, + "loss": 4.094, + "step": 4465 + }, + { + "epoch": 0.04466, + "grad_norm": 0.9898063540458679, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 4466 + }, + { + "epoch": 0.04467, + "grad_norm": 0.9611609578132629, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 4467 + }, + { + "epoch": 0.04468, + "grad_norm": 0.7379047870635986, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 4468 + }, + { + "epoch": 0.04469, + "grad_norm": 0.7446435689926147, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 4469 + }, + { + "epoch": 0.0447, + "grad_norm": 0.7013334631919861, + "learning_rate": 0.003, + "loss": 4.105, + "step": 4470 + }, + { + "epoch": 0.04471, + "grad_norm": 0.7607833743095398, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 4471 + }, + { + "epoch": 0.04472, + "grad_norm": 0.8320324420928955, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 4472 + }, + { + "epoch": 0.04473, + "grad_norm": 0.8373231291770935, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 4473 + }, + { + "epoch": 0.04474, + "grad_norm": 0.6635037660598755, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 4474 + }, + { + "epoch": 0.04475, + "grad_norm": 0.6026391386985779, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 4475 + }, + { + "epoch": 0.04476, + "grad_norm": 0.6115633249282837, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 4476 + }, + { + "epoch": 0.04477, + "grad_norm": 0.7230142951011658, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 4477 + }, + { + "epoch": 0.04478, + "grad_norm": 0.7399475574493408, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 4478 + }, + { + "epoch": 0.04479, + "grad_norm": 0.6423735022544861, + "learning_rate": 0.003, + "loss": 4.073, + "step": 4479 + }, + { + "epoch": 0.0448, + "grad_norm": 0.6744748950004578, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 4480 + }, + { + "epoch": 0.04481, + "grad_norm": 0.8687645196914673, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 4481 + }, + { + "epoch": 0.04482, + "grad_norm": 0.9180449843406677, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 4482 + }, + { + "epoch": 0.04483, + "grad_norm": 0.9672601819038391, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 4483 + }, + { + "epoch": 0.04484, + "grad_norm": 0.9588479399681091, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 4484 + }, + { + "epoch": 0.04485, + "grad_norm": 0.8815595507621765, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 4485 + }, + { + "epoch": 0.04486, + "grad_norm": 0.7194766402244568, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 4486 + }, + { + "epoch": 0.04487, + "grad_norm": 0.6366438269615173, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 4487 + }, + { + "epoch": 0.04488, + "grad_norm": 0.7553848028182983, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 4488 + }, + { + "epoch": 0.04489, + "grad_norm": 0.818514347076416, + "learning_rate": 0.003, + "loss": 4.116, + "step": 4489 + }, + { + "epoch": 0.0449, + "grad_norm": 0.8752609491348267, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 4490 + }, + { + "epoch": 0.04491, + "grad_norm": 0.8151006698608398, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 4491 + }, + { + "epoch": 0.04492, + "grad_norm": 0.852803111076355, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 4492 + }, + { + "epoch": 0.04493, + "grad_norm": 0.8379290699958801, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 4493 + }, + { + "epoch": 0.04494, + "grad_norm": 0.7042667269706726, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 4494 + }, + { + "epoch": 0.04495, + "grad_norm": 0.625234842300415, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 4495 + }, + { + "epoch": 0.04496, + "grad_norm": 0.666206955909729, + "learning_rate": 0.003, + "loss": 4.131, + "step": 4496 + }, + { + "epoch": 0.04497, + "grad_norm": 0.6894402503967285, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 4497 + }, + { + "epoch": 0.04498, + "grad_norm": 0.6315938234329224, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 4498 + }, + { + "epoch": 0.04499, + "grad_norm": 0.6740808486938477, + "learning_rate": 0.003, + "loss": 4.054, + "step": 4499 + }, + { + "epoch": 0.045, + "grad_norm": 0.6926952600479126, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 4500 + }, + { + "epoch": 0.04501, + "grad_norm": 0.7954657673835754, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 4501 + }, + { + "epoch": 0.04502, + "grad_norm": 0.9196973443031311, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 4502 + }, + { + "epoch": 0.04503, + "grad_norm": 0.7730826139450073, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 4503 + }, + { + "epoch": 0.04504, + "grad_norm": 0.5750721096992493, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 4504 + }, + { + "epoch": 0.04505, + "grad_norm": 0.5613157749176025, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 4505 + }, + { + "epoch": 0.04506, + "grad_norm": 0.6329727172851562, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 4506 + }, + { + "epoch": 0.04507, + "grad_norm": 0.5682700872421265, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 4507 + }, + { + "epoch": 0.04508, + "grad_norm": 0.584396481513977, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 4508 + }, + { + "epoch": 0.04509, + "grad_norm": 0.5912066698074341, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 4509 + }, + { + "epoch": 0.0451, + "grad_norm": 0.6018584966659546, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 4510 + }, + { + "epoch": 0.04511, + "grad_norm": 0.5022575855255127, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 4511 + }, + { + "epoch": 0.04512, + "grad_norm": 0.48740530014038086, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 4512 + }, + { + "epoch": 0.04513, + "grad_norm": 0.5775710940361023, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 4513 + }, + { + "epoch": 0.04514, + "grad_norm": 0.5699050426483154, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 4514 + }, + { + "epoch": 0.04515, + "grad_norm": 0.5856686234474182, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 4515 + }, + { + "epoch": 0.04516, + "grad_norm": 0.7566087245941162, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 4516 + }, + { + "epoch": 0.04517, + "grad_norm": 0.8498972654342651, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 4517 + }, + { + "epoch": 0.04518, + "grad_norm": 0.8767289519309998, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 4518 + }, + { + "epoch": 0.04519, + "grad_norm": 0.8576000332832336, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 4519 + }, + { + "epoch": 0.0452, + "grad_norm": 0.6596296429634094, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 4520 + }, + { + "epoch": 0.04521, + "grad_norm": 0.5750494599342346, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 4521 + }, + { + "epoch": 0.04522, + "grad_norm": 0.6795118451118469, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 4522 + }, + { + "epoch": 0.04523, + "grad_norm": 0.9297488331794739, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 4523 + }, + { + "epoch": 0.04524, + "grad_norm": 0.9432056546211243, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 4524 + }, + { + "epoch": 0.04525, + "grad_norm": 0.8285968899726868, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 4525 + }, + { + "epoch": 0.04526, + "grad_norm": 0.757490873336792, + "learning_rate": 0.003, + "loss": 4.101, + "step": 4526 + }, + { + "epoch": 0.04527, + "grad_norm": 0.8935887813568115, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 4527 + }, + { + "epoch": 0.04528, + "grad_norm": 0.8422304391860962, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 4528 + }, + { + "epoch": 0.04529, + "grad_norm": 0.7896732687950134, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 4529 + }, + { + "epoch": 0.0453, + "grad_norm": 0.7451812624931335, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 4530 + }, + { + "epoch": 0.04531, + "grad_norm": 0.6093108654022217, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 4531 + }, + { + "epoch": 0.04532, + "grad_norm": 0.61775141954422, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 4532 + }, + { + "epoch": 0.04533, + "grad_norm": 0.6170506477355957, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 4533 + }, + { + "epoch": 0.04534, + "grad_norm": 0.7349357604980469, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 4534 + }, + { + "epoch": 0.04535, + "grad_norm": 0.7496950626373291, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 4535 + }, + { + "epoch": 0.04536, + "grad_norm": 0.7089481949806213, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 4536 + }, + { + "epoch": 0.04537, + "grad_norm": 0.7460286617279053, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 4537 + }, + { + "epoch": 0.04538, + "grad_norm": 0.7358576059341431, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 4538 + }, + { + "epoch": 0.04539, + "grad_norm": 0.8624926805496216, + "learning_rate": 0.003, + "loss": 4.069, + "step": 4539 + }, + { + "epoch": 0.0454, + "grad_norm": 0.8797523379325867, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 4540 + }, + { + "epoch": 0.04541, + "grad_norm": 0.7332608103752136, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 4541 + }, + { + "epoch": 0.04542, + "grad_norm": 0.6648062467575073, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 4542 + }, + { + "epoch": 0.04543, + "grad_norm": 0.6967626810073853, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 4543 + }, + { + "epoch": 0.04544, + "grad_norm": 0.8820528984069824, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 4544 + }, + { + "epoch": 0.04545, + "grad_norm": 1.0358401536941528, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 4545 + }, + { + "epoch": 0.04546, + "grad_norm": 0.9551036357879639, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 4546 + }, + { + "epoch": 0.04547, + "grad_norm": 0.8078659176826477, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 4547 + }, + { + "epoch": 0.04548, + "grad_norm": 0.7178813219070435, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 4548 + }, + { + "epoch": 0.04549, + "grad_norm": 0.6956031918525696, + "learning_rate": 0.003, + "loss": 4.086, + "step": 4549 + }, + { + "epoch": 0.0455, + "grad_norm": 0.7719703912734985, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 4550 + }, + { + "epoch": 0.04551, + "grad_norm": 0.8239540457725525, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 4551 + }, + { + "epoch": 0.04552, + "grad_norm": 0.6435329914093018, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 4552 + }, + { + "epoch": 0.04553, + "grad_norm": 0.624059796333313, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 4553 + }, + { + "epoch": 0.04554, + "grad_norm": 0.6374339461326599, + "learning_rate": 0.003, + "loss": 4.081, + "step": 4554 + }, + { + "epoch": 0.04555, + "grad_norm": 0.6004091501235962, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 4555 + }, + { + "epoch": 0.04556, + "grad_norm": 0.6625860333442688, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 4556 + }, + { + "epoch": 0.04557, + "grad_norm": 0.6363006830215454, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 4557 + }, + { + "epoch": 0.04558, + "grad_norm": 0.5139217972755432, + "learning_rate": 0.003, + "loss": 4.102, + "step": 4558 + }, + { + "epoch": 0.04559, + "grad_norm": 0.5389745831489563, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 4559 + }, + { + "epoch": 0.0456, + "grad_norm": 0.5130912065505981, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 4560 + }, + { + "epoch": 0.04561, + "grad_norm": 0.4410141706466675, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 4561 + }, + { + "epoch": 0.04562, + "grad_norm": 0.4584052860736847, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 4562 + }, + { + "epoch": 0.04563, + "grad_norm": 0.4929635226726532, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 4563 + }, + { + "epoch": 0.04564, + "grad_norm": 0.5389020442962646, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 4564 + }, + { + "epoch": 0.04565, + "grad_norm": 0.5814055800437927, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 4565 + }, + { + "epoch": 0.04566, + "grad_norm": 0.691055953502655, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 4566 + }, + { + "epoch": 0.04567, + "grad_norm": 0.7630618810653687, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 4567 + }, + { + "epoch": 0.04568, + "grad_norm": 0.6998851895332336, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 4568 + }, + { + "epoch": 0.04569, + "grad_norm": 0.641377866268158, + "learning_rate": 0.003, + "loss": 4.085, + "step": 4569 + }, + { + "epoch": 0.0457, + "grad_norm": 0.6208799481391907, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 4570 + }, + { + "epoch": 0.04571, + "grad_norm": 0.6723332405090332, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 4571 + }, + { + "epoch": 0.04572, + "grad_norm": 0.7085137367248535, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 4572 + }, + { + "epoch": 0.04573, + "grad_norm": 0.5809813141822815, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 4573 + }, + { + "epoch": 0.04574, + "grad_norm": 0.638818621635437, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 4574 + }, + { + "epoch": 0.04575, + "grad_norm": 0.6738321781158447, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 4575 + }, + { + "epoch": 0.04576, + "grad_norm": 0.6550054550170898, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 4576 + }, + { + "epoch": 0.04577, + "grad_norm": 0.6867873072624207, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 4577 + }, + { + "epoch": 0.04578, + "grad_norm": 0.7424322962760925, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 4578 + }, + { + "epoch": 0.04579, + "grad_norm": 1.0748231410980225, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 4579 + }, + { + "epoch": 0.0458, + "grad_norm": 1.1119617223739624, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 4580 + }, + { + "epoch": 0.04581, + "grad_norm": 0.816440999507904, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 4581 + }, + { + "epoch": 0.04582, + "grad_norm": 0.9225358963012695, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 4582 + }, + { + "epoch": 0.04583, + "grad_norm": 1.3155148029327393, + "learning_rate": 0.003, + "loss": 4.11, + "step": 4583 + }, + { + "epoch": 0.04584, + "grad_norm": 0.9258817434310913, + "learning_rate": 0.003, + "loss": 4.103, + "step": 4584 + }, + { + "epoch": 0.04585, + "grad_norm": 1.0143473148345947, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 4585 + }, + { + "epoch": 0.04586, + "grad_norm": 0.9881197214126587, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 4586 + }, + { + "epoch": 0.04587, + "grad_norm": 0.8126780390739441, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 4587 + }, + { + "epoch": 0.04588, + "grad_norm": 0.7166712880134583, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 4588 + }, + { + "epoch": 0.04589, + "grad_norm": 0.652772068977356, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 4589 + }, + { + "epoch": 0.0459, + "grad_norm": 0.671996533870697, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 4590 + }, + { + "epoch": 0.04591, + "grad_norm": 0.7832691669464111, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 4591 + }, + { + "epoch": 0.04592, + "grad_norm": 0.861909031867981, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 4592 + }, + { + "epoch": 0.04593, + "grad_norm": 0.8999555706977844, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 4593 + }, + { + "epoch": 0.04594, + "grad_norm": 0.7495047450065613, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 4594 + }, + { + "epoch": 0.04595, + "grad_norm": 0.6235268712043762, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 4595 + }, + { + "epoch": 0.04596, + "grad_norm": 0.6757757067680359, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 4596 + }, + { + "epoch": 0.04597, + "grad_norm": 0.7834587693214417, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 4597 + }, + { + "epoch": 0.04598, + "grad_norm": 0.7589462399482727, + "learning_rate": 0.003, + "loss": 4.1431, + "step": 4598 + }, + { + "epoch": 0.04599, + "grad_norm": 0.7294259667396545, + "learning_rate": 0.003, + "loss": 4.088, + "step": 4599 + }, + { + "epoch": 0.046, + "grad_norm": 0.5923947095870972, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 4600 + }, + { + "epoch": 0.04601, + "grad_norm": 0.5730067491531372, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 4601 + }, + { + "epoch": 0.04602, + "grad_norm": 0.5987418293952942, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 4602 + }, + { + "epoch": 0.04603, + "grad_norm": 0.6320177912712097, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 4603 + }, + { + "epoch": 0.04604, + "grad_norm": 0.6222275495529175, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 4604 + }, + { + "epoch": 0.04605, + "grad_norm": 0.5284239053726196, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 4605 + }, + { + "epoch": 0.04606, + "grad_norm": 0.3622366189956665, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 4606 + }, + { + "epoch": 0.04607, + "grad_norm": 0.41870078444480896, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 4607 + }, + { + "epoch": 0.04608, + "grad_norm": 0.4503072500228882, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 4608 + }, + { + "epoch": 0.04609, + "grad_norm": 0.4834275245666504, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 4609 + }, + { + "epoch": 0.0461, + "grad_norm": 0.543554961681366, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 4610 + }, + { + "epoch": 0.04611, + "grad_norm": 0.6071869134902954, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 4611 + }, + { + "epoch": 0.04612, + "grad_norm": 0.6595013737678528, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 4612 + }, + { + "epoch": 0.04613, + "grad_norm": 0.6050468683242798, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 4613 + }, + { + "epoch": 0.04614, + "grad_norm": 0.5637344717979431, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 4614 + }, + { + "epoch": 0.04615, + "grad_norm": 0.5728361010551453, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 4615 + }, + { + "epoch": 0.04616, + "grad_norm": 0.6123324036598206, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 4616 + }, + { + "epoch": 0.04617, + "grad_norm": 0.6856390833854675, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 4617 + }, + { + "epoch": 0.04618, + "grad_norm": 0.7527903318405151, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 4618 + }, + { + "epoch": 0.04619, + "grad_norm": 0.8059439659118652, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 4619 + }, + { + "epoch": 0.0462, + "grad_norm": 0.925609290599823, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 4620 + }, + { + "epoch": 0.04621, + "grad_norm": 1.0445528030395508, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 4621 + }, + { + "epoch": 0.04622, + "grad_norm": 1.04862642288208, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 4622 + }, + { + "epoch": 0.04623, + "grad_norm": 0.8834772706031799, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 4623 + }, + { + "epoch": 0.04624, + "grad_norm": 0.9410740733146667, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 4624 + }, + { + "epoch": 0.04625, + "grad_norm": 0.6689598560333252, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 4625 + }, + { + "epoch": 0.04626, + "grad_norm": 0.6782405376434326, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 4626 + }, + { + "epoch": 0.04627, + "grad_norm": 0.5961948037147522, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 4627 + }, + { + "epoch": 0.04628, + "grad_norm": 0.8130809664726257, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 4628 + }, + { + "epoch": 0.04629, + "grad_norm": 0.8763521313667297, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 4629 + }, + { + "epoch": 0.0463, + "grad_norm": 0.911468505859375, + "learning_rate": 0.003, + "loss": 4.112, + "step": 4630 + }, + { + "epoch": 0.04631, + "grad_norm": 1.0289772748947144, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 4631 + }, + { + "epoch": 0.04632, + "grad_norm": 0.940280556678772, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 4632 + }, + { + "epoch": 0.04633, + "grad_norm": 0.7662308812141418, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 4633 + }, + { + "epoch": 0.04634, + "grad_norm": 0.7879605889320374, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 4634 + }, + { + "epoch": 0.04635, + "grad_norm": 0.8211584687232971, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 4635 + }, + { + "epoch": 0.04636, + "grad_norm": 0.8329793810844421, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 4636 + }, + { + "epoch": 0.04637, + "grad_norm": 0.7591750621795654, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 4637 + }, + { + "epoch": 0.04638, + "grad_norm": 0.8332642912864685, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 4638 + }, + { + "epoch": 0.04639, + "grad_norm": 0.9349274039268494, + "learning_rate": 0.003, + "loss": 4.087, + "step": 4639 + }, + { + "epoch": 0.0464, + "grad_norm": 1.0191375017166138, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 4640 + }, + { + "epoch": 0.04641, + "grad_norm": 0.9005967974662781, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 4641 + }, + { + "epoch": 0.04642, + "grad_norm": 0.7704114317893982, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 4642 + }, + { + "epoch": 0.04643, + "grad_norm": 0.6552042365074158, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 4643 + }, + { + "epoch": 0.04644, + "grad_norm": 0.5903872847557068, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 4644 + }, + { + "epoch": 0.04645, + "grad_norm": 0.5108548998832703, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 4645 + }, + { + "epoch": 0.04646, + "grad_norm": 0.5088199973106384, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 4646 + }, + { + "epoch": 0.04647, + "grad_norm": 0.5300918221473694, + "learning_rate": 0.003, + "loss": 4.078, + "step": 4647 + }, + { + "epoch": 0.04648, + "grad_norm": 0.4461018443107605, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 4648 + }, + { + "epoch": 0.04649, + "grad_norm": 0.4164982736110687, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 4649 + }, + { + "epoch": 0.0465, + "grad_norm": 0.4256885349750519, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 4650 + }, + { + "epoch": 0.04651, + "grad_norm": 0.4963768720626831, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 4651 + }, + { + "epoch": 0.04652, + "grad_norm": 0.6006476879119873, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 4652 + }, + { + "epoch": 0.04653, + "grad_norm": 0.6713421940803528, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 4653 + }, + { + "epoch": 0.04654, + "grad_norm": 0.7249282002449036, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 4654 + }, + { + "epoch": 0.04655, + "grad_norm": 0.6726330518722534, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 4655 + }, + { + "epoch": 0.04656, + "grad_norm": 0.536870002746582, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 4656 + }, + { + "epoch": 0.04657, + "grad_norm": 0.5610942244529724, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 4657 + }, + { + "epoch": 0.04658, + "grad_norm": 0.7012815475463867, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 4658 + }, + { + "epoch": 0.04659, + "grad_norm": 0.7252200245857239, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 4659 + }, + { + "epoch": 0.0466, + "grad_norm": 0.6837133765220642, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 4660 + }, + { + "epoch": 0.04661, + "grad_norm": 0.6772074103355408, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 4661 + }, + { + "epoch": 0.04662, + "grad_norm": 0.6671146154403687, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 4662 + }, + { + "epoch": 0.04663, + "grad_norm": 0.6181881427764893, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 4663 + }, + { + "epoch": 0.04664, + "grad_norm": 0.6068332195281982, + "learning_rate": 0.003, + "loss": 4.063, + "step": 4664 + }, + { + "epoch": 0.04665, + "grad_norm": 0.5939993858337402, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 4665 + }, + { + "epoch": 0.04666, + "grad_norm": 0.6058046221733093, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 4666 + }, + { + "epoch": 0.04667, + "grad_norm": 0.6686825752258301, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 4667 + }, + { + "epoch": 0.04668, + "grad_norm": 0.7094378471374512, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 4668 + }, + { + "epoch": 0.04669, + "grad_norm": 0.8139551281929016, + "learning_rate": 0.003, + "loss": 4.098, + "step": 4669 + }, + { + "epoch": 0.0467, + "grad_norm": 1.0420807600021362, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 4670 + }, + { + "epoch": 0.04671, + "grad_norm": 1.1174323558807373, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 4671 + }, + { + "epoch": 0.04672, + "grad_norm": 0.8469231724739075, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 4672 + }, + { + "epoch": 0.04673, + "grad_norm": 0.8073984980583191, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 4673 + }, + { + "epoch": 0.04674, + "grad_norm": 0.76691734790802, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 4674 + }, + { + "epoch": 0.04675, + "grad_norm": 0.8686924576759338, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 4675 + }, + { + "epoch": 0.04676, + "grad_norm": 1.0538427829742432, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 4676 + }, + { + "epoch": 0.04677, + "grad_norm": 1.003646731376648, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 4677 + }, + { + "epoch": 0.04678, + "grad_norm": 0.7191857695579529, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 4678 + }, + { + "epoch": 0.04679, + "grad_norm": 0.6493638753890991, + "learning_rate": 0.003, + "loss": 4.095, + "step": 4679 + }, + { + "epoch": 0.0468, + "grad_norm": 0.6600551605224609, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 4680 + }, + { + "epoch": 0.04681, + "grad_norm": 0.7209051251411438, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 4681 + }, + { + "epoch": 0.04682, + "grad_norm": 0.8192289471626282, + "learning_rate": 0.003, + "loss": 4.088, + "step": 4682 + }, + { + "epoch": 0.04683, + "grad_norm": 0.7203931212425232, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 4683 + }, + { + "epoch": 0.04684, + "grad_norm": 0.6532163023948669, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 4684 + }, + { + "epoch": 0.04685, + "grad_norm": 0.6800267100334167, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 4685 + }, + { + "epoch": 0.04686, + "grad_norm": 0.6990789771080017, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 4686 + }, + { + "epoch": 0.04687, + "grad_norm": 0.6912499666213989, + "learning_rate": 0.003, + "loss": 4.119, + "step": 4687 + }, + { + "epoch": 0.04688, + "grad_norm": 0.7285837531089783, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 4688 + }, + { + "epoch": 0.04689, + "grad_norm": 0.6683054566383362, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 4689 + }, + { + "epoch": 0.0469, + "grad_norm": 0.5229889154434204, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 4690 + }, + { + "epoch": 0.04691, + "grad_norm": 0.49395525455474854, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 4691 + }, + { + "epoch": 0.04692, + "grad_norm": 0.46230220794677734, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 4692 + }, + { + "epoch": 0.04693, + "grad_norm": 0.4435535967350006, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 4693 + }, + { + "epoch": 0.04694, + "grad_norm": 0.43420031666755676, + "learning_rate": 0.003, + "loss": 4.063, + "step": 4694 + }, + { + "epoch": 0.04695, + "grad_norm": 0.4660513699054718, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 4695 + }, + { + "epoch": 0.04696, + "grad_norm": 0.5974222421646118, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 4696 + }, + { + "epoch": 0.04697, + "grad_norm": 0.7868569493293762, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 4697 + }, + { + "epoch": 0.04698, + "grad_norm": 0.9865483641624451, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 4698 + }, + { + "epoch": 0.04699, + "grad_norm": 0.9736619591712952, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 4699 + }, + { + "epoch": 0.047, + "grad_norm": 0.7273622155189514, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 4700 + }, + { + "epoch": 0.04701, + "grad_norm": 0.6724737882614136, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 4701 + }, + { + "epoch": 0.04702, + "grad_norm": 0.8422785997390747, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 4702 + }, + { + "epoch": 0.04703, + "grad_norm": 0.743380606174469, + "learning_rate": 0.003, + "loss": 4.081, + "step": 4703 + }, + { + "epoch": 0.04704, + "grad_norm": 0.6584486365318298, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 4704 + }, + { + "epoch": 0.04705, + "grad_norm": 0.6731492877006531, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 4705 + }, + { + "epoch": 0.04706, + "grad_norm": 0.6525211334228516, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 4706 + }, + { + "epoch": 0.04707, + "grad_norm": 0.6952141523361206, + "learning_rate": 0.003, + "loss": 4.091, + "step": 4707 + }, + { + "epoch": 0.04708, + "grad_norm": 0.6215411424636841, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 4708 + }, + { + "epoch": 0.04709, + "grad_norm": 0.5248569250106812, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 4709 + }, + { + "epoch": 0.0471, + "grad_norm": 0.516987681388855, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 4710 + }, + { + "epoch": 0.04711, + "grad_norm": 0.513450562953949, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 4711 + }, + { + "epoch": 0.04712, + "grad_norm": 0.5270048975944519, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 4712 + }, + { + "epoch": 0.04713, + "grad_norm": 0.5057416558265686, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 4713 + }, + { + "epoch": 0.04714, + "grad_norm": 0.5090026259422302, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 4714 + }, + { + "epoch": 0.04715, + "grad_norm": 0.5532310009002686, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 4715 + }, + { + "epoch": 0.04716, + "grad_norm": 0.6358587145805359, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 4716 + }, + { + "epoch": 0.04717, + "grad_norm": 0.7907416224479675, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 4717 + }, + { + "epoch": 0.04718, + "grad_norm": 0.845724880695343, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 4718 + }, + { + "epoch": 0.04719, + "grad_norm": 0.8444932699203491, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 4719 + }, + { + "epoch": 0.0472, + "grad_norm": 0.8913478255271912, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 4720 + }, + { + "epoch": 0.04721, + "grad_norm": 0.9782090187072754, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 4721 + }, + { + "epoch": 0.04722, + "grad_norm": 1.1551313400268555, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 4722 + }, + { + "epoch": 0.04723, + "grad_norm": 1.027269721031189, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 4723 + }, + { + "epoch": 0.04724, + "grad_norm": 0.9385249018669128, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 4724 + }, + { + "epoch": 0.04725, + "grad_norm": 1.0505073070526123, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 4725 + }, + { + "epoch": 0.04726, + "grad_norm": 1.0679757595062256, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 4726 + }, + { + "epoch": 0.04727, + "grad_norm": 0.9189696907997131, + "learning_rate": 0.003, + "loss": 4.092, + "step": 4727 + }, + { + "epoch": 0.04728, + "grad_norm": 0.9319504499435425, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 4728 + }, + { + "epoch": 0.04729, + "grad_norm": 0.9617239832878113, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 4729 + }, + { + "epoch": 0.0473, + "grad_norm": 0.8712969422340393, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 4730 + }, + { + "epoch": 0.04731, + "grad_norm": 0.8123112320899963, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 4731 + }, + { + "epoch": 0.04732, + "grad_norm": 0.7513419985771179, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 4732 + }, + { + "epoch": 0.04733, + "grad_norm": 0.7118027210235596, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 4733 + }, + { + "epoch": 0.04734, + "grad_norm": 0.6488009691238403, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 4734 + }, + { + "epoch": 0.04735, + "grad_norm": 0.613771378993988, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 4735 + }, + { + "epoch": 0.04736, + "grad_norm": 0.6943199038505554, + "learning_rate": 0.003, + "loss": 4.104, + "step": 4736 + }, + { + "epoch": 0.04737, + "grad_norm": 0.9270548224449158, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 4737 + }, + { + "epoch": 0.04738, + "grad_norm": 1.0554243326187134, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 4738 + }, + { + "epoch": 0.04739, + "grad_norm": 1.0856826305389404, + "learning_rate": 0.003, + "loss": 4.058, + "step": 4739 + }, + { + "epoch": 0.0474, + "grad_norm": 0.664472758769989, + "learning_rate": 0.003, + "loss": 4.118, + "step": 4740 + }, + { + "epoch": 0.04741, + "grad_norm": 0.576934814453125, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 4741 + }, + { + "epoch": 0.04742, + "grad_norm": 0.6367508172988892, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 4742 + }, + { + "epoch": 0.04743, + "grad_norm": 0.669205904006958, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 4743 + }, + { + "epoch": 0.04744, + "grad_norm": 0.67984938621521, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 4744 + }, + { + "epoch": 0.04745, + "grad_norm": 0.6410228610038757, + "learning_rate": 0.003, + "loss": 4.073, + "step": 4745 + }, + { + "epoch": 0.04746, + "grad_norm": 0.6202641725540161, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 4746 + }, + { + "epoch": 0.04747, + "grad_norm": 0.6701365113258362, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 4747 + }, + { + "epoch": 0.04748, + "grad_norm": 0.5763520002365112, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 4748 + }, + { + "epoch": 0.04749, + "grad_norm": 0.4907684326171875, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 4749 + }, + { + "epoch": 0.0475, + "grad_norm": 0.507577121257782, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 4750 + }, + { + "epoch": 0.04751, + "grad_norm": 0.47070327401161194, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 4751 + }, + { + "epoch": 0.04752, + "grad_norm": 0.37678685784339905, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 4752 + }, + { + "epoch": 0.04753, + "grad_norm": 0.3762926459312439, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 4753 + }, + { + "epoch": 0.04754, + "grad_norm": 0.38185784220695496, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 4754 + }, + { + "epoch": 0.04755, + "grad_norm": 0.39762863516807556, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 4755 + }, + { + "epoch": 0.04756, + "grad_norm": 0.42808109521865845, + "learning_rate": 0.003, + "loss": 4.073, + "step": 4756 + }, + { + "epoch": 0.04757, + "grad_norm": 0.4888444244861603, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 4757 + }, + { + "epoch": 0.04758, + "grad_norm": 0.6120839715003967, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 4758 + }, + { + "epoch": 0.04759, + "grad_norm": 0.7273629307746887, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 4759 + }, + { + "epoch": 0.0476, + "grad_norm": 0.7623408436775208, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 4760 + }, + { + "epoch": 0.04761, + "grad_norm": 0.7425439357757568, + "learning_rate": 0.003, + "loss": 4.067, + "step": 4761 + }, + { + "epoch": 0.04762, + "grad_norm": 0.6609148383140564, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 4762 + }, + { + "epoch": 0.04763, + "grad_norm": 0.5869941711425781, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 4763 + }, + { + "epoch": 0.04764, + "grad_norm": 0.6543363928794861, + "learning_rate": 0.003, + "loss": 4.055, + "step": 4764 + }, + { + "epoch": 0.04765, + "grad_norm": 0.6935657858848572, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 4765 + }, + { + "epoch": 0.04766, + "grad_norm": 0.6691122651100159, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 4766 + }, + { + "epoch": 0.04767, + "grad_norm": 0.7805852890014648, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 4767 + }, + { + "epoch": 0.04768, + "grad_norm": 0.8912307024002075, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 4768 + }, + { + "epoch": 0.04769, + "grad_norm": 1.0201588869094849, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 4769 + }, + { + "epoch": 0.0477, + "grad_norm": 1.0079450607299805, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 4770 + }, + { + "epoch": 0.04771, + "grad_norm": 0.9444756507873535, + "learning_rate": 0.003, + "loss": 4.103, + "step": 4771 + }, + { + "epoch": 0.04772, + "grad_norm": 0.9170507788658142, + "learning_rate": 0.003, + "loss": 4.1, + "step": 4772 + }, + { + "epoch": 0.04773, + "grad_norm": 0.7705607414245605, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 4773 + }, + { + "epoch": 0.04774, + "grad_norm": 0.7739026546478271, + "learning_rate": 0.003, + "loss": 4.103, + "step": 4774 + }, + { + "epoch": 0.04775, + "grad_norm": 0.8026602268218994, + "learning_rate": 0.003, + "loss": 4.101, + "step": 4775 + }, + { + "epoch": 0.04776, + "grad_norm": 0.9427769780158997, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 4776 + }, + { + "epoch": 0.04777, + "grad_norm": 1.033246397972107, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 4777 + }, + { + "epoch": 0.04778, + "grad_norm": 0.9891014695167542, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 4778 + }, + { + "epoch": 0.04779, + "grad_norm": 0.8989018201828003, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 4779 + }, + { + "epoch": 0.0478, + "grad_norm": 0.7684260010719299, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 4780 + }, + { + "epoch": 0.04781, + "grad_norm": 0.6779473423957825, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 4781 + }, + { + "epoch": 0.04782, + "grad_norm": 0.7075977325439453, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 4782 + }, + { + "epoch": 0.04783, + "grad_norm": 0.7358579635620117, + "learning_rate": 0.003, + "loss": 4.082, + "step": 4783 + }, + { + "epoch": 0.04784, + "grad_norm": 0.8492463231086731, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 4784 + }, + { + "epoch": 0.04785, + "grad_norm": 1.0335021018981934, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 4785 + }, + { + "epoch": 0.04786, + "grad_norm": 1.014879584312439, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 4786 + }, + { + "epoch": 0.04787, + "grad_norm": 0.7597166299819946, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 4787 + }, + { + "epoch": 0.04788, + "grad_norm": 0.7490633130073547, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 4788 + }, + { + "epoch": 0.04789, + "grad_norm": 0.7627069354057312, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 4789 + }, + { + "epoch": 0.0479, + "grad_norm": 0.7988529801368713, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 4790 + }, + { + "epoch": 0.04791, + "grad_norm": 0.7073328495025635, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 4791 + }, + { + "epoch": 0.04792, + "grad_norm": 0.5642073750495911, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 4792 + }, + { + "epoch": 0.04793, + "grad_norm": 0.4884117543697357, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 4793 + }, + { + "epoch": 0.04794, + "grad_norm": 0.5512744188308716, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 4794 + }, + { + "epoch": 0.04795, + "grad_norm": 0.6215276718139648, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 4795 + }, + { + "epoch": 0.04796, + "grad_norm": 0.5437957644462585, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 4796 + }, + { + "epoch": 0.04797, + "grad_norm": 0.475530207157135, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 4797 + }, + { + "epoch": 0.04798, + "grad_norm": 0.4487910866737366, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 4798 + }, + { + "epoch": 0.04799, + "grad_norm": 0.4571172297000885, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 4799 + }, + { + "epoch": 0.048, + "grad_norm": 0.4549039602279663, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 4800 + }, + { + "epoch": 0.04801, + "grad_norm": 0.4293798804283142, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 4801 + }, + { + "epoch": 0.04802, + "grad_norm": 0.3992469012737274, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 4802 + }, + { + "epoch": 0.04803, + "grad_norm": 0.39434272050857544, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 4803 + }, + { + "epoch": 0.04804, + "grad_norm": 0.4348972737789154, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 4804 + }, + { + "epoch": 0.04805, + "grad_norm": 0.4008714556694031, + "learning_rate": 0.003, + "loss": 4.086, + "step": 4805 + }, + { + "epoch": 0.04806, + "grad_norm": 0.4113296568393707, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 4806 + }, + { + "epoch": 0.04807, + "grad_norm": 0.4709857106208801, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 4807 + }, + { + "epoch": 0.04808, + "grad_norm": 0.6104065179824829, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 4808 + }, + { + "epoch": 0.04809, + "grad_norm": 0.9111732244491577, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 4809 + }, + { + "epoch": 0.0481, + "grad_norm": 1.2555900812149048, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 4810 + }, + { + "epoch": 0.04811, + "grad_norm": 0.6506162881851196, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 4811 + }, + { + "epoch": 0.04812, + "grad_norm": 0.5703248977661133, + "learning_rate": 0.003, + "loss": 4.084, + "step": 4812 + }, + { + "epoch": 0.04813, + "grad_norm": 0.7647963166236877, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 4813 + }, + { + "epoch": 0.04814, + "grad_norm": 0.8564413189888, + "learning_rate": 0.003, + "loss": 4.033, + "step": 4814 + }, + { + "epoch": 0.04815, + "grad_norm": 0.8082033395767212, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 4815 + }, + { + "epoch": 0.04816, + "grad_norm": 0.6857293248176575, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 4816 + }, + { + "epoch": 0.04817, + "grad_norm": 0.5996972918510437, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 4817 + }, + { + "epoch": 0.04818, + "grad_norm": 0.6385634541511536, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 4818 + }, + { + "epoch": 0.04819, + "grad_norm": 0.6788516044616699, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 4819 + }, + { + "epoch": 0.0482, + "grad_norm": 0.6623933911323547, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 4820 + }, + { + "epoch": 0.04821, + "grad_norm": 0.6214218735694885, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 4821 + }, + { + "epoch": 0.04822, + "grad_norm": 0.6991559267044067, + "learning_rate": 0.003, + "loss": 4.075, + "step": 4822 + }, + { + "epoch": 0.04823, + "grad_norm": 0.7806605100631714, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 4823 + }, + { + "epoch": 0.04824, + "grad_norm": 0.7918847799301147, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 4824 + }, + { + "epoch": 0.04825, + "grad_norm": 0.6949120163917542, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 4825 + }, + { + "epoch": 0.04826, + "grad_norm": 0.6233998537063599, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 4826 + }, + { + "epoch": 0.04827, + "grad_norm": 0.6551531553268433, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 4827 + }, + { + "epoch": 0.04828, + "grad_norm": 0.8566862344741821, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 4828 + }, + { + "epoch": 0.04829, + "grad_norm": 1.0621153116226196, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 4829 + }, + { + "epoch": 0.0483, + "grad_norm": 1.08018159866333, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 4830 + }, + { + "epoch": 0.04831, + "grad_norm": 0.8102112412452698, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 4831 + }, + { + "epoch": 0.04832, + "grad_norm": 0.5810506939888, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 4832 + }, + { + "epoch": 0.04833, + "grad_norm": 0.6366890072822571, + "learning_rate": 0.003, + "loss": 4.085, + "step": 4833 + }, + { + "epoch": 0.04834, + "grad_norm": 0.7475446462631226, + "learning_rate": 0.003, + "loss": 4.058, + "step": 4834 + }, + { + "epoch": 0.04835, + "grad_norm": 0.8804463148117065, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 4835 + }, + { + "epoch": 0.04836, + "grad_norm": 0.9185641407966614, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 4836 + }, + { + "epoch": 0.04837, + "grad_norm": 0.7357785105705261, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 4837 + }, + { + "epoch": 0.04838, + "grad_norm": 0.6829032897949219, + "learning_rate": 0.003, + "loss": 4.103, + "step": 4838 + }, + { + "epoch": 0.04839, + "grad_norm": 0.7056359052658081, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 4839 + }, + { + "epoch": 0.0484, + "grad_norm": 0.8075634837150574, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 4840 + }, + { + "epoch": 0.04841, + "grad_norm": 0.8837736248970032, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 4841 + }, + { + "epoch": 0.04842, + "grad_norm": 0.8638008832931519, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 4842 + }, + { + "epoch": 0.04843, + "grad_norm": 0.7591082453727722, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 4843 + }, + { + "epoch": 0.04844, + "grad_norm": 0.6942276954650879, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 4844 + }, + { + "epoch": 0.04845, + "grad_norm": 0.7003408670425415, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 4845 + }, + { + "epoch": 0.04846, + "grad_norm": 0.7513200640678406, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 4846 + }, + { + "epoch": 0.04847, + "grad_norm": 0.8167382478713989, + "learning_rate": 0.003, + "loss": 4.066, + "step": 4847 + }, + { + "epoch": 0.04848, + "grad_norm": 0.8485641479492188, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 4848 + }, + { + "epoch": 0.04849, + "grad_norm": 0.7200870513916016, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 4849 + }, + { + "epoch": 0.0485, + "grad_norm": 0.6998422145843506, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 4850 + }, + { + "epoch": 0.04851, + "grad_norm": 0.6962882280349731, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 4851 + }, + { + "epoch": 0.04852, + "grad_norm": 0.6214674115180969, + "learning_rate": 0.003, + "loss": 4.031, + "step": 4852 + }, + { + "epoch": 0.04853, + "grad_norm": 0.6190800666809082, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 4853 + }, + { + "epoch": 0.04854, + "grad_norm": 0.5581333637237549, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 4854 + }, + { + "epoch": 0.04855, + "grad_norm": 0.4471603035926819, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 4855 + }, + { + "epoch": 0.04856, + "grad_norm": 0.4486013650894165, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 4856 + }, + { + "epoch": 0.04857, + "grad_norm": 0.4567798376083374, + "learning_rate": 0.003, + "loss": 4.101, + "step": 4857 + }, + { + "epoch": 0.04858, + "grad_norm": 0.4150858223438263, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 4858 + }, + { + "epoch": 0.04859, + "grad_norm": 0.428250789642334, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 4859 + }, + { + "epoch": 0.0486, + "grad_norm": 0.44279322028160095, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 4860 + }, + { + "epoch": 0.04861, + "grad_norm": 0.5173445343971252, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 4861 + }, + { + "epoch": 0.04862, + "grad_norm": 0.6446529030799866, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 4862 + }, + { + "epoch": 0.04863, + "grad_norm": 0.7789918184280396, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 4863 + }, + { + "epoch": 0.04864, + "grad_norm": 0.8804915547370911, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 4864 + }, + { + "epoch": 0.04865, + "grad_norm": 0.8473218679428101, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 4865 + }, + { + "epoch": 0.04866, + "grad_norm": 0.7426813840866089, + "learning_rate": 0.003, + "loss": 4.101, + "step": 4866 + }, + { + "epoch": 0.04867, + "grad_norm": 0.708594799041748, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 4867 + }, + { + "epoch": 0.04868, + "grad_norm": 0.8590058088302612, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 4868 + }, + { + "epoch": 0.04869, + "grad_norm": 0.9592583179473877, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 4869 + }, + { + "epoch": 0.0487, + "grad_norm": 0.8780485391616821, + "learning_rate": 0.003, + "loss": 4.089, + "step": 4870 + }, + { + "epoch": 0.04871, + "grad_norm": 0.7498546838760376, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 4871 + }, + { + "epoch": 0.04872, + "grad_norm": 0.7363294363021851, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 4872 + }, + { + "epoch": 0.04873, + "grad_norm": 0.9078441262245178, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 4873 + }, + { + "epoch": 0.04874, + "grad_norm": 0.9841904044151306, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 4874 + }, + { + "epoch": 0.04875, + "grad_norm": 0.8496304154396057, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 4875 + }, + { + "epoch": 0.04876, + "grad_norm": 0.7967992424964905, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 4876 + }, + { + "epoch": 0.04877, + "grad_norm": 0.8723610639572144, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 4877 + }, + { + "epoch": 0.04878, + "grad_norm": 0.8831827044487, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 4878 + }, + { + "epoch": 0.04879, + "grad_norm": 0.8523681163787842, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 4879 + }, + { + "epoch": 0.0488, + "grad_norm": 0.8125349283218384, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 4880 + }, + { + "epoch": 0.04881, + "grad_norm": 0.7172712087631226, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 4881 + }, + { + "epoch": 0.04882, + "grad_norm": 0.8336440920829773, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 4882 + }, + { + "epoch": 0.04883, + "grad_norm": 1.029686450958252, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 4883 + }, + { + "epoch": 0.04884, + "grad_norm": 1.030254602432251, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 4884 + }, + { + "epoch": 0.04885, + "grad_norm": 0.8340651392936707, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 4885 + }, + { + "epoch": 0.04886, + "grad_norm": 0.7567631006240845, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 4886 + }, + { + "epoch": 0.04887, + "grad_norm": 0.7703821659088135, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 4887 + }, + { + "epoch": 0.04888, + "grad_norm": 0.7452254295349121, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 4888 + }, + { + "epoch": 0.04889, + "grad_norm": 0.819527804851532, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 4889 + }, + { + "epoch": 0.0489, + "grad_norm": 0.7526473999023438, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 4890 + }, + { + "epoch": 0.04891, + "grad_norm": 0.6946446299552917, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 4891 + }, + { + "epoch": 0.04892, + "grad_norm": 0.7302446365356445, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 4892 + }, + { + "epoch": 0.04893, + "grad_norm": 0.8050813674926758, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 4893 + }, + { + "epoch": 0.04894, + "grad_norm": 0.788234293460846, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 4894 + }, + { + "epoch": 0.04895, + "grad_norm": 0.7052075266838074, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 4895 + }, + { + "epoch": 0.04896, + "grad_norm": 0.6905450820922852, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 4896 + }, + { + "epoch": 0.04897, + "grad_norm": 0.7363834381103516, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 4897 + }, + { + "epoch": 0.04898, + "grad_norm": 0.7060516476631165, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 4898 + }, + { + "epoch": 0.04899, + "grad_norm": 0.6093109846115112, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 4899 + }, + { + "epoch": 0.049, + "grad_norm": 0.5717660784721375, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 4900 + }, + { + "epoch": 0.04901, + "grad_norm": 0.5589826703071594, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 4901 + }, + { + "epoch": 0.04902, + "grad_norm": 0.595604419708252, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 4902 + }, + { + "epoch": 0.04903, + "grad_norm": 0.5431516170501709, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 4903 + }, + { + "epoch": 0.04904, + "grad_norm": 0.49542325735092163, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 4904 + }, + { + "epoch": 0.04905, + "grad_norm": 0.5120609998703003, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 4905 + }, + { + "epoch": 0.04906, + "grad_norm": 0.5707340240478516, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 4906 + }, + { + "epoch": 0.04907, + "grad_norm": 0.6552608609199524, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 4907 + }, + { + "epoch": 0.04908, + "grad_norm": 0.6238669157028198, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 4908 + }, + { + "epoch": 0.04909, + "grad_norm": 0.6134541034698486, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 4909 + }, + { + "epoch": 0.0491, + "grad_norm": 0.6873631477355957, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 4910 + }, + { + "epoch": 0.04911, + "grad_norm": 0.7277111411094666, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 4911 + }, + { + "epoch": 0.04912, + "grad_norm": 0.6708826422691345, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 4912 + }, + { + "epoch": 0.04913, + "grad_norm": 0.5802143812179565, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 4913 + }, + { + "epoch": 0.04914, + "grad_norm": 0.5891876220703125, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 4914 + }, + { + "epoch": 0.04915, + "grad_norm": 0.6115869879722595, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 4915 + }, + { + "epoch": 0.04916, + "grad_norm": 0.6224087476730347, + "learning_rate": 0.003, + "loss": 4.073, + "step": 4916 + }, + { + "epoch": 0.04917, + "grad_norm": 0.5933270454406738, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 4917 + }, + { + "epoch": 0.04918, + "grad_norm": 0.6610130071640015, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 4918 + }, + { + "epoch": 0.04919, + "grad_norm": 0.6908468008041382, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 4919 + }, + { + "epoch": 0.0492, + "grad_norm": 0.699761152267456, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 4920 + }, + { + "epoch": 0.04921, + "grad_norm": 0.6772635579109192, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 4921 + }, + { + "epoch": 0.04922, + "grad_norm": 0.6279475092887878, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 4922 + }, + { + "epoch": 0.04923, + "grad_norm": 0.5486379861831665, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 4923 + }, + { + "epoch": 0.04924, + "grad_norm": 0.5325474739074707, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 4924 + }, + { + "epoch": 0.04925, + "grad_norm": 0.6036094427108765, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 4925 + }, + { + "epoch": 0.04926, + "grad_norm": 0.6497531533241272, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 4926 + }, + { + "epoch": 0.04927, + "grad_norm": 0.5924155116081238, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 4927 + }, + { + "epoch": 0.04928, + "grad_norm": 0.6325770616531372, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 4928 + }, + { + "epoch": 0.04929, + "grad_norm": 0.6553009152412415, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 4929 + }, + { + "epoch": 0.0493, + "grad_norm": 0.7570836544036865, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 4930 + }, + { + "epoch": 0.04931, + "grad_norm": 0.8154568672180176, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 4931 + }, + { + "epoch": 0.04932, + "grad_norm": 0.7909042239189148, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 4932 + }, + { + "epoch": 0.04933, + "grad_norm": 0.7257623672485352, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 4933 + }, + { + "epoch": 0.04934, + "grad_norm": 0.7812435030937195, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 4934 + }, + { + "epoch": 0.04935, + "grad_norm": 0.8619406819343567, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 4935 + }, + { + "epoch": 0.04936, + "grad_norm": 0.9272763133049011, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 4936 + }, + { + "epoch": 0.04937, + "grad_norm": 0.9949960112571716, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 4937 + }, + { + "epoch": 0.04938, + "grad_norm": 0.9583383202552795, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 4938 + }, + { + "epoch": 0.04939, + "grad_norm": 0.9466757774353027, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 4939 + }, + { + "epoch": 0.0494, + "grad_norm": 0.9211159944534302, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 4940 + }, + { + "epoch": 0.04941, + "grad_norm": 0.7683899402618408, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 4941 + }, + { + "epoch": 0.04942, + "grad_norm": 0.7320684194564819, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 4942 + }, + { + "epoch": 0.04943, + "grad_norm": 0.7409505844116211, + "learning_rate": 0.003, + "loss": 4.109, + "step": 4943 + }, + { + "epoch": 0.04944, + "grad_norm": 0.6583126187324524, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 4944 + }, + { + "epoch": 0.04945, + "grad_norm": 0.7039183974266052, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 4945 + }, + { + "epoch": 0.04946, + "grad_norm": 0.7162462472915649, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 4946 + }, + { + "epoch": 0.04947, + "grad_norm": 0.6111107468605042, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 4947 + }, + { + "epoch": 0.04948, + "grad_norm": 0.6708353161811829, + "learning_rate": 0.003, + "loss": 4.083, + "step": 4948 + }, + { + "epoch": 0.04949, + "grad_norm": 0.9382305145263672, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 4949 + }, + { + "epoch": 0.0495, + "grad_norm": 1.2174804210662842, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 4950 + }, + { + "epoch": 0.04951, + "grad_norm": 0.701120913028717, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 4951 + }, + { + "epoch": 0.04952, + "grad_norm": 0.6446220278739929, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 4952 + }, + { + "epoch": 0.04953, + "grad_norm": 0.7424911856651306, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 4953 + }, + { + "epoch": 0.04954, + "grad_norm": 0.6806883215904236, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 4954 + }, + { + "epoch": 0.04955, + "grad_norm": 0.699410617351532, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 4955 + }, + { + "epoch": 0.04956, + "grad_norm": 0.6064649224281311, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 4956 + }, + { + "epoch": 0.04957, + "grad_norm": 0.5583900809288025, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 4957 + }, + { + "epoch": 0.04958, + "grad_norm": 0.5232359170913696, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 4958 + }, + { + "epoch": 0.04959, + "grad_norm": 0.4963710308074951, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 4959 + }, + { + "epoch": 0.0496, + "grad_norm": 0.5652657151222229, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 4960 + }, + { + "epoch": 0.04961, + "grad_norm": 0.6090538501739502, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 4961 + }, + { + "epoch": 0.04962, + "grad_norm": 0.6194426417350769, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 4962 + }, + { + "epoch": 0.04963, + "grad_norm": 0.6753765344619751, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 4963 + }, + { + "epoch": 0.04964, + "grad_norm": 0.733012855052948, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 4964 + }, + { + "epoch": 0.04965, + "grad_norm": 0.7559370398521423, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 4965 + }, + { + "epoch": 0.04966, + "grad_norm": 0.75950026512146, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 4966 + }, + { + "epoch": 0.04967, + "grad_norm": 0.697681725025177, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 4967 + }, + { + "epoch": 0.04968, + "grad_norm": 0.7198742628097534, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 4968 + }, + { + "epoch": 0.04969, + "grad_norm": 0.6972935795783997, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 4969 + }, + { + "epoch": 0.0497, + "grad_norm": 0.6653467416763306, + "learning_rate": 0.003, + "loss": 4.074, + "step": 4970 + }, + { + "epoch": 0.04971, + "grad_norm": 0.6616604328155518, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 4971 + }, + { + "epoch": 0.04972, + "grad_norm": 0.6987863183021545, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 4972 + }, + { + "epoch": 0.04973, + "grad_norm": 0.7355426549911499, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 4973 + }, + { + "epoch": 0.04974, + "grad_norm": 0.8188806176185608, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 4974 + }, + { + "epoch": 0.04975, + "grad_norm": 0.8495981693267822, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 4975 + }, + { + "epoch": 0.04976, + "grad_norm": 0.9018936157226562, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 4976 + }, + { + "epoch": 0.04977, + "grad_norm": 0.8570343852043152, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 4977 + }, + { + "epoch": 0.04978, + "grad_norm": 0.8940666317939758, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 4978 + }, + { + "epoch": 0.04979, + "grad_norm": 0.8460613489151001, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 4979 + }, + { + "epoch": 0.0498, + "grad_norm": 0.6897357702255249, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 4980 + }, + { + "epoch": 0.04981, + "grad_norm": 0.6147297024726868, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 4981 + }, + { + "epoch": 0.04982, + "grad_norm": 0.6292622089385986, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 4982 + }, + { + "epoch": 0.04983, + "grad_norm": 0.6632224321365356, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 4983 + }, + { + "epoch": 0.04984, + "grad_norm": 0.6822007894515991, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 4984 + }, + { + "epoch": 0.04985, + "grad_norm": 0.6867620348930359, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 4985 + }, + { + "epoch": 0.04986, + "grad_norm": 0.7010688185691833, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 4986 + }, + { + "epoch": 0.04987, + "grad_norm": 0.7688935399055481, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 4987 + }, + { + "epoch": 0.04988, + "grad_norm": 0.7406259179115295, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 4988 + }, + { + "epoch": 0.04989, + "grad_norm": 0.7552890181541443, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 4989 + }, + { + "epoch": 0.0499, + "grad_norm": 0.9177321195602417, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 4990 + }, + { + "epoch": 0.04991, + "grad_norm": 1.0167580842971802, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 4991 + }, + { + "epoch": 0.04992, + "grad_norm": 1.1560254096984863, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 4992 + }, + { + "epoch": 0.04993, + "grad_norm": 0.8258231282234192, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 4993 + }, + { + "epoch": 0.04994, + "grad_norm": 0.790297269821167, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 4994 + }, + { + "epoch": 0.04995, + "grad_norm": 0.8769829273223877, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 4995 + }, + { + "epoch": 0.04996, + "grad_norm": 0.8291935324668884, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 4996 + }, + { + "epoch": 0.04997, + "grad_norm": 0.7305048704147339, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 4997 + }, + { + "epoch": 0.04998, + "grad_norm": 0.6434715390205383, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 4998 + }, + { + "epoch": 0.04999, + "grad_norm": 0.6438940167427063, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 4999 + }, + { + "epoch": 0.05, + "grad_norm": 0.6508415937423706, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 5000 + }, + { + "epoch": 0.05001, + "grad_norm": 0.7253802418708801, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 5001 + }, + { + "epoch": 0.05002, + "grad_norm": 0.7797998785972595, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 5002 + }, + { + "epoch": 0.05003, + "grad_norm": 0.8626973628997803, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 5003 + }, + { + "epoch": 0.05004, + "grad_norm": 0.8299462199211121, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 5004 + }, + { + "epoch": 0.05005, + "grad_norm": 0.8039361834526062, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 5005 + }, + { + "epoch": 0.05006, + "grad_norm": 0.7754454612731934, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 5006 + }, + { + "epoch": 0.05007, + "grad_norm": 0.6967061161994934, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 5007 + }, + { + "epoch": 0.05008, + "grad_norm": 0.5528774261474609, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 5008 + }, + { + "epoch": 0.05009, + "grad_norm": 0.5234524607658386, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 5009 + }, + { + "epoch": 0.0501, + "grad_norm": 0.46996673941612244, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 5010 + }, + { + "epoch": 0.05011, + "grad_norm": 0.5573607683181763, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 5011 + }, + { + "epoch": 0.05012, + "grad_norm": 0.5557873249053955, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 5012 + }, + { + "epoch": 0.05013, + "grad_norm": 0.5415859818458557, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 5013 + }, + { + "epoch": 0.05014, + "grad_norm": 0.5040212869644165, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 5014 + }, + { + "epoch": 0.05015, + "grad_norm": 0.5970138907432556, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 5015 + }, + { + "epoch": 0.05016, + "grad_norm": 0.641599714756012, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 5016 + }, + { + "epoch": 0.05017, + "grad_norm": 0.6173446178436279, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 5017 + }, + { + "epoch": 0.05018, + "grad_norm": 0.6326936483383179, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 5018 + }, + { + "epoch": 0.05019, + "grad_norm": 0.6489758491516113, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 5019 + }, + { + "epoch": 0.0502, + "grad_norm": 0.6537885665893555, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 5020 + }, + { + "epoch": 0.05021, + "grad_norm": 0.6637057662010193, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 5021 + }, + { + "epoch": 0.05022, + "grad_norm": 0.6988769769668579, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 5022 + }, + { + "epoch": 0.05023, + "grad_norm": 0.6984631419181824, + "learning_rate": 0.003, + "loss": 4.093, + "step": 5023 + }, + { + "epoch": 0.05024, + "grad_norm": 0.7143153548240662, + "learning_rate": 0.003, + "loss": 4.053, + "step": 5024 + }, + { + "epoch": 0.05025, + "grad_norm": 0.6525802612304688, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 5025 + }, + { + "epoch": 0.05026, + "grad_norm": 0.6056846976280212, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 5026 + }, + { + "epoch": 0.05027, + "grad_norm": 0.6280860304832458, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 5027 + }, + { + "epoch": 0.05028, + "grad_norm": 0.6172341704368591, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 5028 + }, + { + "epoch": 0.05029, + "grad_norm": 0.6776188015937805, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 5029 + }, + { + "epoch": 0.0503, + "grad_norm": 0.7172451615333557, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 5030 + }, + { + "epoch": 0.05031, + "grad_norm": 0.7082464694976807, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 5031 + }, + { + "epoch": 0.05032, + "grad_norm": 0.7410402297973633, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 5032 + }, + { + "epoch": 0.05033, + "grad_norm": 0.7269566059112549, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 5033 + }, + { + "epoch": 0.05034, + "grad_norm": 0.8119834065437317, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 5034 + }, + { + "epoch": 0.05035, + "grad_norm": 0.9735667109489441, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 5035 + }, + { + "epoch": 0.05036, + "grad_norm": 0.886156439781189, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 5036 + }, + { + "epoch": 0.05037, + "grad_norm": 0.9368439316749573, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 5037 + }, + { + "epoch": 0.05038, + "grad_norm": 0.9387019872665405, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 5038 + }, + { + "epoch": 0.05039, + "grad_norm": 1.077192783355713, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 5039 + }, + { + "epoch": 0.0504, + "grad_norm": 0.8130627274513245, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 5040 + }, + { + "epoch": 0.05041, + "grad_norm": 0.8206466436386108, + "learning_rate": 0.003, + "loss": 4.097, + "step": 5041 + }, + { + "epoch": 0.05042, + "grad_norm": 0.9645205140113831, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 5042 + }, + { + "epoch": 0.05043, + "grad_norm": 1.1380254030227661, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 5043 + }, + { + "epoch": 0.05044, + "grad_norm": 1.002665638923645, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 5044 + }, + { + "epoch": 0.05045, + "grad_norm": 0.9848024845123291, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 5045 + }, + { + "epoch": 0.05046, + "grad_norm": 0.8786163330078125, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 5046 + }, + { + "epoch": 0.05047, + "grad_norm": 0.8327427506446838, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 5047 + }, + { + "epoch": 0.05048, + "grad_norm": 0.7884167432785034, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 5048 + }, + { + "epoch": 0.05049, + "grad_norm": 0.8283893465995789, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 5049 + }, + { + "epoch": 0.0505, + "grad_norm": 0.7522302865982056, + "learning_rate": 0.003, + "loss": 4.1483, + "step": 5050 + }, + { + "epoch": 0.05051, + "grad_norm": 0.7367852926254272, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 5051 + }, + { + "epoch": 0.05052, + "grad_norm": 0.6346535682678223, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 5052 + }, + { + "epoch": 0.05053, + "grad_norm": 0.6007921099662781, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 5053 + }, + { + "epoch": 0.05054, + "grad_norm": 0.5848135948181152, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 5054 + }, + { + "epoch": 0.05055, + "grad_norm": 0.6019494533538818, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 5055 + }, + { + "epoch": 0.05056, + "grad_norm": 0.6064556241035461, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 5056 + }, + { + "epoch": 0.05057, + "grad_norm": 0.6536866426467896, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 5057 + }, + { + "epoch": 0.05058, + "grad_norm": 0.6580118536949158, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 5058 + }, + { + "epoch": 0.05059, + "grad_norm": 0.5559451580047607, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 5059 + }, + { + "epoch": 0.0506, + "grad_norm": 0.512411892414093, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 5060 + }, + { + "epoch": 0.05061, + "grad_norm": 0.5732488632202148, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5061 + }, + { + "epoch": 0.05062, + "grad_norm": 0.6876088380813599, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 5062 + }, + { + "epoch": 0.05063, + "grad_norm": 0.8171137571334839, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 5063 + }, + { + "epoch": 0.05064, + "grad_norm": 0.9967246055603027, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 5064 + }, + { + "epoch": 0.05065, + "grad_norm": 0.9538764953613281, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 5065 + }, + { + "epoch": 0.05066, + "grad_norm": 0.7090097069740295, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 5066 + }, + { + "epoch": 0.05067, + "grad_norm": 0.62126225233078, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 5067 + }, + { + "epoch": 0.05068, + "grad_norm": 0.6797899603843689, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 5068 + }, + { + "epoch": 0.05069, + "grad_norm": 0.7215432524681091, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 5069 + }, + { + "epoch": 0.0507, + "grad_norm": 0.7022911906242371, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 5070 + }, + { + "epoch": 0.05071, + "grad_norm": 0.6767652034759521, + "learning_rate": 0.003, + "loss": 4.083, + "step": 5071 + }, + { + "epoch": 0.05072, + "grad_norm": 0.6579228043556213, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 5072 + }, + { + "epoch": 0.05073, + "grad_norm": 0.6343953609466553, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 5073 + }, + { + "epoch": 0.05074, + "grad_norm": 0.6530939340591431, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 5074 + }, + { + "epoch": 0.05075, + "grad_norm": 0.5952531099319458, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 5075 + }, + { + "epoch": 0.05076, + "grad_norm": 0.5853413939476013, + "learning_rate": 0.003, + "loss": 4.089, + "step": 5076 + }, + { + "epoch": 0.05077, + "grad_norm": 0.6744263768196106, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 5077 + }, + { + "epoch": 0.05078, + "grad_norm": 0.7323231101036072, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 5078 + }, + { + "epoch": 0.05079, + "grad_norm": 0.7203153967857361, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 5079 + }, + { + "epoch": 0.0508, + "grad_norm": 0.7360239624977112, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 5080 + }, + { + "epoch": 0.05081, + "grad_norm": 0.7595191597938538, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 5081 + }, + { + "epoch": 0.05082, + "grad_norm": 0.7678768634796143, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 5082 + }, + { + "epoch": 0.05083, + "grad_norm": 0.7995537519454956, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 5083 + }, + { + "epoch": 0.05084, + "grad_norm": 0.8082243800163269, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 5084 + }, + { + "epoch": 0.05085, + "grad_norm": 0.8145765662193298, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 5085 + }, + { + "epoch": 0.05086, + "grad_norm": 0.8743909001350403, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 5086 + }, + { + "epoch": 0.05087, + "grad_norm": 0.986849844455719, + "learning_rate": 0.003, + "loss": 4.085, + "step": 5087 + }, + { + "epoch": 0.05088, + "grad_norm": 0.9816634058952332, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 5088 + }, + { + "epoch": 0.05089, + "grad_norm": 0.962716817855835, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 5089 + }, + { + "epoch": 0.0509, + "grad_norm": 0.8741410970687866, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 5090 + }, + { + "epoch": 0.05091, + "grad_norm": 0.7220102548599243, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 5091 + }, + { + "epoch": 0.05092, + "grad_norm": 0.718839704990387, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 5092 + }, + { + "epoch": 0.05093, + "grad_norm": 0.8266152143478394, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 5093 + }, + { + "epoch": 0.05094, + "grad_norm": 0.8446786999702454, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 5094 + }, + { + "epoch": 0.05095, + "grad_norm": 0.7700090408325195, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 5095 + }, + { + "epoch": 0.05096, + "grad_norm": 0.715460479259491, + "learning_rate": 0.003, + "loss": 4.078, + "step": 5096 + }, + { + "epoch": 0.05097, + "grad_norm": 0.671999454498291, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 5097 + }, + { + "epoch": 0.05098, + "grad_norm": 0.6299563050270081, + "learning_rate": 0.003, + "loss": 4.063, + "step": 5098 + }, + { + "epoch": 0.05099, + "grad_norm": 0.5293504595756531, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 5099 + }, + { + "epoch": 0.051, + "grad_norm": 0.5206742882728577, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 5100 + }, + { + "epoch": 0.05101, + "grad_norm": 0.5910031199455261, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 5101 + }, + { + "epoch": 0.05102, + "grad_norm": 0.655666172504425, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 5102 + }, + { + "epoch": 0.05103, + "grad_norm": 0.7769567966461182, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 5103 + }, + { + "epoch": 0.05104, + "grad_norm": 0.8552154898643494, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 5104 + }, + { + "epoch": 0.05105, + "grad_norm": 0.8374227285385132, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 5105 + }, + { + "epoch": 0.05106, + "grad_norm": 0.6673815250396729, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 5106 + }, + { + "epoch": 0.05107, + "grad_norm": 0.5111753940582275, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 5107 + }, + { + "epoch": 0.05108, + "grad_norm": 0.5464461445808411, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 5108 + }, + { + "epoch": 0.05109, + "grad_norm": 0.5811792016029358, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 5109 + }, + { + "epoch": 0.0511, + "grad_norm": 0.5256453156471252, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 5110 + }, + { + "epoch": 0.05111, + "grad_norm": 0.5240735411643982, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 5111 + }, + { + "epoch": 0.05112, + "grad_norm": 0.5387896299362183, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 5112 + }, + { + "epoch": 0.05113, + "grad_norm": 0.6552562117576599, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 5113 + }, + { + "epoch": 0.05114, + "grad_norm": 0.6873152852058411, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 5114 + }, + { + "epoch": 0.05115, + "grad_norm": 0.6076439619064331, + "learning_rate": 0.003, + "loss": 4.075, + "step": 5115 + }, + { + "epoch": 0.05116, + "grad_norm": 0.536746621131897, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 5116 + }, + { + "epoch": 0.05117, + "grad_norm": 0.5262901186943054, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 5117 + }, + { + "epoch": 0.05118, + "grad_norm": 0.5599290132522583, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 5118 + }, + { + "epoch": 0.05119, + "grad_norm": 0.5657877922058105, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 5119 + }, + { + "epoch": 0.0512, + "grad_norm": 0.5846731066703796, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 5120 + }, + { + "epoch": 0.05121, + "grad_norm": 0.6840034127235413, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 5121 + }, + { + "epoch": 0.05122, + "grad_norm": 0.9214706420898438, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 5122 + }, + { + "epoch": 0.05123, + "grad_norm": 0.996700644493103, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 5123 + }, + { + "epoch": 0.05124, + "grad_norm": 0.9406014680862427, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 5124 + }, + { + "epoch": 0.05125, + "grad_norm": 0.7598673105239868, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 5125 + }, + { + "epoch": 0.05126, + "grad_norm": 0.6382818818092346, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 5126 + }, + { + "epoch": 0.05127, + "grad_norm": 0.7045958042144775, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 5127 + }, + { + "epoch": 0.05128, + "grad_norm": 0.7263686656951904, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 5128 + }, + { + "epoch": 0.05129, + "grad_norm": 0.7894842624664307, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 5129 + }, + { + "epoch": 0.0513, + "grad_norm": 0.7886913418769836, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 5130 + }, + { + "epoch": 0.05131, + "grad_norm": 0.791679322719574, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 5131 + }, + { + "epoch": 0.05132, + "grad_norm": 0.7831953763961792, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 5132 + }, + { + "epoch": 0.05133, + "grad_norm": 0.9040250778198242, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 5133 + }, + { + "epoch": 0.05134, + "grad_norm": 0.9008460640907288, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 5134 + }, + { + "epoch": 0.05135, + "grad_norm": 0.8832536339759827, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 5135 + }, + { + "epoch": 0.05136, + "grad_norm": 0.886307418346405, + "learning_rate": 0.003, + "loss": 4.052, + "step": 5136 + }, + { + "epoch": 0.05137, + "grad_norm": 0.9245304465293884, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 5137 + }, + { + "epoch": 0.05138, + "grad_norm": 1.0758816003799438, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 5138 + }, + { + "epoch": 0.05139, + "grad_norm": 0.9390636086463928, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 5139 + }, + { + "epoch": 0.0514, + "grad_norm": 0.8630328178405762, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 5140 + }, + { + "epoch": 0.05141, + "grad_norm": 0.6845744252204895, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 5141 + }, + { + "epoch": 0.05142, + "grad_norm": 0.7005607485771179, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 5142 + }, + { + "epoch": 0.05143, + "grad_norm": 0.6236844062805176, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 5143 + }, + { + "epoch": 0.05144, + "grad_norm": 0.5919453501701355, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 5144 + }, + { + "epoch": 0.05145, + "grad_norm": 0.6221011281013489, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 5145 + }, + { + "epoch": 0.05146, + "grad_norm": 0.6235561370849609, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 5146 + }, + { + "epoch": 0.05147, + "grad_norm": 0.6597413420677185, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 5147 + }, + { + "epoch": 0.05148, + "grad_norm": 0.7255842089653015, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 5148 + }, + { + "epoch": 0.05149, + "grad_norm": 0.7855421900749207, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 5149 + }, + { + "epoch": 0.0515, + "grad_norm": 0.6787280440330505, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 5150 + }, + { + "epoch": 0.05151, + "grad_norm": 0.5714637041091919, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 5151 + }, + { + "epoch": 0.05152, + "grad_norm": 0.6426389813423157, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 5152 + }, + { + "epoch": 0.05153, + "grad_norm": 0.6792200207710266, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 5153 + }, + { + "epoch": 0.05154, + "grad_norm": 0.6108073592185974, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 5154 + }, + { + "epoch": 0.05155, + "grad_norm": 0.6600965261459351, + "learning_rate": 0.003, + "loss": 4.08, + "step": 5155 + }, + { + "epoch": 0.05156, + "grad_norm": 0.6948540210723877, + "learning_rate": 0.003, + "loss": 4.104, + "step": 5156 + }, + { + "epoch": 0.05157, + "grad_norm": 0.7809789776802063, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 5157 + }, + { + "epoch": 0.05158, + "grad_norm": 0.80213862657547, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 5158 + }, + { + "epoch": 0.05159, + "grad_norm": 0.7642487287521362, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 5159 + }, + { + "epoch": 0.0516, + "grad_norm": 0.7955514192581177, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 5160 + }, + { + "epoch": 0.05161, + "grad_norm": 0.5677341222763062, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 5161 + }, + { + "epoch": 0.05162, + "grad_norm": 0.5948107242584229, + "learning_rate": 0.003, + "loss": 4.063, + "step": 5162 + }, + { + "epoch": 0.05163, + "grad_norm": 0.7498951554298401, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 5163 + }, + { + "epoch": 0.05164, + "grad_norm": 0.98694908618927, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 5164 + }, + { + "epoch": 0.05165, + "grad_norm": 1.0628858804702759, + "learning_rate": 0.003, + "loss": 4.093, + "step": 5165 + }, + { + "epoch": 0.05166, + "grad_norm": 0.817200243473053, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 5166 + }, + { + "epoch": 0.05167, + "grad_norm": 0.6983803510665894, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 5167 + }, + { + "epoch": 0.05168, + "grad_norm": 0.6873238682746887, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 5168 + }, + { + "epoch": 0.05169, + "grad_norm": 0.6252221465110779, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 5169 + }, + { + "epoch": 0.0517, + "grad_norm": 0.6794670224189758, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 5170 + }, + { + "epoch": 0.05171, + "grad_norm": 0.6629231572151184, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 5171 + }, + { + "epoch": 0.05172, + "grad_norm": 0.7229346036911011, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 5172 + }, + { + "epoch": 0.05173, + "grad_norm": 0.8697673678398132, + "learning_rate": 0.003, + "loss": 4.098, + "step": 5173 + }, + { + "epoch": 0.05174, + "grad_norm": 1.057416319847107, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 5174 + }, + { + "epoch": 0.05175, + "grad_norm": 0.8897588849067688, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 5175 + }, + { + "epoch": 0.05176, + "grad_norm": 0.6620845794677734, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 5176 + }, + { + "epoch": 0.05177, + "grad_norm": 0.659085750579834, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 5177 + }, + { + "epoch": 0.05178, + "grad_norm": 0.7169133424758911, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 5178 + }, + { + "epoch": 0.05179, + "grad_norm": 0.6649430394172668, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 5179 + }, + { + "epoch": 0.0518, + "grad_norm": 0.5839197635650635, + "learning_rate": 0.003, + "loss": 4.067, + "step": 5180 + }, + { + "epoch": 0.05181, + "grad_norm": 0.44185495376586914, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 5181 + }, + { + "epoch": 0.05182, + "grad_norm": 0.4659267067909241, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 5182 + }, + { + "epoch": 0.05183, + "grad_norm": 0.5202658772468567, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 5183 + }, + { + "epoch": 0.05184, + "grad_norm": 0.5452793836593628, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 5184 + }, + { + "epoch": 0.05185, + "grad_norm": 0.6658215522766113, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 5185 + }, + { + "epoch": 0.05186, + "grad_norm": 0.7893977165222168, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 5186 + }, + { + "epoch": 0.05187, + "grad_norm": 0.8065964579582214, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 5187 + }, + { + "epoch": 0.05188, + "grad_norm": 0.7456889748573303, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 5188 + }, + { + "epoch": 0.05189, + "grad_norm": 0.7495638132095337, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 5189 + }, + { + "epoch": 0.0519, + "grad_norm": 0.8337745666503906, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5190 + }, + { + "epoch": 0.05191, + "grad_norm": 0.8566077351570129, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 5191 + }, + { + "epoch": 0.05192, + "grad_norm": 0.805960476398468, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 5192 + }, + { + "epoch": 0.05193, + "grad_norm": 0.789035439491272, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 5193 + }, + { + "epoch": 0.05194, + "grad_norm": 0.7752846479415894, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 5194 + }, + { + "epoch": 0.05195, + "grad_norm": 0.711561381816864, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 5195 + }, + { + "epoch": 0.05196, + "grad_norm": 0.6495727896690369, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 5196 + }, + { + "epoch": 0.05197, + "grad_norm": 0.6552587747573853, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 5197 + }, + { + "epoch": 0.05198, + "grad_norm": 0.5346404314041138, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 5198 + }, + { + "epoch": 0.05199, + "grad_norm": 0.5881133079528809, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 5199 + }, + { + "epoch": 0.052, + "grad_norm": 0.5984171032905579, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 5200 + }, + { + "epoch": 0.05201, + "grad_norm": 0.5811743140220642, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 5201 + }, + { + "epoch": 0.05202, + "grad_norm": 0.5741499662399292, + "learning_rate": 0.003, + "loss": 4.058, + "step": 5202 + }, + { + "epoch": 0.05203, + "grad_norm": 0.5631682276725769, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 5203 + }, + { + "epoch": 0.05204, + "grad_norm": 0.6046068668365479, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 5204 + }, + { + "epoch": 0.05205, + "grad_norm": 0.6720539331436157, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 5205 + }, + { + "epoch": 0.05206, + "grad_norm": 0.8187111020088196, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 5206 + }, + { + "epoch": 0.05207, + "grad_norm": 0.9796744585037231, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 5207 + }, + { + "epoch": 0.05208, + "grad_norm": 1.0159255266189575, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 5208 + }, + { + "epoch": 0.05209, + "grad_norm": 0.9315694570541382, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 5209 + }, + { + "epoch": 0.0521, + "grad_norm": 0.932378351688385, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 5210 + }, + { + "epoch": 0.05211, + "grad_norm": 0.8560777902603149, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 5211 + }, + { + "epoch": 0.05212, + "grad_norm": 0.7157209515571594, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 5212 + }, + { + "epoch": 0.05213, + "grad_norm": 0.7000187635421753, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 5213 + }, + { + "epoch": 0.05214, + "grad_norm": 0.7313051223754883, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 5214 + }, + { + "epoch": 0.05215, + "grad_norm": 0.7454226016998291, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 5215 + }, + { + "epoch": 0.05216, + "grad_norm": 0.7446316480636597, + "learning_rate": 0.003, + "loss": 4.074, + "step": 5216 + }, + { + "epoch": 0.05217, + "grad_norm": 0.892437219619751, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 5217 + }, + { + "epoch": 0.05218, + "grad_norm": 1.0080622434616089, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 5218 + }, + { + "epoch": 0.05219, + "grad_norm": 0.969513475894928, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 5219 + }, + { + "epoch": 0.0522, + "grad_norm": 0.8957400918006897, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 5220 + }, + { + "epoch": 0.05221, + "grad_norm": 0.7468663454055786, + "learning_rate": 0.003, + "loss": 4.077, + "step": 5221 + }, + { + "epoch": 0.05222, + "grad_norm": 0.6890023946762085, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 5222 + }, + { + "epoch": 0.05223, + "grad_norm": 0.7769709825515747, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 5223 + }, + { + "epoch": 0.05224, + "grad_norm": 0.8365834355354309, + "learning_rate": 0.003, + "loss": 4.079, + "step": 5224 + }, + { + "epoch": 0.05225, + "grad_norm": 0.8886056542396545, + "learning_rate": 0.003, + "loss": 4.106, + "step": 5225 + }, + { + "epoch": 0.05226, + "grad_norm": 1.014122486114502, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 5226 + }, + { + "epoch": 0.05227, + "grad_norm": 0.9814178943634033, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 5227 + }, + { + "epoch": 0.05228, + "grad_norm": 0.8787015080451965, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 5228 + }, + { + "epoch": 0.05229, + "grad_norm": 0.7741072177886963, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 5229 + }, + { + "epoch": 0.0523, + "grad_norm": 0.6090134978294373, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 5230 + }, + { + "epoch": 0.05231, + "grad_norm": 0.588998556137085, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 5231 + }, + { + "epoch": 0.05232, + "grad_norm": 0.5598298907279968, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 5232 + }, + { + "epoch": 0.05233, + "grad_norm": 0.5243165493011475, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 5233 + }, + { + "epoch": 0.05234, + "grad_norm": 0.41240203380584717, + "learning_rate": 0.003, + "loss": 4.099, + "step": 5234 + }, + { + "epoch": 0.05235, + "grad_norm": 0.4857303500175476, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 5235 + }, + { + "epoch": 0.05236, + "grad_norm": 0.5221879482269287, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 5236 + }, + { + "epoch": 0.05237, + "grad_norm": 0.5270381569862366, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 5237 + }, + { + "epoch": 0.05238, + "grad_norm": 0.6028867959976196, + "learning_rate": 0.003, + "loss": 4.053, + "step": 5238 + }, + { + "epoch": 0.05239, + "grad_norm": 0.6225501298904419, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 5239 + }, + { + "epoch": 0.0524, + "grad_norm": 0.6203356385231018, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 5240 + }, + { + "epoch": 0.05241, + "grad_norm": 0.6311340928077698, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 5241 + }, + { + "epoch": 0.05242, + "grad_norm": 0.6222201585769653, + "learning_rate": 0.003, + "loss": 4.084, + "step": 5242 + }, + { + "epoch": 0.05243, + "grad_norm": 0.6350955963134766, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 5243 + }, + { + "epoch": 0.05244, + "grad_norm": 0.6730124354362488, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 5244 + }, + { + "epoch": 0.05245, + "grad_norm": 0.8762339353561401, + "learning_rate": 0.003, + "loss": 4.066, + "step": 5245 + }, + { + "epoch": 0.05246, + "grad_norm": 1.018054485321045, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 5246 + }, + { + "epoch": 0.05247, + "grad_norm": 1.032390832901001, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 5247 + }, + { + "epoch": 0.05248, + "grad_norm": 0.6803662180900574, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 5248 + }, + { + "epoch": 0.05249, + "grad_norm": 0.7293293476104736, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 5249 + }, + { + "epoch": 0.0525, + "grad_norm": 1.0431487560272217, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 5250 + }, + { + "epoch": 0.05251, + "grad_norm": 0.9717729687690735, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 5251 + }, + { + "epoch": 0.05252, + "grad_norm": 0.8074532747268677, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 5252 + }, + { + "epoch": 0.05253, + "grad_norm": 0.8047144412994385, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 5253 + }, + { + "epoch": 0.05254, + "grad_norm": 0.7531102895736694, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 5254 + }, + { + "epoch": 0.05255, + "grad_norm": 0.6377032995223999, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 5255 + }, + { + "epoch": 0.05256, + "grad_norm": 0.5738644003868103, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 5256 + }, + { + "epoch": 0.05257, + "grad_norm": 0.5616352558135986, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 5257 + }, + { + "epoch": 0.05258, + "grad_norm": 0.5875840783119202, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 5258 + }, + { + "epoch": 0.05259, + "grad_norm": 0.5328964591026306, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 5259 + }, + { + "epoch": 0.0526, + "grad_norm": 0.5470474362373352, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 5260 + }, + { + "epoch": 0.05261, + "grad_norm": 0.5530418753623962, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 5261 + }, + { + "epoch": 0.05262, + "grad_norm": 0.543959379196167, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 5262 + }, + { + "epoch": 0.05263, + "grad_norm": 0.5292184352874756, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 5263 + }, + { + "epoch": 0.05264, + "grad_norm": 0.5867078900337219, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 5264 + }, + { + "epoch": 0.05265, + "grad_norm": 0.5979050397872925, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 5265 + }, + { + "epoch": 0.05266, + "grad_norm": 0.5827270746231079, + "learning_rate": 0.003, + "loss": 4.037, + "step": 5266 + }, + { + "epoch": 0.05267, + "grad_norm": 0.7350538969039917, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 5267 + }, + { + "epoch": 0.05268, + "grad_norm": 0.8645079731941223, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 5268 + }, + { + "epoch": 0.05269, + "grad_norm": 0.8950510025024414, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 5269 + }, + { + "epoch": 0.0527, + "grad_norm": 0.8910822868347168, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 5270 + }, + { + "epoch": 0.05271, + "grad_norm": 0.7200420498847961, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 5271 + }, + { + "epoch": 0.05272, + "grad_norm": 0.6258223652839661, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 5272 + }, + { + "epoch": 0.05273, + "grad_norm": 0.7092558741569519, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 5273 + }, + { + "epoch": 0.05274, + "grad_norm": 0.6717981696128845, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 5274 + }, + { + "epoch": 0.05275, + "grad_norm": 0.7079071402549744, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 5275 + }, + { + "epoch": 0.05276, + "grad_norm": 0.6641931533813477, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 5276 + }, + { + "epoch": 0.05277, + "grad_norm": 0.6190426349639893, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 5277 + }, + { + "epoch": 0.05278, + "grad_norm": 0.5594713687896729, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 5278 + }, + { + "epoch": 0.05279, + "grad_norm": 0.6031233668327332, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 5279 + }, + { + "epoch": 0.0528, + "grad_norm": 0.678126871585846, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 5280 + }, + { + "epoch": 0.05281, + "grad_norm": 0.6510528326034546, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 5281 + }, + { + "epoch": 0.05282, + "grad_norm": 0.6001152992248535, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 5282 + }, + { + "epoch": 0.05283, + "grad_norm": 0.5790733098983765, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 5283 + }, + { + "epoch": 0.05284, + "grad_norm": 0.6320419907569885, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 5284 + }, + { + "epoch": 0.05285, + "grad_norm": 0.7830936312675476, + "learning_rate": 0.003, + "loss": 4.034, + "step": 5285 + }, + { + "epoch": 0.05286, + "grad_norm": 0.9895585775375366, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 5286 + }, + { + "epoch": 0.05287, + "grad_norm": 1.0884960889816284, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 5287 + }, + { + "epoch": 0.05288, + "grad_norm": 0.9570357203483582, + "learning_rate": 0.003, + "loss": 4.063, + "step": 5288 + }, + { + "epoch": 0.05289, + "grad_norm": 0.9439939260482788, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 5289 + }, + { + "epoch": 0.0529, + "grad_norm": 0.8485848903656006, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 5290 + }, + { + "epoch": 0.05291, + "grad_norm": 0.9338999390602112, + "learning_rate": 0.003, + "loss": 4.104, + "step": 5291 + }, + { + "epoch": 0.05292, + "grad_norm": 0.8072062730789185, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 5292 + }, + { + "epoch": 0.05293, + "grad_norm": 0.8015543818473816, + "learning_rate": 0.003, + "loss": 4.095, + "step": 5293 + }, + { + "epoch": 0.05294, + "grad_norm": 0.819395124912262, + "learning_rate": 0.003, + "loss": 4.081, + "step": 5294 + }, + { + "epoch": 0.05295, + "grad_norm": 0.92290198802948, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 5295 + }, + { + "epoch": 0.05296, + "grad_norm": 0.8833643794059753, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 5296 + }, + { + "epoch": 0.05297, + "grad_norm": 0.7732718586921692, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 5297 + }, + { + "epoch": 0.05298, + "grad_norm": 0.7301531434059143, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 5298 + }, + { + "epoch": 0.05299, + "grad_norm": 0.6377935409545898, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 5299 + }, + { + "epoch": 0.053, + "grad_norm": 0.653924286365509, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 5300 + }, + { + "epoch": 0.05301, + "grad_norm": 0.6310106515884399, + "learning_rate": 0.003, + "loss": 4.095, + "step": 5301 + }, + { + "epoch": 0.05302, + "grad_norm": 0.5855343341827393, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 5302 + }, + { + "epoch": 0.05303, + "grad_norm": 0.6311571598052979, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 5303 + }, + { + "epoch": 0.05304, + "grad_norm": 0.6711182594299316, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 5304 + }, + { + "epoch": 0.05305, + "grad_norm": 0.8134438991546631, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 5305 + }, + { + "epoch": 0.05306, + "grad_norm": 0.8880993723869324, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 5306 + }, + { + "epoch": 0.05307, + "grad_norm": 0.8936927318572998, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 5307 + }, + { + "epoch": 0.05308, + "grad_norm": 0.8056278824806213, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 5308 + }, + { + "epoch": 0.05309, + "grad_norm": 0.6631887555122375, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 5309 + }, + { + "epoch": 0.0531, + "grad_norm": 0.6476406455039978, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 5310 + }, + { + "epoch": 0.05311, + "grad_norm": 0.7253839373588562, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 5311 + }, + { + "epoch": 0.05312, + "grad_norm": 0.7280350923538208, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 5312 + }, + { + "epoch": 0.05313, + "grad_norm": 0.7820828557014465, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 5313 + }, + { + "epoch": 0.05314, + "grad_norm": 0.782772958278656, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 5314 + }, + { + "epoch": 0.05315, + "grad_norm": 0.7675919532775879, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 5315 + }, + { + "epoch": 0.05316, + "grad_norm": 0.7182729244232178, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 5316 + }, + { + "epoch": 0.05317, + "grad_norm": 0.6802192330360413, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 5317 + }, + { + "epoch": 0.05318, + "grad_norm": 0.6332565546035767, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 5318 + }, + { + "epoch": 0.05319, + "grad_norm": 0.5553032159805298, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 5319 + }, + { + "epoch": 0.0532, + "grad_norm": 0.5743739008903503, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 5320 + }, + { + "epoch": 0.05321, + "grad_norm": 0.5735093951225281, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 5321 + }, + { + "epoch": 0.05322, + "grad_norm": 0.5387066006660461, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 5322 + }, + { + "epoch": 0.05323, + "grad_norm": 0.5573674440383911, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 5323 + }, + { + "epoch": 0.05324, + "grad_norm": 0.5417758226394653, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 5324 + }, + { + "epoch": 0.05325, + "grad_norm": 0.5368819832801819, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 5325 + }, + { + "epoch": 0.05326, + "grad_norm": 0.6424285173416138, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 5326 + }, + { + "epoch": 0.05327, + "grad_norm": 0.6725877523422241, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 5327 + }, + { + "epoch": 0.05328, + "grad_norm": 0.8542432188987732, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 5328 + }, + { + "epoch": 0.05329, + "grad_norm": 1.01321542263031, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 5329 + }, + { + "epoch": 0.0533, + "grad_norm": 0.9700354933738708, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 5330 + }, + { + "epoch": 0.05331, + "grad_norm": 0.8577306270599365, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 5331 + }, + { + "epoch": 0.05332, + "grad_norm": 0.9007319808006287, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 5332 + }, + { + "epoch": 0.05333, + "grad_norm": 0.7996346950531006, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 5333 + }, + { + "epoch": 0.05334, + "grad_norm": 0.7633571624755859, + "learning_rate": 0.003, + "loss": 4.061, + "step": 5334 + }, + { + "epoch": 0.05335, + "grad_norm": 0.7068867683410645, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 5335 + }, + { + "epoch": 0.05336, + "grad_norm": 0.6439290046691895, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 5336 + }, + { + "epoch": 0.05337, + "grad_norm": 0.5971997380256653, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 5337 + }, + { + "epoch": 0.05338, + "grad_norm": 0.6028808355331421, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 5338 + }, + { + "epoch": 0.05339, + "grad_norm": 0.5407232046127319, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 5339 + }, + { + "epoch": 0.0534, + "grad_norm": 0.5392054915428162, + "learning_rate": 0.003, + "loss": 4.073, + "step": 5340 + }, + { + "epoch": 0.05341, + "grad_norm": 0.5244172215461731, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 5341 + }, + { + "epoch": 0.05342, + "grad_norm": 0.5212708115577698, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 5342 + }, + { + "epoch": 0.05343, + "grad_norm": 0.531294047832489, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 5343 + }, + { + "epoch": 0.05344, + "grad_norm": 0.4932909309864044, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 5344 + }, + { + "epoch": 0.05345, + "grad_norm": 0.5027388334274292, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 5345 + }, + { + "epoch": 0.05346, + "grad_norm": 0.6342618465423584, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 5346 + }, + { + "epoch": 0.05347, + "grad_norm": 0.8551032543182373, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 5347 + }, + { + "epoch": 0.05348, + "grad_norm": 1.2991005182266235, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 5348 + }, + { + "epoch": 0.05349, + "grad_norm": 0.8248401284217834, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5349 + }, + { + "epoch": 0.0535, + "grad_norm": 0.6131325960159302, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 5350 + }, + { + "epoch": 0.05351, + "grad_norm": 0.7400457859039307, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 5351 + }, + { + "epoch": 0.05352, + "grad_norm": 0.8651459813117981, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 5352 + }, + { + "epoch": 0.05353, + "grad_norm": 0.7526074051856995, + "learning_rate": 0.003, + "loss": 4.07, + "step": 5353 + }, + { + "epoch": 0.05354, + "grad_norm": 0.6973598599433899, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 5354 + }, + { + "epoch": 0.05355, + "grad_norm": 0.7028447985649109, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 5355 + }, + { + "epoch": 0.05356, + "grad_norm": 0.7545419335365295, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 5356 + }, + { + "epoch": 0.05357, + "grad_norm": 0.7842448949813843, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 5357 + }, + { + "epoch": 0.05358, + "grad_norm": 0.8168363571166992, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 5358 + }, + { + "epoch": 0.05359, + "grad_norm": 0.8426969647407532, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 5359 + }, + { + "epoch": 0.0536, + "grad_norm": 0.7467427849769592, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5360 + }, + { + "epoch": 0.05361, + "grad_norm": 0.7218089699745178, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 5361 + }, + { + "epoch": 0.05362, + "grad_norm": 0.8307386636734009, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 5362 + }, + { + "epoch": 0.05363, + "grad_norm": 0.8872959017753601, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 5363 + }, + { + "epoch": 0.05364, + "grad_norm": 0.7643839716911316, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 5364 + }, + { + "epoch": 0.05365, + "grad_norm": 0.7082914113998413, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 5365 + }, + { + "epoch": 0.05366, + "grad_norm": 0.8015276789665222, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 5366 + }, + { + "epoch": 0.05367, + "grad_norm": 0.8754156827926636, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 5367 + }, + { + "epoch": 0.05368, + "grad_norm": 1.0555367469787598, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 5368 + }, + { + "epoch": 0.05369, + "grad_norm": 1.0798786878585815, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 5369 + }, + { + "epoch": 0.0537, + "grad_norm": 0.8246363997459412, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 5370 + }, + { + "epoch": 0.05371, + "grad_norm": 0.5798512697219849, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 5371 + }, + { + "epoch": 0.05372, + "grad_norm": 0.6000420451164246, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 5372 + }, + { + "epoch": 0.05373, + "grad_norm": 0.6344287395477295, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 5373 + }, + { + "epoch": 0.05374, + "grad_norm": 0.7042878270149231, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 5374 + }, + { + "epoch": 0.05375, + "grad_norm": 0.728354275226593, + "learning_rate": 0.003, + "loss": 4.063, + "step": 5375 + }, + { + "epoch": 0.05376, + "grad_norm": 0.6129562854766846, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 5376 + }, + { + "epoch": 0.05377, + "grad_norm": 0.4737764000892639, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 5377 + }, + { + "epoch": 0.05378, + "grad_norm": 0.593817949295044, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 5378 + }, + { + "epoch": 0.05379, + "grad_norm": 0.6451365351676941, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 5379 + }, + { + "epoch": 0.0538, + "grad_norm": 0.5840117335319519, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 5380 + }, + { + "epoch": 0.05381, + "grad_norm": 0.5710886716842651, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 5381 + }, + { + "epoch": 0.05382, + "grad_norm": 0.5268121957778931, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 5382 + }, + { + "epoch": 0.05383, + "grad_norm": 0.4619216322898865, + "learning_rate": 0.003, + "loss": 4.074, + "step": 5383 + }, + { + "epoch": 0.05384, + "grad_norm": 0.5457313060760498, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 5384 + }, + { + "epoch": 0.05385, + "grad_norm": 0.5551273226737976, + "learning_rate": 0.003, + "loss": 4.082, + "step": 5385 + }, + { + "epoch": 0.05386, + "grad_norm": 0.6332549452781677, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 5386 + }, + { + "epoch": 0.05387, + "grad_norm": 0.7257327437400818, + "learning_rate": 0.003, + "loss": 4.053, + "step": 5387 + }, + { + "epoch": 0.05388, + "grad_norm": 0.7519972920417786, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 5388 + }, + { + "epoch": 0.05389, + "grad_norm": 0.7855745553970337, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 5389 + }, + { + "epoch": 0.0539, + "grad_norm": 0.8400787115097046, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 5390 + }, + { + "epoch": 0.05391, + "grad_norm": 0.8107534050941467, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 5391 + }, + { + "epoch": 0.05392, + "grad_norm": 0.6744571328163147, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 5392 + }, + { + "epoch": 0.05393, + "grad_norm": 0.5944811105728149, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 5393 + }, + { + "epoch": 0.05394, + "grad_norm": 0.581836998462677, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 5394 + }, + { + "epoch": 0.05395, + "grad_norm": 0.5920911431312561, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 5395 + }, + { + "epoch": 0.05396, + "grad_norm": 0.6112498641014099, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 5396 + }, + { + "epoch": 0.05397, + "grad_norm": 0.6823821067810059, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 5397 + }, + { + "epoch": 0.05398, + "grad_norm": 0.807902991771698, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 5398 + }, + { + "epoch": 0.05399, + "grad_norm": 1.0339542627334595, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 5399 + }, + { + "epoch": 0.054, + "grad_norm": 1.1419845819473267, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 5400 + }, + { + "epoch": 0.05401, + "grad_norm": 0.7827608585357666, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 5401 + }, + { + "epoch": 0.05402, + "grad_norm": 0.725298285484314, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 5402 + }, + { + "epoch": 0.05403, + "grad_norm": 0.7886972427368164, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 5403 + }, + { + "epoch": 0.05404, + "grad_norm": 0.8297346234321594, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 5404 + }, + { + "epoch": 0.05405, + "grad_norm": 0.8761243224143982, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 5405 + }, + { + "epoch": 0.05406, + "grad_norm": 0.9909228086471558, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 5406 + }, + { + "epoch": 0.05407, + "grad_norm": 1.0708101987838745, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 5407 + }, + { + "epoch": 0.05408, + "grad_norm": 0.9255216717720032, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 5408 + }, + { + "epoch": 0.05409, + "grad_norm": 1.0283777713775635, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 5409 + }, + { + "epoch": 0.0541, + "grad_norm": 1.0340330600738525, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 5410 + }, + { + "epoch": 0.05411, + "grad_norm": 0.8593031167984009, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 5411 + }, + { + "epoch": 0.05412, + "grad_norm": 0.7811865210533142, + "learning_rate": 0.003, + "loss": 4.09, + "step": 5412 + }, + { + "epoch": 0.05413, + "grad_norm": 0.8573576211929321, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 5413 + }, + { + "epoch": 0.05414, + "grad_norm": 0.9427306652069092, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 5414 + }, + { + "epoch": 0.05415, + "grad_norm": 1.2074553966522217, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 5415 + }, + { + "epoch": 0.05416, + "grad_norm": 0.827156662940979, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 5416 + }, + { + "epoch": 0.05417, + "grad_norm": 0.715948760509491, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 5417 + }, + { + "epoch": 0.05418, + "grad_norm": 0.6932428479194641, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 5418 + }, + { + "epoch": 0.05419, + "grad_norm": 0.849570095539093, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 5419 + }, + { + "epoch": 0.0542, + "grad_norm": 0.8560599088668823, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 5420 + }, + { + "epoch": 0.05421, + "grad_norm": 0.7644525170326233, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 5421 + }, + { + "epoch": 0.05422, + "grad_norm": 0.652097761631012, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 5422 + }, + { + "epoch": 0.05423, + "grad_norm": 0.5640384554862976, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 5423 + }, + { + "epoch": 0.05424, + "grad_norm": 0.5313403010368347, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 5424 + }, + { + "epoch": 0.05425, + "grad_norm": 0.6055113077163696, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 5425 + }, + { + "epoch": 0.05426, + "grad_norm": 0.5395570993423462, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 5426 + }, + { + "epoch": 0.05427, + "grad_norm": 0.5673665404319763, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 5427 + }, + { + "epoch": 0.05428, + "grad_norm": 0.6410355567932129, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 5428 + }, + { + "epoch": 0.05429, + "grad_norm": 0.7815173268318176, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 5429 + }, + { + "epoch": 0.0543, + "grad_norm": 0.8199576735496521, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 5430 + }, + { + "epoch": 0.05431, + "grad_norm": 0.7338206768035889, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 5431 + }, + { + "epoch": 0.05432, + "grad_norm": 0.6335470676422119, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 5432 + }, + { + "epoch": 0.05433, + "grad_norm": 0.599968433380127, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 5433 + }, + { + "epoch": 0.05434, + "grad_norm": 0.5905585885047913, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 5434 + }, + { + "epoch": 0.05435, + "grad_norm": 0.6752657890319824, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 5435 + }, + { + "epoch": 0.05436, + "grad_norm": 0.6819540858268738, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 5436 + }, + { + "epoch": 0.05437, + "grad_norm": 0.6730635762214661, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 5437 + }, + { + "epoch": 0.05438, + "grad_norm": 0.7070807814598083, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 5438 + }, + { + "epoch": 0.05439, + "grad_norm": 0.722388744354248, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 5439 + }, + { + "epoch": 0.0544, + "grad_norm": 0.6570700407028198, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 5440 + }, + { + "epoch": 0.05441, + "grad_norm": 0.6882500648498535, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 5441 + }, + { + "epoch": 0.05442, + "grad_norm": 0.7646239995956421, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 5442 + }, + { + "epoch": 0.05443, + "grad_norm": 0.7554492354393005, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 5443 + }, + { + "epoch": 0.05444, + "grad_norm": 0.7602279782295227, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 5444 + }, + { + "epoch": 0.05445, + "grad_norm": 0.6765475273132324, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 5445 + }, + { + "epoch": 0.05446, + "grad_norm": 0.5565710067749023, + "learning_rate": 0.003, + "loss": 4.054, + "step": 5446 + }, + { + "epoch": 0.05447, + "grad_norm": 0.49256643652915955, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 5447 + }, + { + "epoch": 0.05448, + "grad_norm": 0.5195604562759399, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 5448 + }, + { + "epoch": 0.05449, + "grad_norm": 0.5942506194114685, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 5449 + }, + { + "epoch": 0.0545, + "grad_norm": 0.6499919891357422, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 5450 + }, + { + "epoch": 0.05451, + "grad_norm": 0.6596571803092957, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 5451 + }, + { + "epoch": 0.05452, + "grad_norm": 0.6069369316101074, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 5452 + }, + { + "epoch": 0.05453, + "grad_norm": 0.6619586944580078, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 5453 + }, + { + "epoch": 0.05454, + "grad_norm": 0.6113542318344116, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 5454 + }, + { + "epoch": 0.05455, + "grad_norm": 0.598733127117157, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 5455 + }, + { + "epoch": 0.05456, + "grad_norm": 0.5969057679176331, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 5456 + }, + { + "epoch": 0.05457, + "grad_norm": 0.6789122819900513, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 5457 + }, + { + "epoch": 0.05458, + "grad_norm": 0.805001974105835, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 5458 + }, + { + "epoch": 0.05459, + "grad_norm": 1.0944715738296509, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 5459 + }, + { + "epoch": 0.0546, + "grad_norm": 0.9038564562797546, + "learning_rate": 0.003, + "loss": 4.07, + "step": 5460 + }, + { + "epoch": 0.05461, + "grad_norm": 0.8081420660018921, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 5461 + }, + { + "epoch": 0.05462, + "grad_norm": 0.7987942099571228, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 5462 + }, + { + "epoch": 0.05463, + "grad_norm": 0.8367160558700562, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 5463 + }, + { + "epoch": 0.05464, + "grad_norm": 0.8023653626441956, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 5464 + }, + { + "epoch": 0.05465, + "grad_norm": 0.8824062347412109, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 5465 + }, + { + "epoch": 0.05466, + "grad_norm": 0.8047217130661011, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 5466 + }, + { + "epoch": 0.05467, + "grad_norm": 0.6920082569122314, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 5467 + }, + { + "epoch": 0.05468, + "grad_norm": 0.7836238145828247, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 5468 + }, + { + "epoch": 0.05469, + "grad_norm": 1.0192773342132568, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 5469 + }, + { + "epoch": 0.0547, + "grad_norm": 1.280296802520752, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 5470 + }, + { + "epoch": 0.05471, + "grad_norm": 0.6577931046485901, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5471 + }, + { + "epoch": 0.05472, + "grad_norm": 0.7552001476287842, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 5472 + }, + { + "epoch": 0.05473, + "grad_norm": 0.7697224020957947, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 5473 + }, + { + "epoch": 0.05474, + "grad_norm": 0.8115640878677368, + "learning_rate": 0.003, + "loss": 4.064, + "step": 5474 + }, + { + "epoch": 0.05475, + "grad_norm": 0.8436885476112366, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 5475 + }, + { + "epoch": 0.05476, + "grad_norm": 0.9373152256011963, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 5476 + }, + { + "epoch": 0.05477, + "grad_norm": 0.8886646032333374, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 5477 + }, + { + "epoch": 0.05478, + "grad_norm": 0.7719184756278992, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 5478 + }, + { + "epoch": 0.05479, + "grad_norm": 0.7790704965591431, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 5479 + }, + { + "epoch": 0.0548, + "grad_norm": 0.7557684779167175, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 5480 + }, + { + "epoch": 0.05481, + "grad_norm": 0.761655867099762, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 5481 + }, + { + "epoch": 0.05482, + "grad_norm": 0.7313814163208008, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 5482 + }, + { + "epoch": 0.05483, + "grad_norm": 0.7328376770019531, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 5483 + }, + { + "epoch": 0.05484, + "grad_norm": 0.6982470750808716, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 5484 + }, + { + "epoch": 0.05485, + "grad_norm": 0.5700132846832275, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 5485 + }, + { + "epoch": 0.05486, + "grad_norm": 0.5434563755989075, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 5486 + }, + { + "epoch": 0.05487, + "grad_norm": 0.48218417167663574, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 5487 + }, + { + "epoch": 0.05488, + "grad_norm": 0.5017120242118835, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 5488 + }, + { + "epoch": 0.05489, + "grad_norm": 0.48504316806793213, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 5489 + }, + { + "epoch": 0.0549, + "grad_norm": 0.5177329778671265, + "learning_rate": 0.003, + "loss": 4.067, + "step": 5490 + }, + { + "epoch": 0.05491, + "grad_norm": 0.6282394528388977, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 5491 + }, + { + "epoch": 0.05492, + "grad_norm": 0.8392156958580017, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 5492 + }, + { + "epoch": 0.05493, + "grad_norm": 1.1088998317718506, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 5493 + }, + { + "epoch": 0.05494, + "grad_norm": 1.0195647478103638, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 5494 + }, + { + "epoch": 0.05495, + "grad_norm": 0.8553412556648254, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 5495 + }, + { + "epoch": 0.05496, + "grad_norm": 0.7595451474189758, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 5496 + }, + { + "epoch": 0.05497, + "grad_norm": 0.7024115920066833, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 5497 + }, + { + "epoch": 0.05498, + "grad_norm": 0.5727024674415588, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 5498 + }, + { + "epoch": 0.05499, + "grad_norm": 0.5514757037162781, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 5499 + }, + { + "epoch": 0.055, + "grad_norm": 0.6886269450187683, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 5500 + }, + { + "epoch": 0.05501, + "grad_norm": 0.8767561912536621, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 5501 + }, + { + "epoch": 0.05502, + "grad_norm": 0.9685607552528381, + "learning_rate": 0.003, + "loss": 4.06, + "step": 5502 + }, + { + "epoch": 0.05503, + "grad_norm": 0.8435690402984619, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 5503 + }, + { + "epoch": 0.05504, + "grad_norm": 0.6953498721122742, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 5504 + }, + { + "epoch": 0.05505, + "grad_norm": 0.6831181645393372, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 5505 + }, + { + "epoch": 0.05506, + "grad_norm": 0.6843089461326599, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 5506 + }, + { + "epoch": 0.05507, + "grad_norm": 0.5160518884658813, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 5507 + }, + { + "epoch": 0.05508, + "grad_norm": 0.5106964111328125, + "learning_rate": 0.003, + "loss": 4.069, + "step": 5508 + }, + { + "epoch": 0.05509, + "grad_norm": 0.49985066056251526, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 5509 + }, + { + "epoch": 0.0551, + "grad_norm": 0.511665403842926, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 5510 + }, + { + "epoch": 0.05511, + "grad_norm": 0.4913787543773651, + "learning_rate": 0.003, + "loss": 4.07, + "step": 5511 + }, + { + "epoch": 0.05512, + "grad_norm": 0.5255283713340759, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 5512 + }, + { + "epoch": 0.05513, + "grad_norm": 0.5358270406723022, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 5513 + }, + { + "epoch": 0.05514, + "grad_norm": 0.4987005293369293, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 5514 + }, + { + "epoch": 0.05515, + "grad_norm": 0.43703991174697876, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 5515 + }, + { + "epoch": 0.05516, + "grad_norm": 0.45214366912841797, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 5516 + }, + { + "epoch": 0.05517, + "grad_norm": 0.4733821749687195, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 5517 + }, + { + "epoch": 0.05518, + "grad_norm": 0.5811468958854675, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 5518 + }, + { + "epoch": 0.05519, + "grad_norm": 0.6899372339248657, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 5519 + }, + { + "epoch": 0.0552, + "grad_norm": 0.7010800838470459, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 5520 + }, + { + "epoch": 0.05521, + "grad_norm": 0.7988923788070679, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 5521 + }, + { + "epoch": 0.05522, + "grad_norm": 0.8185820579528809, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 5522 + }, + { + "epoch": 0.05523, + "grad_norm": 0.6712361574172974, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 5523 + }, + { + "epoch": 0.05524, + "grad_norm": 0.8089494705200195, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 5524 + }, + { + "epoch": 0.05525, + "grad_norm": 0.9261167049407959, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 5525 + }, + { + "epoch": 0.05526, + "grad_norm": 0.9133365154266357, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 5526 + }, + { + "epoch": 0.05527, + "grad_norm": 0.8450931310653687, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 5527 + }, + { + "epoch": 0.05528, + "grad_norm": 0.8984810709953308, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 5528 + }, + { + "epoch": 0.05529, + "grad_norm": 0.9497979283332825, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 5529 + }, + { + "epoch": 0.0553, + "grad_norm": 1.1338160037994385, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 5530 + }, + { + "epoch": 0.05531, + "grad_norm": 0.8146209120750427, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 5531 + }, + { + "epoch": 0.05532, + "grad_norm": 0.6237087249755859, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 5532 + }, + { + "epoch": 0.05533, + "grad_norm": 0.5327185392379761, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 5533 + }, + { + "epoch": 0.05534, + "grad_norm": 0.6118891835212708, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 5534 + }, + { + "epoch": 0.05535, + "grad_norm": 0.6608664989471436, + "learning_rate": 0.003, + "loss": 4.088, + "step": 5535 + }, + { + "epoch": 0.05536, + "grad_norm": 0.7102659344673157, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 5536 + }, + { + "epoch": 0.05537, + "grad_norm": 0.684587836265564, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 5537 + }, + { + "epoch": 0.05538, + "grad_norm": 0.7321231961250305, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 5538 + }, + { + "epoch": 0.05539, + "grad_norm": 0.798923909664154, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 5539 + }, + { + "epoch": 0.0554, + "grad_norm": 0.93647700548172, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 5540 + }, + { + "epoch": 0.05541, + "grad_norm": 1.2185319662094116, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 5541 + }, + { + "epoch": 0.05542, + "grad_norm": 0.9062533378601074, + "learning_rate": 0.003, + "loss": 4.074, + "step": 5542 + }, + { + "epoch": 0.05543, + "grad_norm": 0.7092111706733704, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 5543 + }, + { + "epoch": 0.05544, + "grad_norm": 0.5959237217903137, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 5544 + }, + { + "epoch": 0.05545, + "grad_norm": 0.6062164306640625, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 5545 + }, + { + "epoch": 0.05546, + "grad_norm": 0.64149010181427, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 5546 + }, + { + "epoch": 0.05547, + "grad_norm": 0.6897364258766174, + "learning_rate": 0.003, + "loss": 4.071, + "step": 5547 + }, + { + "epoch": 0.05548, + "grad_norm": 0.7177988886833191, + "learning_rate": 0.003, + "loss": 4.046, + "step": 5548 + }, + { + "epoch": 0.05549, + "grad_norm": 0.7397828102111816, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 5549 + }, + { + "epoch": 0.0555, + "grad_norm": 0.7000327110290527, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 5550 + }, + { + "epoch": 0.05551, + "grad_norm": 0.657970666885376, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 5551 + }, + { + "epoch": 0.05552, + "grad_norm": 0.5877078771591187, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 5552 + }, + { + "epoch": 0.05553, + "grad_norm": 0.6410927176475525, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 5553 + }, + { + "epoch": 0.05554, + "grad_norm": 0.7099154591560364, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 5554 + }, + { + "epoch": 0.05555, + "grad_norm": 0.9199838042259216, + "learning_rate": 0.003, + "loss": 4.109, + "step": 5555 + }, + { + "epoch": 0.05556, + "grad_norm": 1.068846344947815, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 5556 + }, + { + "epoch": 0.05557, + "grad_norm": 1.1641470193862915, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 5557 + }, + { + "epoch": 0.05558, + "grad_norm": 0.830019474029541, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 5558 + }, + { + "epoch": 0.05559, + "grad_norm": 0.6737862825393677, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 5559 + }, + { + "epoch": 0.0556, + "grad_norm": 0.725947916507721, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 5560 + }, + { + "epoch": 0.05561, + "grad_norm": 0.8011854887008667, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 5561 + }, + { + "epoch": 0.05562, + "grad_norm": 0.900126039981842, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5562 + }, + { + "epoch": 0.05563, + "grad_norm": 0.9443784952163696, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 5563 + }, + { + "epoch": 0.05564, + "grad_norm": 0.8231141567230225, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 5564 + }, + { + "epoch": 0.05565, + "grad_norm": 0.6643379330635071, + "learning_rate": 0.003, + "loss": 4.103, + "step": 5565 + }, + { + "epoch": 0.05566, + "grad_norm": 0.6313121318817139, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 5566 + }, + { + "epoch": 0.05567, + "grad_norm": 0.651742160320282, + "learning_rate": 0.003, + "loss": 4.062, + "step": 5567 + }, + { + "epoch": 0.05568, + "grad_norm": 0.6961485743522644, + "learning_rate": 0.003, + "loss": 4.095, + "step": 5568 + }, + { + "epoch": 0.05569, + "grad_norm": 0.7175015211105347, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 5569 + }, + { + "epoch": 0.0557, + "grad_norm": 0.8055482506752014, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 5570 + }, + { + "epoch": 0.05571, + "grad_norm": 0.8521894812583923, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 5571 + }, + { + "epoch": 0.05572, + "grad_norm": 0.8425765037536621, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 5572 + }, + { + "epoch": 0.05573, + "grad_norm": 0.8626205325126648, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 5573 + }, + { + "epoch": 0.05574, + "grad_norm": 0.9191294312477112, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 5574 + }, + { + "epoch": 0.05575, + "grad_norm": 0.7949203252792358, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 5575 + }, + { + "epoch": 0.05576, + "grad_norm": 0.6600616574287415, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 5576 + }, + { + "epoch": 0.05577, + "grad_norm": 0.5764843821525574, + "learning_rate": 0.003, + "loss": 4.064, + "step": 5577 + }, + { + "epoch": 0.05578, + "grad_norm": 0.5156270265579224, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 5578 + }, + { + "epoch": 0.05579, + "grad_norm": 0.5665589570999146, + "learning_rate": 0.003, + "loss": 4.078, + "step": 5579 + }, + { + "epoch": 0.0558, + "grad_norm": 0.5314322710037231, + "learning_rate": 0.003, + "loss": 4.068, + "step": 5580 + }, + { + "epoch": 0.05581, + "grad_norm": 0.5085921287536621, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 5581 + }, + { + "epoch": 0.05582, + "grad_norm": 0.5511303544044495, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 5582 + }, + { + "epoch": 0.05583, + "grad_norm": 0.5478542447090149, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 5583 + }, + { + "epoch": 0.05584, + "grad_norm": 0.601093590259552, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 5584 + }, + { + "epoch": 0.05585, + "grad_norm": 0.6396297812461853, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 5585 + }, + { + "epoch": 0.05586, + "grad_norm": 0.5573460459709167, + "learning_rate": 0.003, + "loss": 4.074, + "step": 5586 + }, + { + "epoch": 0.05587, + "grad_norm": 0.5427320599555969, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 5587 + }, + { + "epoch": 0.05588, + "grad_norm": 0.5215794444084167, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 5588 + }, + { + "epoch": 0.05589, + "grad_norm": 0.5945990085601807, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 5589 + }, + { + "epoch": 0.0559, + "grad_norm": 0.7637835144996643, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 5590 + }, + { + "epoch": 0.05591, + "grad_norm": 0.8685312271118164, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 5591 + }, + { + "epoch": 0.05592, + "grad_norm": 1.0447856187820435, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 5592 + }, + { + "epoch": 0.05593, + "grad_norm": 0.9590107798576355, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 5593 + }, + { + "epoch": 0.05594, + "grad_norm": 0.6435539722442627, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 5594 + }, + { + "epoch": 0.05595, + "grad_norm": 0.6978614330291748, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 5595 + }, + { + "epoch": 0.05596, + "grad_norm": 0.8227618932723999, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 5596 + }, + { + "epoch": 0.05597, + "grad_norm": 0.8983748555183411, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 5597 + }, + { + "epoch": 0.05598, + "grad_norm": 0.864307165145874, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 5598 + }, + { + "epoch": 0.05599, + "grad_norm": 0.7344439029693604, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 5599 + }, + { + "epoch": 0.056, + "grad_norm": 0.5893545746803284, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 5600 + }, + { + "epoch": 0.05601, + "grad_norm": 0.6228591203689575, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 5601 + }, + { + "epoch": 0.05602, + "grad_norm": 0.7110965847969055, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 5602 + }, + { + "epoch": 0.05603, + "grad_norm": 0.6936246752738953, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 5603 + }, + { + "epoch": 0.05604, + "grad_norm": 0.696966290473938, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 5604 + }, + { + "epoch": 0.05605, + "grad_norm": 0.7240187525749207, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 5605 + }, + { + "epoch": 0.05606, + "grad_norm": 0.807013750076294, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 5606 + }, + { + "epoch": 0.05607, + "grad_norm": 0.8161320090293884, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 5607 + }, + { + "epoch": 0.05608, + "grad_norm": 0.78641277551651, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 5608 + }, + { + "epoch": 0.05609, + "grad_norm": 0.6392374038696289, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 5609 + }, + { + "epoch": 0.0561, + "grad_norm": 0.7331043481826782, + "learning_rate": 0.003, + "loss": 4.101, + "step": 5610 + }, + { + "epoch": 0.05611, + "grad_norm": 0.7653400301933289, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 5611 + }, + { + "epoch": 0.05612, + "grad_norm": 0.8886222243309021, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 5612 + }, + { + "epoch": 0.05613, + "grad_norm": 1.155055284500122, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 5613 + }, + { + "epoch": 0.05614, + "grad_norm": 0.8169151544570923, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 5614 + }, + { + "epoch": 0.05615, + "grad_norm": 0.7616031765937805, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 5615 + }, + { + "epoch": 0.05616, + "grad_norm": 0.7899324297904968, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 5616 + }, + { + "epoch": 0.05617, + "grad_norm": 0.7144830226898193, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 5617 + }, + { + "epoch": 0.05618, + "grad_norm": 0.6747182011604309, + "learning_rate": 0.003, + "loss": 4.059, + "step": 5618 + }, + { + "epoch": 0.05619, + "grad_norm": 0.6562235951423645, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 5619 + }, + { + "epoch": 0.0562, + "grad_norm": 0.5861763954162598, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 5620 + }, + { + "epoch": 0.05621, + "grad_norm": 0.5951058864593506, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 5621 + }, + { + "epoch": 0.05622, + "grad_norm": 0.5614970922470093, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 5622 + }, + { + "epoch": 0.05623, + "grad_norm": 0.5402902364730835, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 5623 + }, + { + "epoch": 0.05624, + "grad_norm": 0.5211602449417114, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 5624 + }, + { + "epoch": 0.05625, + "grad_norm": 0.5597396492958069, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 5625 + }, + { + "epoch": 0.05626, + "grad_norm": 0.6586695909500122, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 5626 + }, + { + "epoch": 0.05627, + "grad_norm": 0.8000897765159607, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 5627 + }, + { + "epoch": 0.05628, + "grad_norm": 0.8259612321853638, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 5628 + }, + { + "epoch": 0.05629, + "grad_norm": 0.9033710956573486, + "learning_rate": 0.003, + "loss": 4.081, + "step": 5629 + }, + { + "epoch": 0.0563, + "grad_norm": 0.9907213449478149, + "learning_rate": 0.003, + "loss": 4.066, + "step": 5630 + }, + { + "epoch": 0.05631, + "grad_norm": 0.9617007970809937, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 5631 + }, + { + "epoch": 0.05632, + "grad_norm": 1.084578275680542, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 5632 + }, + { + "epoch": 0.05633, + "grad_norm": 0.8903836607933044, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 5633 + }, + { + "epoch": 0.05634, + "grad_norm": 0.8303310871124268, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 5634 + }, + { + "epoch": 0.05635, + "grad_norm": 0.8191870450973511, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 5635 + }, + { + "epoch": 0.05636, + "grad_norm": 0.8257316946983337, + "learning_rate": 0.003, + "loss": 4.077, + "step": 5636 + }, + { + "epoch": 0.05637, + "grad_norm": 0.8293499946594238, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 5637 + }, + { + "epoch": 0.05638, + "grad_norm": 0.7823120355606079, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 5638 + }, + { + "epoch": 0.05639, + "grad_norm": 0.7697476148605347, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 5639 + }, + { + "epoch": 0.0564, + "grad_norm": 0.783143937587738, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 5640 + }, + { + "epoch": 0.05641, + "grad_norm": 0.7607929110527039, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 5641 + }, + { + "epoch": 0.05642, + "grad_norm": 0.8684077858924866, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 5642 + }, + { + "epoch": 0.05643, + "grad_norm": 0.8539713025093079, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 5643 + }, + { + "epoch": 0.05644, + "grad_norm": 0.7667094469070435, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 5644 + }, + { + "epoch": 0.05645, + "grad_norm": 0.6266912817955017, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 5645 + }, + { + "epoch": 0.05646, + "grad_norm": 0.5951451659202576, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 5646 + }, + { + "epoch": 0.05647, + "grad_norm": 0.5667705535888672, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 5647 + }, + { + "epoch": 0.05648, + "grad_norm": 0.5914098024368286, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 5648 + }, + { + "epoch": 0.05649, + "grad_norm": 0.6961567401885986, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 5649 + }, + { + "epoch": 0.0565, + "grad_norm": 0.6270180940628052, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 5650 + }, + { + "epoch": 0.05651, + "grad_norm": 0.5370350480079651, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 5651 + }, + { + "epoch": 0.05652, + "grad_norm": 0.43584731221199036, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 5652 + }, + { + "epoch": 0.05653, + "grad_norm": 0.4690995514392853, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 5653 + }, + { + "epoch": 0.05654, + "grad_norm": 0.48374971747398376, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 5654 + }, + { + "epoch": 0.05655, + "grad_norm": 0.6023554801940918, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 5655 + }, + { + "epoch": 0.05656, + "grad_norm": 0.8215942978858948, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 5656 + }, + { + "epoch": 0.05657, + "grad_norm": 1.086574912071228, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 5657 + }, + { + "epoch": 0.05658, + "grad_norm": 0.9434396028518677, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 5658 + }, + { + "epoch": 0.05659, + "grad_norm": 0.8089240789413452, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 5659 + }, + { + "epoch": 0.0566, + "grad_norm": 0.6663325428962708, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 5660 + }, + { + "epoch": 0.05661, + "grad_norm": 0.7506017088890076, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 5661 + }, + { + "epoch": 0.05662, + "grad_norm": 0.8057636618614197, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 5662 + }, + { + "epoch": 0.05663, + "grad_norm": 0.7259244322776794, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 5663 + }, + { + "epoch": 0.05664, + "grad_norm": 0.6917450428009033, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 5664 + }, + { + "epoch": 0.05665, + "grad_norm": 0.7202029824256897, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 5665 + }, + { + "epoch": 0.05666, + "grad_norm": 0.7679628133773804, + "learning_rate": 0.003, + "loss": 4.075, + "step": 5666 + }, + { + "epoch": 0.05667, + "grad_norm": 0.725118100643158, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 5667 + }, + { + "epoch": 0.05668, + "grad_norm": 0.6143912672996521, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 5668 + }, + { + "epoch": 0.05669, + "grad_norm": 0.5653799772262573, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 5669 + }, + { + "epoch": 0.0567, + "grad_norm": 0.6119061708450317, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 5670 + }, + { + "epoch": 0.05671, + "grad_norm": 0.735202968120575, + "learning_rate": 0.003, + "loss": 4.059, + "step": 5671 + }, + { + "epoch": 0.05672, + "grad_norm": 0.7487754225730896, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 5672 + }, + { + "epoch": 0.05673, + "grad_norm": 0.7744391560554504, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 5673 + }, + { + "epoch": 0.05674, + "grad_norm": 0.7870099544525146, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 5674 + }, + { + "epoch": 0.05675, + "grad_norm": 0.888359785079956, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 5675 + }, + { + "epoch": 0.05676, + "grad_norm": 0.7705642580986023, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 5676 + }, + { + "epoch": 0.05677, + "grad_norm": 0.5689957737922668, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 5677 + }, + { + "epoch": 0.05678, + "grad_norm": 0.5393115282058716, + "learning_rate": 0.003, + "loss": 4.061, + "step": 5678 + }, + { + "epoch": 0.05679, + "grad_norm": 0.5655098557472229, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 5679 + }, + { + "epoch": 0.0568, + "grad_norm": 0.6075883507728577, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 5680 + }, + { + "epoch": 0.05681, + "grad_norm": 0.6243870854377747, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 5681 + }, + { + "epoch": 0.05682, + "grad_norm": 0.6489701867103577, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 5682 + }, + { + "epoch": 0.05683, + "grad_norm": 0.6358552575111389, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 5683 + }, + { + "epoch": 0.05684, + "grad_norm": 0.6328252553939819, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 5684 + }, + { + "epoch": 0.05685, + "grad_norm": 0.596019446849823, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 5685 + }, + { + "epoch": 0.05686, + "grad_norm": 0.6575425863265991, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 5686 + }, + { + "epoch": 0.05687, + "grad_norm": 0.6724589467048645, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 5687 + }, + { + "epoch": 0.05688, + "grad_norm": 0.6699223518371582, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 5688 + }, + { + "epoch": 0.05689, + "grad_norm": 0.7112768292427063, + "learning_rate": 0.003, + "loss": 4.038, + "step": 5689 + }, + { + "epoch": 0.0569, + "grad_norm": 0.8006166815757751, + "learning_rate": 0.003, + "loss": 4.067, + "step": 5690 + }, + { + "epoch": 0.05691, + "grad_norm": 0.9695540070533752, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 5691 + }, + { + "epoch": 0.05692, + "grad_norm": 1.068497896194458, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 5692 + }, + { + "epoch": 0.05693, + "grad_norm": 0.9201273918151855, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 5693 + }, + { + "epoch": 0.05694, + "grad_norm": 0.8582456111907959, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 5694 + }, + { + "epoch": 0.05695, + "grad_norm": 0.933272123336792, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 5695 + }, + { + "epoch": 0.05696, + "grad_norm": 0.8945035338401794, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 5696 + }, + { + "epoch": 0.05697, + "grad_norm": 0.8932650685310364, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 5697 + }, + { + "epoch": 0.05698, + "grad_norm": 1.038320541381836, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 5698 + }, + { + "epoch": 0.05699, + "grad_norm": 1.0705705881118774, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 5699 + }, + { + "epoch": 0.057, + "grad_norm": 0.9729557037353516, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 5700 + }, + { + "epoch": 0.05701, + "grad_norm": 0.9241772294044495, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 5701 + }, + { + "epoch": 0.05702, + "grad_norm": 0.9331616759300232, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 5702 + }, + { + "epoch": 0.05703, + "grad_norm": 0.8024580478668213, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 5703 + }, + { + "epoch": 0.05704, + "grad_norm": 0.726856529712677, + "learning_rate": 0.003, + "loss": 4.078, + "step": 5704 + }, + { + "epoch": 0.05705, + "grad_norm": 0.6963912844657898, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 5705 + }, + { + "epoch": 0.05706, + "grad_norm": 0.6920854449272156, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 5706 + }, + { + "epoch": 0.05707, + "grad_norm": 0.7907909154891968, + "learning_rate": 0.003, + "loss": 4.075, + "step": 5707 + }, + { + "epoch": 0.05708, + "grad_norm": 0.8948379755020142, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 5708 + }, + { + "epoch": 0.05709, + "grad_norm": 0.7752033472061157, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 5709 + }, + { + "epoch": 0.0571, + "grad_norm": 0.577285885810852, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 5710 + }, + { + "epoch": 0.05711, + "grad_norm": 0.5138233304023743, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 5711 + }, + { + "epoch": 0.05712, + "grad_norm": 0.5642086267471313, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 5712 + }, + { + "epoch": 0.05713, + "grad_norm": 0.6007573008537292, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 5713 + }, + { + "epoch": 0.05714, + "grad_norm": 0.5792896151542664, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 5714 + }, + { + "epoch": 0.05715, + "grad_norm": 0.551459550857544, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 5715 + }, + { + "epoch": 0.05716, + "grad_norm": 0.6014549136161804, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 5716 + }, + { + "epoch": 0.05717, + "grad_norm": 0.6075748801231384, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 5717 + }, + { + "epoch": 0.05718, + "grad_norm": 0.7294437885284424, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 5718 + }, + { + "epoch": 0.05719, + "grad_norm": 0.8543415069580078, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 5719 + }, + { + "epoch": 0.0572, + "grad_norm": 0.8952844738960266, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 5720 + }, + { + "epoch": 0.05721, + "grad_norm": 0.8221727013587952, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 5721 + }, + { + "epoch": 0.05722, + "grad_norm": 0.8208548426628113, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 5722 + }, + { + "epoch": 0.05723, + "grad_norm": 0.8748874664306641, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 5723 + }, + { + "epoch": 0.05724, + "grad_norm": 0.747161328792572, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 5724 + }, + { + "epoch": 0.05725, + "grad_norm": 0.6817445158958435, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 5725 + }, + { + "epoch": 0.05726, + "grad_norm": 0.6044837832450867, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 5726 + }, + { + "epoch": 0.05727, + "grad_norm": 0.5440679788589478, + "learning_rate": 0.003, + "loss": 4.049, + "step": 5727 + }, + { + "epoch": 0.05728, + "grad_norm": 0.5482798218727112, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 5728 + }, + { + "epoch": 0.05729, + "grad_norm": 0.5878225564956665, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 5729 + }, + { + "epoch": 0.0573, + "grad_norm": 0.5776148438453674, + "learning_rate": 0.003, + "loss": 4.03, + "step": 5730 + }, + { + "epoch": 0.05731, + "grad_norm": 0.6275573968887329, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 5731 + }, + { + "epoch": 0.05732, + "grad_norm": 0.6557134985923767, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 5732 + }, + { + "epoch": 0.05733, + "grad_norm": 0.6577096581459045, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 5733 + }, + { + "epoch": 0.05734, + "grad_norm": 0.7086383104324341, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 5734 + }, + { + "epoch": 0.05735, + "grad_norm": 0.7373378276824951, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 5735 + }, + { + "epoch": 0.05736, + "grad_norm": 0.8890331983566284, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 5736 + }, + { + "epoch": 0.05737, + "grad_norm": 1.07709801197052, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 5737 + }, + { + "epoch": 0.05738, + "grad_norm": 0.7901797294616699, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 5738 + }, + { + "epoch": 0.05739, + "grad_norm": 0.596919596195221, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 5739 + }, + { + "epoch": 0.0574, + "grad_norm": 0.7254593372344971, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 5740 + }, + { + "epoch": 0.05741, + "grad_norm": 0.7549036741256714, + "learning_rate": 0.003, + "loss": 4.046, + "step": 5741 + }, + { + "epoch": 0.05742, + "grad_norm": 0.8220978379249573, + "learning_rate": 0.003, + "loss": 4.065, + "step": 5742 + }, + { + "epoch": 0.05743, + "grad_norm": 0.8430297374725342, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 5743 + }, + { + "epoch": 0.05744, + "grad_norm": 0.9336887001991272, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 5744 + }, + { + "epoch": 0.05745, + "grad_norm": 0.9208829402923584, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 5745 + }, + { + "epoch": 0.05746, + "grad_norm": 0.8885453939437866, + "learning_rate": 0.003, + "loss": 4.069, + "step": 5746 + }, + { + "epoch": 0.05747, + "grad_norm": 0.8862046003341675, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 5747 + }, + { + "epoch": 0.05748, + "grad_norm": 0.8300103545188904, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 5748 + }, + { + "epoch": 0.05749, + "grad_norm": 0.747201681137085, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 5749 + }, + { + "epoch": 0.0575, + "grad_norm": 0.7727035880088806, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 5750 + }, + { + "epoch": 0.05751, + "grad_norm": 0.7511826753616333, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 5751 + }, + { + "epoch": 0.05752, + "grad_norm": 0.9084669947624207, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 5752 + }, + { + "epoch": 0.05753, + "grad_norm": 0.8014945983886719, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 5753 + }, + { + "epoch": 0.05754, + "grad_norm": 0.8008480072021484, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 5754 + }, + { + "epoch": 0.05755, + "grad_norm": 0.8061231970787048, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 5755 + }, + { + "epoch": 0.05756, + "grad_norm": 0.8867525458335876, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 5756 + }, + { + "epoch": 0.05757, + "grad_norm": 0.9424989223480225, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 5757 + }, + { + "epoch": 0.05758, + "grad_norm": 0.9214147329330444, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 5758 + }, + { + "epoch": 0.05759, + "grad_norm": 0.8139206767082214, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 5759 + }, + { + "epoch": 0.0576, + "grad_norm": 0.7159584760665894, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 5760 + }, + { + "epoch": 0.05761, + "grad_norm": 0.7513906955718994, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 5761 + }, + { + "epoch": 0.05762, + "grad_norm": 0.7571491003036499, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 5762 + }, + { + "epoch": 0.05763, + "grad_norm": 0.7423854470252991, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 5763 + }, + { + "epoch": 0.05764, + "grad_norm": 0.7205909490585327, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 5764 + }, + { + "epoch": 0.05765, + "grad_norm": 0.672808051109314, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 5765 + }, + { + "epoch": 0.05766, + "grad_norm": 0.5956676006317139, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 5766 + }, + { + "epoch": 0.05767, + "grad_norm": 0.6225743889808655, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 5767 + }, + { + "epoch": 0.05768, + "grad_norm": 0.6613940000534058, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 5768 + }, + { + "epoch": 0.05769, + "grad_norm": 0.6825332641601562, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 5769 + }, + { + "epoch": 0.0577, + "grad_norm": 0.6207984089851379, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 5770 + }, + { + "epoch": 0.05771, + "grad_norm": 0.6129540801048279, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 5771 + }, + { + "epoch": 0.05772, + "grad_norm": 0.6390023827552795, + "learning_rate": 0.003, + "loss": 4.084, + "step": 5772 + }, + { + "epoch": 0.05773, + "grad_norm": 0.7322221994400024, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 5773 + }, + { + "epoch": 0.05774, + "grad_norm": 0.7655969262123108, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 5774 + }, + { + "epoch": 0.05775, + "grad_norm": 0.7013616561889648, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 5775 + }, + { + "epoch": 0.05776, + "grad_norm": 0.5979982018470764, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 5776 + }, + { + "epoch": 0.05777, + "grad_norm": 0.4852938652038574, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 5777 + }, + { + "epoch": 0.05778, + "grad_norm": 0.5266335010528564, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 5778 + }, + { + "epoch": 0.05779, + "grad_norm": 0.676627516746521, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 5779 + }, + { + "epoch": 0.0578, + "grad_norm": 0.8605179190635681, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 5780 + }, + { + "epoch": 0.05781, + "grad_norm": 0.9018716216087341, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 5781 + }, + { + "epoch": 0.05782, + "grad_norm": 0.7411509156227112, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 5782 + }, + { + "epoch": 0.05783, + "grad_norm": 0.5312531590461731, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 5783 + }, + { + "epoch": 0.05784, + "grad_norm": 0.5841469764709473, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 5784 + }, + { + "epoch": 0.05785, + "grad_norm": 0.7773861289024353, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 5785 + }, + { + "epoch": 0.05786, + "grad_norm": 0.7710773944854736, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 5786 + }, + { + "epoch": 0.05787, + "grad_norm": 0.6817703247070312, + "learning_rate": 0.003, + "loss": 4.039, + "step": 5787 + }, + { + "epoch": 0.05788, + "grad_norm": 0.7051790952682495, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 5788 + }, + { + "epoch": 0.05789, + "grad_norm": 0.7617840766906738, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 5789 + }, + { + "epoch": 0.0579, + "grad_norm": 0.7429543733596802, + "learning_rate": 0.003, + "loss": 4.079, + "step": 5790 + }, + { + "epoch": 0.05791, + "grad_norm": 0.6779160499572754, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 5791 + }, + { + "epoch": 0.05792, + "grad_norm": 0.7390137314796448, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 5792 + }, + { + "epoch": 0.05793, + "grad_norm": 0.64161217212677, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 5793 + }, + { + "epoch": 0.05794, + "grad_norm": 0.6376305818557739, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 5794 + }, + { + "epoch": 0.05795, + "grad_norm": 0.5993477702140808, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 5795 + }, + { + "epoch": 0.05796, + "grad_norm": 0.6392378211021423, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 5796 + }, + { + "epoch": 0.05797, + "grad_norm": 0.6526309251785278, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 5797 + }, + { + "epoch": 0.05798, + "grad_norm": 0.7773505449295044, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 5798 + }, + { + "epoch": 0.05799, + "grad_norm": 0.9118677377700806, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 5799 + }, + { + "epoch": 0.058, + "grad_norm": 1.0093053579330444, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 5800 + }, + { + "epoch": 0.05801, + "grad_norm": 1.0082321166992188, + "learning_rate": 0.003, + "loss": 4.068, + "step": 5801 + }, + { + "epoch": 0.05802, + "grad_norm": 0.8237224221229553, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 5802 + }, + { + "epoch": 0.05803, + "grad_norm": 0.8256529569625854, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 5803 + }, + { + "epoch": 0.05804, + "grad_norm": 0.8300235867500305, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 5804 + }, + { + "epoch": 0.05805, + "grad_norm": 0.8762092590332031, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 5805 + }, + { + "epoch": 0.05806, + "grad_norm": 0.8307374715805054, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 5806 + }, + { + "epoch": 0.05807, + "grad_norm": 0.8506011962890625, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 5807 + }, + { + "epoch": 0.05808, + "grad_norm": 0.8518326282501221, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 5808 + }, + { + "epoch": 0.05809, + "grad_norm": 0.9622186422348022, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 5809 + }, + { + "epoch": 0.0581, + "grad_norm": 1.2263989448547363, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 5810 + }, + { + "epoch": 0.05811, + "grad_norm": 0.9196377396583557, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 5811 + }, + { + "epoch": 0.05812, + "grad_norm": 0.8726242184638977, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 5812 + }, + { + "epoch": 0.05813, + "grad_norm": 0.8841713070869446, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 5813 + }, + { + "epoch": 0.05814, + "grad_norm": 0.851186215877533, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 5814 + }, + { + "epoch": 0.05815, + "grad_norm": 0.8571911454200745, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 5815 + }, + { + "epoch": 0.05816, + "grad_norm": 0.7161019444465637, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 5816 + }, + { + "epoch": 0.05817, + "grad_norm": 0.7098151445388794, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 5817 + }, + { + "epoch": 0.05818, + "grad_norm": 0.6854085922241211, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 5818 + }, + { + "epoch": 0.05819, + "grad_norm": 0.6578537225723267, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 5819 + }, + { + "epoch": 0.0582, + "grad_norm": 0.5946629643440247, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 5820 + }, + { + "epoch": 0.05821, + "grad_norm": 0.5087273716926575, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 5821 + }, + { + "epoch": 0.05822, + "grad_norm": 0.4627154767513275, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 5822 + }, + { + "epoch": 0.05823, + "grad_norm": 0.3847285509109497, + "learning_rate": 0.003, + "loss": 4.071, + "step": 5823 + }, + { + "epoch": 0.05824, + "grad_norm": 0.3624761998653412, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 5824 + }, + { + "epoch": 0.05825, + "grad_norm": 0.3458106219768524, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 5825 + }, + { + "epoch": 0.05826, + "grad_norm": 0.3547622561454773, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 5826 + }, + { + "epoch": 0.05827, + "grad_norm": 0.3819022476673126, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 5827 + }, + { + "epoch": 0.05828, + "grad_norm": 0.38096553087234497, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 5828 + }, + { + "epoch": 0.05829, + "grad_norm": 0.41201043128967285, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 5829 + }, + { + "epoch": 0.0583, + "grad_norm": 0.5885247588157654, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 5830 + }, + { + "epoch": 0.05831, + "grad_norm": 0.8928015828132629, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 5831 + }, + { + "epoch": 0.05832, + "grad_norm": 1.3094123601913452, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 5832 + }, + { + "epoch": 0.05833, + "grad_norm": 0.48254647850990295, + "learning_rate": 0.003, + "loss": 4.061, + "step": 5833 + }, + { + "epoch": 0.05834, + "grad_norm": 0.7129554748535156, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 5834 + }, + { + "epoch": 0.05835, + "grad_norm": 1.0598682165145874, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 5835 + }, + { + "epoch": 0.05836, + "grad_norm": 0.7417224645614624, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 5836 + }, + { + "epoch": 0.05837, + "grad_norm": 0.6337145566940308, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 5837 + }, + { + "epoch": 0.05838, + "grad_norm": 0.7507103085517883, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 5838 + }, + { + "epoch": 0.05839, + "grad_norm": 0.9548166990280151, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 5839 + }, + { + "epoch": 0.0584, + "grad_norm": 0.9799003005027771, + "learning_rate": 0.003, + "loss": 4.106, + "step": 5840 + }, + { + "epoch": 0.05841, + "grad_norm": 0.9687362909317017, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 5841 + }, + { + "epoch": 0.05842, + "grad_norm": 0.8212959170341492, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 5842 + }, + { + "epoch": 0.05843, + "grad_norm": 0.8485821485519409, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 5843 + }, + { + "epoch": 0.05844, + "grad_norm": 1.0464115142822266, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 5844 + }, + { + "epoch": 0.05845, + "grad_norm": 0.9341375231742859, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 5845 + }, + { + "epoch": 0.05846, + "grad_norm": 0.9866040349006653, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 5846 + }, + { + "epoch": 0.05847, + "grad_norm": 0.9109794497489929, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 5847 + }, + { + "epoch": 0.05848, + "grad_norm": 0.715682864189148, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 5848 + }, + { + "epoch": 0.05849, + "grad_norm": 0.618606686592102, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 5849 + }, + { + "epoch": 0.0585, + "grad_norm": 0.6357572078704834, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 5850 + }, + { + "epoch": 0.05851, + "grad_norm": 0.6797095537185669, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 5851 + }, + { + "epoch": 0.05852, + "grad_norm": 0.6348204612731934, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 5852 + }, + { + "epoch": 0.05853, + "grad_norm": 0.5872939229011536, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 5853 + }, + { + "epoch": 0.05854, + "grad_norm": 0.6542626619338989, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 5854 + }, + { + "epoch": 0.05855, + "grad_norm": 0.6466689705848694, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 5855 + }, + { + "epoch": 0.05856, + "grad_norm": 0.7588222622871399, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 5856 + }, + { + "epoch": 0.05857, + "grad_norm": 1.0768864154815674, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 5857 + }, + { + "epoch": 0.05858, + "grad_norm": 0.9665427803993225, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 5858 + }, + { + "epoch": 0.05859, + "grad_norm": 0.7883841395378113, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 5859 + }, + { + "epoch": 0.0586, + "grad_norm": 0.7918685078620911, + "learning_rate": 0.003, + "loss": 4.087, + "step": 5860 + }, + { + "epoch": 0.05861, + "grad_norm": 0.711283802986145, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 5861 + }, + { + "epoch": 0.05862, + "grad_norm": 0.7785149812698364, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 5862 + }, + { + "epoch": 0.05863, + "grad_norm": 0.6729006171226501, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 5863 + }, + { + "epoch": 0.05864, + "grad_norm": 0.5089790225028992, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 5864 + }, + { + "epoch": 0.05865, + "grad_norm": 0.5055457949638367, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 5865 + }, + { + "epoch": 0.05866, + "grad_norm": 0.5680795311927795, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 5866 + }, + { + "epoch": 0.05867, + "grad_norm": 0.6734050512313843, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 5867 + }, + { + "epoch": 0.05868, + "grad_norm": 0.68412184715271, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 5868 + }, + { + "epoch": 0.05869, + "grad_norm": 0.6876507997512817, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 5869 + }, + { + "epoch": 0.0587, + "grad_norm": 0.6638537645339966, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 5870 + }, + { + "epoch": 0.05871, + "grad_norm": 0.6519478559494019, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 5871 + }, + { + "epoch": 0.05872, + "grad_norm": 0.5919815897941589, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 5872 + }, + { + "epoch": 0.05873, + "grad_norm": 0.5140731930732727, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 5873 + }, + { + "epoch": 0.05874, + "grad_norm": 0.5663528442382812, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 5874 + }, + { + "epoch": 0.05875, + "grad_norm": 0.6782262921333313, + "learning_rate": 0.003, + "loss": 4.068, + "step": 5875 + }, + { + "epoch": 0.05876, + "grad_norm": 0.8224900364875793, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 5876 + }, + { + "epoch": 0.05877, + "grad_norm": 0.8717034459114075, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 5877 + }, + { + "epoch": 0.05878, + "grad_norm": 0.8781859874725342, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 5878 + }, + { + "epoch": 0.05879, + "grad_norm": 0.6580458879470825, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 5879 + }, + { + "epoch": 0.0588, + "grad_norm": 0.7042397856712341, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 5880 + }, + { + "epoch": 0.05881, + "grad_norm": 0.6979352831840515, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 5881 + }, + { + "epoch": 0.05882, + "grad_norm": 0.7222563624382019, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 5882 + }, + { + "epoch": 0.05883, + "grad_norm": 0.7145103216171265, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 5883 + }, + { + "epoch": 0.05884, + "grad_norm": 0.6383862495422363, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 5884 + }, + { + "epoch": 0.05885, + "grad_norm": 0.5817567706108093, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 5885 + }, + { + "epoch": 0.05886, + "grad_norm": 0.6241205930709839, + "learning_rate": 0.003, + "loss": 4.069, + "step": 5886 + }, + { + "epoch": 0.05887, + "grad_norm": 0.702506959438324, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 5887 + }, + { + "epoch": 0.05888, + "grad_norm": 0.6908167600631714, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 5888 + }, + { + "epoch": 0.05889, + "grad_norm": 0.7523607015609741, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 5889 + }, + { + "epoch": 0.0589, + "grad_norm": 0.8569447994232178, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 5890 + }, + { + "epoch": 0.05891, + "grad_norm": 1.0577572584152222, + "learning_rate": 0.003, + "loss": 4.064, + "step": 5891 + }, + { + "epoch": 0.05892, + "grad_norm": 0.9264476895332336, + "learning_rate": 0.003, + "loss": 4.095, + "step": 5892 + }, + { + "epoch": 0.05893, + "grad_norm": 0.7721839547157288, + "learning_rate": 0.003, + "loss": 4.067, + "step": 5893 + }, + { + "epoch": 0.05894, + "grad_norm": 0.7444570660591125, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 5894 + }, + { + "epoch": 0.05895, + "grad_norm": 0.7716593146324158, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 5895 + }, + { + "epoch": 0.05896, + "grad_norm": 0.834312915802002, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 5896 + }, + { + "epoch": 0.05897, + "grad_norm": 0.9077129364013672, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 5897 + }, + { + "epoch": 0.05898, + "grad_norm": 0.7383908629417419, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 5898 + }, + { + "epoch": 0.05899, + "grad_norm": 0.7159144878387451, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 5899 + }, + { + "epoch": 0.059, + "grad_norm": 0.8070700168609619, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 5900 + }, + { + "epoch": 0.05901, + "grad_norm": 0.9503213763237, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 5901 + }, + { + "epoch": 0.05902, + "grad_norm": 0.9452478885650635, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 5902 + }, + { + "epoch": 0.05903, + "grad_norm": 0.9364940524101257, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 5903 + }, + { + "epoch": 0.05904, + "grad_norm": 0.710010290145874, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 5904 + }, + { + "epoch": 0.05905, + "grad_norm": 0.7259859442710876, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 5905 + }, + { + "epoch": 0.05906, + "grad_norm": 0.7474007606506348, + "learning_rate": 0.003, + "loss": 4.068, + "step": 5906 + }, + { + "epoch": 0.05907, + "grad_norm": 0.7143236994743347, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 5907 + }, + { + "epoch": 0.05908, + "grad_norm": 0.846874475479126, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 5908 + }, + { + "epoch": 0.05909, + "grad_norm": 0.8068088889122009, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 5909 + }, + { + "epoch": 0.0591, + "grad_norm": 0.681771993637085, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 5910 + }, + { + "epoch": 0.05911, + "grad_norm": 0.7519658207893372, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 5911 + }, + { + "epoch": 0.05912, + "grad_norm": 0.7764798998832703, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 5912 + }, + { + "epoch": 0.05913, + "grad_norm": 0.757898211479187, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 5913 + }, + { + "epoch": 0.05914, + "grad_norm": 0.7228046655654907, + "learning_rate": 0.003, + "loss": 4.088, + "step": 5914 + }, + { + "epoch": 0.05915, + "grad_norm": 0.6196589469909668, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 5915 + }, + { + "epoch": 0.05916, + "grad_norm": 0.5313118696212769, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 5916 + }, + { + "epoch": 0.05917, + "grad_norm": 0.536577045917511, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 5917 + }, + { + "epoch": 0.05918, + "grad_norm": 0.5481544733047485, + "learning_rate": 0.003, + "loss": 4.059, + "step": 5918 + }, + { + "epoch": 0.05919, + "grad_norm": 0.6215094327926636, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 5919 + }, + { + "epoch": 0.0592, + "grad_norm": 0.6950171589851379, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 5920 + }, + { + "epoch": 0.05921, + "grad_norm": 0.7951127290725708, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 5921 + }, + { + "epoch": 0.05922, + "grad_norm": 0.9624848961830139, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 5922 + }, + { + "epoch": 0.05923, + "grad_norm": 1.0542023181915283, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 5923 + }, + { + "epoch": 0.05924, + "grad_norm": 0.8465194702148438, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 5924 + }, + { + "epoch": 0.05925, + "grad_norm": 0.7096172571182251, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 5925 + }, + { + "epoch": 0.05926, + "grad_norm": 0.7524508833885193, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 5926 + }, + { + "epoch": 0.05927, + "grad_norm": 0.6621628403663635, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 5927 + }, + { + "epoch": 0.05928, + "grad_norm": 0.5687286853790283, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 5928 + }, + { + "epoch": 0.05929, + "grad_norm": 0.4693851172924042, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 5929 + }, + { + "epoch": 0.0593, + "grad_norm": 0.4943394064903259, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 5930 + }, + { + "epoch": 0.05931, + "grad_norm": 0.5767983198165894, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 5931 + }, + { + "epoch": 0.05932, + "grad_norm": 0.7448903918266296, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 5932 + }, + { + "epoch": 0.05933, + "grad_norm": 0.9883130788803101, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 5933 + }, + { + "epoch": 0.05934, + "grad_norm": 1.1292710304260254, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 5934 + }, + { + "epoch": 0.05935, + "grad_norm": 0.6654629707336426, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 5935 + }, + { + "epoch": 0.05936, + "grad_norm": 0.7037827372550964, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 5936 + }, + { + "epoch": 0.05937, + "grad_norm": 0.928344190120697, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 5937 + }, + { + "epoch": 0.05938, + "grad_norm": 0.9594924449920654, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 5938 + }, + { + "epoch": 0.05939, + "grad_norm": 0.8684089183807373, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 5939 + }, + { + "epoch": 0.0594, + "grad_norm": 0.7445632815361023, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 5940 + }, + { + "epoch": 0.05941, + "grad_norm": 0.7455309629440308, + "learning_rate": 0.003, + "loss": 4.088, + "step": 5941 + }, + { + "epoch": 0.05942, + "grad_norm": 0.6535532474517822, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 5942 + }, + { + "epoch": 0.05943, + "grad_norm": 0.7083020210266113, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 5943 + }, + { + "epoch": 0.05944, + "grad_norm": 0.5781853795051575, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 5944 + }, + { + "epoch": 0.05945, + "grad_norm": 0.4758516550064087, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 5945 + }, + { + "epoch": 0.05946, + "grad_norm": 0.43648025393486023, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 5946 + }, + { + "epoch": 0.05947, + "grad_norm": 0.40230774879455566, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 5947 + }, + { + "epoch": 0.05948, + "grad_norm": 0.43578827381134033, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 5948 + }, + { + "epoch": 0.05949, + "grad_norm": 0.4666615426540375, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 5949 + }, + { + "epoch": 0.0595, + "grad_norm": 0.5350773334503174, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 5950 + }, + { + "epoch": 0.05951, + "grad_norm": 0.7080516815185547, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 5951 + }, + { + "epoch": 0.05952, + "grad_norm": 0.7631043791770935, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 5952 + }, + { + "epoch": 0.05953, + "grad_norm": 0.7321561574935913, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 5953 + }, + { + "epoch": 0.05954, + "grad_norm": 0.7224215269088745, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 5954 + }, + { + "epoch": 0.05955, + "grad_norm": 0.6791732311248779, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 5955 + }, + { + "epoch": 0.05956, + "grad_norm": 0.6897500157356262, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 5956 + }, + { + "epoch": 0.05957, + "grad_norm": 0.9236707091331482, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 5957 + }, + { + "epoch": 0.05958, + "grad_norm": 1.0184153318405151, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 5958 + }, + { + "epoch": 0.05959, + "grad_norm": 0.9222090840339661, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 5959 + }, + { + "epoch": 0.0596, + "grad_norm": 0.823390007019043, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 5960 + }, + { + "epoch": 0.05961, + "grad_norm": 0.85313880443573, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 5961 + }, + { + "epoch": 0.05962, + "grad_norm": 0.9738097786903381, + "learning_rate": 0.003, + "loss": 4.073, + "step": 5962 + }, + { + "epoch": 0.05963, + "grad_norm": 0.9151560664176941, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 5963 + }, + { + "epoch": 0.05964, + "grad_norm": 0.9065539240837097, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 5964 + }, + { + "epoch": 0.05965, + "grad_norm": 0.7944349050521851, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 5965 + }, + { + "epoch": 0.05966, + "grad_norm": 0.7729898691177368, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 5966 + }, + { + "epoch": 0.05967, + "grad_norm": 0.8545308113098145, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 5967 + }, + { + "epoch": 0.05968, + "grad_norm": 0.9394176006317139, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 5968 + }, + { + "epoch": 0.05969, + "grad_norm": 0.8269079327583313, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 5969 + }, + { + "epoch": 0.0597, + "grad_norm": 0.7771570086479187, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 5970 + }, + { + "epoch": 0.05971, + "grad_norm": 0.7587796449661255, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 5971 + }, + { + "epoch": 0.05972, + "grad_norm": 0.8089188933372498, + "learning_rate": 0.003, + "loss": 4.097, + "step": 5972 + }, + { + "epoch": 0.05973, + "grad_norm": 0.7736026048660278, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 5973 + }, + { + "epoch": 0.05974, + "grad_norm": 0.7009705305099487, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 5974 + }, + { + "epoch": 0.05975, + "grad_norm": 0.6236057281494141, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 5975 + }, + { + "epoch": 0.05976, + "grad_norm": 0.6456733345985413, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 5976 + }, + { + "epoch": 0.05977, + "grad_norm": 0.5764129757881165, + "learning_rate": 0.003, + "loss": 4.076, + "step": 5977 + }, + { + "epoch": 0.05978, + "grad_norm": 0.5077314972877502, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 5978 + }, + { + "epoch": 0.05979, + "grad_norm": 0.4247589111328125, + "learning_rate": 0.003, + "loss": 4.065, + "step": 5979 + }, + { + "epoch": 0.0598, + "grad_norm": 0.42051416635513306, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 5980 + }, + { + "epoch": 0.05981, + "grad_norm": 0.42619195580482483, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 5981 + }, + { + "epoch": 0.05982, + "grad_norm": 0.5673835873603821, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 5982 + }, + { + "epoch": 0.05983, + "grad_norm": 0.9145967364311218, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 5983 + }, + { + "epoch": 0.05984, + "grad_norm": 1.3005692958831787, + "learning_rate": 0.003, + "loss": 4.06, + "step": 5984 + }, + { + "epoch": 0.05985, + "grad_norm": 0.5140734910964966, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 5985 + }, + { + "epoch": 0.05986, + "grad_norm": 0.6605982184410095, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 5986 + }, + { + "epoch": 0.05987, + "grad_norm": 0.8068457841873169, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 5987 + }, + { + "epoch": 0.05988, + "grad_norm": 0.7576589584350586, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 5988 + }, + { + "epoch": 0.05989, + "grad_norm": 0.6234875321388245, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 5989 + }, + { + "epoch": 0.0599, + "grad_norm": 0.5436728596687317, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 5990 + }, + { + "epoch": 0.05991, + "grad_norm": 0.5909923911094666, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 5991 + }, + { + "epoch": 0.05992, + "grad_norm": 0.6115368008613586, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 5992 + }, + { + "epoch": 0.05993, + "grad_norm": 0.5838632583618164, + "learning_rate": 0.003, + "loss": 4.061, + "step": 5993 + }, + { + "epoch": 0.05994, + "grad_norm": 0.6138798594474792, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 5994 + }, + { + "epoch": 0.05995, + "grad_norm": 0.6672993302345276, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 5995 + }, + { + "epoch": 0.05996, + "grad_norm": 0.7187176942825317, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 5996 + }, + { + "epoch": 0.05997, + "grad_norm": 0.7012836933135986, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 5997 + }, + { + "epoch": 0.05998, + "grad_norm": 0.6116321086883545, + "learning_rate": 0.003, + "loss": 4.083, + "step": 5998 + }, + { + "epoch": 0.05999, + "grad_norm": 0.5162574648857117, + "learning_rate": 0.003, + "loss": 4.047, + "step": 5999 + }, + { + "epoch": 0.06, + "grad_norm": 0.6102678179740906, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 6000 + }, + { + "epoch": 0.06001, + "grad_norm": 0.5848944187164307, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 6001 + }, + { + "epoch": 0.06002, + "grad_norm": 0.5892003178596497, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 6002 + }, + { + "epoch": 0.06003, + "grad_norm": 0.5941994190216064, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 6003 + }, + { + "epoch": 0.06004, + "grad_norm": 0.6040821075439453, + "learning_rate": 0.003, + "loss": 4.041, + "step": 6004 + }, + { + "epoch": 0.06005, + "grad_norm": 0.6764587163925171, + "learning_rate": 0.003, + "loss": 4.06, + "step": 6005 + }, + { + "epoch": 0.06006, + "grad_norm": 0.689892590045929, + "learning_rate": 0.003, + "loss": 4.041, + "step": 6006 + }, + { + "epoch": 0.06007, + "grad_norm": 0.6645123958587646, + "learning_rate": 0.003, + "loss": 4.089, + "step": 6007 + }, + { + "epoch": 0.06008, + "grad_norm": 0.780821681022644, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 6008 + }, + { + "epoch": 0.06009, + "grad_norm": 1.0214614868164062, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 6009 + }, + { + "epoch": 0.0601, + "grad_norm": 1.1233888864517212, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 6010 + }, + { + "epoch": 0.06011, + "grad_norm": 0.8268139958381653, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 6011 + }, + { + "epoch": 0.06012, + "grad_norm": 0.8503698706626892, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 6012 + }, + { + "epoch": 0.06013, + "grad_norm": 0.9249141216278076, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 6013 + }, + { + "epoch": 0.06014, + "grad_norm": 1.1101765632629395, + "learning_rate": 0.003, + "loss": 4.076, + "step": 6014 + }, + { + "epoch": 0.06015, + "grad_norm": 1.0137884616851807, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 6015 + }, + { + "epoch": 0.06016, + "grad_norm": 0.7965986132621765, + "learning_rate": 0.003, + "loss": 4.089, + "step": 6016 + }, + { + "epoch": 0.06017, + "grad_norm": 0.7965149879455566, + "learning_rate": 0.003, + "loss": 4.092, + "step": 6017 + }, + { + "epoch": 0.06018, + "grad_norm": 0.8309733271598816, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 6018 + }, + { + "epoch": 0.06019, + "grad_norm": 0.768108606338501, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 6019 + }, + { + "epoch": 0.0602, + "grad_norm": 0.8096131086349487, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 6020 + }, + { + "epoch": 0.06021, + "grad_norm": 0.8794146776199341, + "learning_rate": 0.003, + "loss": 4.092, + "step": 6021 + }, + { + "epoch": 0.06022, + "grad_norm": 0.9591919183731079, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 6022 + }, + { + "epoch": 0.06023, + "grad_norm": 0.8065183162689209, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 6023 + }, + { + "epoch": 0.06024, + "grad_norm": 0.8542725443840027, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 6024 + }, + { + "epoch": 0.06025, + "grad_norm": 0.8307952284812927, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 6025 + }, + { + "epoch": 0.06026, + "grad_norm": 0.8478344678878784, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 6026 + }, + { + "epoch": 0.06027, + "grad_norm": 0.8286115527153015, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 6027 + }, + { + "epoch": 0.06028, + "grad_norm": 0.9184243083000183, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 6028 + }, + { + "epoch": 0.06029, + "grad_norm": 0.8957465887069702, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 6029 + }, + { + "epoch": 0.0603, + "grad_norm": 1.0010573863983154, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 6030 + }, + { + "epoch": 0.06031, + "grad_norm": 1.060820460319519, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 6031 + }, + { + "epoch": 0.06032, + "grad_norm": 0.9341599941253662, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 6032 + }, + { + "epoch": 0.06033, + "grad_norm": 0.7799586653709412, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 6033 + }, + { + "epoch": 0.06034, + "grad_norm": 0.6523258686065674, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 6034 + }, + { + "epoch": 0.06035, + "grad_norm": 0.4869402050971985, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 6035 + }, + { + "epoch": 0.06036, + "grad_norm": 0.5165879726409912, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 6036 + }, + { + "epoch": 0.06037, + "grad_norm": 0.5209839940071106, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 6037 + }, + { + "epoch": 0.06038, + "grad_norm": 0.5267767310142517, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 6038 + }, + { + "epoch": 0.06039, + "grad_norm": 0.5402923822402954, + "learning_rate": 0.003, + "loss": 4.038, + "step": 6039 + }, + { + "epoch": 0.0604, + "grad_norm": 0.5481833815574646, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 6040 + }, + { + "epoch": 0.06041, + "grad_norm": 0.5473806262016296, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 6041 + }, + { + "epoch": 0.06042, + "grad_norm": 0.5405387878417969, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 6042 + }, + { + "epoch": 0.06043, + "grad_norm": 0.5141624808311462, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 6043 + }, + { + "epoch": 0.06044, + "grad_norm": 0.4394950270652771, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 6044 + }, + { + "epoch": 0.06045, + "grad_norm": 0.49056580662727356, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 6045 + }, + { + "epoch": 0.06046, + "grad_norm": 0.5915701389312744, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 6046 + }, + { + "epoch": 0.06047, + "grad_norm": 0.7667026519775391, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 6047 + }, + { + "epoch": 0.06048, + "grad_norm": 0.9280398488044739, + "learning_rate": 0.003, + "loss": 4.046, + "step": 6048 + }, + { + "epoch": 0.06049, + "grad_norm": 0.8987364768981934, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 6049 + }, + { + "epoch": 0.0605, + "grad_norm": 0.6798592209815979, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 6050 + }, + { + "epoch": 0.06051, + "grad_norm": 0.5418064594268799, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 6051 + }, + { + "epoch": 0.06052, + "grad_norm": 0.6874141693115234, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 6052 + }, + { + "epoch": 0.06053, + "grad_norm": 0.7224387526512146, + "learning_rate": 0.003, + "loss": 4.034, + "step": 6053 + }, + { + "epoch": 0.06054, + "grad_norm": 0.6871939897537231, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 6054 + }, + { + "epoch": 0.06055, + "grad_norm": 0.5547940135002136, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 6055 + }, + { + "epoch": 0.06056, + "grad_norm": 0.46671125292778015, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 6056 + }, + { + "epoch": 0.06057, + "grad_norm": 0.5383384823799133, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 6057 + }, + { + "epoch": 0.06058, + "grad_norm": 0.5768425464630127, + "learning_rate": 0.003, + "loss": 4.046, + "step": 6058 + }, + { + "epoch": 0.06059, + "grad_norm": 0.6451521515846252, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 6059 + }, + { + "epoch": 0.0606, + "grad_norm": 0.7362764477729797, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 6060 + }, + { + "epoch": 0.06061, + "grad_norm": 0.7581610083580017, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 6061 + }, + { + "epoch": 0.06062, + "grad_norm": 0.7688402533531189, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 6062 + }, + { + "epoch": 0.06063, + "grad_norm": 0.7899247407913208, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 6063 + }, + { + "epoch": 0.06064, + "grad_norm": 0.6973926424980164, + "learning_rate": 0.003, + "loss": 4.089, + "step": 6064 + }, + { + "epoch": 0.06065, + "grad_norm": 0.6818444132804871, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 6065 + }, + { + "epoch": 0.06066, + "grad_norm": 0.7340236902236938, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 6066 + }, + { + "epoch": 0.06067, + "grad_norm": 0.8874813914299011, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 6067 + }, + { + "epoch": 0.06068, + "grad_norm": 0.9855170845985413, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 6068 + }, + { + "epoch": 0.06069, + "grad_norm": 1.1523756980895996, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 6069 + }, + { + "epoch": 0.0607, + "grad_norm": 0.873957097530365, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 6070 + }, + { + "epoch": 0.06071, + "grad_norm": 0.848626971244812, + "learning_rate": 0.003, + "loss": 4.052, + "step": 6071 + }, + { + "epoch": 0.06072, + "grad_norm": 0.830126166343689, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 6072 + }, + { + "epoch": 0.06073, + "grad_norm": 0.7734904289245605, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 6073 + }, + { + "epoch": 0.06074, + "grad_norm": 0.8461311459541321, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 6074 + }, + { + "epoch": 0.06075, + "grad_norm": 1.0410598516464233, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 6075 + }, + { + "epoch": 0.06076, + "grad_norm": 1.1896835565567017, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 6076 + }, + { + "epoch": 0.06077, + "grad_norm": 0.8349819779396057, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 6077 + }, + { + "epoch": 0.06078, + "grad_norm": 0.7226845026016235, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 6078 + }, + { + "epoch": 0.06079, + "grad_norm": 0.7609370350837708, + "learning_rate": 0.003, + "loss": 4.075, + "step": 6079 + }, + { + "epoch": 0.0608, + "grad_norm": 0.7053298354148865, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 6080 + }, + { + "epoch": 0.06081, + "grad_norm": 0.6266418099403381, + "learning_rate": 0.003, + "loss": 4.049, + "step": 6081 + }, + { + "epoch": 0.06082, + "grad_norm": 0.5934705138206482, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 6082 + }, + { + "epoch": 0.06083, + "grad_norm": 0.6792622804641724, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 6083 + }, + { + "epoch": 0.06084, + "grad_norm": 0.7161068916320801, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 6084 + }, + { + "epoch": 0.06085, + "grad_norm": 0.6832570433616638, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 6085 + }, + { + "epoch": 0.06086, + "grad_norm": 0.6206290125846863, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 6086 + }, + { + "epoch": 0.06087, + "grad_norm": 0.5676894783973694, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 6087 + }, + { + "epoch": 0.06088, + "grad_norm": 0.6253965497016907, + "learning_rate": 0.003, + "loss": 4.049, + "step": 6088 + }, + { + "epoch": 0.06089, + "grad_norm": 0.5663153529167175, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 6089 + }, + { + "epoch": 0.0609, + "grad_norm": 0.5186519026756287, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 6090 + }, + { + "epoch": 0.06091, + "grad_norm": 0.5059489011764526, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 6091 + }, + { + "epoch": 0.06092, + "grad_norm": 0.4492594599723816, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 6092 + }, + { + "epoch": 0.06093, + "grad_norm": 0.48771554231643677, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 6093 + }, + { + "epoch": 0.06094, + "grad_norm": 0.4371291399002075, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 6094 + }, + { + "epoch": 0.06095, + "grad_norm": 0.4618400037288666, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 6095 + }, + { + "epoch": 0.06096, + "grad_norm": 0.42540499567985535, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 6096 + }, + { + "epoch": 0.06097, + "grad_norm": 0.45694103837013245, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 6097 + }, + { + "epoch": 0.06098, + "grad_norm": 0.5427758693695068, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 6098 + }, + { + "epoch": 0.06099, + "grad_norm": 0.578697919845581, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 6099 + }, + { + "epoch": 0.061, + "grad_norm": 0.7686467170715332, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 6100 + }, + { + "epoch": 0.06101, + "grad_norm": 1.0454940795898438, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 6101 + }, + { + "epoch": 0.06102, + "grad_norm": 1.1773312091827393, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 6102 + }, + { + "epoch": 0.06103, + "grad_norm": 0.805872917175293, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 6103 + }, + { + "epoch": 0.06104, + "grad_norm": 0.7373183369636536, + "learning_rate": 0.003, + "loss": 4.055, + "step": 6104 + }, + { + "epoch": 0.06105, + "grad_norm": 0.6697652339935303, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 6105 + }, + { + "epoch": 0.06106, + "grad_norm": 0.6978040337562561, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 6106 + }, + { + "epoch": 0.06107, + "grad_norm": 0.6311635375022888, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 6107 + }, + { + "epoch": 0.06108, + "grad_norm": 0.6601365208625793, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 6108 + }, + { + "epoch": 0.06109, + "grad_norm": 0.5691725015640259, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 6109 + }, + { + "epoch": 0.0611, + "grad_norm": 0.5813072919845581, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 6110 + }, + { + "epoch": 0.06111, + "grad_norm": 0.5668128728866577, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 6111 + }, + { + "epoch": 0.06112, + "grad_norm": 0.5596557855606079, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 6112 + }, + { + "epoch": 0.06113, + "grad_norm": 0.6563438773155212, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 6113 + }, + { + "epoch": 0.06114, + "grad_norm": 0.7970550060272217, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 6114 + }, + { + "epoch": 0.06115, + "grad_norm": 0.8884240388870239, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 6115 + }, + { + "epoch": 0.06116, + "grad_norm": 1.06717848777771, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 6116 + }, + { + "epoch": 0.06117, + "grad_norm": 1.105543613433838, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 6117 + }, + { + "epoch": 0.06118, + "grad_norm": 0.8822982907295227, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 6118 + }, + { + "epoch": 0.06119, + "grad_norm": 0.8931971192359924, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 6119 + }, + { + "epoch": 0.0612, + "grad_norm": 0.9181088209152222, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 6120 + }, + { + "epoch": 0.06121, + "grad_norm": 0.8232288360595703, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 6121 + }, + { + "epoch": 0.06122, + "grad_norm": 0.662312924861908, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 6122 + }, + { + "epoch": 0.06123, + "grad_norm": 0.7112892270088196, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 6123 + }, + { + "epoch": 0.06124, + "grad_norm": 0.7430440783500671, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 6124 + }, + { + "epoch": 0.06125, + "grad_norm": 0.8577669262886047, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 6125 + }, + { + "epoch": 0.06126, + "grad_norm": 0.9018658399581909, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 6126 + }, + { + "epoch": 0.06127, + "grad_norm": 0.9262766242027283, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 6127 + }, + { + "epoch": 0.06128, + "grad_norm": 0.9801653027534485, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 6128 + }, + { + "epoch": 0.06129, + "grad_norm": 0.9130566120147705, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 6129 + }, + { + "epoch": 0.0613, + "grad_norm": 0.7119159698486328, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 6130 + }, + { + "epoch": 0.06131, + "grad_norm": 0.721723198890686, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 6131 + }, + { + "epoch": 0.06132, + "grad_norm": 0.6934512257575989, + "learning_rate": 0.003, + "loss": 4.073, + "step": 6132 + }, + { + "epoch": 0.06133, + "grad_norm": 0.8672846555709839, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 6133 + }, + { + "epoch": 0.06134, + "grad_norm": 1.000223159790039, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 6134 + }, + { + "epoch": 0.06135, + "grad_norm": 1.0614702701568604, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 6135 + }, + { + "epoch": 0.06136, + "grad_norm": 0.8382878303527832, + "learning_rate": 0.003, + "loss": 4.071, + "step": 6136 + }, + { + "epoch": 0.06137, + "grad_norm": 0.7506641745567322, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 6137 + }, + { + "epoch": 0.06138, + "grad_norm": 0.7175465822219849, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 6138 + }, + { + "epoch": 0.06139, + "grad_norm": 0.7327171564102173, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 6139 + }, + { + "epoch": 0.0614, + "grad_norm": 0.7524228096008301, + "learning_rate": 0.003, + "loss": 4.038, + "step": 6140 + }, + { + "epoch": 0.06141, + "grad_norm": 0.8113182783126831, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 6141 + }, + { + "epoch": 0.06142, + "grad_norm": 0.8624609708786011, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 6142 + }, + { + "epoch": 0.06143, + "grad_norm": 0.849567711353302, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 6143 + }, + { + "epoch": 0.06144, + "grad_norm": 0.8006253838539124, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 6144 + }, + { + "epoch": 0.06145, + "grad_norm": 0.6653422713279724, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 6145 + }, + { + "epoch": 0.06146, + "grad_norm": 0.5989100337028503, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 6146 + }, + { + "epoch": 0.06147, + "grad_norm": 0.6381635069847107, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 6147 + }, + { + "epoch": 0.06148, + "grad_norm": 0.6760680079460144, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 6148 + }, + { + "epoch": 0.06149, + "grad_norm": 0.7898040413856506, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 6149 + }, + { + "epoch": 0.0615, + "grad_norm": 0.8890254497528076, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 6150 + }, + { + "epoch": 0.06151, + "grad_norm": 0.7810690999031067, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 6151 + }, + { + "epoch": 0.06152, + "grad_norm": 0.7254818677902222, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 6152 + }, + { + "epoch": 0.06153, + "grad_norm": 0.7281543612480164, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 6153 + }, + { + "epoch": 0.06154, + "grad_norm": 0.7368656992912292, + "learning_rate": 0.003, + "loss": 4.026, + "step": 6154 + }, + { + "epoch": 0.06155, + "grad_norm": 0.6993001699447632, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 6155 + }, + { + "epoch": 0.06156, + "grad_norm": 0.5923306345939636, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 6156 + }, + { + "epoch": 0.06157, + "grad_norm": 0.45892608165740967, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 6157 + }, + { + "epoch": 0.06158, + "grad_norm": 0.41562631726264954, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 6158 + }, + { + "epoch": 0.06159, + "grad_norm": 0.5908848643302917, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 6159 + }, + { + "epoch": 0.0616, + "grad_norm": 0.7091723084449768, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 6160 + }, + { + "epoch": 0.06161, + "grad_norm": 0.7991834282875061, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 6161 + }, + { + "epoch": 0.06162, + "grad_norm": 0.8020153045654297, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 6162 + }, + { + "epoch": 0.06163, + "grad_norm": 0.6712086200714111, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 6163 + }, + { + "epoch": 0.06164, + "grad_norm": 0.6103070378303528, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 6164 + }, + { + "epoch": 0.06165, + "grad_norm": 0.7152314782142639, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 6165 + }, + { + "epoch": 0.06166, + "grad_norm": 0.8041547536849976, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 6166 + }, + { + "epoch": 0.06167, + "grad_norm": 0.7833403944969177, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 6167 + }, + { + "epoch": 0.06168, + "grad_norm": 0.7658514976501465, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 6168 + }, + { + "epoch": 0.06169, + "grad_norm": 0.6437884569168091, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 6169 + }, + { + "epoch": 0.0617, + "grad_norm": 0.5259288549423218, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 6170 + }, + { + "epoch": 0.06171, + "grad_norm": 0.6789013147354126, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 6171 + }, + { + "epoch": 0.06172, + "grad_norm": 0.6907355189323425, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 6172 + }, + { + "epoch": 0.06173, + "grad_norm": 0.8013864755630493, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 6173 + }, + { + "epoch": 0.06174, + "grad_norm": 0.7729394435882568, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 6174 + }, + { + "epoch": 0.06175, + "grad_norm": 0.6916416883468628, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 6175 + }, + { + "epoch": 0.06176, + "grad_norm": 0.7110108137130737, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 6176 + }, + { + "epoch": 0.06177, + "grad_norm": 0.7319554686546326, + "learning_rate": 0.003, + "loss": 4.058, + "step": 6177 + }, + { + "epoch": 0.06178, + "grad_norm": 0.7097826600074768, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 6178 + }, + { + "epoch": 0.06179, + "grad_norm": 0.6155682802200317, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 6179 + }, + { + "epoch": 0.0618, + "grad_norm": 0.6252792477607727, + "learning_rate": 0.003, + "loss": 4.059, + "step": 6180 + }, + { + "epoch": 0.06181, + "grad_norm": 0.6055827140808105, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 6181 + }, + { + "epoch": 0.06182, + "grad_norm": 0.6797299981117249, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 6182 + }, + { + "epoch": 0.06183, + "grad_norm": 0.740317702293396, + "learning_rate": 0.003, + "loss": 4.05, + "step": 6183 + }, + { + "epoch": 0.06184, + "grad_norm": 0.9291132092475891, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 6184 + }, + { + "epoch": 0.06185, + "grad_norm": 0.8949309587478638, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 6185 + }, + { + "epoch": 0.06186, + "grad_norm": 0.703565239906311, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 6186 + }, + { + "epoch": 0.06187, + "grad_norm": 0.7581156492233276, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 6187 + }, + { + "epoch": 0.06188, + "grad_norm": 0.7350788712501526, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 6188 + }, + { + "epoch": 0.06189, + "grad_norm": 0.7255872488021851, + "learning_rate": 0.003, + "loss": 4.045, + "step": 6189 + }, + { + "epoch": 0.0619, + "grad_norm": 0.8740925788879395, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 6190 + }, + { + "epoch": 0.06191, + "grad_norm": 1.0470556020736694, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 6191 + }, + { + "epoch": 0.06192, + "grad_norm": 0.9209544658660889, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 6192 + }, + { + "epoch": 0.06193, + "grad_norm": 0.8478721976280212, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 6193 + }, + { + "epoch": 0.06194, + "grad_norm": 0.9317265748977661, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 6194 + }, + { + "epoch": 0.06195, + "grad_norm": 0.90468829870224, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 6195 + }, + { + "epoch": 0.06196, + "grad_norm": 0.8252323269844055, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 6196 + }, + { + "epoch": 0.06197, + "grad_norm": 0.7629353404045105, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 6197 + }, + { + "epoch": 0.06198, + "grad_norm": 0.7873762845993042, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 6198 + }, + { + "epoch": 0.06199, + "grad_norm": 0.9207159876823425, + "learning_rate": 0.003, + "loss": 4.096, + "step": 6199 + }, + { + "epoch": 0.062, + "grad_norm": 0.9485105872154236, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 6200 + }, + { + "epoch": 0.06201, + "grad_norm": 0.8545713424682617, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 6201 + }, + { + "epoch": 0.06202, + "grad_norm": 0.7046213150024414, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 6202 + }, + { + "epoch": 0.06203, + "grad_norm": 0.7020225524902344, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 6203 + }, + { + "epoch": 0.06204, + "grad_norm": 0.7241159677505493, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 6204 + }, + { + "epoch": 0.06205, + "grad_norm": 0.6984315514564514, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 6205 + }, + { + "epoch": 0.06206, + "grad_norm": 0.7261056900024414, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 6206 + }, + { + "epoch": 0.06207, + "grad_norm": 0.7522352337837219, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 6207 + }, + { + "epoch": 0.06208, + "grad_norm": 0.7981951832771301, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 6208 + }, + { + "epoch": 0.06209, + "grad_norm": 0.9211035370826721, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 6209 + }, + { + "epoch": 0.0621, + "grad_norm": 1.0060456991195679, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 6210 + }, + { + "epoch": 0.06211, + "grad_norm": 0.993887722492218, + "learning_rate": 0.003, + "loss": 4.077, + "step": 6211 + }, + { + "epoch": 0.06212, + "grad_norm": 0.9443503618240356, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 6212 + }, + { + "epoch": 0.06213, + "grad_norm": 0.9290711879730225, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 6213 + }, + { + "epoch": 0.06214, + "grad_norm": 0.8242254257202148, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 6214 + }, + { + "epoch": 0.06215, + "grad_norm": 0.7361489534378052, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 6215 + }, + { + "epoch": 0.06216, + "grad_norm": 0.720862627029419, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 6216 + }, + { + "epoch": 0.06217, + "grad_norm": 0.612246572971344, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 6217 + }, + { + "epoch": 0.06218, + "grad_norm": 0.43744581937789917, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 6218 + }, + { + "epoch": 0.06219, + "grad_norm": 0.46168094873428345, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 6219 + }, + { + "epoch": 0.0622, + "grad_norm": 0.5252222418785095, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 6220 + }, + { + "epoch": 0.06221, + "grad_norm": 0.6057314276695251, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 6221 + }, + { + "epoch": 0.06222, + "grad_norm": 0.6545212268829346, + "learning_rate": 0.003, + "loss": 4.062, + "step": 6222 + }, + { + "epoch": 0.06223, + "grad_norm": 0.6369545459747314, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 6223 + }, + { + "epoch": 0.06224, + "grad_norm": 0.69230717420578, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 6224 + }, + { + "epoch": 0.06225, + "grad_norm": 0.7959716320037842, + "learning_rate": 0.003, + "loss": 4.067, + "step": 6225 + }, + { + "epoch": 0.06226, + "grad_norm": 0.9329782724380493, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 6226 + }, + { + "epoch": 0.06227, + "grad_norm": 0.9592517018318176, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 6227 + }, + { + "epoch": 0.06228, + "grad_norm": 0.8127028942108154, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 6228 + }, + { + "epoch": 0.06229, + "grad_norm": 0.5880432724952698, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 6229 + }, + { + "epoch": 0.0623, + "grad_norm": 0.5531348586082458, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 6230 + }, + { + "epoch": 0.06231, + "grad_norm": 0.6375777721405029, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 6231 + }, + { + "epoch": 0.06232, + "grad_norm": 0.6568375825881958, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 6232 + }, + { + "epoch": 0.06233, + "grad_norm": 0.6916172504425049, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 6233 + }, + { + "epoch": 0.06234, + "grad_norm": 0.6829954981803894, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 6234 + }, + { + "epoch": 0.06235, + "grad_norm": 0.6556631326675415, + "learning_rate": 0.003, + "loss": 4.051, + "step": 6235 + }, + { + "epoch": 0.06236, + "grad_norm": 0.6663537621498108, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 6236 + }, + { + "epoch": 0.06237, + "grad_norm": 0.6312961578369141, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 6237 + }, + { + "epoch": 0.06238, + "grad_norm": 0.5864769220352173, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 6238 + }, + { + "epoch": 0.06239, + "grad_norm": 0.607796311378479, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 6239 + }, + { + "epoch": 0.0624, + "grad_norm": 0.6244125962257385, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 6240 + }, + { + "epoch": 0.06241, + "grad_norm": 0.5208604335784912, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 6241 + }, + { + "epoch": 0.06242, + "grad_norm": 0.570060670375824, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 6242 + }, + { + "epoch": 0.06243, + "grad_norm": 0.5775870084762573, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 6243 + }, + { + "epoch": 0.06244, + "grad_norm": 0.6506668329238892, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 6244 + }, + { + "epoch": 0.06245, + "grad_norm": 0.7539433836936951, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 6245 + }, + { + "epoch": 0.06246, + "grad_norm": 0.9184788465499878, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 6246 + }, + { + "epoch": 0.06247, + "grad_norm": 1.0040006637573242, + "learning_rate": 0.003, + "loss": 4.066, + "step": 6247 + }, + { + "epoch": 0.06248, + "grad_norm": 0.9109286069869995, + "learning_rate": 0.003, + "loss": 4.064, + "step": 6248 + }, + { + "epoch": 0.06249, + "grad_norm": 0.7625318765640259, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 6249 + }, + { + "epoch": 0.0625, + "grad_norm": 0.6695582270622253, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 6250 + }, + { + "epoch": 0.06251, + "grad_norm": 0.6111177802085876, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 6251 + }, + { + "epoch": 0.06252, + "grad_norm": 0.6622089743614197, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 6252 + }, + { + "epoch": 0.06253, + "grad_norm": 0.6060909628868103, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 6253 + }, + { + "epoch": 0.06254, + "grad_norm": 0.5808566212654114, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 6254 + }, + { + "epoch": 0.06255, + "grad_norm": 0.5286803245544434, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 6255 + }, + { + "epoch": 0.06256, + "grad_norm": 0.5413640141487122, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 6256 + }, + { + "epoch": 0.06257, + "grad_norm": 0.6904173493385315, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 6257 + }, + { + "epoch": 0.06258, + "grad_norm": 0.7716007828712463, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 6258 + }, + { + "epoch": 0.06259, + "grad_norm": 0.9134862422943115, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 6259 + }, + { + "epoch": 0.0626, + "grad_norm": 0.9206618070602417, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 6260 + }, + { + "epoch": 0.06261, + "grad_norm": 0.776254415512085, + "learning_rate": 0.003, + "loss": 4.049, + "step": 6261 + }, + { + "epoch": 0.06262, + "grad_norm": 0.7170522212982178, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 6262 + }, + { + "epoch": 0.06263, + "grad_norm": 0.7091271281242371, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 6263 + }, + { + "epoch": 0.06264, + "grad_norm": 0.7496375441551208, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 6264 + }, + { + "epoch": 0.06265, + "grad_norm": 0.8993237018585205, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 6265 + }, + { + "epoch": 0.06266, + "grad_norm": 0.9477318525314331, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 6266 + }, + { + "epoch": 0.06267, + "grad_norm": 1.139681339263916, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 6267 + }, + { + "epoch": 0.06268, + "grad_norm": 1.0855191946029663, + "learning_rate": 0.003, + "loss": 4.055, + "step": 6268 + }, + { + "epoch": 0.06269, + "grad_norm": 0.8395910859107971, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 6269 + }, + { + "epoch": 0.0627, + "grad_norm": 0.9118239879608154, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 6270 + }, + { + "epoch": 0.06271, + "grad_norm": 0.8857484459877014, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 6271 + }, + { + "epoch": 0.06272, + "grad_norm": 0.8165472745895386, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 6272 + }, + { + "epoch": 0.06273, + "grad_norm": 0.8503358960151672, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 6273 + }, + { + "epoch": 0.06274, + "grad_norm": 0.9717170000076294, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 6274 + }, + { + "epoch": 0.06275, + "grad_norm": 0.9205477833747864, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 6275 + }, + { + "epoch": 0.06276, + "grad_norm": 0.8088390827178955, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 6276 + }, + { + "epoch": 0.06277, + "grad_norm": 0.6606585383415222, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 6277 + }, + { + "epoch": 0.06278, + "grad_norm": 0.5426492094993591, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 6278 + }, + { + "epoch": 0.06279, + "grad_norm": 0.5785556435585022, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 6279 + }, + { + "epoch": 0.0628, + "grad_norm": 0.6185020208358765, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 6280 + }, + { + "epoch": 0.06281, + "grad_norm": 0.671174943447113, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 6281 + }, + { + "epoch": 0.06282, + "grad_norm": 0.7437264323234558, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 6282 + }, + { + "epoch": 0.06283, + "grad_norm": 0.8574413657188416, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 6283 + }, + { + "epoch": 0.06284, + "grad_norm": 0.8529795408248901, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 6284 + }, + { + "epoch": 0.06285, + "grad_norm": 0.6762763261795044, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 6285 + }, + { + "epoch": 0.06286, + "grad_norm": 0.6096449494361877, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 6286 + }, + { + "epoch": 0.06287, + "grad_norm": 0.625961184501648, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 6287 + }, + { + "epoch": 0.06288, + "grad_norm": 0.5842055678367615, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 6288 + }, + { + "epoch": 0.06289, + "grad_norm": 0.5878425240516663, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 6289 + }, + { + "epoch": 0.0629, + "grad_norm": 0.5359255075454712, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 6290 + }, + { + "epoch": 0.06291, + "grad_norm": 0.5622683167457581, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 6291 + }, + { + "epoch": 0.06292, + "grad_norm": 0.6570186614990234, + "learning_rate": 0.003, + "loss": 4.05, + "step": 6292 + }, + { + "epoch": 0.06293, + "grad_norm": 0.709516704082489, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 6293 + }, + { + "epoch": 0.06294, + "grad_norm": 0.709321141242981, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 6294 + }, + { + "epoch": 0.06295, + "grad_norm": 0.702165424823761, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 6295 + }, + { + "epoch": 0.06296, + "grad_norm": 0.5981850028038025, + "learning_rate": 0.003, + "loss": 4.079, + "step": 6296 + }, + { + "epoch": 0.06297, + "grad_norm": 0.5463345050811768, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 6297 + }, + { + "epoch": 0.06298, + "grad_norm": 0.5999792218208313, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 6298 + }, + { + "epoch": 0.06299, + "grad_norm": 0.7810006737709045, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 6299 + }, + { + "epoch": 0.063, + "grad_norm": 0.8830216526985168, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 6300 + }, + { + "epoch": 0.06301, + "grad_norm": 0.9055691361427307, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 6301 + }, + { + "epoch": 0.06302, + "grad_norm": 0.9385949969291687, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 6302 + }, + { + "epoch": 0.06303, + "grad_norm": 0.7440510988235474, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 6303 + }, + { + "epoch": 0.06304, + "grad_norm": 0.6754160523414612, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 6304 + }, + { + "epoch": 0.06305, + "grad_norm": 0.7607747316360474, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 6305 + }, + { + "epoch": 0.06306, + "grad_norm": 0.8998769521713257, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 6306 + }, + { + "epoch": 0.06307, + "grad_norm": 0.9299960136413574, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 6307 + }, + { + "epoch": 0.06308, + "grad_norm": 0.8129688501358032, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 6308 + }, + { + "epoch": 0.06309, + "grad_norm": 0.7249149084091187, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 6309 + }, + { + "epoch": 0.0631, + "grad_norm": 0.827510416507721, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 6310 + }, + { + "epoch": 0.06311, + "grad_norm": 0.8704838752746582, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 6311 + }, + { + "epoch": 0.06312, + "grad_norm": 0.8416996598243713, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 6312 + }, + { + "epoch": 0.06313, + "grad_norm": 0.7403684854507446, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 6313 + }, + { + "epoch": 0.06314, + "grad_norm": 0.6789849400520325, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 6314 + }, + { + "epoch": 0.06315, + "grad_norm": 0.6259597539901733, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 6315 + }, + { + "epoch": 0.06316, + "grad_norm": 0.5615050196647644, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 6316 + }, + { + "epoch": 0.06317, + "grad_norm": 0.6137986183166504, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 6317 + }, + { + "epoch": 0.06318, + "grad_norm": 0.7117832899093628, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 6318 + }, + { + "epoch": 0.06319, + "grad_norm": 0.9356580972671509, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 6319 + }, + { + "epoch": 0.0632, + "grad_norm": 1.1936414241790771, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 6320 + }, + { + "epoch": 0.06321, + "grad_norm": 0.993614673614502, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 6321 + }, + { + "epoch": 0.06322, + "grad_norm": 0.857085645198822, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 6322 + }, + { + "epoch": 0.06323, + "grad_norm": 0.8355134725570679, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 6323 + }, + { + "epoch": 0.06324, + "grad_norm": 0.8579593896865845, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 6324 + }, + { + "epoch": 0.06325, + "grad_norm": 0.8484706878662109, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 6325 + }, + { + "epoch": 0.06326, + "grad_norm": 0.689569354057312, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 6326 + }, + { + "epoch": 0.06327, + "grad_norm": 0.5412641763687134, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 6327 + }, + { + "epoch": 0.06328, + "grad_norm": 0.5388414859771729, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 6328 + }, + { + "epoch": 0.06329, + "grad_norm": 0.5315625071525574, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 6329 + }, + { + "epoch": 0.0633, + "grad_norm": 0.4952152669429779, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 6330 + }, + { + "epoch": 0.06331, + "grad_norm": 0.5569480657577515, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 6331 + }, + { + "epoch": 0.06332, + "grad_norm": 0.5562824606895447, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 6332 + }, + { + "epoch": 0.06333, + "grad_norm": 0.6068150997161865, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 6333 + }, + { + "epoch": 0.06334, + "grad_norm": 0.7447565197944641, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 6334 + }, + { + "epoch": 0.06335, + "grad_norm": 0.7546894550323486, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 6335 + }, + { + "epoch": 0.06336, + "grad_norm": 0.8086697459220886, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 6336 + }, + { + "epoch": 0.06337, + "grad_norm": 0.8379433751106262, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 6337 + }, + { + "epoch": 0.06338, + "grad_norm": 0.8580986857414246, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 6338 + }, + { + "epoch": 0.06339, + "grad_norm": 0.7837094664573669, + "learning_rate": 0.003, + "loss": 4.041, + "step": 6339 + }, + { + "epoch": 0.0634, + "grad_norm": 0.6586499810218811, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 6340 + }, + { + "epoch": 0.06341, + "grad_norm": 0.6290658116340637, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 6341 + }, + { + "epoch": 0.06342, + "grad_norm": 0.6797477006912231, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 6342 + }, + { + "epoch": 0.06343, + "grad_norm": 0.64532470703125, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 6343 + }, + { + "epoch": 0.06344, + "grad_norm": 0.5823304653167725, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 6344 + }, + { + "epoch": 0.06345, + "grad_norm": 0.6109977960586548, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 6345 + }, + { + "epoch": 0.06346, + "grad_norm": 0.72659832239151, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 6346 + }, + { + "epoch": 0.06347, + "grad_norm": 0.9619790315628052, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 6347 + }, + { + "epoch": 0.06348, + "grad_norm": 0.9899624586105347, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 6348 + }, + { + "epoch": 0.06349, + "grad_norm": 0.9189885854721069, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 6349 + }, + { + "epoch": 0.0635, + "grad_norm": 0.8876418471336365, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 6350 + }, + { + "epoch": 0.06351, + "grad_norm": 0.7306442260742188, + "learning_rate": 0.003, + "loss": 4.071, + "step": 6351 + }, + { + "epoch": 0.06352, + "grad_norm": 0.6585521697998047, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 6352 + }, + { + "epoch": 0.06353, + "grad_norm": 0.5543274879455566, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 6353 + }, + { + "epoch": 0.06354, + "grad_norm": 0.5330720543861389, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 6354 + }, + { + "epoch": 0.06355, + "grad_norm": 0.5855516791343689, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 6355 + }, + { + "epoch": 0.06356, + "grad_norm": 0.5944929122924805, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 6356 + }, + { + "epoch": 0.06357, + "grad_norm": 0.5652433633804321, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 6357 + }, + { + "epoch": 0.06358, + "grad_norm": 0.5926536917686462, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 6358 + }, + { + "epoch": 0.06359, + "grad_norm": 0.6881373524665833, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 6359 + }, + { + "epoch": 0.0636, + "grad_norm": 0.8648481369018555, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 6360 + }, + { + "epoch": 0.06361, + "grad_norm": 0.8661714196205139, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 6361 + }, + { + "epoch": 0.06362, + "grad_norm": 0.7547315955162048, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 6362 + }, + { + "epoch": 0.06363, + "grad_norm": 0.6218990683555603, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 6363 + }, + { + "epoch": 0.06364, + "grad_norm": 0.6733616590499878, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 6364 + }, + { + "epoch": 0.06365, + "grad_norm": 0.848198652267456, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 6365 + }, + { + "epoch": 0.06366, + "grad_norm": 1.0001298189163208, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 6366 + }, + { + "epoch": 0.06367, + "grad_norm": 1.0065937042236328, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 6367 + }, + { + "epoch": 0.06368, + "grad_norm": 0.7900940179824829, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 6368 + }, + { + "epoch": 0.06369, + "grad_norm": 0.6981412768363953, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 6369 + }, + { + "epoch": 0.0637, + "grad_norm": 0.7063578367233276, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 6370 + }, + { + "epoch": 0.06371, + "grad_norm": 0.7151117920875549, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 6371 + }, + { + "epoch": 0.06372, + "grad_norm": 0.6243462562561035, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 6372 + }, + { + "epoch": 0.06373, + "grad_norm": 0.6759061813354492, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 6373 + }, + { + "epoch": 0.06374, + "grad_norm": 0.7649587988853455, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 6374 + }, + { + "epoch": 0.06375, + "grad_norm": 0.8787922859191895, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 6375 + }, + { + "epoch": 0.06376, + "grad_norm": 0.8679553866386414, + "learning_rate": 0.003, + "loss": 4.067, + "step": 6376 + }, + { + "epoch": 0.06377, + "grad_norm": 0.7838705778121948, + "learning_rate": 0.003, + "loss": 4.036, + "step": 6377 + }, + { + "epoch": 0.06378, + "grad_norm": 0.6770272850990295, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 6378 + }, + { + "epoch": 0.06379, + "grad_norm": 0.7824772000312805, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 6379 + }, + { + "epoch": 0.0638, + "grad_norm": 0.9476855993270874, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 6380 + }, + { + "epoch": 0.06381, + "grad_norm": 0.968838095664978, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 6381 + }, + { + "epoch": 0.06382, + "grad_norm": 0.7842888832092285, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 6382 + }, + { + "epoch": 0.06383, + "grad_norm": 0.6376921534538269, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 6383 + }, + { + "epoch": 0.06384, + "grad_norm": 0.5353909134864807, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 6384 + }, + { + "epoch": 0.06385, + "grad_norm": 0.5485050082206726, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 6385 + }, + { + "epoch": 0.06386, + "grad_norm": 0.6523105502128601, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 6386 + }, + { + "epoch": 0.06387, + "grad_norm": 0.7976499795913696, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 6387 + }, + { + "epoch": 0.06388, + "grad_norm": 0.8359301686286926, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 6388 + }, + { + "epoch": 0.06389, + "grad_norm": 0.7813435792922974, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 6389 + }, + { + "epoch": 0.0639, + "grad_norm": 0.7872916460037231, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 6390 + }, + { + "epoch": 0.06391, + "grad_norm": 0.7407468557357788, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 6391 + }, + { + "epoch": 0.06392, + "grad_norm": 0.8617271184921265, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 6392 + }, + { + "epoch": 0.06393, + "grad_norm": 0.8929271101951599, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 6393 + }, + { + "epoch": 0.06394, + "grad_norm": 0.8621600270271301, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 6394 + }, + { + "epoch": 0.06395, + "grad_norm": 0.8182836771011353, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 6395 + }, + { + "epoch": 0.06396, + "grad_norm": 0.644528329372406, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 6396 + }, + { + "epoch": 0.06397, + "grad_norm": 0.5910365581512451, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 6397 + }, + { + "epoch": 0.06398, + "grad_norm": 0.6243775486946106, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 6398 + }, + { + "epoch": 0.06399, + "grad_norm": 0.6281679272651672, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 6399 + }, + { + "epoch": 0.064, + "grad_norm": 0.7241624593734741, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 6400 + }, + { + "epoch": 0.06401, + "grad_norm": 0.8978641033172607, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 6401 + }, + { + "epoch": 0.06402, + "grad_norm": 0.8401793837547302, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 6402 + }, + { + "epoch": 0.06403, + "grad_norm": 0.704843282699585, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 6403 + }, + { + "epoch": 0.06404, + "grad_norm": 0.7263625264167786, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 6404 + }, + { + "epoch": 0.06405, + "grad_norm": 0.7337961196899414, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 6405 + }, + { + "epoch": 0.06406, + "grad_norm": 0.729682445526123, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 6406 + }, + { + "epoch": 0.06407, + "grad_norm": 0.7377498149871826, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 6407 + }, + { + "epoch": 0.06408, + "grad_norm": 0.7378864884376526, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 6408 + }, + { + "epoch": 0.06409, + "grad_norm": 0.6734524369239807, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 6409 + }, + { + "epoch": 0.0641, + "grad_norm": 0.5793408751487732, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 6410 + }, + { + "epoch": 0.06411, + "grad_norm": 0.5833590626716614, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 6411 + }, + { + "epoch": 0.06412, + "grad_norm": 0.6013709902763367, + "learning_rate": 0.003, + "loss": 4.044, + "step": 6412 + }, + { + "epoch": 0.06413, + "grad_norm": 0.7732909321784973, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 6413 + }, + { + "epoch": 0.06414, + "grad_norm": 0.8160001039505005, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 6414 + }, + { + "epoch": 0.06415, + "grad_norm": 0.7408420443534851, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 6415 + }, + { + "epoch": 0.06416, + "grad_norm": 0.7382277846336365, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 6416 + }, + { + "epoch": 0.06417, + "grad_norm": 0.7491039037704468, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 6417 + }, + { + "epoch": 0.06418, + "grad_norm": 0.8163646459579468, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 6418 + }, + { + "epoch": 0.06419, + "grad_norm": 0.7435489296913147, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 6419 + }, + { + "epoch": 0.0642, + "grad_norm": 0.604588508605957, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 6420 + }, + { + "epoch": 0.06421, + "grad_norm": 0.532421350479126, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 6421 + }, + { + "epoch": 0.06422, + "grad_norm": 0.5848860144615173, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 6422 + }, + { + "epoch": 0.06423, + "grad_norm": 0.6077694296836853, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 6423 + }, + { + "epoch": 0.06424, + "grad_norm": 0.5964341163635254, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 6424 + }, + { + "epoch": 0.06425, + "grad_norm": 0.5771605968475342, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 6425 + }, + { + "epoch": 0.06426, + "grad_norm": 0.5373570919036865, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 6426 + }, + { + "epoch": 0.06427, + "grad_norm": 0.417835533618927, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 6427 + }, + { + "epoch": 0.06428, + "grad_norm": 0.4400598406791687, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 6428 + }, + { + "epoch": 0.06429, + "grad_norm": 0.437053918838501, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 6429 + }, + { + "epoch": 0.0643, + "grad_norm": 0.5374695658683777, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 6430 + }, + { + "epoch": 0.06431, + "grad_norm": 0.7007521390914917, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 6431 + }, + { + "epoch": 0.06432, + "grad_norm": 0.9336084723472595, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 6432 + }, + { + "epoch": 0.06433, + "grad_norm": 1.0574723482131958, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 6433 + }, + { + "epoch": 0.06434, + "grad_norm": 0.8598960041999817, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 6434 + }, + { + "epoch": 0.06435, + "grad_norm": 0.8860291838645935, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 6435 + }, + { + "epoch": 0.06436, + "grad_norm": 0.9345943927764893, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 6436 + }, + { + "epoch": 0.06437, + "grad_norm": 1.1661456823349, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 6437 + }, + { + "epoch": 0.06438, + "grad_norm": 1.0386587381362915, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 6438 + }, + { + "epoch": 0.06439, + "grad_norm": 1.1419264078140259, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 6439 + }, + { + "epoch": 0.0644, + "grad_norm": 1.1448736190795898, + "learning_rate": 0.003, + "loss": 4.077, + "step": 6440 + }, + { + "epoch": 0.06441, + "grad_norm": 1.128818154335022, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 6441 + }, + { + "epoch": 0.06442, + "grad_norm": 0.9655861854553223, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 6442 + }, + { + "epoch": 0.06443, + "grad_norm": 0.9556558132171631, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 6443 + }, + { + "epoch": 0.06444, + "grad_norm": 0.9255332350730896, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 6444 + }, + { + "epoch": 0.06445, + "grad_norm": 1.0121527910232544, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 6445 + }, + { + "epoch": 0.06446, + "grad_norm": 0.8829848170280457, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 6446 + }, + { + "epoch": 0.06447, + "grad_norm": 0.8952497839927673, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 6447 + }, + { + "epoch": 0.06448, + "grad_norm": 0.9387702345848083, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 6448 + }, + { + "epoch": 0.06449, + "grad_norm": 0.8153852224349976, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 6449 + }, + { + "epoch": 0.0645, + "grad_norm": 0.814096212387085, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 6450 + }, + { + "epoch": 0.06451, + "grad_norm": 0.7601085901260376, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 6451 + }, + { + "epoch": 0.06452, + "grad_norm": 0.7431148886680603, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 6452 + }, + { + "epoch": 0.06453, + "grad_norm": 0.7803151607513428, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 6453 + }, + { + "epoch": 0.06454, + "grad_norm": 0.7259302735328674, + "learning_rate": 0.003, + "loss": 4.061, + "step": 6454 + }, + { + "epoch": 0.06455, + "grad_norm": 0.6496488451957703, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 6455 + }, + { + "epoch": 0.06456, + "grad_norm": 0.719208836555481, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 6456 + }, + { + "epoch": 0.06457, + "grad_norm": 0.8120383620262146, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 6457 + }, + { + "epoch": 0.06458, + "grad_norm": 0.9060261845588684, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 6458 + }, + { + "epoch": 0.06459, + "grad_norm": 0.9373885989189148, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 6459 + }, + { + "epoch": 0.0646, + "grad_norm": 0.9116479158401489, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 6460 + }, + { + "epoch": 0.06461, + "grad_norm": 0.7550241947174072, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 6461 + }, + { + "epoch": 0.06462, + "grad_norm": 0.5618889927864075, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 6462 + }, + { + "epoch": 0.06463, + "grad_norm": 0.5237634181976318, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 6463 + }, + { + "epoch": 0.06464, + "grad_norm": 0.5218420624732971, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 6464 + }, + { + "epoch": 0.06465, + "grad_norm": 0.6775352954864502, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 6465 + }, + { + "epoch": 0.06466, + "grad_norm": 0.7687344551086426, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 6466 + }, + { + "epoch": 0.06467, + "grad_norm": 0.7261245250701904, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 6467 + }, + { + "epoch": 0.06468, + "grad_norm": 0.5774818062782288, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 6468 + }, + { + "epoch": 0.06469, + "grad_norm": 0.49446791410446167, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 6469 + }, + { + "epoch": 0.0647, + "grad_norm": 0.5678610801696777, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 6470 + }, + { + "epoch": 0.06471, + "grad_norm": 0.586567223072052, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 6471 + }, + { + "epoch": 0.06472, + "grad_norm": 0.5532718896865845, + "learning_rate": 0.003, + "loss": 4.057, + "step": 6472 + }, + { + "epoch": 0.06473, + "grad_norm": 0.5075469613075256, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 6473 + }, + { + "epoch": 0.06474, + "grad_norm": 0.4478394389152527, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 6474 + }, + { + "epoch": 0.06475, + "grad_norm": 0.5158814191818237, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 6475 + }, + { + "epoch": 0.06476, + "grad_norm": 0.5592943429946899, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 6476 + }, + { + "epoch": 0.06477, + "grad_norm": 0.6047767400741577, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 6477 + }, + { + "epoch": 0.06478, + "grad_norm": 0.637370765209198, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 6478 + }, + { + "epoch": 0.06479, + "grad_norm": 0.6284873485565186, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 6479 + }, + { + "epoch": 0.0648, + "grad_norm": 0.7605395913124084, + "learning_rate": 0.003, + "loss": 4.037, + "step": 6480 + }, + { + "epoch": 0.06481, + "grad_norm": 0.7899333238601685, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 6481 + }, + { + "epoch": 0.06482, + "grad_norm": 0.8250449299812317, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 6482 + }, + { + "epoch": 0.06483, + "grad_norm": 0.7915768623352051, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 6483 + }, + { + "epoch": 0.06484, + "grad_norm": 0.7572276592254639, + "learning_rate": 0.003, + "loss": 4.054, + "step": 6484 + }, + { + "epoch": 0.06485, + "grad_norm": 0.75547856092453, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 6485 + }, + { + "epoch": 0.06486, + "grad_norm": 0.7201529145240784, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 6486 + }, + { + "epoch": 0.06487, + "grad_norm": 0.6418141722679138, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 6487 + }, + { + "epoch": 0.06488, + "grad_norm": 0.5657222270965576, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 6488 + }, + { + "epoch": 0.06489, + "grad_norm": 0.5981627106666565, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 6489 + }, + { + "epoch": 0.0649, + "grad_norm": 0.6097392439842224, + "learning_rate": 0.003, + "loss": 4.042, + "step": 6490 + }, + { + "epoch": 0.06491, + "grad_norm": 0.6370054483413696, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 6491 + }, + { + "epoch": 0.06492, + "grad_norm": 0.6754859685897827, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 6492 + }, + { + "epoch": 0.06493, + "grad_norm": 0.6854249835014343, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 6493 + }, + { + "epoch": 0.06494, + "grad_norm": 0.6282760500907898, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 6494 + }, + { + "epoch": 0.06495, + "grad_norm": 0.5374482274055481, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 6495 + }, + { + "epoch": 0.06496, + "grad_norm": 0.5668441653251648, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 6496 + }, + { + "epoch": 0.06497, + "grad_norm": 0.5741870403289795, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 6497 + }, + { + "epoch": 0.06498, + "grad_norm": 0.6695168018341064, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 6498 + }, + { + "epoch": 0.06499, + "grad_norm": 0.7330256104469299, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 6499 + }, + { + "epoch": 0.065, + "grad_norm": 0.9226296544075012, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 6500 + }, + { + "epoch": 0.06501, + "grad_norm": 1.1895571947097778, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 6501 + }, + { + "epoch": 0.06502, + "grad_norm": 0.9606831073760986, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 6502 + }, + { + "epoch": 0.06503, + "grad_norm": 0.7914904952049255, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 6503 + }, + { + "epoch": 0.06504, + "grad_norm": 0.8200267553329468, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 6504 + }, + { + "epoch": 0.06505, + "grad_norm": 0.8769012093544006, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 6505 + }, + { + "epoch": 0.06506, + "grad_norm": 1.0376636981964111, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 6506 + }, + { + "epoch": 0.06507, + "grad_norm": 1.039513349533081, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 6507 + }, + { + "epoch": 0.06508, + "grad_norm": 0.9043866395950317, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 6508 + }, + { + "epoch": 0.06509, + "grad_norm": 0.9053654670715332, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 6509 + }, + { + "epoch": 0.0651, + "grad_norm": 0.7699447274208069, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 6510 + }, + { + "epoch": 0.06511, + "grad_norm": 0.7321429252624512, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 6511 + }, + { + "epoch": 0.06512, + "grad_norm": 0.7014932036399841, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 6512 + }, + { + "epoch": 0.06513, + "grad_norm": 0.8323968648910522, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 6513 + }, + { + "epoch": 0.06514, + "grad_norm": 0.9952423572540283, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 6514 + }, + { + "epoch": 0.06515, + "grad_norm": 1.039555311203003, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 6515 + }, + { + "epoch": 0.06516, + "grad_norm": 0.7747417688369751, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 6516 + }, + { + "epoch": 0.06517, + "grad_norm": 0.6039211750030518, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 6517 + }, + { + "epoch": 0.06518, + "grad_norm": 0.619925856590271, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 6518 + }, + { + "epoch": 0.06519, + "grad_norm": 0.635144054889679, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 6519 + }, + { + "epoch": 0.0652, + "grad_norm": 0.5885560512542725, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 6520 + }, + { + "epoch": 0.06521, + "grad_norm": 0.5701271891593933, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 6521 + }, + { + "epoch": 0.06522, + "grad_norm": 0.6156445145606995, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 6522 + }, + { + "epoch": 0.06523, + "grad_norm": 0.5962189435958862, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 6523 + }, + { + "epoch": 0.06524, + "grad_norm": 0.5860252380371094, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 6524 + }, + { + "epoch": 0.06525, + "grad_norm": 0.5954103469848633, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 6525 + }, + { + "epoch": 0.06526, + "grad_norm": 0.5705258250236511, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 6526 + }, + { + "epoch": 0.06527, + "grad_norm": 0.6064693927764893, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 6527 + }, + { + "epoch": 0.06528, + "grad_norm": 0.6207301020622253, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 6528 + }, + { + "epoch": 0.06529, + "grad_norm": 0.5855225920677185, + "learning_rate": 0.003, + "loss": 4.049, + "step": 6529 + }, + { + "epoch": 0.0653, + "grad_norm": 0.6270435452461243, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 6530 + }, + { + "epoch": 0.06531, + "grad_norm": 0.6729288697242737, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 6531 + }, + { + "epoch": 0.06532, + "grad_norm": 0.8055482506752014, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 6532 + }, + { + "epoch": 0.06533, + "grad_norm": 0.8493422865867615, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 6533 + }, + { + "epoch": 0.06534, + "grad_norm": 1.0013922452926636, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 6534 + }, + { + "epoch": 0.06535, + "grad_norm": 1.09293794631958, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 6535 + }, + { + "epoch": 0.06536, + "grad_norm": 0.7350423336029053, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 6536 + }, + { + "epoch": 0.06537, + "grad_norm": 0.589043915271759, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 6537 + }, + { + "epoch": 0.06538, + "grad_norm": 0.6554220914840698, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 6538 + }, + { + "epoch": 0.06539, + "grad_norm": 0.6359267830848694, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 6539 + }, + { + "epoch": 0.0654, + "grad_norm": 0.6657839417457581, + "learning_rate": 0.003, + "loss": 4.089, + "step": 6540 + }, + { + "epoch": 0.06541, + "grad_norm": 0.6628341674804688, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 6541 + }, + { + "epoch": 0.06542, + "grad_norm": 0.6489399671554565, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 6542 + }, + { + "epoch": 0.06543, + "grad_norm": 0.6529117226600647, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 6543 + }, + { + "epoch": 0.06544, + "grad_norm": 0.6327089667320251, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 6544 + }, + { + "epoch": 0.06545, + "grad_norm": 0.6792625784873962, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 6545 + }, + { + "epoch": 0.06546, + "grad_norm": 0.6849966645240784, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 6546 + }, + { + "epoch": 0.06547, + "grad_norm": 0.673355758190155, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 6547 + }, + { + "epoch": 0.06548, + "grad_norm": 0.5450528860092163, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 6548 + }, + { + "epoch": 0.06549, + "grad_norm": 0.519288957118988, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 6549 + }, + { + "epoch": 0.0655, + "grad_norm": 0.5287633538246155, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 6550 + }, + { + "epoch": 0.06551, + "grad_norm": 0.56468665599823, + "learning_rate": 0.003, + "loss": 4.027, + "step": 6551 + }, + { + "epoch": 0.06552, + "grad_norm": 0.6360222101211548, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 6552 + }, + { + "epoch": 0.06553, + "grad_norm": 0.6669138073921204, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 6553 + }, + { + "epoch": 0.06554, + "grad_norm": 0.7647295594215393, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 6554 + }, + { + "epoch": 0.06555, + "grad_norm": 0.9226347208023071, + "learning_rate": 0.003, + "loss": 4.052, + "step": 6555 + }, + { + "epoch": 0.06556, + "grad_norm": 0.9914687871932983, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 6556 + }, + { + "epoch": 0.06557, + "grad_norm": 1.0823801755905151, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 6557 + }, + { + "epoch": 0.06558, + "grad_norm": 0.888142466545105, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 6558 + }, + { + "epoch": 0.06559, + "grad_norm": 0.8824571371078491, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 6559 + }, + { + "epoch": 0.0656, + "grad_norm": 0.9318628907203674, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 6560 + }, + { + "epoch": 0.06561, + "grad_norm": 1.0599497556686401, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 6561 + }, + { + "epoch": 0.06562, + "grad_norm": 0.9844862818717957, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 6562 + }, + { + "epoch": 0.06563, + "grad_norm": 0.9434862732887268, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 6563 + }, + { + "epoch": 0.06564, + "grad_norm": 0.8977039456367493, + "learning_rate": 0.003, + "loss": 4.08, + "step": 6564 + }, + { + "epoch": 0.06565, + "grad_norm": 1.0686181783676147, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 6565 + }, + { + "epoch": 0.06566, + "grad_norm": 0.949355959892273, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 6566 + }, + { + "epoch": 0.06567, + "grad_norm": 1.2126977443695068, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 6567 + }, + { + "epoch": 0.06568, + "grad_norm": 0.8228438496589661, + "learning_rate": 0.003, + "loss": 4.101, + "step": 6568 + }, + { + "epoch": 0.06569, + "grad_norm": 0.7816138863563538, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 6569 + }, + { + "epoch": 0.0657, + "grad_norm": 0.7952476143836975, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 6570 + }, + { + "epoch": 0.06571, + "grad_norm": 0.8013068437576294, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 6571 + }, + { + "epoch": 0.06572, + "grad_norm": 0.7705895304679871, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 6572 + }, + { + "epoch": 0.06573, + "grad_norm": 0.9079228043556213, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 6573 + }, + { + "epoch": 0.06574, + "grad_norm": 0.9015453457832336, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 6574 + }, + { + "epoch": 0.06575, + "grad_norm": 0.8765079975128174, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 6575 + }, + { + "epoch": 0.06576, + "grad_norm": 0.9798871874809265, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 6576 + }, + { + "epoch": 0.06577, + "grad_norm": 0.9194977283477783, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 6577 + }, + { + "epoch": 0.06578, + "grad_norm": 0.8564810752868652, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 6578 + }, + { + "epoch": 0.06579, + "grad_norm": 0.9468756318092346, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 6579 + }, + { + "epoch": 0.0658, + "grad_norm": 1.0715759992599487, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 6580 + }, + { + "epoch": 0.06581, + "grad_norm": 1.0081336498260498, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 6581 + }, + { + "epoch": 0.06582, + "grad_norm": 1.1618338823318481, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 6582 + }, + { + "epoch": 0.06583, + "grad_norm": 0.7572551369667053, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 6583 + }, + { + "epoch": 0.06584, + "grad_norm": 0.699311375617981, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 6584 + }, + { + "epoch": 0.06585, + "grad_norm": 0.7904641628265381, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 6585 + }, + { + "epoch": 0.06586, + "grad_norm": 0.7364819049835205, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 6586 + }, + { + "epoch": 0.06587, + "grad_norm": 0.6441203355789185, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 6587 + }, + { + "epoch": 0.06588, + "grad_norm": 0.6132922172546387, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 6588 + }, + { + "epoch": 0.06589, + "grad_norm": 0.5162426829338074, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 6589 + }, + { + "epoch": 0.0659, + "grad_norm": 0.4835585057735443, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 6590 + }, + { + "epoch": 0.06591, + "grad_norm": 0.4747855067253113, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 6591 + }, + { + "epoch": 0.06592, + "grad_norm": 0.4445967674255371, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 6592 + }, + { + "epoch": 0.06593, + "grad_norm": 0.419645220041275, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 6593 + }, + { + "epoch": 0.06594, + "grad_norm": 0.4333605468273163, + "learning_rate": 0.003, + "loss": 4.065, + "step": 6594 + }, + { + "epoch": 0.06595, + "grad_norm": 0.4344804584980011, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 6595 + }, + { + "epoch": 0.06596, + "grad_norm": 0.4777514636516571, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 6596 + }, + { + "epoch": 0.06597, + "grad_norm": 0.4417304992675781, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 6597 + }, + { + "epoch": 0.06598, + "grad_norm": 0.4197903573513031, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 6598 + }, + { + "epoch": 0.06599, + "grad_norm": 0.447392076253891, + "learning_rate": 0.003, + "loss": 4.056, + "step": 6599 + }, + { + "epoch": 0.066, + "grad_norm": 0.43983280658721924, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 6600 + }, + { + "epoch": 0.06601, + "grad_norm": 0.41866225004196167, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 6601 + }, + { + "epoch": 0.06602, + "grad_norm": 0.42563024163246155, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 6602 + }, + { + "epoch": 0.06603, + "grad_norm": 0.4792834520339966, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 6603 + }, + { + "epoch": 0.06604, + "grad_norm": 0.5046424269676208, + "learning_rate": 0.003, + "loss": 4.034, + "step": 6604 + }, + { + "epoch": 0.06605, + "grad_norm": 0.4766392111778259, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 6605 + }, + { + "epoch": 0.06606, + "grad_norm": 0.4675769507884979, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 6606 + }, + { + "epoch": 0.06607, + "grad_norm": 0.5015403032302856, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 6607 + }, + { + "epoch": 0.06608, + "grad_norm": 0.5450870990753174, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 6608 + }, + { + "epoch": 0.06609, + "grad_norm": 0.5614660978317261, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 6609 + }, + { + "epoch": 0.0661, + "grad_norm": 0.6893346905708313, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 6610 + }, + { + "epoch": 0.06611, + "grad_norm": 1.0388861894607544, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 6611 + }, + { + "epoch": 0.06612, + "grad_norm": 1.5366320610046387, + "learning_rate": 0.003, + "loss": 4.057, + "step": 6612 + }, + { + "epoch": 0.06613, + "grad_norm": 0.6268520355224609, + "learning_rate": 0.003, + "loss": 4.055, + "step": 6613 + }, + { + "epoch": 0.06614, + "grad_norm": 0.7668797373771667, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 6614 + }, + { + "epoch": 0.06615, + "grad_norm": 0.9136865735054016, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 6615 + }, + { + "epoch": 0.06616, + "grad_norm": 0.9507898688316345, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 6616 + }, + { + "epoch": 0.06617, + "grad_norm": 0.9568789005279541, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 6617 + }, + { + "epoch": 0.06618, + "grad_norm": 0.8532642126083374, + "learning_rate": 0.003, + "loss": 4.087, + "step": 6618 + }, + { + "epoch": 0.06619, + "grad_norm": 0.8649228811264038, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 6619 + }, + { + "epoch": 0.0662, + "grad_norm": 0.9151912927627563, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 6620 + }, + { + "epoch": 0.06621, + "grad_norm": 0.9457108378410339, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 6621 + }, + { + "epoch": 0.06622, + "grad_norm": 0.9539183378219604, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 6622 + }, + { + "epoch": 0.06623, + "grad_norm": 1.0284817218780518, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 6623 + }, + { + "epoch": 0.06624, + "grad_norm": 0.9398470520973206, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 6624 + }, + { + "epoch": 0.06625, + "grad_norm": 0.9005268812179565, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 6625 + }, + { + "epoch": 0.06626, + "grad_norm": 0.9820789694786072, + "learning_rate": 0.003, + "loss": 4.095, + "step": 6626 + }, + { + "epoch": 0.06627, + "grad_norm": 1.002232551574707, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 6627 + }, + { + "epoch": 0.06628, + "grad_norm": 1.0087987184524536, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 6628 + }, + { + "epoch": 0.06629, + "grad_norm": 0.9122500419616699, + "learning_rate": 0.003, + "loss": 4.084, + "step": 6629 + }, + { + "epoch": 0.0663, + "grad_norm": 0.9479754567146301, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 6630 + }, + { + "epoch": 0.06631, + "grad_norm": 0.840352475643158, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 6631 + }, + { + "epoch": 0.06632, + "grad_norm": 0.8523590564727783, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 6632 + }, + { + "epoch": 0.06633, + "grad_norm": 0.9672338366508484, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 6633 + }, + { + "epoch": 0.06634, + "grad_norm": 1.0756926536560059, + "learning_rate": 0.003, + "loss": 4.095, + "step": 6634 + }, + { + "epoch": 0.06635, + "grad_norm": 0.8541210293769836, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 6635 + }, + { + "epoch": 0.06636, + "grad_norm": 0.7803227305412292, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 6636 + }, + { + "epoch": 0.06637, + "grad_norm": 0.7811536192893982, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 6637 + }, + { + "epoch": 0.06638, + "grad_norm": 0.7923469543457031, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 6638 + }, + { + "epoch": 0.06639, + "grad_norm": 0.7449505925178528, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 6639 + }, + { + "epoch": 0.0664, + "grad_norm": 0.8193292617797852, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 6640 + }, + { + "epoch": 0.06641, + "grad_norm": 0.731978178024292, + "learning_rate": 0.003, + "loss": 4.076, + "step": 6641 + }, + { + "epoch": 0.06642, + "grad_norm": 0.6337401270866394, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 6642 + }, + { + "epoch": 0.06643, + "grad_norm": 0.633114218711853, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 6643 + }, + { + "epoch": 0.06644, + "grad_norm": 0.6180500388145447, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 6644 + }, + { + "epoch": 0.06645, + "grad_norm": 0.6271146535873413, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 6645 + }, + { + "epoch": 0.06646, + "grad_norm": 0.6926038265228271, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 6646 + }, + { + "epoch": 0.06647, + "grad_norm": 0.6170114278793335, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 6647 + }, + { + "epoch": 0.06648, + "grad_norm": 0.6361090540885925, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 6648 + }, + { + "epoch": 0.06649, + "grad_norm": 0.5704471468925476, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 6649 + }, + { + "epoch": 0.0665, + "grad_norm": 0.5242335200309753, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 6650 + }, + { + "epoch": 0.06651, + "grad_norm": 0.5608069896697998, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 6651 + }, + { + "epoch": 0.06652, + "grad_norm": 0.4782668650150299, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 6652 + }, + { + "epoch": 0.06653, + "grad_norm": 0.5027245879173279, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 6653 + }, + { + "epoch": 0.06654, + "grad_norm": 0.5486674308776855, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 6654 + }, + { + "epoch": 0.06655, + "grad_norm": 0.5735544562339783, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 6655 + }, + { + "epoch": 0.06656, + "grad_norm": 0.6207435131072998, + "learning_rate": 0.003, + "loss": 4.032, + "step": 6656 + }, + { + "epoch": 0.06657, + "grad_norm": 0.676517128944397, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 6657 + }, + { + "epoch": 0.06658, + "grad_norm": 0.7929190993309021, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 6658 + }, + { + "epoch": 0.06659, + "grad_norm": 0.9687618613243103, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 6659 + }, + { + "epoch": 0.0666, + "grad_norm": 1.0018961429595947, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 6660 + }, + { + "epoch": 0.06661, + "grad_norm": 0.6997591257095337, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 6661 + }, + { + "epoch": 0.06662, + "grad_norm": 0.7275426387786865, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 6662 + }, + { + "epoch": 0.06663, + "grad_norm": 0.895223081111908, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 6663 + }, + { + "epoch": 0.06664, + "grad_norm": 0.8222059011459351, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 6664 + }, + { + "epoch": 0.06665, + "grad_norm": 0.6808730363845825, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 6665 + }, + { + "epoch": 0.06666, + "grad_norm": 0.5744432806968689, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 6666 + }, + { + "epoch": 0.06667, + "grad_norm": 0.6136099100112915, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 6667 + }, + { + "epoch": 0.06668, + "grad_norm": 0.620444118976593, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 6668 + }, + { + "epoch": 0.06669, + "grad_norm": 0.5014281272888184, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 6669 + }, + { + "epoch": 0.0667, + "grad_norm": 0.5208943486213684, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 6670 + }, + { + "epoch": 0.06671, + "grad_norm": 0.6109919548034668, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 6671 + }, + { + "epoch": 0.06672, + "grad_norm": 0.604210376739502, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 6672 + }, + { + "epoch": 0.06673, + "grad_norm": 0.5179533958435059, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 6673 + }, + { + "epoch": 0.06674, + "grad_norm": 0.49810075759887695, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 6674 + }, + { + "epoch": 0.06675, + "grad_norm": 0.480003297328949, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 6675 + }, + { + "epoch": 0.06676, + "grad_norm": 0.4583708345890045, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 6676 + }, + { + "epoch": 0.06677, + "grad_norm": 0.5014526844024658, + "learning_rate": 0.003, + "loss": 4.061, + "step": 6677 + }, + { + "epoch": 0.06678, + "grad_norm": 0.6773420572280884, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 6678 + }, + { + "epoch": 0.06679, + "grad_norm": 0.7450854182243347, + "learning_rate": 0.003, + "loss": 4.05, + "step": 6679 + }, + { + "epoch": 0.0668, + "grad_norm": 0.7281231880187988, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 6680 + }, + { + "epoch": 0.06681, + "grad_norm": 0.7678727507591248, + "learning_rate": 0.003, + "loss": 4.043, + "step": 6681 + }, + { + "epoch": 0.06682, + "grad_norm": 0.8374806642532349, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 6682 + }, + { + "epoch": 0.06683, + "grad_norm": 0.6994584798812866, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 6683 + }, + { + "epoch": 0.06684, + "grad_norm": 0.8886795043945312, + "learning_rate": 0.003, + "loss": 4.069, + "step": 6684 + }, + { + "epoch": 0.06685, + "grad_norm": 0.9739590883255005, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 6685 + }, + { + "epoch": 0.06686, + "grad_norm": 1.0791521072387695, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 6686 + }, + { + "epoch": 0.06687, + "grad_norm": 1.1976380348205566, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 6687 + }, + { + "epoch": 0.06688, + "grad_norm": 0.9513477683067322, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 6688 + }, + { + "epoch": 0.06689, + "grad_norm": 0.8864545822143555, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 6689 + }, + { + "epoch": 0.0669, + "grad_norm": 0.7956053614616394, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 6690 + }, + { + "epoch": 0.06691, + "grad_norm": 0.8103273510932922, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 6691 + }, + { + "epoch": 0.06692, + "grad_norm": 0.8394555449485779, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 6692 + }, + { + "epoch": 0.06693, + "grad_norm": 0.8771018385887146, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 6693 + }, + { + "epoch": 0.06694, + "grad_norm": 0.8685076236724854, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 6694 + }, + { + "epoch": 0.06695, + "grad_norm": 0.8724049925804138, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 6695 + }, + { + "epoch": 0.06696, + "grad_norm": 0.8524837493896484, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 6696 + }, + { + "epoch": 0.06697, + "grad_norm": 0.7606315016746521, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 6697 + }, + { + "epoch": 0.06698, + "grad_norm": 0.7235579490661621, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 6698 + }, + { + "epoch": 0.06699, + "grad_norm": 0.7316949963569641, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 6699 + }, + { + "epoch": 0.067, + "grad_norm": 0.6315829753875732, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 6700 + }, + { + "epoch": 0.06701, + "grad_norm": 0.631214439868927, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 6701 + }, + { + "epoch": 0.06702, + "grad_norm": 0.6383786201477051, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 6702 + }, + { + "epoch": 0.06703, + "grad_norm": 0.6043766736984253, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 6703 + }, + { + "epoch": 0.06704, + "grad_norm": 0.5771951675415039, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 6704 + }, + { + "epoch": 0.06705, + "grad_norm": 0.62209153175354, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 6705 + }, + { + "epoch": 0.06706, + "grad_norm": 0.7116777300834656, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 6706 + }, + { + "epoch": 0.06707, + "grad_norm": 0.8620831370353699, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 6707 + }, + { + "epoch": 0.06708, + "grad_norm": 0.8954139351844788, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 6708 + }, + { + "epoch": 0.06709, + "grad_norm": 0.8915477991104126, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 6709 + }, + { + "epoch": 0.0671, + "grad_norm": 0.8220002055168152, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 6710 + }, + { + "epoch": 0.06711, + "grad_norm": 0.6552330255508423, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 6711 + }, + { + "epoch": 0.06712, + "grad_norm": 0.6547254323959351, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 6712 + }, + { + "epoch": 0.06713, + "grad_norm": 0.7258485555648804, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 6713 + }, + { + "epoch": 0.06714, + "grad_norm": 0.8777948617935181, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 6714 + }, + { + "epoch": 0.06715, + "grad_norm": 0.9278170466423035, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 6715 + }, + { + "epoch": 0.06716, + "grad_norm": 0.7775757312774658, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 6716 + }, + { + "epoch": 0.06717, + "grad_norm": 0.7415366172790527, + "learning_rate": 0.003, + "loss": 4.08, + "step": 6717 + }, + { + "epoch": 0.06718, + "grad_norm": 0.7367812991142273, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 6718 + }, + { + "epoch": 0.06719, + "grad_norm": 0.7651417255401611, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 6719 + }, + { + "epoch": 0.0672, + "grad_norm": 0.7453299760818481, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 6720 + }, + { + "epoch": 0.06721, + "grad_norm": 0.6883461475372314, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 6721 + }, + { + "epoch": 0.06722, + "grad_norm": 0.7742420434951782, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 6722 + }, + { + "epoch": 0.06723, + "grad_norm": 0.7198421359062195, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 6723 + }, + { + "epoch": 0.06724, + "grad_norm": 0.6653247475624084, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 6724 + }, + { + "epoch": 0.06725, + "grad_norm": 0.6466822028160095, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 6725 + }, + { + "epoch": 0.06726, + "grad_norm": 0.5745895504951477, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 6726 + }, + { + "epoch": 0.06727, + "grad_norm": 0.530097246170044, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 6727 + }, + { + "epoch": 0.06728, + "grad_norm": 0.5702248811721802, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 6728 + }, + { + "epoch": 0.06729, + "grad_norm": 0.7315114140510559, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 6729 + }, + { + "epoch": 0.0673, + "grad_norm": 0.9155875444412231, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 6730 + }, + { + "epoch": 0.06731, + "grad_norm": 1.1372816562652588, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 6731 + }, + { + "epoch": 0.06732, + "grad_norm": 0.925939679145813, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 6732 + }, + { + "epoch": 0.06733, + "grad_norm": 0.6862448453903198, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 6733 + }, + { + "epoch": 0.06734, + "grad_norm": 0.622930109500885, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 6734 + }, + { + "epoch": 0.06735, + "grad_norm": 0.6733621954917908, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 6735 + }, + { + "epoch": 0.06736, + "grad_norm": 0.683295488357544, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 6736 + }, + { + "epoch": 0.06737, + "grad_norm": 0.5437440872192383, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 6737 + }, + { + "epoch": 0.06738, + "grad_norm": 0.5390477180480957, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 6738 + }, + { + "epoch": 0.06739, + "grad_norm": 0.4995149075984955, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 6739 + }, + { + "epoch": 0.0674, + "grad_norm": 0.5221118927001953, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 6740 + }, + { + "epoch": 0.06741, + "grad_norm": 0.6219124794006348, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 6741 + }, + { + "epoch": 0.06742, + "grad_norm": 0.7121551632881165, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 6742 + }, + { + "epoch": 0.06743, + "grad_norm": 0.8330824971199036, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 6743 + }, + { + "epoch": 0.06744, + "grad_norm": 0.7893618941307068, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 6744 + }, + { + "epoch": 0.06745, + "grad_norm": 0.6968402862548828, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 6745 + }, + { + "epoch": 0.06746, + "grad_norm": 0.7016124725341797, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 6746 + }, + { + "epoch": 0.06747, + "grad_norm": 0.8388985395431519, + "learning_rate": 0.003, + "loss": 4.054, + "step": 6747 + }, + { + "epoch": 0.06748, + "grad_norm": 0.8102107644081116, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 6748 + }, + { + "epoch": 0.06749, + "grad_norm": 0.6827043890953064, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 6749 + }, + { + "epoch": 0.0675, + "grad_norm": 0.7439062595367432, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 6750 + }, + { + "epoch": 0.06751, + "grad_norm": 0.7768245935440063, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 6751 + }, + { + "epoch": 0.06752, + "grad_norm": 0.8553226590156555, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 6752 + }, + { + "epoch": 0.06753, + "grad_norm": 0.9639877080917358, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 6753 + }, + { + "epoch": 0.06754, + "grad_norm": 0.9903431534767151, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 6754 + }, + { + "epoch": 0.06755, + "grad_norm": 0.9414642453193665, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 6755 + }, + { + "epoch": 0.06756, + "grad_norm": 1.0175652503967285, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 6756 + }, + { + "epoch": 0.06757, + "grad_norm": 1.0221658945083618, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 6757 + }, + { + "epoch": 0.06758, + "grad_norm": 0.9344669580459595, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 6758 + }, + { + "epoch": 0.06759, + "grad_norm": 0.824642539024353, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 6759 + }, + { + "epoch": 0.0676, + "grad_norm": 0.8851695656776428, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 6760 + }, + { + "epoch": 0.06761, + "grad_norm": 0.8390560150146484, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 6761 + }, + { + "epoch": 0.06762, + "grad_norm": 0.8119285702705383, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 6762 + }, + { + "epoch": 0.06763, + "grad_norm": 0.8722089529037476, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 6763 + }, + { + "epoch": 0.06764, + "grad_norm": 0.8680243492126465, + "learning_rate": 0.003, + "loss": 4.049, + "step": 6764 + }, + { + "epoch": 0.06765, + "grad_norm": 0.7568630576133728, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 6765 + }, + { + "epoch": 0.06766, + "grad_norm": 0.6181029677391052, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 6766 + }, + { + "epoch": 0.06767, + "grad_norm": 0.6267876029014587, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 6767 + }, + { + "epoch": 0.06768, + "grad_norm": 0.6331607103347778, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 6768 + }, + { + "epoch": 0.06769, + "grad_norm": 0.5892623662948608, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 6769 + }, + { + "epoch": 0.0677, + "grad_norm": 0.6032274961471558, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 6770 + }, + { + "epoch": 0.06771, + "grad_norm": 0.6521531343460083, + "learning_rate": 0.003, + "loss": 4.048, + "step": 6771 + }, + { + "epoch": 0.06772, + "grad_norm": 0.5568339824676514, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 6772 + }, + { + "epoch": 0.06773, + "grad_norm": 0.4971171021461487, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 6773 + }, + { + "epoch": 0.06774, + "grad_norm": 0.47084400057792664, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 6774 + }, + { + "epoch": 0.06775, + "grad_norm": 0.48115456104278564, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 6775 + }, + { + "epoch": 0.06776, + "grad_norm": 0.4709925055503845, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 6776 + }, + { + "epoch": 0.06777, + "grad_norm": 0.5252442359924316, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 6777 + }, + { + "epoch": 0.06778, + "grad_norm": 0.503268837928772, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 6778 + }, + { + "epoch": 0.06779, + "grad_norm": 0.5064627528190613, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 6779 + }, + { + "epoch": 0.0678, + "grad_norm": 0.45777052640914917, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 6780 + }, + { + "epoch": 0.06781, + "grad_norm": 0.5485308766365051, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 6781 + }, + { + "epoch": 0.06782, + "grad_norm": 0.7215536236763, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 6782 + }, + { + "epoch": 0.06783, + "grad_norm": 0.9196093082427979, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 6783 + }, + { + "epoch": 0.06784, + "grad_norm": 1.2304813861846924, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 6784 + }, + { + "epoch": 0.06785, + "grad_norm": 0.6971254348754883, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 6785 + }, + { + "epoch": 0.06786, + "grad_norm": 0.6344663500785828, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 6786 + }, + { + "epoch": 0.06787, + "grad_norm": 0.7827377915382385, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 6787 + }, + { + "epoch": 0.06788, + "grad_norm": 0.8472016453742981, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 6788 + }, + { + "epoch": 0.06789, + "grad_norm": 0.8765668869018555, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 6789 + }, + { + "epoch": 0.0679, + "grad_norm": 0.8723348379135132, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 6790 + }, + { + "epoch": 0.06791, + "grad_norm": 0.7648828625679016, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 6791 + }, + { + "epoch": 0.06792, + "grad_norm": 0.7784323692321777, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 6792 + }, + { + "epoch": 0.06793, + "grad_norm": 0.7629876732826233, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 6793 + }, + { + "epoch": 0.06794, + "grad_norm": 0.7403256297111511, + "learning_rate": 0.003, + "loss": 4.078, + "step": 6794 + }, + { + "epoch": 0.06795, + "grad_norm": 0.7046683430671692, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 6795 + }, + { + "epoch": 0.06796, + "grad_norm": 0.7108621597290039, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 6796 + }, + { + "epoch": 0.06797, + "grad_norm": 0.7049769759178162, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 6797 + }, + { + "epoch": 0.06798, + "grad_norm": 0.7718124389648438, + "learning_rate": 0.003, + "loss": 4.051, + "step": 6798 + }, + { + "epoch": 0.06799, + "grad_norm": 0.8259626030921936, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 6799 + }, + { + "epoch": 0.068, + "grad_norm": 0.7455199956893921, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 6800 + }, + { + "epoch": 0.06801, + "grad_norm": 0.6687690615653992, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 6801 + }, + { + "epoch": 0.06802, + "grad_norm": 0.663851797580719, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 6802 + }, + { + "epoch": 0.06803, + "grad_norm": 0.726151168346405, + "learning_rate": 0.003, + "loss": 4.054, + "step": 6803 + }, + { + "epoch": 0.06804, + "grad_norm": 0.7056896686553955, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 6804 + }, + { + "epoch": 0.06805, + "grad_norm": 0.7536913156509399, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 6805 + }, + { + "epoch": 0.06806, + "grad_norm": 1.025210976600647, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 6806 + }, + { + "epoch": 0.06807, + "grad_norm": 1.3468793630599976, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 6807 + }, + { + "epoch": 0.06808, + "grad_norm": 0.5080417990684509, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 6808 + }, + { + "epoch": 0.06809, + "grad_norm": 0.7991033792495728, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 6809 + }, + { + "epoch": 0.0681, + "grad_norm": 0.9151610136032104, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 6810 + }, + { + "epoch": 0.06811, + "grad_norm": 0.8046905994415283, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 6811 + }, + { + "epoch": 0.06812, + "grad_norm": 0.7898740172386169, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 6812 + }, + { + "epoch": 0.06813, + "grad_norm": 0.7531576752662659, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 6813 + }, + { + "epoch": 0.06814, + "grad_norm": 0.7203744649887085, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 6814 + }, + { + "epoch": 0.06815, + "grad_norm": 0.793602705001831, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 6815 + }, + { + "epoch": 0.06816, + "grad_norm": 0.7131784558296204, + "learning_rate": 0.003, + "loss": 4.074, + "step": 6816 + }, + { + "epoch": 0.06817, + "grad_norm": 0.6014184951782227, + "learning_rate": 0.003, + "loss": 4.042, + "step": 6817 + }, + { + "epoch": 0.06818, + "grad_norm": 0.5902239084243774, + "learning_rate": 0.003, + "loss": 4.062, + "step": 6818 + }, + { + "epoch": 0.06819, + "grad_norm": 0.6026824712753296, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 6819 + }, + { + "epoch": 0.0682, + "grad_norm": 0.6438929438591003, + "learning_rate": 0.003, + "loss": 4.084, + "step": 6820 + }, + { + "epoch": 0.06821, + "grad_norm": 0.6496386528015137, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 6821 + }, + { + "epoch": 0.06822, + "grad_norm": 0.6901900768280029, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 6822 + }, + { + "epoch": 0.06823, + "grad_norm": 0.7220962047576904, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 6823 + }, + { + "epoch": 0.06824, + "grad_norm": 0.7161133289337158, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 6824 + }, + { + "epoch": 0.06825, + "grad_norm": 0.7654991149902344, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 6825 + }, + { + "epoch": 0.06826, + "grad_norm": 0.7452932000160217, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 6826 + }, + { + "epoch": 0.06827, + "grad_norm": 0.7094281911849976, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 6827 + }, + { + "epoch": 0.06828, + "grad_norm": 0.8043527603149414, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 6828 + }, + { + "epoch": 0.06829, + "grad_norm": 0.7790755033493042, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 6829 + }, + { + "epoch": 0.0683, + "grad_norm": 0.9166526794433594, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 6830 + }, + { + "epoch": 0.06831, + "grad_norm": 1.1532281637191772, + "learning_rate": 0.003, + "loss": 4.091, + "step": 6831 + }, + { + "epoch": 0.06832, + "grad_norm": 0.878596842288971, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 6832 + }, + { + "epoch": 0.06833, + "grad_norm": 0.6871094107627869, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 6833 + }, + { + "epoch": 0.06834, + "grad_norm": 0.5933970212936401, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 6834 + }, + { + "epoch": 0.06835, + "grad_norm": 0.635981023311615, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 6835 + }, + { + "epoch": 0.06836, + "grad_norm": 0.6953107714653015, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 6836 + }, + { + "epoch": 0.06837, + "grad_norm": 0.7208017706871033, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 6837 + }, + { + "epoch": 0.06838, + "grad_norm": 0.6934213638305664, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 6838 + }, + { + "epoch": 0.06839, + "grad_norm": 0.6325958371162415, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 6839 + }, + { + "epoch": 0.0684, + "grad_norm": 0.592176616191864, + "learning_rate": 0.003, + "loss": 4.058, + "step": 6840 + }, + { + "epoch": 0.06841, + "grad_norm": 0.5508002042770386, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 6841 + }, + { + "epoch": 0.06842, + "grad_norm": 0.5809416770935059, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 6842 + }, + { + "epoch": 0.06843, + "grad_norm": 0.5343518257141113, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 6843 + }, + { + "epoch": 0.06844, + "grad_norm": 0.6117638349533081, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 6844 + }, + { + "epoch": 0.06845, + "grad_norm": 0.6104193329811096, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 6845 + }, + { + "epoch": 0.06846, + "grad_norm": 0.6597003936767578, + "learning_rate": 0.003, + "loss": 4.047, + "step": 6846 + }, + { + "epoch": 0.06847, + "grad_norm": 0.907615065574646, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 6847 + }, + { + "epoch": 0.06848, + "grad_norm": 1.0650358200073242, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 6848 + }, + { + "epoch": 0.06849, + "grad_norm": 0.9667098522186279, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 6849 + }, + { + "epoch": 0.0685, + "grad_norm": 0.7904394865036011, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 6850 + }, + { + "epoch": 0.06851, + "grad_norm": 0.6749697327613831, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 6851 + }, + { + "epoch": 0.06852, + "grad_norm": 0.7591043710708618, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 6852 + }, + { + "epoch": 0.06853, + "grad_norm": 0.723334550857544, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 6853 + }, + { + "epoch": 0.06854, + "grad_norm": 0.6678939461708069, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 6854 + }, + { + "epoch": 0.06855, + "grad_norm": 0.735528290271759, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 6855 + }, + { + "epoch": 0.06856, + "grad_norm": 0.802253246307373, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 6856 + }, + { + "epoch": 0.06857, + "grad_norm": 0.872907817363739, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 6857 + }, + { + "epoch": 0.06858, + "grad_norm": 0.8424869775772095, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 6858 + }, + { + "epoch": 0.06859, + "grad_norm": 0.8607826232910156, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 6859 + }, + { + "epoch": 0.0686, + "grad_norm": 0.8124536871910095, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 6860 + }, + { + "epoch": 0.06861, + "grad_norm": 0.7242681384086609, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 6861 + }, + { + "epoch": 0.06862, + "grad_norm": 0.6859086155891418, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 6862 + }, + { + "epoch": 0.06863, + "grad_norm": 0.6288987994194031, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 6863 + }, + { + "epoch": 0.06864, + "grad_norm": 0.589547336101532, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 6864 + }, + { + "epoch": 0.06865, + "grad_norm": 0.5942288637161255, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 6865 + }, + { + "epoch": 0.06866, + "grad_norm": 0.6688255071640015, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 6866 + }, + { + "epoch": 0.06867, + "grad_norm": 0.7595317363739014, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 6867 + }, + { + "epoch": 0.06868, + "grad_norm": 1.0023117065429688, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 6868 + }, + { + "epoch": 0.06869, + "grad_norm": 1.171606421470642, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 6869 + }, + { + "epoch": 0.0687, + "grad_norm": 0.6550358533859253, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 6870 + }, + { + "epoch": 0.06871, + "grad_norm": 0.5838268995285034, + "learning_rate": 0.003, + "loss": 4.031, + "step": 6871 + }, + { + "epoch": 0.06872, + "grad_norm": 0.6378373503684998, + "learning_rate": 0.003, + "loss": 4.035, + "step": 6872 + }, + { + "epoch": 0.06873, + "grad_norm": 0.7861857414245605, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 6873 + }, + { + "epoch": 0.06874, + "grad_norm": 0.7961738109588623, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 6874 + }, + { + "epoch": 0.06875, + "grad_norm": 0.7601486444473267, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 6875 + }, + { + "epoch": 0.06876, + "grad_norm": 0.7805059552192688, + "learning_rate": 0.003, + "loss": 4.069, + "step": 6876 + }, + { + "epoch": 0.06877, + "grad_norm": 0.75244140625, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 6877 + }, + { + "epoch": 0.06878, + "grad_norm": 0.6880764365196228, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 6878 + }, + { + "epoch": 0.06879, + "grad_norm": 0.6558811664581299, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 6879 + }, + { + "epoch": 0.0688, + "grad_norm": 0.6576889157295227, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 6880 + }, + { + "epoch": 0.06881, + "grad_norm": 0.6587197780609131, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 6881 + }, + { + "epoch": 0.06882, + "grad_norm": 0.6243919730186462, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 6882 + }, + { + "epoch": 0.06883, + "grad_norm": 0.6536480188369751, + "learning_rate": 0.003, + "loss": 4.068, + "step": 6883 + }, + { + "epoch": 0.06884, + "grad_norm": 0.6172704696655273, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 6884 + }, + { + "epoch": 0.06885, + "grad_norm": 0.6676458716392517, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 6885 + }, + { + "epoch": 0.06886, + "grad_norm": 0.761619508266449, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 6886 + }, + { + "epoch": 0.06887, + "grad_norm": 0.8133268356323242, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 6887 + }, + { + "epoch": 0.06888, + "grad_norm": 0.9232909679412842, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 6888 + }, + { + "epoch": 0.06889, + "grad_norm": 0.880495011806488, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 6889 + }, + { + "epoch": 0.0689, + "grad_norm": 0.8015482425689697, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 6890 + }, + { + "epoch": 0.06891, + "grad_norm": 0.7700675129890442, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 6891 + }, + { + "epoch": 0.06892, + "grad_norm": 0.8113595843315125, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 6892 + }, + { + "epoch": 0.06893, + "grad_norm": 0.7549485564231873, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 6893 + }, + { + "epoch": 0.06894, + "grad_norm": 0.8220582604408264, + "learning_rate": 0.003, + "loss": 4.039, + "step": 6894 + }, + { + "epoch": 0.06895, + "grad_norm": 0.8129627704620361, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 6895 + }, + { + "epoch": 0.06896, + "grad_norm": 0.6896864771842957, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 6896 + }, + { + "epoch": 0.06897, + "grad_norm": 0.6018128991127014, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 6897 + }, + { + "epoch": 0.06898, + "grad_norm": 0.6657570004463196, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 6898 + }, + { + "epoch": 0.06899, + "grad_norm": 0.7586705088615417, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 6899 + }, + { + "epoch": 0.069, + "grad_norm": 0.6607414484024048, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 6900 + }, + { + "epoch": 0.06901, + "grad_norm": 0.7915401458740234, + "learning_rate": 0.003, + "loss": 4.024, + "step": 6901 + }, + { + "epoch": 0.06902, + "grad_norm": 0.8700487017631531, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 6902 + }, + { + "epoch": 0.06903, + "grad_norm": 1.0735187530517578, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 6903 + }, + { + "epoch": 0.06904, + "grad_norm": 0.9046341776847839, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 6904 + }, + { + "epoch": 0.06905, + "grad_norm": 0.7316793203353882, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 6905 + }, + { + "epoch": 0.06906, + "grad_norm": 0.5604428648948669, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 6906 + }, + { + "epoch": 0.06907, + "grad_norm": 0.564428448677063, + "learning_rate": 0.003, + "loss": 4.038, + "step": 6907 + }, + { + "epoch": 0.06908, + "grad_norm": 0.6861115097999573, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 6908 + }, + { + "epoch": 0.06909, + "grad_norm": 0.7282766699790955, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 6909 + }, + { + "epoch": 0.0691, + "grad_norm": 0.6792446970939636, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 6910 + }, + { + "epoch": 0.06911, + "grad_norm": 0.6289098858833313, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 6911 + }, + { + "epoch": 0.06912, + "grad_norm": 0.5979729890823364, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 6912 + }, + { + "epoch": 0.06913, + "grad_norm": 0.7078564167022705, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 6913 + }, + { + "epoch": 0.06914, + "grad_norm": 0.8100939393043518, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 6914 + }, + { + "epoch": 0.06915, + "grad_norm": 0.893566906452179, + "learning_rate": 0.003, + "loss": 4.083, + "step": 6915 + }, + { + "epoch": 0.06916, + "grad_norm": 0.9428468942642212, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 6916 + }, + { + "epoch": 0.06917, + "grad_norm": 0.8902822732925415, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 6917 + }, + { + "epoch": 0.06918, + "grad_norm": 0.9391528367996216, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 6918 + }, + { + "epoch": 0.06919, + "grad_norm": 0.8996408581733704, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 6919 + }, + { + "epoch": 0.0692, + "grad_norm": 0.8599734902381897, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 6920 + }, + { + "epoch": 0.06921, + "grad_norm": 0.8355299234390259, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 6921 + }, + { + "epoch": 0.06922, + "grad_norm": 0.812288224697113, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 6922 + }, + { + "epoch": 0.06923, + "grad_norm": 0.9053574204444885, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 6923 + }, + { + "epoch": 0.06924, + "grad_norm": 1.0341438055038452, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 6924 + }, + { + "epoch": 0.06925, + "grad_norm": 1.2150890827178955, + "learning_rate": 0.003, + "loss": 4.095, + "step": 6925 + }, + { + "epoch": 0.06926, + "grad_norm": 0.722688615322113, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 6926 + }, + { + "epoch": 0.06927, + "grad_norm": 0.6441181302070618, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 6927 + }, + { + "epoch": 0.06928, + "grad_norm": 0.6667382121086121, + "learning_rate": 0.003, + "loss": 4.047, + "step": 6928 + }, + { + "epoch": 0.06929, + "grad_norm": 0.6884969472885132, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 6929 + }, + { + "epoch": 0.0693, + "grad_norm": 0.7797659039497375, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 6930 + }, + { + "epoch": 0.06931, + "grad_norm": 0.7904036641120911, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 6931 + }, + { + "epoch": 0.06932, + "grad_norm": 0.7277095317840576, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 6932 + }, + { + "epoch": 0.06933, + "grad_norm": 0.6969214677810669, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 6933 + }, + { + "epoch": 0.06934, + "grad_norm": 0.7735503911972046, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 6934 + }, + { + "epoch": 0.06935, + "grad_norm": 0.8755772113800049, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 6935 + }, + { + "epoch": 0.06936, + "grad_norm": 0.9457864761352539, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 6936 + }, + { + "epoch": 0.06937, + "grad_norm": 1.0101808309555054, + "learning_rate": 0.003, + "loss": 4.058, + "step": 6937 + }, + { + "epoch": 0.06938, + "grad_norm": 0.7283269762992859, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 6938 + }, + { + "epoch": 0.06939, + "grad_norm": 0.5417292714118958, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 6939 + }, + { + "epoch": 0.0694, + "grad_norm": 0.5455449819564819, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 6940 + }, + { + "epoch": 0.06941, + "grad_norm": 0.603982150554657, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 6941 + }, + { + "epoch": 0.06942, + "grad_norm": 0.6116390228271484, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 6942 + }, + { + "epoch": 0.06943, + "grad_norm": 0.5719929337501526, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 6943 + }, + { + "epoch": 0.06944, + "grad_norm": 0.554226279258728, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 6944 + }, + { + "epoch": 0.06945, + "grad_norm": 0.6480233669281006, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 6945 + }, + { + "epoch": 0.06946, + "grad_norm": 0.6147804856300354, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 6946 + }, + { + "epoch": 0.06947, + "grad_norm": 0.6045358777046204, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 6947 + }, + { + "epoch": 0.06948, + "grad_norm": 0.671883225440979, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 6948 + }, + { + "epoch": 0.06949, + "grad_norm": 0.7604190707206726, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 6949 + }, + { + "epoch": 0.0695, + "grad_norm": 0.8634862303733826, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 6950 + }, + { + "epoch": 0.06951, + "grad_norm": 0.8246872425079346, + "learning_rate": 0.003, + "loss": 4.071, + "step": 6951 + }, + { + "epoch": 0.06952, + "grad_norm": 0.7395027875900269, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 6952 + }, + { + "epoch": 0.06953, + "grad_norm": 0.7343699336051941, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 6953 + }, + { + "epoch": 0.06954, + "grad_norm": 0.7108721733093262, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 6954 + }, + { + "epoch": 0.06955, + "grad_norm": 0.6825199723243713, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 6955 + }, + { + "epoch": 0.06956, + "grad_norm": 0.7615572810173035, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 6956 + }, + { + "epoch": 0.06957, + "grad_norm": 0.8274344801902771, + "learning_rate": 0.003, + "loss": 4.049, + "step": 6957 + }, + { + "epoch": 0.06958, + "grad_norm": 0.8784217238426208, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 6958 + }, + { + "epoch": 0.06959, + "grad_norm": 0.7803829908370972, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 6959 + }, + { + "epoch": 0.0696, + "grad_norm": 0.8395748138427734, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 6960 + }, + { + "epoch": 0.06961, + "grad_norm": 0.7431996464729309, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 6961 + }, + { + "epoch": 0.06962, + "grad_norm": 0.7766883373260498, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 6962 + }, + { + "epoch": 0.06963, + "grad_norm": 0.8304595947265625, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 6963 + }, + { + "epoch": 0.06964, + "grad_norm": 0.8011763095855713, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 6964 + }, + { + "epoch": 0.06965, + "grad_norm": 0.6499873995780945, + "learning_rate": 0.003, + "loss": 4.039, + "step": 6965 + }, + { + "epoch": 0.06966, + "grad_norm": 0.5533931851387024, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 6966 + }, + { + "epoch": 0.06967, + "grad_norm": 0.6004241704940796, + "learning_rate": 0.003, + "loss": 4.043, + "step": 6967 + }, + { + "epoch": 0.06968, + "grad_norm": 0.6045736074447632, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 6968 + }, + { + "epoch": 0.06969, + "grad_norm": 0.6887038946151733, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 6969 + }, + { + "epoch": 0.0697, + "grad_norm": 0.5907984972000122, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 6970 + }, + { + "epoch": 0.06971, + "grad_norm": 0.5247988700866699, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 6971 + }, + { + "epoch": 0.06972, + "grad_norm": 0.5534252524375916, + "learning_rate": 0.003, + "loss": 4.005, + "step": 6972 + }, + { + "epoch": 0.06973, + "grad_norm": 0.6565967798233032, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 6973 + }, + { + "epoch": 0.06974, + "grad_norm": 0.7747914791107178, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 6974 + }, + { + "epoch": 0.06975, + "grad_norm": 0.8217228055000305, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 6975 + }, + { + "epoch": 0.06976, + "grad_norm": 0.9226599931716919, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 6976 + }, + { + "epoch": 0.06977, + "grad_norm": 1.0128273963928223, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 6977 + }, + { + "epoch": 0.06978, + "grad_norm": 1.0614304542541504, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 6978 + }, + { + "epoch": 0.06979, + "grad_norm": 0.8185182809829712, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 6979 + }, + { + "epoch": 0.0698, + "grad_norm": 0.7523839473724365, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 6980 + }, + { + "epoch": 0.06981, + "grad_norm": 0.809232234954834, + "learning_rate": 0.003, + "loss": 4.04, + "step": 6981 + }, + { + "epoch": 0.06982, + "grad_norm": 0.8641067147254944, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 6982 + }, + { + "epoch": 0.06983, + "grad_norm": 0.8018274903297424, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 6983 + }, + { + "epoch": 0.06984, + "grad_norm": 0.8893591165542603, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 6984 + }, + { + "epoch": 0.06985, + "grad_norm": 1.0270733833312988, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 6985 + }, + { + "epoch": 0.06986, + "grad_norm": 0.9915245175361633, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 6986 + }, + { + "epoch": 0.06987, + "grad_norm": 0.9925371408462524, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 6987 + }, + { + "epoch": 0.06988, + "grad_norm": 1.083008885383606, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 6988 + }, + { + "epoch": 0.06989, + "grad_norm": 1.013920783996582, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 6989 + }, + { + "epoch": 0.0699, + "grad_norm": 0.8849502205848694, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 6990 + }, + { + "epoch": 0.06991, + "grad_norm": 0.7566224932670593, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 6991 + }, + { + "epoch": 0.06992, + "grad_norm": 0.6404983997344971, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 6992 + }, + { + "epoch": 0.06993, + "grad_norm": 0.5881397128105164, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 6993 + }, + { + "epoch": 0.06994, + "grad_norm": 0.6447122097015381, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 6994 + }, + { + "epoch": 0.06995, + "grad_norm": 0.5685395002365112, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 6995 + }, + { + "epoch": 0.06996, + "grad_norm": 0.5681016445159912, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 6996 + }, + { + "epoch": 0.06997, + "grad_norm": 0.6152894496917725, + "learning_rate": 0.003, + "loss": 4.03, + "step": 6997 + }, + { + "epoch": 0.06998, + "grad_norm": 0.6342592835426331, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 6998 + }, + { + "epoch": 0.06999, + "grad_norm": 0.6017648577690125, + "learning_rate": 0.003, + "loss": 4.032, + "step": 6999 + }, + { + "epoch": 0.07, + "grad_norm": 0.6009829640388489, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 7000 + }, + { + "epoch": 0.07001, + "grad_norm": 0.5869970321655273, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 7001 + }, + { + "epoch": 0.07002, + "grad_norm": 0.5607997179031372, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 7002 + }, + { + "epoch": 0.07003, + "grad_norm": 0.5857222080230713, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 7003 + }, + { + "epoch": 0.07004, + "grad_norm": 0.6829317212104797, + "learning_rate": 0.003, + "loss": 4.019, + "step": 7004 + }, + { + "epoch": 0.07005, + "grad_norm": 0.7422044277191162, + "learning_rate": 0.003, + "loss": 4.045, + "step": 7005 + }, + { + "epoch": 0.07006, + "grad_norm": 0.7209708094596863, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 7006 + }, + { + "epoch": 0.07007, + "grad_norm": 0.6910662055015564, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 7007 + }, + { + "epoch": 0.07008, + "grad_norm": 0.724881649017334, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 7008 + }, + { + "epoch": 0.07009, + "grad_norm": 0.7626492977142334, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 7009 + }, + { + "epoch": 0.0701, + "grad_norm": 0.7107502222061157, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 7010 + }, + { + "epoch": 0.07011, + "grad_norm": 0.6228822469711304, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 7011 + }, + { + "epoch": 0.07012, + "grad_norm": 0.6056353449821472, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 7012 + }, + { + "epoch": 0.07013, + "grad_norm": 0.5881388783454895, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 7013 + }, + { + "epoch": 0.07014, + "grad_norm": 0.5778508186340332, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 7014 + }, + { + "epoch": 0.07015, + "grad_norm": 0.6298239231109619, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 7015 + }, + { + "epoch": 0.07016, + "grad_norm": 0.6628670692443848, + "learning_rate": 0.003, + "loss": 4.036, + "step": 7016 + }, + { + "epoch": 0.07017, + "grad_norm": 0.7037574052810669, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 7017 + }, + { + "epoch": 0.07018, + "grad_norm": 0.7001218199729919, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 7018 + }, + { + "epoch": 0.07019, + "grad_norm": 0.7447029948234558, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 7019 + }, + { + "epoch": 0.0702, + "grad_norm": 0.7939612865447998, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 7020 + }, + { + "epoch": 0.07021, + "grad_norm": 1.0586154460906982, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 7021 + }, + { + "epoch": 0.07022, + "grad_norm": 0.9932177662849426, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 7022 + }, + { + "epoch": 0.07023, + "grad_norm": 0.7585009932518005, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 7023 + }, + { + "epoch": 0.07024, + "grad_norm": 0.583694338798523, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 7024 + }, + { + "epoch": 0.07025, + "grad_norm": 0.5445935130119324, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 7025 + }, + { + "epoch": 0.07026, + "grad_norm": 0.5929006934165955, + "learning_rate": 0.003, + "loss": 4.025, + "step": 7026 + }, + { + "epoch": 0.07027, + "grad_norm": 0.6477934122085571, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 7027 + }, + { + "epoch": 0.07028, + "grad_norm": 0.8614700436592102, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 7028 + }, + { + "epoch": 0.07029, + "grad_norm": 1.0163159370422363, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 7029 + }, + { + "epoch": 0.0703, + "grad_norm": 0.9567646980285645, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 7030 + }, + { + "epoch": 0.07031, + "grad_norm": 0.8252813816070557, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 7031 + }, + { + "epoch": 0.07032, + "grad_norm": 0.8402538895606995, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 7032 + }, + { + "epoch": 0.07033, + "grad_norm": 0.9886388182640076, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 7033 + }, + { + "epoch": 0.07034, + "grad_norm": 0.9284382462501526, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 7034 + }, + { + "epoch": 0.07035, + "grad_norm": 0.8074457049369812, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 7035 + }, + { + "epoch": 0.07036, + "grad_norm": 0.8691509962081909, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 7036 + }, + { + "epoch": 0.07037, + "grad_norm": 0.9332064986228943, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 7037 + }, + { + "epoch": 0.07038, + "grad_norm": 0.9712073802947998, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 7038 + }, + { + "epoch": 0.07039, + "grad_norm": 0.8838890194892883, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 7039 + }, + { + "epoch": 0.0704, + "grad_norm": 0.9608893394470215, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 7040 + }, + { + "epoch": 0.07041, + "grad_norm": 0.7875691056251526, + "learning_rate": 0.003, + "loss": 4.058, + "step": 7041 + }, + { + "epoch": 0.07042, + "grad_norm": 0.6502549648284912, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 7042 + }, + { + "epoch": 0.07043, + "grad_norm": 0.6589664816856384, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 7043 + }, + { + "epoch": 0.07044, + "grad_norm": 0.7496007084846497, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 7044 + }, + { + "epoch": 0.07045, + "grad_norm": 0.6947444677352905, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 7045 + }, + { + "epoch": 0.07046, + "grad_norm": 0.619331955909729, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 7046 + }, + { + "epoch": 0.07047, + "grad_norm": 0.6577057838439941, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 7047 + }, + { + "epoch": 0.07048, + "grad_norm": 0.727730393409729, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 7048 + }, + { + "epoch": 0.07049, + "grad_norm": 0.8019500374794006, + "learning_rate": 0.003, + "loss": 4.052, + "step": 7049 + }, + { + "epoch": 0.0705, + "grad_norm": 0.8129701614379883, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 7050 + }, + { + "epoch": 0.07051, + "grad_norm": 0.9163272976875305, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 7051 + }, + { + "epoch": 0.07052, + "grad_norm": 1.0532561540603638, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 7052 + }, + { + "epoch": 0.07053, + "grad_norm": 0.8734548687934875, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 7053 + }, + { + "epoch": 0.07054, + "grad_norm": 0.8844286799430847, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 7054 + }, + { + "epoch": 0.07055, + "grad_norm": 0.8699377179145813, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 7055 + }, + { + "epoch": 0.07056, + "grad_norm": 0.8266685605049133, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 7056 + }, + { + "epoch": 0.07057, + "grad_norm": 0.7403209805488586, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 7057 + }, + { + "epoch": 0.07058, + "grad_norm": 0.7143754959106445, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 7058 + }, + { + "epoch": 0.07059, + "grad_norm": 0.7726312875747681, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 7059 + }, + { + "epoch": 0.0706, + "grad_norm": 0.7227030396461487, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 7060 + }, + { + "epoch": 0.07061, + "grad_norm": 0.7146661281585693, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 7061 + }, + { + "epoch": 0.07062, + "grad_norm": 0.7332003712654114, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 7062 + }, + { + "epoch": 0.07063, + "grad_norm": 0.7866206765174866, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 7063 + }, + { + "epoch": 0.07064, + "grad_norm": 0.8742010593414307, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 7064 + }, + { + "epoch": 0.07065, + "grad_norm": 0.7553660869598389, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 7065 + }, + { + "epoch": 0.07066, + "grad_norm": 0.6427456140518188, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 7066 + }, + { + "epoch": 0.07067, + "grad_norm": 0.561661422252655, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 7067 + }, + { + "epoch": 0.07068, + "grad_norm": 0.49639126658439636, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 7068 + }, + { + "epoch": 0.07069, + "grad_norm": 0.5630697011947632, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 7069 + }, + { + "epoch": 0.0707, + "grad_norm": 0.6418008208274841, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 7070 + }, + { + "epoch": 0.07071, + "grad_norm": 0.5743558406829834, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 7071 + }, + { + "epoch": 0.07072, + "grad_norm": 0.6298477053642273, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 7072 + }, + { + "epoch": 0.07073, + "grad_norm": 0.6862906813621521, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 7073 + }, + { + "epoch": 0.07074, + "grad_norm": 0.7200958728790283, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 7074 + }, + { + "epoch": 0.07075, + "grad_norm": 0.8121984601020813, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 7075 + }, + { + "epoch": 0.07076, + "grad_norm": 1.0043666362762451, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 7076 + }, + { + "epoch": 0.07077, + "grad_norm": 1.0342673063278198, + "learning_rate": 0.003, + "loss": 4.045, + "step": 7077 + }, + { + "epoch": 0.07078, + "grad_norm": 0.8044958710670471, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 7078 + }, + { + "epoch": 0.07079, + "grad_norm": 0.5679816007614136, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 7079 + }, + { + "epoch": 0.0708, + "grad_norm": 0.6305522918701172, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 7080 + }, + { + "epoch": 0.07081, + "grad_norm": 0.6851279139518738, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7081 + }, + { + "epoch": 0.07082, + "grad_norm": 0.7641652822494507, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 7082 + }, + { + "epoch": 0.07083, + "grad_norm": 0.7183335423469543, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 7083 + }, + { + "epoch": 0.07084, + "grad_norm": 0.6829813718795776, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 7084 + }, + { + "epoch": 0.07085, + "grad_norm": 0.8056393265724182, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 7085 + }, + { + "epoch": 0.07086, + "grad_norm": 0.8371883034706116, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 7086 + }, + { + "epoch": 0.07087, + "grad_norm": 0.8727524876594543, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 7087 + }, + { + "epoch": 0.07088, + "grad_norm": 0.9062744379043579, + "learning_rate": 0.003, + "loss": 4.024, + "step": 7088 + }, + { + "epoch": 0.07089, + "grad_norm": 0.7283234000205994, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 7089 + }, + { + "epoch": 0.0709, + "grad_norm": 0.6992966532707214, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 7090 + }, + { + "epoch": 0.07091, + "grad_norm": 0.6640942692756653, + "learning_rate": 0.003, + "loss": 4.039, + "step": 7091 + }, + { + "epoch": 0.07092, + "grad_norm": 0.5844820737838745, + "learning_rate": 0.003, + "loss": 4.038, + "step": 7092 + }, + { + "epoch": 0.07093, + "grad_norm": 0.5870398879051208, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 7093 + }, + { + "epoch": 0.07094, + "grad_norm": 0.635111391544342, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 7094 + }, + { + "epoch": 0.07095, + "grad_norm": 0.7052547335624695, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 7095 + }, + { + "epoch": 0.07096, + "grad_norm": 0.8631133437156677, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 7096 + }, + { + "epoch": 0.07097, + "grad_norm": 0.8540601134300232, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 7097 + }, + { + "epoch": 0.07098, + "grad_norm": 0.6860929727554321, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 7098 + }, + { + "epoch": 0.07099, + "grad_norm": 0.6758686304092407, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 7099 + }, + { + "epoch": 0.071, + "grad_norm": 0.6959379315376282, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 7100 + }, + { + "epoch": 0.07101, + "grad_norm": 0.6961118578910828, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 7101 + }, + { + "epoch": 0.07102, + "grad_norm": 0.7694392204284668, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 7102 + }, + { + "epoch": 0.07103, + "grad_norm": 0.8063876032829285, + "learning_rate": 0.003, + "loss": 4.067, + "step": 7103 + }, + { + "epoch": 0.07104, + "grad_norm": 0.7437668442726135, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 7104 + }, + { + "epoch": 0.07105, + "grad_norm": 0.7525627613067627, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 7105 + }, + { + "epoch": 0.07106, + "grad_norm": 0.6717851758003235, + "learning_rate": 0.003, + "loss": 4.057, + "step": 7106 + }, + { + "epoch": 0.07107, + "grad_norm": 0.6691828966140747, + "learning_rate": 0.003, + "loss": 4.059, + "step": 7107 + }, + { + "epoch": 0.07108, + "grad_norm": 0.5766210556030273, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 7108 + }, + { + "epoch": 0.07109, + "grad_norm": 0.503217875957489, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 7109 + }, + { + "epoch": 0.0711, + "grad_norm": 0.5127413272857666, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 7110 + }, + { + "epoch": 0.07111, + "grad_norm": 0.6104854941368103, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 7111 + }, + { + "epoch": 0.07112, + "grad_norm": 0.7156305313110352, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 7112 + }, + { + "epoch": 0.07113, + "grad_norm": 0.9921921491622925, + "learning_rate": 0.003, + "loss": 4.082, + "step": 7113 + }, + { + "epoch": 0.07114, + "grad_norm": 1.3377934694290161, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 7114 + }, + { + "epoch": 0.07115, + "grad_norm": 0.8231159448623657, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 7115 + }, + { + "epoch": 0.07116, + "grad_norm": 0.8249427676200867, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 7116 + }, + { + "epoch": 0.07117, + "grad_norm": 0.7995317578315735, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 7117 + }, + { + "epoch": 0.07118, + "grad_norm": 0.8792988657951355, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 7118 + }, + { + "epoch": 0.07119, + "grad_norm": 0.870874285697937, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 7119 + }, + { + "epoch": 0.0712, + "grad_norm": 0.8077058792114258, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 7120 + }, + { + "epoch": 0.07121, + "grad_norm": 0.7849383354187012, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 7121 + }, + { + "epoch": 0.07122, + "grad_norm": 0.6466601490974426, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 7122 + }, + { + "epoch": 0.07123, + "grad_norm": 0.6387749910354614, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 7123 + }, + { + "epoch": 0.07124, + "grad_norm": 0.6785484552383423, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 7124 + }, + { + "epoch": 0.07125, + "grad_norm": 0.7016876935958862, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 7125 + }, + { + "epoch": 0.07126, + "grad_norm": 0.6375389695167542, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 7126 + }, + { + "epoch": 0.07127, + "grad_norm": 0.6511231660842896, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 7127 + }, + { + "epoch": 0.07128, + "grad_norm": 0.6708038449287415, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 7128 + }, + { + "epoch": 0.07129, + "grad_norm": 0.81505286693573, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 7129 + }, + { + "epoch": 0.0713, + "grad_norm": 0.9627731442451477, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 7130 + }, + { + "epoch": 0.07131, + "grad_norm": 1.073629379272461, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 7131 + }, + { + "epoch": 0.07132, + "grad_norm": 0.7830027341842651, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 7132 + }, + { + "epoch": 0.07133, + "grad_norm": 0.6530002951622009, + "learning_rate": 0.003, + "loss": 4.08, + "step": 7133 + }, + { + "epoch": 0.07134, + "grad_norm": 0.6852246522903442, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 7134 + }, + { + "epoch": 0.07135, + "grad_norm": 0.646599292755127, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 7135 + }, + { + "epoch": 0.07136, + "grad_norm": 0.6449059247970581, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 7136 + }, + { + "epoch": 0.07137, + "grad_norm": 0.76864093542099, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 7137 + }, + { + "epoch": 0.07138, + "grad_norm": 0.853398859500885, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 7138 + }, + { + "epoch": 0.07139, + "grad_norm": 0.786717414855957, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 7139 + }, + { + "epoch": 0.0714, + "grad_norm": 0.6665356755256653, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 7140 + }, + { + "epoch": 0.07141, + "grad_norm": 0.5583412051200867, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 7141 + }, + { + "epoch": 0.07142, + "grad_norm": 0.5142695307731628, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 7142 + }, + { + "epoch": 0.07143, + "grad_norm": 0.5608883500099182, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 7143 + }, + { + "epoch": 0.07144, + "grad_norm": 0.5874366760253906, + "learning_rate": 0.003, + "loss": 4.021, + "step": 7144 + }, + { + "epoch": 0.07145, + "grad_norm": 0.7359901070594788, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 7145 + }, + { + "epoch": 0.07146, + "grad_norm": 0.8212249875068665, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 7146 + }, + { + "epoch": 0.07147, + "grad_norm": 0.8431135416030884, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 7147 + }, + { + "epoch": 0.07148, + "grad_norm": 0.7992448806762695, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 7148 + }, + { + "epoch": 0.07149, + "grad_norm": 0.7589306235313416, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 7149 + }, + { + "epoch": 0.0715, + "grad_norm": 0.9306392669677734, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 7150 + }, + { + "epoch": 0.07151, + "grad_norm": 0.9924577474594116, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 7151 + }, + { + "epoch": 0.07152, + "grad_norm": 0.883642315864563, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 7152 + }, + { + "epoch": 0.07153, + "grad_norm": 0.7332523465156555, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 7153 + }, + { + "epoch": 0.07154, + "grad_norm": 0.7003593444824219, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 7154 + }, + { + "epoch": 0.07155, + "grad_norm": 0.6480372548103333, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 7155 + }, + { + "epoch": 0.07156, + "grad_norm": 0.681715190410614, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 7156 + }, + { + "epoch": 0.07157, + "grad_norm": 0.7040699124336243, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 7157 + }, + { + "epoch": 0.07158, + "grad_norm": 0.6674701571464539, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 7158 + }, + { + "epoch": 0.07159, + "grad_norm": 0.7054368257522583, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 7159 + }, + { + "epoch": 0.0716, + "grad_norm": 0.8435150384902954, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 7160 + }, + { + "epoch": 0.07161, + "grad_norm": 0.7803497314453125, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 7161 + }, + { + "epoch": 0.07162, + "grad_norm": 0.9659714698791504, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 7162 + }, + { + "epoch": 0.07163, + "grad_norm": 1.1324249505996704, + "learning_rate": 0.003, + "loss": 4.065, + "step": 7163 + }, + { + "epoch": 0.07164, + "grad_norm": 0.8073561191558838, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 7164 + }, + { + "epoch": 0.07165, + "grad_norm": 0.6374465823173523, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 7165 + }, + { + "epoch": 0.07166, + "grad_norm": 0.6178154945373535, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 7166 + }, + { + "epoch": 0.07167, + "grad_norm": 0.683618426322937, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 7167 + }, + { + "epoch": 0.07168, + "grad_norm": 0.8149320483207703, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 7168 + }, + { + "epoch": 0.07169, + "grad_norm": 1.0499589443206787, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 7169 + }, + { + "epoch": 0.0717, + "grad_norm": 0.8370349407196045, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 7170 + }, + { + "epoch": 0.07171, + "grad_norm": 0.7322642803192139, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 7171 + }, + { + "epoch": 0.07172, + "grad_norm": 0.7073912620544434, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 7172 + }, + { + "epoch": 0.07173, + "grad_norm": 0.7349017858505249, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 7173 + }, + { + "epoch": 0.07174, + "grad_norm": 0.7687479853630066, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 7174 + }, + { + "epoch": 0.07175, + "grad_norm": 0.778110682964325, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 7175 + }, + { + "epoch": 0.07176, + "grad_norm": 0.7889383435249329, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 7176 + }, + { + "epoch": 0.07177, + "grad_norm": 0.7536366581916809, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 7177 + }, + { + "epoch": 0.07178, + "grad_norm": 0.5942474603652954, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 7178 + }, + { + "epoch": 0.07179, + "grad_norm": 0.607769787311554, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 7179 + }, + { + "epoch": 0.0718, + "grad_norm": 0.6560423970222473, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 7180 + }, + { + "epoch": 0.07181, + "grad_norm": 0.7127732038497925, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 7181 + }, + { + "epoch": 0.07182, + "grad_norm": 0.9144400954246521, + "learning_rate": 0.003, + "loss": 4.063, + "step": 7182 + }, + { + "epoch": 0.07183, + "grad_norm": 1.0960463285446167, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 7183 + }, + { + "epoch": 0.07184, + "grad_norm": 0.7976519465446472, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 7184 + }, + { + "epoch": 0.07185, + "grad_norm": 0.6410822868347168, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 7185 + }, + { + "epoch": 0.07186, + "grad_norm": 0.5536119937896729, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 7186 + }, + { + "epoch": 0.07187, + "grad_norm": 0.562323808670044, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 7187 + }, + { + "epoch": 0.07188, + "grad_norm": 0.5706769227981567, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 7188 + }, + { + "epoch": 0.07189, + "grad_norm": 0.6065142154693604, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 7189 + }, + { + "epoch": 0.0719, + "grad_norm": 0.6539909839630127, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 7190 + }, + { + "epoch": 0.07191, + "grad_norm": 0.716647207736969, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 7191 + }, + { + "epoch": 0.07192, + "grad_norm": 0.7328898310661316, + "learning_rate": 0.003, + "loss": 4.028, + "step": 7192 + }, + { + "epoch": 0.07193, + "grad_norm": 0.7779203653335571, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 7193 + }, + { + "epoch": 0.07194, + "grad_norm": 0.7531735897064209, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 7194 + }, + { + "epoch": 0.07195, + "grad_norm": 0.7268919944763184, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 7195 + }, + { + "epoch": 0.07196, + "grad_norm": 0.7242192625999451, + "learning_rate": 0.003, + "loss": 4.037, + "step": 7196 + }, + { + "epoch": 0.07197, + "grad_norm": 0.7115669846534729, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 7197 + }, + { + "epoch": 0.07198, + "grad_norm": 0.6028633117675781, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 7198 + }, + { + "epoch": 0.07199, + "grad_norm": 0.6346673965454102, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 7199 + }, + { + "epoch": 0.072, + "grad_norm": 0.6983155012130737, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 7200 + }, + { + "epoch": 0.07201, + "grad_norm": 0.6601306796073914, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 7201 + }, + { + "epoch": 0.07202, + "grad_norm": 0.7019174098968506, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 7202 + }, + { + "epoch": 0.07203, + "grad_norm": 0.7765396237373352, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 7203 + }, + { + "epoch": 0.07204, + "grad_norm": 0.8035591244697571, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 7204 + }, + { + "epoch": 0.07205, + "grad_norm": 0.8030667901039124, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 7205 + }, + { + "epoch": 0.07206, + "grad_norm": 0.7670336365699768, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 7206 + }, + { + "epoch": 0.07207, + "grad_norm": 0.8267470598220825, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 7207 + }, + { + "epoch": 0.07208, + "grad_norm": 0.9432637691497803, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 7208 + }, + { + "epoch": 0.07209, + "grad_norm": 1.0053582191467285, + "learning_rate": 0.003, + "loss": 4.05, + "step": 7209 + }, + { + "epoch": 0.0721, + "grad_norm": 0.8599228262901306, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 7210 + }, + { + "epoch": 0.07211, + "grad_norm": 0.814187228679657, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 7211 + }, + { + "epoch": 0.07212, + "grad_norm": 0.7470682859420776, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 7212 + }, + { + "epoch": 0.07213, + "grad_norm": 0.6890302300453186, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 7213 + }, + { + "epoch": 0.07214, + "grad_norm": 0.6987749934196472, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 7214 + }, + { + "epoch": 0.07215, + "grad_norm": 0.6948664784431458, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 7215 + }, + { + "epoch": 0.07216, + "grad_norm": 0.739776074886322, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 7216 + }, + { + "epoch": 0.07217, + "grad_norm": 0.7527135014533997, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 7217 + }, + { + "epoch": 0.07218, + "grad_norm": 0.9124861359596252, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 7218 + }, + { + "epoch": 0.07219, + "grad_norm": 1.1901919841766357, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 7219 + }, + { + "epoch": 0.0722, + "grad_norm": 0.7887942790985107, + "learning_rate": 0.003, + "loss": 4.063, + "step": 7220 + }, + { + "epoch": 0.07221, + "grad_norm": 0.667094886302948, + "learning_rate": 0.003, + "loss": 4.061, + "step": 7221 + }, + { + "epoch": 0.07222, + "grad_norm": 0.6238069534301758, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 7222 + }, + { + "epoch": 0.07223, + "grad_norm": 0.6278685331344604, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 7223 + }, + { + "epoch": 0.07224, + "grad_norm": 0.5980105400085449, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 7224 + }, + { + "epoch": 0.07225, + "grad_norm": 0.5822548270225525, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 7225 + }, + { + "epoch": 0.07226, + "grad_norm": 0.705129086971283, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 7226 + }, + { + "epoch": 0.07227, + "grad_norm": 0.8607189655303955, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 7227 + }, + { + "epoch": 0.07228, + "grad_norm": 1.0843205451965332, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 7228 + }, + { + "epoch": 0.07229, + "grad_norm": 0.9284544587135315, + "learning_rate": 0.003, + "loss": 4.051, + "step": 7229 + }, + { + "epoch": 0.0723, + "grad_norm": 0.7482180595397949, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 7230 + }, + { + "epoch": 0.07231, + "grad_norm": 0.7119454741477966, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 7231 + }, + { + "epoch": 0.07232, + "grad_norm": 0.8325626850128174, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 7232 + }, + { + "epoch": 0.07233, + "grad_norm": 0.847053050994873, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 7233 + }, + { + "epoch": 0.07234, + "grad_norm": 0.8324984908103943, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 7234 + }, + { + "epoch": 0.07235, + "grad_norm": 0.9077214598655701, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 7235 + }, + { + "epoch": 0.07236, + "grad_norm": 0.7818212509155273, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 7236 + }, + { + "epoch": 0.07237, + "grad_norm": 0.7284092307090759, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 7237 + }, + { + "epoch": 0.07238, + "grad_norm": 0.706311047077179, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 7238 + }, + { + "epoch": 0.07239, + "grad_norm": 0.8178088068962097, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 7239 + }, + { + "epoch": 0.0724, + "grad_norm": 0.8817880749702454, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 7240 + }, + { + "epoch": 0.07241, + "grad_norm": 0.8495557308197021, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 7241 + }, + { + "epoch": 0.07242, + "grad_norm": 0.766780436038971, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 7242 + }, + { + "epoch": 0.07243, + "grad_norm": 0.7193701863288879, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 7243 + }, + { + "epoch": 0.07244, + "grad_norm": 0.6461489200592041, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 7244 + }, + { + "epoch": 0.07245, + "grad_norm": 0.5884445905685425, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 7245 + }, + { + "epoch": 0.07246, + "grad_norm": 0.5750178694725037, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 7246 + }, + { + "epoch": 0.07247, + "grad_norm": 0.6441517472267151, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 7247 + }, + { + "epoch": 0.07248, + "grad_norm": 0.6365780830383301, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 7248 + }, + { + "epoch": 0.07249, + "grad_norm": 0.6930205821990967, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 7249 + }, + { + "epoch": 0.0725, + "grad_norm": 0.7784088850021362, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 7250 + }, + { + "epoch": 0.07251, + "grad_norm": 0.9219940900802612, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 7251 + }, + { + "epoch": 0.07252, + "grad_norm": 1.1105650663375854, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 7252 + }, + { + "epoch": 0.07253, + "grad_norm": 0.7862104177474976, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 7253 + }, + { + "epoch": 0.07254, + "grad_norm": 0.6501257419586182, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 7254 + }, + { + "epoch": 0.07255, + "grad_norm": 0.7006444334983826, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 7255 + }, + { + "epoch": 0.07256, + "grad_norm": 0.7784441113471985, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 7256 + }, + { + "epoch": 0.07257, + "grad_norm": 0.7117348313331604, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 7257 + }, + { + "epoch": 0.07258, + "grad_norm": 0.6730650663375854, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 7258 + }, + { + "epoch": 0.07259, + "grad_norm": 0.6756656765937805, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 7259 + }, + { + "epoch": 0.0726, + "grad_norm": 0.7697281837463379, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 7260 + }, + { + "epoch": 0.07261, + "grad_norm": 0.7414706945419312, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 7261 + }, + { + "epoch": 0.07262, + "grad_norm": 0.8930975794792175, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 7262 + }, + { + "epoch": 0.07263, + "grad_norm": 0.9588114023208618, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 7263 + }, + { + "epoch": 0.07264, + "grad_norm": 0.765650749206543, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 7264 + }, + { + "epoch": 0.07265, + "grad_norm": 0.6993110775947571, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 7265 + }, + { + "epoch": 0.07266, + "grad_norm": 0.6595311164855957, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 7266 + }, + { + "epoch": 0.07267, + "grad_norm": 0.6637097001075745, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 7267 + }, + { + "epoch": 0.07268, + "grad_norm": 0.6798438429832458, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 7268 + }, + { + "epoch": 0.07269, + "grad_norm": 0.7596190571784973, + "learning_rate": 0.003, + "loss": 4.052, + "step": 7269 + }, + { + "epoch": 0.0727, + "grad_norm": 0.7761021852493286, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 7270 + }, + { + "epoch": 0.07271, + "grad_norm": 0.8009915947914124, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 7271 + }, + { + "epoch": 0.07272, + "grad_norm": 1.0279133319854736, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 7272 + }, + { + "epoch": 0.07273, + "grad_norm": 1.1666834354400635, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 7273 + }, + { + "epoch": 0.07274, + "grad_norm": 0.6210851073265076, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 7274 + }, + { + "epoch": 0.07275, + "grad_norm": 0.6968746781349182, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 7275 + }, + { + "epoch": 0.07276, + "grad_norm": 0.9396815299987793, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 7276 + }, + { + "epoch": 0.07277, + "grad_norm": 0.9510892033576965, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 7277 + }, + { + "epoch": 0.07278, + "grad_norm": 0.8797145485877991, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 7278 + }, + { + "epoch": 0.07279, + "grad_norm": 0.7459223866462708, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 7279 + }, + { + "epoch": 0.0728, + "grad_norm": 0.7744358777999878, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 7280 + }, + { + "epoch": 0.07281, + "grad_norm": 0.7015716433525085, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 7281 + }, + { + "epoch": 0.07282, + "grad_norm": 0.7247535586357117, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 7282 + }, + { + "epoch": 0.07283, + "grad_norm": 0.726428210735321, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 7283 + }, + { + "epoch": 0.07284, + "grad_norm": 0.7073158621788025, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 7284 + }, + { + "epoch": 0.07285, + "grad_norm": 0.682987630367279, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 7285 + }, + { + "epoch": 0.07286, + "grad_norm": 0.6815469264984131, + "learning_rate": 0.003, + "loss": 4.033, + "step": 7286 + }, + { + "epoch": 0.07287, + "grad_norm": 0.7032678127288818, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 7287 + }, + { + "epoch": 0.07288, + "grad_norm": 0.6728482246398926, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 7288 + }, + { + "epoch": 0.07289, + "grad_norm": 0.6951834559440613, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 7289 + }, + { + "epoch": 0.0729, + "grad_norm": 0.638177752494812, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 7290 + }, + { + "epoch": 0.07291, + "grad_norm": 0.6703484654426575, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 7291 + }, + { + "epoch": 0.07292, + "grad_norm": 0.6980002522468567, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 7292 + }, + { + "epoch": 0.07293, + "grad_norm": 0.6970522403717041, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 7293 + }, + { + "epoch": 0.07294, + "grad_norm": 0.6544967293739319, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 7294 + }, + { + "epoch": 0.07295, + "grad_norm": 0.6716231107711792, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 7295 + }, + { + "epoch": 0.07296, + "grad_norm": 0.8019461631774902, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 7296 + }, + { + "epoch": 0.07297, + "grad_norm": 0.9077995419502258, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 7297 + }, + { + "epoch": 0.07298, + "grad_norm": 0.8515841364860535, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 7298 + }, + { + "epoch": 0.07299, + "grad_norm": 0.6916643977165222, + "learning_rate": 0.003, + "loss": 4.039, + "step": 7299 + }, + { + "epoch": 0.073, + "grad_norm": 0.6934566497802734, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 7300 + }, + { + "epoch": 0.07301, + "grad_norm": 0.6328970193862915, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 7301 + }, + { + "epoch": 0.07302, + "grad_norm": 0.6315169930458069, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 7302 + }, + { + "epoch": 0.07303, + "grad_norm": 0.731637179851532, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 7303 + }, + { + "epoch": 0.07304, + "grad_norm": 0.7277827858924866, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 7304 + }, + { + "epoch": 0.07305, + "grad_norm": 0.7023566961288452, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 7305 + }, + { + "epoch": 0.07306, + "grad_norm": 0.6430983543395996, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 7306 + }, + { + "epoch": 0.07307, + "grad_norm": 0.708091676235199, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 7307 + }, + { + "epoch": 0.07308, + "grad_norm": 0.9044761657714844, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 7308 + }, + { + "epoch": 0.07309, + "grad_norm": 1.0247414112091064, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 7309 + }, + { + "epoch": 0.0731, + "grad_norm": 1.082109808921814, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 7310 + }, + { + "epoch": 0.07311, + "grad_norm": 0.8486955165863037, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 7311 + }, + { + "epoch": 0.07312, + "grad_norm": 0.9376227259635925, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 7312 + }, + { + "epoch": 0.07313, + "grad_norm": 0.7661151885986328, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 7313 + }, + { + "epoch": 0.07314, + "grad_norm": 0.7679696679115295, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 7314 + }, + { + "epoch": 0.07315, + "grad_norm": 0.7783799767494202, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 7315 + }, + { + "epoch": 0.07316, + "grad_norm": 0.7211359739303589, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 7316 + }, + { + "epoch": 0.07317, + "grad_norm": 0.6663383841514587, + "learning_rate": 0.003, + "loss": 4.053, + "step": 7317 + }, + { + "epoch": 0.07318, + "grad_norm": 0.5428328514099121, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 7318 + }, + { + "epoch": 0.07319, + "grad_norm": 0.5540145039558411, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 7319 + }, + { + "epoch": 0.0732, + "grad_norm": 0.6541301608085632, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 7320 + }, + { + "epoch": 0.07321, + "grad_norm": 0.8049659729003906, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 7321 + }, + { + "epoch": 0.07322, + "grad_norm": 0.926133394241333, + "learning_rate": 0.003, + "loss": 4.065, + "step": 7322 + }, + { + "epoch": 0.07323, + "grad_norm": 0.8449392318725586, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 7323 + }, + { + "epoch": 0.07324, + "grad_norm": 0.7204037308692932, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 7324 + }, + { + "epoch": 0.07325, + "grad_norm": 0.6747505068778992, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 7325 + }, + { + "epoch": 0.07326, + "grad_norm": 0.6664218306541443, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 7326 + }, + { + "epoch": 0.07327, + "grad_norm": 0.781737208366394, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 7327 + }, + { + "epoch": 0.07328, + "grad_norm": 0.8926900625228882, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 7328 + }, + { + "epoch": 0.07329, + "grad_norm": 0.911156177520752, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 7329 + }, + { + "epoch": 0.0733, + "grad_norm": 0.7268263101577759, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 7330 + }, + { + "epoch": 0.07331, + "grad_norm": 0.6712400317192078, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 7331 + }, + { + "epoch": 0.07332, + "grad_norm": 0.5922292470932007, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 7332 + }, + { + "epoch": 0.07333, + "grad_norm": 0.6073193550109863, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 7333 + }, + { + "epoch": 0.07334, + "grad_norm": 0.5941932201385498, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 7334 + }, + { + "epoch": 0.07335, + "grad_norm": 0.6539658904075623, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 7335 + }, + { + "epoch": 0.07336, + "grad_norm": 0.6258936524391174, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 7336 + }, + { + "epoch": 0.07337, + "grad_norm": 0.6982399225234985, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 7337 + }, + { + "epoch": 0.07338, + "grad_norm": 0.7497730851173401, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 7338 + }, + { + "epoch": 0.07339, + "grad_norm": 0.8154767751693726, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 7339 + }, + { + "epoch": 0.0734, + "grad_norm": 0.8396859765052795, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 7340 + }, + { + "epoch": 0.07341, + "grad_norm": 0.8484392762184143, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 7341 + }, + { + "epoch": 0.07342, + "grad_norm": 0.769058108329773, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 7342 + }, + { + "epoch": 0.07343, + "grad_norm": 0.6709862351417542, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 7343 + }, + { + "epoch": 0.07344, + "grad_norm": 0.6540923714637756, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 7344 + }, + { + "epoch": 0.07345, + "grad_norm": 0.6276072859764099, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 7345 + }, + { + "epoch": 0.07346, + "grad_norm": 0.6179201006889343, + "learning_rate": 0.003, + "loss": 4.011, + "step": 7346 + }, + { + "epoch": 0.07347, + "grad_norm": 0.5960707664489746, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 7347 + }, + { + "epoch": 0.07348, + "grad_norm": 0.5133612751960754, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 7348 + }, + { + "epoch": 0.07349, + "grad_norm": 0.47574296593666077, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 7349 + }, + { + "epoch": 0.0735, + "grad_norm": 0.5420452952384949, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 7350 + }, + { + "epoch": 0.07351, + "grad_norm": 0.6298367977142334, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 7351 + }, + { + "epoch": 0.07352, + "grad_norm": 0.6977446675300598, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 7352 + }, + { + "epoch": 0.07353, + "grad_norm": 0.7407834529876709, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 7353 + }, + { + "epoch": 0.07354, + "grad_norm": 0.8189613819122314, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 7354 + }, + { + "epoch": 0.07355, + "grad_norm": 0.9231334924697876, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 7355 + }, + { + "epoch": 0.07356, + "grad_norm": 0.90891432762146, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 7356 + }, + { + "epoch": 0.07357, + "grad_norm": 0.8406992554664612, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 7357 + }, + { + "epoch": 0.07358, + "grad_norm": 0.8127254843711853, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 7358 + }, + { + "epoch": 0.07359, + "grad_norm": 0.8756438493728638, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 7359 + }, + { + "epoch": 0.0736, + "grad_norm": 0.9333860278129578, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 7360 + }, + { + "epoch": 0.07361, + "grad_norm": 0.9368488788604736, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 7361 + }, + { + "epoch": 0.07362, + "grad_norm": 0.827850878238678, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 7362 + }, + { + "epoch": 0.07363, + "grad_norm": 0.7587769031524658, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 7363 + }, + { + "epoch": 0.07364, + "grad_norm": 0.881401777267456, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 7364 + }, + { + "epoch": 0.07365, + "grad_norm": 1.1441792249679565, + "learning_rate": 0.003, + "loss": 4.028, + "step": 7365 + }, + { + "epoch": 0.07366, + "grad_norm": 1.050272822380066, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 7366 + }, + { + "epoch": 0.07367, + "grad_norm": 0.9028184413909912, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 7367 + }, + { + "epoch": 0.07368, + "grad_norm": 0.7709707617759705, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 7368 + }, + { + "epoch": 0.07369, + "grad_norm": 0.8372074365615845, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 7369 + }, + { + "epoch": 0.0737, + "grad_norm": 0.8988859057426453, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 7370 + }, + { + "epoch": 0.07371, + "grad_norm": 0.9398235082626343, + "learning_rate": 0.003, + "loss": 4.09, + "step": 7371 + }, + { + "epoch": 0.07372, + "grad_norm": 0.979543924331665, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 7372 + }, + { + "epoch": 0.07373, + "grad_norm": 0.940352201461792, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 7373 + }, + { + "epoch": 0.07374, + "grad_norm": 0.956783652305603, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 7374 + }, + { + "epoch": 0.07375, + "grad_norm": 0.842356264591217, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 7375 + }, + { + "epoch": 0.07376, + "grad_norm": 0.9043733477592468, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 7376 + }, + { + "epoch": 0.07377, + "grad_norm": 0.9078689217567444, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 7377 + }, + { + "epoch": 0.07378, + "grad_norm": 1.0029642581939697, + "learning_rate": 0.003, + "loss": 4.084, + "step": 7378 + }, + { + "epoch": 0.07379, + "grad_norm": 1.116869568824768, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 7379 + }, + { + "epoch": 0.0738, + "grad_norm": 0.9037948250770569, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 7380 + }, + { + "epoch": 0.07381, + "grad_norm": 0.8272243738174438, + "learning_rate": 0.003, + "loss": 4.05, + "step": 7381 + }, + { + "epoch": 0.07382, + "grad_norm": 0.7809496521949768, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 7382 + }, + { + "epoch": 0.07383, + "grad_norm": 0.915755033493042, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 7383 + }, + { + "epoch": 0.07384, + "grad_norm": 0.9646238684654236, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 7384 + }, + { + "epoch": 0.07385, + "grad_norm": 0.7966357469558716, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 7385 + }, + { + "epoch": 0.07386, + "grad_norm": 0.7612378597259521, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 7386 + }, + { + "epoch": 0.07387, + "grad_norm": 0.7063902616500854, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 7387 + }, + { + "epoch": 0.07388, + "grad_norm": 0.7505108118057251, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 7388 + }, + { + "epoch": 0.07389, + "grad_norm": 0.6540176868438721, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7389 + }, + { + "epoch": 0.0739, + "grad_norm": 0.5906816720962524, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 7390 + }, + { + "epoch": 0.07391, + "grad_norm": 0.6899387240409851, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 7391 + }, + { + "epoch": 0.07392, + "grad_norm": 0.7748689651489258, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 7392 + }, + { + "epoch": 0.07393, + "grad_norm": 0.7996538877487183, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 7393 + }, + { + "epoch": 0.07394, + "grad_norm": 0.7616129517555237, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 7394 + }, + { + "epoch": 0.07395, + "grad_norm": 0.6533021330833435, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 7395 + }, + { + "epoch": 0.07396, + "grad_norm": 0.4827071726322174, + "learning_rate": 0.003, + "loss": 4.045, + "step": 7396 + }, + { + "epoch": 0.07397, + "grad_norm": 0.530000627040863, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 7397 + }, + { + "epoch": 0.07398, + "grad_norm": 0.5404955148696899, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 7398 + }, + { + "epoch": 0.07399, + "grad_norm": 0.5249134302139282, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 7399 + }, + { + "epoch": 0.074, + "grad_norm": 0.6226985454559326, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 7400 + }, + { + "epoch": 0.07401, + "grad_norm": 0.6949280500411987, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 7401 + }, + { + "epoch": 0.07402, + "grad_norm": 0.6506314873695374, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 7402 + }, + { + "epoch": 0.07403, + "grad_norm": 0.5710362792015076, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 7403 + }, + { + "epoch": 0.07404, + "grad_norm": 0.5003657341003418, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 7404 + }, + { + "epoch": 0.07405, + "grad_norm": 0.4554838538169861, + "learning_rate": 0.003, + "loss": 4.034, + "step": 7405 + }, + { + "epoch": 0.07406, + "grad_norm": 0.5003705024719238, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 7406 + }, + { + "epoch": 0.07407, + "grad_norm": 0.5069988369941711, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 7407 + }, + { + "epoch": 0.07408, + "grad_norm": 0.4921700656414032, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 7408 + }, + { + "epoch": 0.07409, + "grad_norm": 0.5235347151756287, + "learning_rate": 0.003, + "loss": 4.03, + "step": 7409 + }, + { + "epoch": 0.0741, + "grad_norm": 0.532010555267334, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 7410 + }, + { + "epoch": 0.07411, + "grad_norm": 0.5434505343437195, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 7411 + }, + { + "epoch": 0.07412, + "grad_norm": 0.6439001560211182, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 7412 + }, + { + "epoch": 0.07413, + "grad_norm": 0.7864586710929871, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 7413 + }, + { + "epoch": 0.07414, + "grad_norm": 1.0050026178359985, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 7414 + }, + { + "epoch": 0.07415, + "grad_norm": 1.035174012184143, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 7415 + }, + { + "epoch": 0.07416, + "grad_norm": 0.7707844376564026, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 7416 + }, + { + "epoch": 0.07417, + "grad_norm": 0.7662503719329834, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 7417 + }, + { + "epoch": 0.07418, + "grad_norm": 0.8928897380828857, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 7418 + }, + { + "epoch": 0.07419, + "grad_norm": 0.8406341671943665, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 7419 + }, + { + "epoch": 0.0742, + "grad_norm": 0.8542524576187134, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 7420 + }, + { + "epoch": 0.07421, + "grad_norm": 0.8937854766845703, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 7421 + }, + { + "epoch": 0.07422, + "grad_norm": 0.9040619730949402, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 7422 + }, + { + "epoch": 0.07423, + "grad_norm": 0.7693426012992859, + "learning_rate": 0.003, + "loss": 4.032, + "step": 7423 + }, + { + "epoch": 0.07424, + "grad_norm": 0.7264552116394043, + "learning_rate": 0.003, + "loss": 4.06, + "step": 7424 + }, + { + "epoch": 0.07425, + "grad_norm": 0.7355172038078308, + "learning_rate": 0.003, + "loss": 4.024, + "step": 7425 + }, + { + "epoch": 0.07426, + "grad_norm": 0.7334502339363098, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 7426 + }, + { + "epoch": 0.07427, + "grad_norm": 0.8754525780677795, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 7427 + }, + { + "epoch": 0.07428, + "grad_norm": 0.9548163414001465, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 7428 + }, + { + "epoch": 0.07429, + "grad_norm": 0.9186201095581055, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 7429 + }, + { + "epoch": 0.0743, + "grad_norm": 1.0932062864303589, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 7430 + }, + { + "epoch": 0.07431, + "grad_norm": 0.7931951284408569, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 7431 + }, + { + "epoch": 0.07432, + "grad_norm": 0.6674275994300842, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 7432 + }, + { + "epoch": 0.07433, + "grad_norm": 0.5529371500015259, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 7433 + }, + { + "epoch": 0.07434, + "grad_norm": 0.5146396160125732, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 7434 + }, + { + "epoch": 0.07435, + "grad_norm": 0.5399970412254333, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 7435 + }, + { + "epoch": 0.07436, + "grad_norm": 0.6538156867027283, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 7436 + }, + { + "epoch": 0.07437, + "grad_norm": 0.8563846349716187, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 7437 + }, + { + "epoch": 0.07438, + "grad_norm": 1.1433502435684204, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 7438 + }, + { + "epoch": 0.07439, + "grad_norm": 0.8334239721298218, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 7439 + }, + { + "epoch": 0.0744, + "grad_norm": 0.6448179483413696, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7440 + }, + { + "epoch": 0.07441, + "grad_norm": 0.7158012986183167, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 7441 + }, + { + "epoch": 0.07442, + "grad_norm": 0.8154329061508179, + "learning_rate": 0.003, + "loss": 4.061, + "step": 7442 + }, + { + "epoch": 0.07443, + "grad_norm": 0.8508999943733215, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 7443 + }, + { + "epoch": 0.07444, + "grad_norm": 0.8207703828811646, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 7444 + }, + { + "epoch": 0.07445, + "grad_norm": 0.7776240110397339, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 7445 + }, + { + "epoch": 0.07446, + "grad_norm": 0.7464262247085571, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 7446 + }, + { + "epoch": 0.07447, + "grad_norm": 0.6316878795623779, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 7447 + }, + { + "epoch": 0.07448, + "grad_norm": 0.6105993390083313, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 7448 + }, + { + "epoch": 0.07449, + "grad_norm": 0.6523858308792114, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 7449 + }, + { + "epoch": 0.0745, + "grad_norm": 0.7932419180870056, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 7450 + }, + { + "epoch": 0.07451, + "grad_norm": 0.736053466796875, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 7451 + }, + { + "epoch": 0.07452, + "grad_norm": 0.7587615251541138, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 7452 + }, + { + "epoch": 0.07453, + "grad_norm": 0.7929425239562988, + "learning_rate": 0.003, + "loss": 4.052, + "step": 7453 + }, + { + "epoch": 0.07454, + "grad_norm": 0.8557668924331665, + "learning_rate": 0.003, + "loss": 4.03, + "step": 7454 + }, + { + "epoch": 0.07455, + "grad_norm": 0.8376034498214722, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 7455 + }, + { + "epoch": 0.07456, + "grad_norm": 0.6638903021812439, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 7456 + }, + { + "epoch": 0.07457, + "grad_norm": 0.5936045050621033, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 7457 + }, + { + "epoch": 0.07458, + "grad_norm": 0.6461426615715027, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 7458 + }, + { + "epoch": 0.07459, + "grad_norm": 0.7057914137840271, + "learning_rate": 0.003, + "loss": 4.031, + "step": 7459 + }, + { + "epoch": 0.0746, + "grad_norm": 0.7673705220222473, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 7460 + }, + { + "epoch": 0.07461, + "grad_norm": 0.7809221744537354, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 7461 + }, + { + "epoch": 0.07462, + "grad_norm": 0.7208760976791382, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 7462 + }, + { + "epoch": 0.07463, + "grad_norm": 0.5994004011154175, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 7463 + }, + { + "epoch": 0.07464, + "grad_norm": 0.6108066439628601, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 7464 + }, + { + "epoch": 0.07465, + "grad_norm": 0.7504144906997681, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 7465 + }, + { + "epoch": 0.07466, + "grad_norm": 0.8534754514694214, + "learning_rate": 0.003, + "loss": 4.063, + "step": 7466 + }, + { + "epoch": 0.07467, + "grad_norm": 0.8829362988471985, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 7467 + }, + { + "epoch": 0.07468, + "grad_norm": 0.8858988881111145, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 7468 + }, + { + "epoch": 0.07469, + "grad_norm": 0.8821552395820618, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 7469 + }, + { + "epoch": 0.0747, + "grad_norm": 0.8428568840026855, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 7470 + }, + { + "epoch": 0.07471, + "grad_norm": 0.7221834659576416, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7471 + }, + { + "epoch": 0.07472, + "grad_norm": 0.6946365833282471, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 7472 + }, + { + "epoch": 0.07473, + "grad_norm": 0.6493435502052307, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 7473 + }, + { + "epoch": 0.07474, + "grad_norm": 0.7220835089683533, + "learning_rate": 0.003, + "loss": 4.04, + "step": 7474 + }, + { + "epoch": 0.07475, + "grad_norm": 0.7649142146110535, + "learning_rate": 0.003, + "loss": 4.054, + "step": 7475 + }, + { + "epoch": 0.07476, + "grad_norm": 0.914003312587738, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 7476 + }, + { + "epoch": 0.07477, + "grad_norm": 1.0200934410095215, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 7477 + }, + { + "epoch": 0.07478, + "grad_norm": 1.0416991710662842, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 7478 + }, + { + "epoch": 0.07479, + "grad_norm": 0.8721848130226135, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 7479 + }, + { + "epoch": 0.0748, + "grad_norm": 0.7051395177841187, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 7480 + }, + { + "epoch": 0.07481, + "grad_norm": 0.6780185103416443, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 7481 + }, + { + "epoch": 0.07482, + "grad_norm": 0.6697372198104858, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 7482 + }, + { + "epoch": 0.07483, + "grad_norm": 0.6301858425140381, + "learning_rate": 0.003, + "loss": 4.07, + "step": 7483 + }, + { + "epoch": 0.07484, + "grad_norm": 0.6258836984634399, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 7484 + }, + { + "epoch": 0.07485, + "grad_norm": 0.667647123336792, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 7485 + }, + { + "epoch": 0.07486, + "grad_norm": 0.8375819325447083, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 7486 + }, + { + "epoch": 0.07487, + "grad_norm": 1.0791035890579224, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 7487 + }, + { + "epoch": 0.07488, + "grad_norm": 0.9067569375038147, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 7488 + }, + { + "epoch": 0.07489, + "grad_norm": 0.6088890433311462, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 7489 + }, + { + "epoch": 0.0749, + "grad_norm": 0.5461151599884033, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 7490 + }, + { + "epoch": 0.07491, + "grad_norm": 0.5697436928749084, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 7491 + }, + { + "epoch": 0.07492, + "grad_norm": 0.5373200178146362, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 7492 + }, + { + "epoch": 0.07493, + "grad_norm": 0.45845916867256165, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 7493 + }, + { + "epoch": 0.07494, + "grad_norm": 0.4908093512058258, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 7494 + }, + { + "epoch": 0.07495, + "grad_norm": 0.4465610682964325, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 7495 + }, + { + "epoch": 0.07496, + "grad_norm": 0.44271135330200195, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 7496 + }, + { + "epoch": 0.07497, + "grad_norm": 0.4760777950286865, + "learning_rate": 0.003, + "loss": 4.037, + "step": 7497 + }, + { + "epoch": 0.07498, + "grad_norm": 0.5491810441017151, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 7498 + }, + { + "epoch": 0.07499, + "grad_norm": 0.5653870105743408, + "learning_rate": 0.003, + "loss": 4.03, + "step": 7499 + }, + { + "epoch": 0.075, + "grad_norm": 0.5969352722167969, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 7500 + }, + { + "epoch": 0.07501, + "grad_norm": 0.6566980481147766, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 7501 + }, + { + "epoch": 0.07502, + "grad_norm": 0.7576538920402527, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 7502 + }, + { + "epoch": 0.07503, + "grad_norm": 0.9570623636245728, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 7503 + }, + { + "epoch": 0.07504, + "grad_norm": 1.1501446962356567, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 7504 + }, + { + "epoch": 0.07505, + "grad_norm": 0.821244478225708, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 7505 + }, + { + "epoch": 0.07506, + "grad_norm": 0.9706292152404785, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 7506 + }, + { + "epoch": 0.07507, + "grad_norm": 1.2102551460266113, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 7507 + }, + { + "epoch": 0.07508, + "grad_norm": 0.8915047645568848, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 7508 + }, + { + "epoch": 0.07509, + "grad_norm": 0.795073390007019, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 7509 + }, + { + "epoch": 0.0751, + "grad_norm": 0.7662889957427979, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 7510 + }, + { + "epoch": 0.07511, + "grad_norm": 0.7474156022071838, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 7511 + }, + { + "epoch": 0.07512, + "grad_norm": 0.8176290392875671, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 7512 + }, + { + "epoch": 0.07513, + "grad_norm": 0.9439889192581177, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 7513 + }, + { + "epoch": 0.07514, + "grad_norm": 1.1200593709945679, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 7514 + }, + { + "epoch": 0.07515, + "grad_norm": 0.9754398465156555, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 7515 + }, + { + "epoch": 0.07516, + "grad_norm": 0.8390057682991028, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 7516 + }, + { + "epoch": 0.07517, + "grad_norm": 0.7585924863815308, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 7517 + }, + { + "epoch": 0.07518, + "grad_norm": 0.8066619038581848, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 7518 + }, + { + "epoch": 0.07519, + "grad_norm": 0.8263963460922241, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 7519 + }, + { + "epoch": 0.0752, + "grad_norm": 0.7621636390686035, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 7520 + }, + { + "epoch": 0.07521, + "grad_norm": 0.8395116329193115, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 7521 + }, + { + "epoch": 0.07522, + "grad_norm": 0.855266273021698, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 7522 + }, + { + "epoch": 0.07523, + "grad_norm": 0.9209402203559875, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 7523 + }, + { + "epoch": 0.07524, + "grad_norm": 0.8674631714820862, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 7524 + }, + { + "epoch": 0.07525, + "grad_norm": 0.9512166380882263, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 7525 + }, + { + "epoch": 0.07526, + "grad_norm": 1.0687352418899536, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 7526 + }, + { + "epoch": 0.07527, + "grad_norm": 0.8810518383979797, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 7527 + }, + { + "epoch": 0.07528, + "grad_norm": 0.8249692320823669, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 7528 + }, + { + "epoch": 0.07529, + "grad_norm": 0.779848039150238, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 7529 + }, + { + "epoch": 0.0753, + "grad_norm": 0.5802583694458008, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 7530 + }, + { + "epoch": 0.07531, + "grad_norm": 0.5488690137863159, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 7531 + }, + { + "epoch": 0.07532, + "grad_norm": 0.6669039130210876, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 7532 + }, + { + "epoch": 0.07533, + "grad_norm": 0.7888681292533875, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 7533 + }, + { + "epoch": 0.07534, + "grad_norm": 0.8699842691421509, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7534 + }, + { + "epoch": 0.07535, + "grad_norm": 0.8488181829452515, + "learning_rate": 0.003, + "loss": 4.059, + "step": 7535 + }, + { + "epoch": 0.07536, + "grad_norm": 0.7273427844047546, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 7536 + }, + { + "epoch": 0.07537, + "grad_norm": 0.6958652138710022, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 7537 + }, + { + "epoch": 0.07538, + "grad_norm": 0.7339373826980591, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 7538 + }, + { + "epoch": 0.07539, + "grad_norm": 0.5761485695838928, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 7539 + }, + { + "epoch": 0.0754, + "grad_norm": 0.6199266314506531, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 7540 + }, + { + "epoch": 0.07541, + "grad_norm": 0.5513209104537964, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 7541 + }, + { + "epoch": 0.07542, + "grad_norm": 0.6161065697669983, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 7542 + }, + { + "epoch": 0.07543, + "grad_norm": 0.7098742723464966, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 7543 + }, + { + "epoch": 0.07544, + "grad_norm": 0.7922773957252502, + "learning_rate": 0.003, + "loss": 4.057, + "step": 7544 + }, + { + "epoch": 0.07545, + "grad_norm": 0.7381566166877747, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 7545 + }, + { + "epoch": 0.07546, + "grad_norm": 0.6755424737930298, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 7546 + }, + { + "epoch": 0.07547, + "grad_norm": 0.8436152338981628, + "learning_rate": 0.003, + "loss": 4.059, + "step": 7547 + }, + { + "epoch": 0.07548, + "grad_norm": 1.0440216064453125, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 7548 + }, + { + "epoch": 0.07549, + "grad_norm": 1.2041999101638794, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 7549 + }, + { + "epoch": 0.0755, + "grad_norm": 0.8359081745147705, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 7550 + }, + { + "epoch": 0.07551, + "grad_norm": 0.7312293648719788, + "learning_rate": 0.003, + "loss": 4.062, + "step": 7551 + }, + { + "epoch": 0.07552, + "grad_norm": 0.6296018362045288, + "learning_rate": 0.003, + "loss": 4.058, + "step": 7552 + }, + { + "epoch": 0.07553, + "grad_norm": 0.6481240391731262, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 7553 + }, + { + "epoch": 0.07554, + "grad_norm": 0.6704967021942139, + "learning_rate": 0.003, + "loss": 4.059, + "step": 7554 + }, + { + "epoch": 0.07555, + "grad_norm": 0.7643706202507019, + "learning_rate": 0.003, + "loss": 4.049, + "step": 7555 + }, + { + "epoch": 0.07556, + "grad_norm": 0.7920114994049072, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 7556 + }, + { + "epoch": 0.07557, + "grad_norm": 0.813525915145874, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 7557 + }, + { + "epoch": 0.07558, + "grad_norm": 0.8534443378448486, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 7558 + }, + { + "epoch": 0.07559, + "grad_norm": 0.8572565317153931, + "learning_rate": 0.003, + "loss": 4.018, + "step": 7559 + }, + { + "epoch": 0.0756, + "grad_norm": 0.8627861142158508, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 7560 + }, + { + "epoch": 0.07561, + "grad_norm": 0.7403004765510559, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 7561 + }, + { + "epoch": 0.07562, + "grad_norm": 0.695635974407196, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 7562 + }, + { + "epoch": 0.07563, + "grad_norm": 0.6318756937980652, + "learning_rate": 0.003, + "loss": 4.047, + "step": 7563 + }, + { + "epoch": 0.07564, + "grad_norm": 0.5857333540916443, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 7564 + }, + { + "epoch": 0.07565, + "grad_norm": 0.5988854169845581, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 7565 + }, + { + "epoch": 0.07566, + "grad_norm": 0.6155543923377991, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 7566 + }, + { + "epoch": 0.07567, + "grad_norm": 0.625502347946167, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 7567 + }, + { + "epoch": 0.07568, + "grad_norm": 0.6525031328201294, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 7568 + }, + { + "epoch": 0.07569, + "grad_norm": 0.705432116985321, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 7569 + }, + { + "epoch": 0.0757, + "grad_norm": 0.7517141103744507, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 7570 + }, + { + "epoch": 0.07571, + "grad_norm": 0.7912932634353638, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 7571 + }, + { + "epoch": 0.07572, + "grad_norm": 0.6804235577583313, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 7572 + }, + { + "epoch": 0.07573, + "grad_norm": 0.5986480712890625, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 7573 + }, + { + "epoch": 0.07574, + "grad_norm": 0.6231833696365356, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 7574 + }, + { + "epoch": 0.07575, + "grad_norm": 0.6790160536766052, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 7575 + }, + { + "epoch": 0.07576, + "grad_norm": 0.657383918762207, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 7576 + }, + { + "epoch": 0.07577, + "grad_norm": 0.5640320777893066, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 7577 + }, + { + "epoch": 0.07578, + "grad_norm": 0.5741853713989258, + "learning_rate": 0.003, + "loss": 4.041, + "step": 7578 + }, + { + "epoch": 0.07579, + "grad_norm": 0.5155695080757141, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 7579 + }, + { + "epoch": 0.0758, + "grad_norm": 0.489566445350647, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 7580 + }, + { + "epoch": 0.07581, + "grad_norm": 0.4907761812210083, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 7581 + }, + { + "epoch": 0.07582, + "grad_norm": 0.6237406134605408, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 7582 + }, + { + "epoch": 0.07583, + "grad_norm": 0.6927093267440796, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 7583 + }, + { + "epoch": 0.07584, + "grad_norm": 0.7526013851165771, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 7584 + }, + { + "epoch": 0.07585, + "grad_norm": 0.7571908235549927, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 7585 + }, + { + "epoch": 0.07586, + "grad_norm": 0.794185996055603, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 7586 + }, + { + "epoch": 0.07587, + "grad_norm": 1.0438753366470337, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 7587 + }, + { + "epoch": 0.07588, + "grad_norm": 1.1041924953460693, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 7588 + }, + { + "epoch": 0.07589, + "grad_norm": 0.8445767164230347, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 7589 + }, + { + "epoch": 0.0759, + "grad_norm": 0.6652212738990784, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 7590 + }, + { + "epoch": 0.07591, + "grad_norm": 0.5634509325027466, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 7591 + }, + { + "epoch": 0.07592, + "grad_norm": 0.7388189435005188, + "learning_rate": 0.003, + "loss": 4.056, + "step": 7592 + }, + { + "epoch": 0.07593, + "grad_norm": 0.841446042060852, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 7593 + }, + { + "epoch": 0.07594, + "grad_norm": 0.9196954965591431, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 7594 + }, + { + "epoch": 0.07595, + "grad_norm": 0.9264878630638123, + "learning_rate": 0.003, + "loss": 4.038, + "step": 7595 + }, + { + "epoch": 0.07596, + "grad_norm": 0.7635065913200378, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 7596 + }, + { + "epoch": 0.07597, + "grad_norm": 0.6822502613067627, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 7597 + }, + { + "epoch": 0.07598, + "grad_norm": 0.7266221642494202, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 7598 + }, + { + "epoch": 0.07599, + "grad_norm": 0.6876087784767151, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 7599 + }, + { + "epoch": 0.076, + "grad_norm": 0.6935887932777405, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 7600 + }, + { + "epoch": 0.07601, + "grad_norm": 0.7845306992530823, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 7601 + }, + { + "epoch": 0.07602, + "grad_norm": 0.7442597150802612, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 7602 + }, + { + "epoch": 0.07603, + "grad_norm": 0.7260087728500366, + "learning_rate": 0.003, + "loss": 4.045, + "step": 7603 + }, + { + "epoch": 0.07604, + "grad_norm": 0.6882612705230713, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 7604 + }, + { + "epoch": 0.07605, + "grad_norm": 0.690847635269165, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 7605 + }, + { + "epoch": 0.07606, + "grad_norm": 0.8980867862701416, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 7606 + }, + { + "epoch": 0.07607, + "grad_norm": 1.1527798175811768, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 7607 + }, + { + "epoch": 0.07608, + "grad_norm": 1.0187749862670898, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 7608 + }, + { + "epoch": 0.07609, + "grad_norm": 0.8751260042190552, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 7609 + }, + { + "epoch": 0.0761, + "grad_norm": 0.7832580804824829, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 7610 + }, + { + "epoch": 0.07611, + "grad_norm": 0.8499090671539307, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 7611 + }, + { + "epoch": 0.07612, + "grad_norm": 0.9208655953407288, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 7612 + }, + { + "epoch": 0.07613, + "grad_norm": 1.014868140220642, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 7613 + }, + { + "epoch": 0.07614, + "grad_norm": 0.9037597179412842, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 7614 + }, + { + "epoch": 0.07615, + "grad_norm": 0.675439715385437, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 7615 + }, + { + "epoch": 0.07616, + "grad_norm": 0.6229849457740784, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 7616 + }, + { + "epoch": 0.07617, + "grad_norm": 0.6365832090377808, + "learning_rate": 0.003, + "loss": 4.014, + "step": 7617 + }, + { + "epoch": 0.07618, + "grad_norm": 0.5526401400566101, + "learning_rate": 0.003, + "loss": 4.079, + "step": 7618 + }, + { + "epoch": 0.07619, + "grad_norm": 0.4938875734806061, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 7619 + }, + { + "epoch": 0.0762, + "grad_norm": 0.5198700428009033, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 7620 + }, + { + "epoch": 0.07621, + "grad_norm": 0.587841272354126, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 7621 + }, + { + "epoch": 0.07622, + "grad_norm": 0.6705599427223206, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 7622 + }, + { + "epoch": 0.07623, + "grad_norm": 0.7022711634635925, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 7623 + }, + { + "epoch": 0.07624, + "grad_norm": 0.7389097809791565, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 7624 + }, + { + "epoch": 0.07625, + "grad_norm": 0.7409002184867859, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 7625 + }, + { + "epoch": 0.07626, + "grad_norm": 0.6842824220657349, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 7626 + }, + { + "epoch": 0.07627, + "grad_norm": 0.635125458240509, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 7627 + }, + { + "epoch": 0.07628, + "grad_norm": 0.49942052364349365, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 7628 + }, + { + "epoch": 0.07629, + "grad_norm": 0.5460994839668274, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 7629 + }, + { + "epoch": 0.0763, + "grad_norm": 0.6246569752693176, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 7630 + }, + { + "epoch": 0.07631, + "grad_norm": 0.7679206728935242, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 7631 + }, + { + "epoch": 0.07632, + "grad_norm": 1.076688528060913, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 7632 + }, + { + "epoch": 0.07633, + "grad_norm": 0.9796765446662903, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 7633 + }, + { + "epoch": 0.07634, + "grad_norm": 0.9849674105644226, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 7634 + }, + { + "epoch": 0.07635, + "grad_norm": 0.8030697107315063, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 7635 + }, + { + "epoch": 0.07636, + "grad_norm": 0.7698909640312195, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 7636 + }, + { + "epoch": 0.07637, + "grad_norm": 1.0407158136367798, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 7637 + }, + { + "epoch": 0.07638, + "grad_norm": 0.8812909722328186, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 7638 + }, + { + "epoch": 0.07639, + "grad_norm": 0.6991906762123108, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 7639 + }, + { + "epoch": 0.0764, + "grad_norm": 0.5978763699531555, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 7640 + }, + { + "epoch": 0.07641, + "grad_norm": 0.631737470626831, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 7641 + }, + { + "epoch": 0.07642, + "grad_norm": 0.6913520097732544, + "learning_rate": 0.003, + "loss": 4.063, + "step": 7642 + }, + { + "epoch": 0.07643, + "grad_norm": 0.7022256851196289, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 7643 + }, + { + "epoch": 0.07644, + "grad_norm": 0.6650403738021851, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 7644 + }, + { + "epoch": 0.07645, + "grad_norm": 0.7082775235176086, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 7645 + }, + { + "epoch": 0.07646, + "grad_norm": 0.7065228819847107, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 7646 + }, + { + "epoch": 0.07647, + "grad_norm": 0.6930513381958008, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 7647 + }, + { + "epoch": 0.07648, + "grad_norm": 0.7207313776016235, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 7648 + }, + { + "epoch": 0.07649, + "grad_norm": 0.6997010111808777, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 7649 + }, + { + "epoch": 0.0765, + "grad_norm": 0.7499129176139832, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 7650 + }, + { + "epoch": 0.07651, + "grad_norm": 0.9787033796310425, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 7651 + }, + { + "epoch": 0.07652, + "grad_norm": 1.251487374305725, + "learning_rate": 0.003, + "loss": 4.051, + "step": 7652 + }, + { + "epoch": 0.07653, + "grad_norm": 0.899669349193573, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 7653 + }, + { + "epoch": 0.07654, + "grad_norm": 0.9999637007713318, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 7654 + }, + { + "epoch": 0.07655, + "grad_norm": 0.9081646800041199, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 7655 + }, + { + "epoch": 0.07656, + "grad_norm": 0.783086359500885, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 7656 + }, + { + "epoch": 0.07657, + "grad_norm": 0.7844822406768799, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 7657 + }, + { + "epoch": 0.07658, + "grad_norm": 0.7419793009757996, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 7658 + }, + { + "epoch": 0.07659, + "grad_norm": 0.715597927570343, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 7659 + }, + { + "epoch": 0.0766, + "grad_norm": 0.6549267768859863, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 7660 + }, + { + "epoch": 0.07661, + "grad_norm": 0.785686731338501, + "learning_rate": 0.003, + "loss": 4.043, + "step": 7661 + }, + { + "epoch": 0.07662, + "grad_norm": 1.0206282138824463, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 7662 + }, + { + "epoch": 0.07663, + "grad_norm": 1.1866017580032349, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 7663 + }, + { + "epoch": 0.07664, + "grad_norm": 0.7022275328636169, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 7664 + }, + { + "epoch": 0.07665, + "grad_norm": 0.6321223974227905, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 7665 + }, + { + "epoch": 0.07666, + "grad_norm": 0.917163074016571, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 7666 + }, + { + "epoch": 0.07667, + "grad_norm": 1.0060826539993286, + "learning_rate": 0.003, + "loss": 4.06, + "step": 7667 + }, + { + "epoch": 0.07668, + "grad_norm": 0.7803816795349121, + "learning_rate": 0.003, + "loss": 4.046, + "step": 7668 + }, + { + "epoch": 0.07669, + "grad_norm": 0.5439775586128235, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 7669 + }, + { + "epoch": 0.0767, + "grad_norm": 0.5985819697380066, + "learning_rate": 0.003, + "loss": 4.023, + "step": 7670 + }, + { + "epoch": 0.07671, + "grad_norm": 0.6868377923965454, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 7671 + }, + { + "epoch": 0.07672, + "grad_norm": 0.7403073906898499, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 7672 + }, + { + "epoch": 0.07673, + "grad_norm": 0.754479169845581, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 7673 + }, + { + "epoch": 0.07674, + "grad_norm": 0.6656348705291748, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 7674 + }, + { + "epoch": 0.07675, + "grad_norm": 0.582493007183075, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 7675 + }, + { + "epoch": 0.07676, + "grad_norm": 0.5362693071365356, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 7676 + }, + { + "epoch": 0.07677, + "grad_norm": 0.5358790755271912, + "learning_rate": 0.003, + "loss": 4.031, + "step": 7677 + }, + { + "epoch": 0.07678, + "grad_norm": 0.5923407673835754, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 7678 + }, + { + "epoch": 0.07679, + "grad_norm": 0.6382786631584167, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 7679 + }, + { + "epoch": 0.0768, + "grad_norm": 0.6146613955497742, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 7680 + }, + { + "epoch": 0.07681, + "grad_norm": 0.5740124583244324, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 7681 + }, + { + "epoch": 0.07682, + "grad_norm": 0.5345171093940735, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 7682 + }, + { + "epoch": 0.07683, + "grad_norm": 0.5641381144523621, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 7683 + }, + { + "epoch": 0.07684, + "grad_norm": 0.6472615599632263, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 7684 + }, + { + "epoch": 0.07685, + "grad_norm": 0.7183139324188232, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 7685 + }, + { + "epoch": 0.07686, + "grad_norm": 0.8126670122146606, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 7686 + }, + { + "epoch": 0.07687, + "grad_norm": 0.8759085536003113, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 7687 + }, + { + "epoch": 0.07688, + "grad_norm": 0.8345987796783447, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 7688 + }, + { + "epoch": 0.07689, + "grad_norm": 0.7326918244361877, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 7689 + }, + { + "epoch": 0.0769, + "grad_norm": 0.7978694438934326, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 7690 + }, + { + "epoch": 0.07691, + "grad_norm": 0.9254701733589172, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 7691 + }, + { + "epoch": 0.07692, + "grad_norm": 1.1151405572891235, + "learning_rate": 0.003, + "loss": 4.071, + "step": 7692 + }, + { + "epoch": 0.07693, + "grad_norm": 0.7501732707023621, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 7693 + }, + { + "epoch": 0.07694, + "grad_norm": 0.6858186721801758, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 7694 + }, + { + "epoch": 0.07695, + "grad_norm": 0.7969062924385071, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 7695 + }, + { + "epoch": 0.07696, + "grad_norm": 0.782137393951416, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 7696 + }, + { + "epoch": 0.07697, + "grad_norm": 0.9599317312240601, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 7697 + }, + { + "epoch": 0.07698, + "grad_norm": 1.1077126264572144, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 7698 + }, + { + "epoch": 0.07699, + "grad_norm": 0.8662140965461731, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 7699 + }, + { + "epoch": 0.077, + "grad_norm": 0.7337355017662048, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 7700 + }, + { + "epoch": 0.07701, + "grad_norm": 0.7257059216499329, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 7701 + }, + { + "epoch": 0.07702, + "grad_norm": 0.6639032959938049, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 7702 + }, + { + "epoch": 0.07703, + "grad_norm": 0.7701454162597656, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 7703 + }, + { + "epoch": 0.07704, + "grad_norm": 0.8333292603492737, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 7704 + }, + { + "epoch": 0.07705, + "grad_norm": 0.71268230676651, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 7705 + }, + { + "epoch": 0.07706, + "grad_norm": 0.7142350077629089, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 7706 + }, + { + "epoch": 0.07707, + "grad_norm": 0.7584391832351685, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 7707 + }, + { + "epoch": 0.07708, + "grad_norm": 0.6686346530914307, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 7708 + }, + { + "epoch": 0.07709, + "grad_norm": 0.6045640707015991, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 7709 + }, + { + "epoch": 0.0771, + "grad_norm": 0.5481950640678406, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 7710 + }, + { + "epoch": 0.07711, + "grad_norm": 0.5819096565246582, + "learning_rate": 0.003, + "loss": 4.028, + "step": 7711 + }, + { + "epoch": 0.07712, + "grad_norm": 0.703336238861084, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 7712 + }, + { + "epoch": 0.07713, + "grad_norm": 0.6952189803123474, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 7713 + }, + { + "epoch": 0.07714, + "grad_norm": 0.7354077100753784, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 7714 + }, + { + "epoch": 0.07715, + "grad_norm": 0.8077874779701233, + "learning_rate": 0.003, + "loss": 4.044, + "step": 7715 + }, + { + "epoch": 0.07716, + "grad_norm": 1.056091070175171, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 7716 + }, + { + "epoch": 0.07717, + "grad_norm": 1.122626543045044, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 7717 + }, + { + "epoch": 0.07718, + "grad_norm": 0.7326860427856445, + "learning_rate": 0.003, + "loss": 4.071, + "step": 7718 + }, + { + "epoch": 0.07719, + "grad_norm": 0.7065683007240295, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 7719 + }, + { + "epoch": 0.0772, + "grad_norm": 0.825425922870636, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 7720 + }, + { + "epoch": 0.07721, + "grad_norm": 0.8575567007064819, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 7721 + }, + { + "epoch": 0.07722, + "grad_norm": 0.8774976134300232, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 7722 + }, + { + "epoch": 0.07723, + "grad_norm": 0.8449294567108154, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 7723 + }, + { + "epoch": 0.07724, + "grad_norm": 0.9954960942268372, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 7724 + }, + { + "epoch": 0.07725, + "grad_norm": 1.0463624000549316, + "learning_rate": 0.003, + "loss": 4.051, + "step": 7725 + }, + { + "epoch": 0.07726, + "grad_norm": 0.8136222958564758, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 7726 + }, + { + "epoch": 0.07727, + "grad_norm": 0.9302545189857483, + "learning_rate": 0.003, + "loss": 4.04, + "step": 7727 + }, + { + "epoch": 0.07728, + "grad_norm": 0.8933360576629639, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 7728 + }, + { + "epoch": 0.07729, + "grad_norm": 0.7947314977645874, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 7729 + }, + { + "epoch": 0.0773, + "grad_norm": 0.8221551179885864, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 7730 + }, + { + "epoch": 0.07731, + "grad_norm": 0.7646525502204895, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 7731 + }, + { + "epoch": 0.07732, + "grad_norm": 0.5452698469161987, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 7732 + }, + { + "epoch": 0.07733, + "grad_norm": 0.5334752202033997, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 7733 + }, + { + "epoch": 0.07734, + "grad_norm": 0.5438298583030701, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 7734 + }, + { + "epoch": 0.07735, + "grad_norm": 0.6537331938743591, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 7735 + }, + { + "epoch": 0.07736, + "grad_norm": 0.762072741985321, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 7736 + }, + { + "epoch": 0.07737, + "grad_norm": 0.9448016285896301, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 7737 + }, + { + "epoch": 0.07738, + "grad_norm": 0.943324863910675, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 7738 + }, + { + "epoch": 0.07739, + "grad_norm": 0.761298418045044, + "learning_rate": 0.003, + "loss": 4.051, + "step": 7739 + }, + { + "epoch": 0.0774, + "grad_norm": 0.6667705178260803, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 7740 + }, + { + "epoch": 0.07741, + "grad_norm": 0.7198627591133118, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 7741 + }, + { + "epoch": 0.07742, + "grad_norm": 0.685585081577301, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 7742 + }, + { + "epoch": 0.07743, + "grad_norm": 0.5970986485481262, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 7743 + }, + { + "epoch": 0.07744, + "grad_norm": 0.5659307837486267, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 7744 + }, + { + "epoch": 0.07745, + "grad_norm": 0.5722436308860779, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 7745 + }, + { + "epoch": 0.07746, + "grad_norm": 0.6215358972549438, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 7746 + }, + { + "epoch": 0.07747, + "grad_norm": 0.638965904712677, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 7747 + }, + { + "epoch": 0.07748, + "grad_norm": 0.5998613834381104, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 7748 + }, + { + "epoch": 0.07749, + "grad_norm": 0.716853141784668, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 7749 + }, + { + "epoch": 0.0775, + "grad_norm": 0.8205824494361877, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 7750 + }, + { + "epoch": 0.07751, + "grad_norm": 0.8468402028083801, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 7751 + }, + { + "epoch": 0.07752, + "grad_norm": 0.8478083610534668, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 7752 + }, + { + "epoch": 0.07753, + "grad_norm": 0.8648585677146912, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 7753 + }, + { + "epoch": 0.07754, + "grad_norm": 0.7602769732475281, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 7754 + }, + { + "epoch": 0.07755, + "grad_norm": 0.7095845341682434, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 7755 + }, + { + "epoch": 0.07756, + "grad_norm": 0.7669501304626465, + "learning_rate": 0.003, + "loss": 4.066, + "step": 7756 + }, + { + "epoch": 0.07757, + "grad_norm": 0.7360534071922302, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 7757 + }, + { + "epoch": 0.07758, + "grad_norm": 0.967415988445282, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 7758 + }, + { + "epoch": 0.07759, + "grad_norm": 1.213719129562378, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 7759 + }, + { + "epoch": 0.0776, + "grad_norm": 0.8443424701690674, + "learning_rate": 0.003, + "loss": 4.024, + "step": 7760 + }, + { + "epoch": 0.07761, + "grad_norm": 0.7246764302253723, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 7761 + }, + { + "epoch": 0.07762, + "grad_norm": 0.8051997423171997, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 7762 + }, + { + "epoch": 0.07763, + "grad_norm": 0.8146281242370605, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 7763 + }, + { + "epoch": 0.07764, + "grad_norm": 0.9019950032234192, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 7764 + }, + { + "epoch": 0.07765, + "grad_norm": 1.0602766275405884, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 7765 + }, + { + "epoch": 0.07766, + "grad_norm": 0.9736347794532776, + "learning_rate": 0.003, + "loss": 4.036, + "step": 7766 + }, + { + "epoch": 0.07767, + "grad_norm": 0.7766427993774414, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 7767 + }, + { + "epoch": 0.07768, + "grad_norm": 0.5827065706253052, + "learning_rate": 0.003, + "loss": 4.042, + "step": 7768 + }, + { + "epoch": 0.07769, + "grad_norm": 0.5819374322891235, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 7769 + }, + { + "epoch": 0.0777, + "grad_norm": 0.6189792156219482, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 7770 + }, + { + "epoch": 0.07771, + "grad_norm": 0.6131955981254578, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 7771 + }, + { + "epoch": 0.07772, + "grad_norm": 0.731395959854126, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 7772 + }, + { + "epoch": 0.07773, + "grad_norm": 0.8028513789176941, + "learning_rate": 0.003, + "loss": 4.019, + "step": 7773 + }, + { + "epoch": 0.07774, + "grad_norm": 0.8190731406211853, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 7774 + }, + { + "epoch": 0.07775, + "grad_norm": 0.8094491362571716, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 7775 + }, + { + "epoch": 0.07776, + "grad_norm": 0.6882286071777344, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 7776 + }, + { + "epoch": 0.07777, + "grad_norm": 0.6559415459632874, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 7777 + }, + { + "epoch": 0.07778, + "grad_norm": 0.7245689630508423, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 7778 + }, + { + "epoch": 0.07779, + "grad_norm": 0.7859965562820435, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 7779 + }, + { + "epoch": 0.0778, + "grad_norm": 0.7404747605323792, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 7780 + }, + { + "epoch": 0.07781, + "grad_norm": 0.6684058308601379, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 7781 + }, + { + "epoch": 0.07782, + "grad_norm": 0.5081612467765808, + "learning_rate": 0.003, + "loss": 4.023, + "step": 7782 + }, + { + "epoch": 0.07783, + "grad_norm": 0.5472040176391602, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 7783 + }, + { + "epoch": 0.07784, + "grad_norm": 0.6857728958129883, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 7784 + }, + { + "epoch": 0.07785, + "grad_norm": 0.8567017316818237, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 7785 + }, + { + "epoch": 0.07786, + "grad_norm": 0.9442057609558105, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 7786 + }, + { + "epoch": 0.07787, + "grad_norm": 0.9159606099128723, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 7787 + }, + { + "epoch": 0.07788, + "grad_norm": 0.9067151546478271, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 7788 + }, + { + "epoch": 0.07789, + "grad_norm": 1.0266743898391724, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 7789 + }, + { + "epoch": 0.0779, + "grad_norm": 1.1035289764404297, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 7790 + }, + { + "epoch": 0.07791, + "grad_norm": 0.7223553657531738, + "learning_rate": 0.003, + "loss": 4.033, + "step": 7791 + }, + { + "epoch": 0.07792, + "grad_norm": 0.5318799018859863, + "learning_rate": 0.003, + "loss": 4.014, + "step": 7792 + }, + { + "epoch": 0.07793, + "grad_norm": 0.6023153066635132, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 7793 + }, + { + "epoch": 0.07794, + "grad_norm": 0.6616175770759583, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 7794 + }, + { + "epoch": 0.07795, + "grad_norm": 0.674656093120575, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 7795 + }, + { + "epoch": 0.07796, + "grad_norm": 0.6360393166542053, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 7796 + }, + { + "epoch": 0.07797, + "grad_norm": 0.6176856160163879, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 7797 + }, + { + "epoch": 0.07798, + "grad_norm": 0.7055824995040894, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 7798 + }, + { + "epoch": 0.07799, + "grad_norm": 0.7401881217956543, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 7799 + }, + { + "epoch": 0.078, + "grad_norm": 0.7410922646522522, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 7800 + }, + { + "epoch": 0.07801, + "grad_norm": 0.8164018392562866, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 7801 + }, + { + "epoch": 0.07802, + "grad_norm": 1.1281871795654297, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 7802 + }, + { + "epoch": 0.07803, + "grad_norm": 0.9571782946586609, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 7803 + }, + { + "epoch": 0.07804, + "grad_norm": 0.8569245338439941, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 7804 + }, + { + "epoch": 0.07805, + "grad_norm": 0.8247978687286377, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 7805 + }, + { + "epoch": 0.07806, + "grad_norm": 0.881071925163269, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 7806 + }, + { + "epoch": 0.07807, + "grad_norm": 0.7734600305557251, + "learning_rate": 0.003, + "loss": 4.071, + "step": 7807 + }, + { + "epoch": 0.07808, + "grad_norm": 0.6630396246910095, + "learning_rate": 0.003, + "loss": 4.033, + "step": 7808 + }, + { + "epoch": 0.07809, + "grad_norm": 0.6479990482330322, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 7809 + }, + { + "epoch": 0.0781, + "grad_norm": 0.646216630935669, + "learning_rate": 0.003, + "loss": 4.029, + "step": 7810 + }, + { + "epoch": 0.07811, + "grad_norm": 0.6451820731163025, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 7811 + }, + { + "epoch": 0.07812, + "grad_norm": 0.7043893933296204, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 7812 + }, + { + "epoch": 0.07813, + "grad_norm": 0.7477477192878723, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 7813 + }, + { + "epoch": 0.07814, + "grad_norm": 0.6975932121276855, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 7814 + }, + { + "epoch": 0.07815, + "grad_norm": 0.6276878714561462, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 7815 + }, + { + "epoch": 0.07816, + "grad_norm": 0.6076654195785522, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 7816 + }, + { + "epoch": 0.07817, + "grad_norm": 0.7753585577011108, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 7817 + }, + { + "epoch": 0.07818, + "grad_norm": 0.9763312935829163, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 7818 + }, + { + "epoch": 0.07819, + "grad_norm": 1.1070810556411743, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 7819 + }, + { + "epoch": 0.0782, + "grad_norm": 0.7738063335418701, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 7820 + }, + { + "epoch": 0.07821, + "grad_norm": 0.6099690198898315, + "learning_rate": 0.003, + "loss": 4.025, + "step": 7821 + }, + { + "epoch": 0.07822, + "grad_norm": 0.6168242692947388, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 7822 + }, + { + "epoch": 0.07823, + "grad_norm": 0.5779638290405273, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 7823 + }, + { + "epoch": 0.07824, + "grad_norm": 0.5537580847740173, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 7824 + }, + { + "epoch": 0.07825, + "grad_norm": 0.5701668858528137, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 7825 + }, + { + "epoch": 0.07826, + "grad_norm": 0.5944071412086487, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 7826 + }, + { + "epoch": 0.07827, + "grad_norm": 0.5485585927963257, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 7827 + }, + { + "epoch": 0.07828, + "grad_norm": 0.6365675330162048, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 7828 + }, + { + "epoch": 0.07829, + "grad_norm": 0.7511796951293945, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 7829 + }, + { + "epoch": 0.0783, + "grad_norm": 0.8600240349769592, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 7830 + }, + { + "epoch": 0.07831, + "grad_norm": 0.8572528958320618, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 7831 + }, + { + "epoch": 0.07832, + "grad_norm": 0.7038732171058655, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 7832 + }, + { + "epoch": 0.07833, + "grad_norm": 0.7847913503646851, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 7833 + }, + { + "epoch": 0.07834, + "grad_norm": 0.860440731048584, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 7834 + }, + { + "epoch": 0.07835, + "grad_norm": 0.8758241534233093, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 7835 + }, + { + "epoch": 0.07836, + "grad_norm": 0.8424054980278015, + "learning_rate": 0.003, + "loss": 4.041, + "step": 7836 + }, + { + "epoch": 0.07837, + "grad_norm": 0.7555618286132812, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 7837 + }, + { + "epoch": 0.07838, + "grad_norm": 0.7149111032485962, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 7838 + }, + { + "epoch": 0.07839, + "grad_norm": 0.6796376705169678, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 7839 + }, + { + "epoch": 0.0784, + "grad_norm": 0.750508725643158, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 7840 + }, + { + "epoch": 0.07841, + "grad_norm": 0.7040848731994629, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 7841 + }, + { + "epoch": 0.07842, + "grad_norm": 0.7227228283882141, + "learning_rate": 0.003, + "loss": 4.027, + "step": 7842 + }, + { + "epoch": 0.07843, + "grad_norm": 0.8472539782524109, + "learning_rate": 0.003, + "loss": 4.045, + "step": 7843 + }, + { + "epoch": 0.07844, + "grad_norm": 0.7912211418151855, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 7844 + }, + { + "epoch": 0.07845, + "grad_norm": 0.7454449534416199, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 7845 + }, + { + "epoch": 0.07846, + "grad_norm": 0.7517043352127075, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 7846 + }, + { + "epoch": 0.07847, + "grad_norm": 0.7840257287025452, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 7847 + }, + { + "epoch": 0.07848, + "grad_norm": 0.9238049983978271, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 7848 + }, + { + "epoch": 0.07849, + "grad_norm": 1.0469474792480469, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 7849 + }, + { + "epoch": 0.0785, + "grad_norm": 1.0180537700653076, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 7850 + }, + { + "epoch": 0.07851, + "grad_norm": 0.8952250480651855, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 7851 + }, + { + "epoch": 0.07852, + "grad_norm": 0.8214434385299683, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 7852 + }, + { + "epoch": 0.07853, + "grad_norm": 0.8445104956626892, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 7853 + }, + { + "epoch": 0.07854, + "grad_norm": 0.9933289885520935, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 7854 + }, + { + "epoch": 0.07855, + "grad_norm": 1.1062895059585571, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 7855 + }, + { + "epoch": 0.07856, + "grad_norm": 0.9354727864265442, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 7856 + }, + { + "epoch": 0.07857, + "grad_norm": 0.9971434473991394, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 7857 + }, + { + "epoch": 0.07858, + "grad_norm": 1.0467644929885864, + "learning_rate": 0.003, + "loss": 4.075, + "step": 7858 + }, + { + "epoch": 0.07859, + "grad_norm": 0.8451358675956726, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 7859 + }, + { + "epoch": 0.0786, + "grad_norm": 0.8124087452888489, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 7860 + }, + { + "epoch": 0.07861, + "grad_norm": 0.8194620013237, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 7861 + }, + { + "epoch": 0.07862, + "grad_norm": 0.7896692156791687, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 7862 + }, + { + "epoch": 0.07863, + "grad_norm": 0.7973822951316833, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 7863 + }, + { + "epoch": 0.07864, + "grad_norm": 0.745921790599823, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 7864 + }, + { + "epoch": 0.07865, + "grad_norm": 0.6143619418144226, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 7865 + }, + { + "epoch": 0.07866, + "grad_norm": 0.5146561861038208, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 7866 + }, + { + "epoch": 0.07867, + "grad_norm": 0.5350759029388428, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 7867 + }, + { + "epoch": 0.07868, + "grad_norm": 0.5663687586784363, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 7868 + }, + { + "epoch": 0.07869, + "grad_norm": 0.6298394799232483, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 7869 + }, + { + "epoch": 0.0787, + "grad_norm": 0.6779914498329163, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 7870 + }, + { + "epoch": 0.07871, + "grad_norm": 0.6579896211624146, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 7871 + }, + { + "epoch": 0.07872, + "grad_norm": 0.5511500239372253, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 7872 + }, + { + "epoch": 0.07873, + "grad_norm": 0.5043216943740845, + "learning_rate": 0.003, + "loss": 4.022, + "step": 7873 + }, + { + "epoch": 0.07874, + "grad_norm": 0.46951040625572205, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 7874 + }, + { + "epoch": 0.07875, + "grad_norm": 0.490416556596756, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 7875 + }, + { + "epoch": 0.07876, + "grad_norm": 0.5800589323043823, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 7876 + }, + { + "epoch": 0.07877, + "grad_norm": 0.6350085139274597, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 7877 + }, + { + "epoch": 0.07878, + "grad_norm": 0.792594313621521, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 7878 + }, + { + "epoch": 0.07879, + "grad_norm": 1.0079691410064697, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 7879 + }, + { + "epoch": 0.0788, + "grad_norm": 1.099412441253662, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 7880 + }, + { + "epoch": 0.07881, + "grad_norm": 0.741230845451355, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 7881 + }, + { + "epoch": 0.07882, + "grad_norm": 0.6595082879066467, + "learning_rate": 0.003, + "loss": 4.033, + "step": 7882 + }, + { + "epoch": 0.07883, + "grad_norm": 0.7658863067626953, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 7883 + }, + { + "epoch": 0.07884, + "grad_norm": 0.826673686504364, + "learning_rate": 0.003, + "loss": 4.041, + "step": 7884 + }, + { + "epoch": 0.07885, + "grad_norm": 0.7689464092254639, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 7885 + }, + { + "epoch": 0.07886, + "grad_norm": 0.5779816508293152, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 7886 + }, + { + "epoch": 0.07887, + "grad_norm": 0.5393301844596863, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 7887 + }, + { + "epoch": 0.07888, + "grad_norm": 0.6773719787597656, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 7888 + }, + { + "epoch": 0.07889, + "grad_norm": 0.7781038284301758, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 7889 + }, + { + "epoch": 0.0789, + "grad_norm": 0.8151108026504517, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 7890 + }, + { + "epoch": 0.07891, + "grad_norm": 0.9264851212501526, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 7891 + }, + { + "epoch": 0.07892, + "grad_norm": 0.8555210828781128, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 7892 + }, + { + "epoch": 0.07893, + "grad_norm": 0.6806337833404541, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 7893 + }, + { + "epoch": 0.07894, + "grad_norm": 0.6658015251159668, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 7894 + }, + { + "epoch": 0.07895, + "grad_norm": 0.5995755791664124, + "learning_rate": 0.003, + "loss": 4.019, + "step": 7895 + }, + { + "epoch": 0.07896, + "grad_norm": 0.6623494625091553, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 7896 + }, + { + "epoch": 0.07897, + "grad_norm": 0.6996875405311584, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 7897 + }, + { + "epoch": 0.07898, + "grad_norm": 0.8288904428482056, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 7898 + }, + { + "epoch": 0.07899, + "grad_norm": 0.8660441040992737, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 7899 + }, + { + "epoch": 0.079, + "grad_norm": 0.9470987319946289, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 7900 + }, + { + "epoch": 0.07901, + "grad_norm": 0.9151023030281067, + "learning_rate": 0.003, + "loss": 4.079, + "step": 7901 + }, + { + "epoch": 0.07902, + "grad_norm": 0.8372184634208679, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 7902 + }, + { + "epoch": 0.07903, + "grad_norm": 0.6515103578567505, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 7903 + }, + { + "epoch": 0.07904, + "grad_norm": 0.7328464388847351, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 7904 + }, + { + "epoch": 0.07905, + "grad_norm": 0.8890835046768188, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 7905 + }, + { + "epoch": 0.07906, + "grad_norm": 1.1531867980957031, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 7906 + }, + { + "epoch": 0.07907, + "grad_norm": 1.1185576915740967, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 7907 + }, + { + "epoch": 0.07908, + "grad_norm": 0.7459601759910583, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 7908 + }, + { + "epoch": 0.07909, + "grad_norm": 0.6596913933753967, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 7909 + }, + { + "epoch": 0.0791, + "grad_norm": 0.7314830422401428, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 7910 + }, + { + "epoch": 0.07911, + "grad_norm": 0.6769105792045593, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 7911 + }, + { + "epoch": 0.07912, + "grad_norm": 0.6970905661582947, + "learning_rate": 0.003, + "loss": 4.032, + "step": 7912 + }, + { + "epoch": 0.07913, + "grad_norm": 0.6443745493888855, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 7913 + }, + { + "epoch": 0.07914, + "grad_norm": 0.5966974496841431, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 7914 + }, + { + "epoch": 0.07915, + "grad_norm": 0.5250338912010193, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 7915 + }, + { + "epoch": 0.07916, + "grad_norm": 0.5534821152687073, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 7916 + }, + { + "epoch": 0.07917, + "grad_norm": 0.4856780767440796, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 7917 + }, + { + "epoch": 0.07918, + "grad_norm": 0.4974648654460907, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 7918 + }, + { + "epoch": 0.07919, + "grad_norm": 0.5157517194747925, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 7919 + }, + { + "epoch": 0.0792, + "grad_norm": 0.5172123312950134, + "learning_rate": 0.003, + "loss": 4.027, + "step": 7920 + }, + { + "epoch": 0.07921, + "grad_norm": 0.4519246816635132, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 7921 + }, + { + "epoch": 0.07922, + "grad_norm": 0.48282045125961304, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 7922 + }, + { + "epoch": 0.07923, + "grad_norm": 0.5496918559074402, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 7923 + }, + { + "epoch": 0.07924, + "grad_norm": 0.6421642899513245, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 7924 + }, + { + "epoch": 0.07925, + "grad_norm": 0.7793181538581848, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 7925 + }, + { + "epoch": 0.07926, + "grad_norm": 0.9132375121116638, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 7926 + }, + { + "epoch": 0.07927, + "grad_norm": 0.9230619668960571, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 7927 + }, + { + "epoch": 0.07928, + "grad_norm": 0.7421871423721313, + "learning_rate": 0.003, + "loss": 4.041, + "step": 7928 + }, + { + "epoch": 0.07929, + "grad_norm": 0.6340886950492859, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 7929 + }, + { + "epoch": 0.0793, + "grad_norm": 0.6207582950592041, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 7930 + }, + { + "epoch": 0.07931, + "grad_norm": 0.698438286781311, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 7931 + }, + { + "epoch": 0.07932, + "grad_norm": 0.932792067527771, + "learning_rate": 0.003, + "loss": 4.054, + "step": 7932 + }, + { + "epoch": 0.07933, + "grad_norm": 0.8964979648590088, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 7933 + }, + { + "epoch": 0.07934, + "grad_norm": 0.8300673365592957, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 7934 + }, + { + "epoch": 0.07935, + "grad_norm": 0.7360064387321472, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 7935 + }, + { + "epoch": 0.07936, + "grad_norm": 0.7895861864089966, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 7936 + }, + { + "epoch": 0.07937, + "grad_norm": 0.9133578538894653, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 7937 + }, + { + "epoch": 0.07938, + "grad_norm": 0.9982143044471741, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 7938 + }, + { + "epoch": 0.07939, + "grad_norm": 1.1373426914215088, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 7939 + }, + { + "epoch": 0.0794, + "grad_norm": 0.8364454507827759, + "learning_rate": 0.003, + "loss": 4.037, + "step": 7940 + }, + { + "epoch": 0.07941, + "grad_norm": 0.8427025675773621, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 7941 + }, + { + "epoch": 0.07942, + "grad_norm": 0.8358206152915955, + "learning_rate": 0.003, + "loss": 4.028, + "step": 7942 + }, + { + "epoch": 0.07943, + "grad_norm": 0.70419842004776, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 7943 + }, + { + "epoch": 0.07944, + "grad_norm": 0.6298567056655884, + "learning_rate": 0.003, + "loss": 4.072, + "step": 7944 + }, + { + "epoch": 0.07945, + "grad_norm": 0.7173011898994446, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 7945 + }, + { + "epoch": 0.07946, + "grad_norm": 0.8936134576797485, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 7946 + }, + { + "epoch": 0.07947, + "grad_norm": 0.9170552492141724, + "learning_rate": 0.003, + "loss": 4.068, + "step": 7947 + }, + { + "epoch": 0.07948, + "grad_norm": 0.9906176924705505, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 7948 + }, + { + "epoch": 0.07949, + "grad_norm": 0.9943937659263611, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 7949 + }, + { + "epoch": 0.0795, + "grad_norm": 0.97963547706604, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 7950 + }, + { + "epoch": 0.07951, + "grad_norm": 0.949016273021698, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 7951 + }, + { + "epoch": 0.07952, + "grad_norm": 0.7665191292762756, + "learning_rate": 0.003, + "loss": 4.059, + "step": 7952 + }, + { + "epoch": 0.07953, + "grad_norm": 0.7401639819145203, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 7953 + }, + { + "epoch": 0.07954, + "grad_norm": 0.8128531575202942, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 7954 + }, + { + "epoch": 0.07955, + "grad_norm": 0.8641282320022583, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 7955 + }, + { + "epoch": 0.07956, + "grad_norm": 0.897680401802063, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 7956 + }, + { + "epoch": 0.07957, + "grad_norm": 0.8444660305976868, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 7957 + }, + { + "epoch": 0.07958, + "grad_norm": 0.890282392501831, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 7958 + }, + { + "epoch": 0.07959, + "grad_norm": 0.8856026530265808, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 7959 + }, + { + "epoch": 0.0796, + "grad_norm": 0.792759358882904, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 7960 + }, + { + "epoch": 0.07961, + "grad_norm": 0.8642300963401794, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 7961 + }, + { + "epoch": 0.07962, + "grad_norm": 0.886371374130249, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 7962 + }, + { + "epoch": 0.07963, + "grad_norm": 0.8674740195274353, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 7963 + }, + { + "epoch": 0.07964, + "grad_norm": 0.9062060713768005, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 7964 + }, + { + "epoch": 0.07965, + "grad_norm": 1.045652151107788, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 7965 + }, + { + "epoch": 0.07966, + "grad_norm": 0.8896709680557251, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 7966 + }, + { + "epoch": 0.07967, + "grad_norm": 0.8535459041595459, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 7967 + }, + { + "epoch": 0.07968, + "grad_norm": 0.7971822023391724, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 7968 + }, + { + "epoch": 0.07969, + "grad_norm": 0.7176892161369324, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 7969 + }, + { + "epoch": 0.0797, + "grad_norm": 0.619379997253418, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 7970 + }, + { + "epoch": 0.07971, + "grad_norm": 0.7199423909187317, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 7971 + }, + { + "epoch": 0.07972, + "grad_norm": 0.7708553671836853, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 7972 + }, + { + "epoch": 0.07973, + "grad_norm": 0.8473473191261292, + "learning_rate": 0.003, + "loss": 4.057, + "step": 7973 + }, + { + "epoch": 0.07974, + "grad_norm": 0.8833823204040527, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 7974 + }, + { + "epoch": 0.07975, + "grad_norm": 0.8675002455711365, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 7975 + }, + { + "epoch": 0.07976, + "grad_norm": 0.7916547656059265, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 7976 + }, + { + "epoch": 0.07977, + "grad_norm": 0.5385777354240417, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 7977 + }, + { + "epoch": 0.07978, + "grad_norm": 0.5118996500968933, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 7978 + }, + { + "epoch": 0.07979, + "grad_norm": 0.5327662229537964, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 7979 + }, + { + "epoch": 0.0798, + "grad_norm": 0.5415238738059998, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 7980 + }, + { + "epoch": 0.07981, + "grad_norm": 0.5292704105377197, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 7981 + }, + { + "epoch": 0.07982, + "grad_norm": 0.5772751569747925, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 7982 + }, + { + "epoch": 0.07983, + "grad_norm": 0.6106157302856445, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 7983 + }, + { + "epoch": 0.07984, + "grad_norm": 0.7016780376434326, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 7984 + }, + { + "epoch": 0.07985, + "grad_norm": 0.841801643371582, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 7985 + }, + { + "epoch": 0.07986, + "grad_norm": 0.817077100276947, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 7986 + }, + { + "epoch": 0.07987, + "grad_norm": 0.6411316394805908, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 7987 + }, + { + "epoch": 0.07988, + "grad_norm": 0.6220775246620178, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 7988 + }, + { + "epoch": 0.07989, + "grad_norm": 0.5362831354141235, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 7989 + }, + { + "epoch": 0.0799, + "grad_norm": 0.5406444072723389, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 7990 + }, + { + "epoch": 0.07991, + "grad_norm": 0.5161384344100952, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 7991 + }, + { + "epoch": 0.07992, + "grad_norm": 0.45497363805770874, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 7992 + }, + { + "epoch": 0.07993, + "grad_norm": 0.4724503755569458, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 7993 + }, + { + "epoch": 0.07994, + "grad_norm": 0.4843752980232239, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 7994 + }, + { + "epoch": 0.07995, + "grad_norm": 0.5137373805046082, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 7995 + }, + { + "epoch": 0.07996, + "grad_norm": 0.5532665252685547, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 7996 + }, + { + "epoch": 0.07997, + "grad_norm": 0.6154145002365112, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 7997 + }, + { + "epoch": 0.07998, + "grad_norm": 0.7569679021835327, + "learning_rate": 0.003, + "loss": 4.062, + "step": 7998 + }, + { + "epoch": 0.07999, + "grad_norm": 0.9815395474433899, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 7999 + }, + { + "epoch": 0.08, + "grad_norm": 1.1601155996322632, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 8000 + }, + { + "epoch": 0.08001, + "grad_norm": 0.6949255466461182, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 8001 + }, + { + "epoch": 0.08002, + "grad_norm": 0.6404258012771606, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 8002 + }, + { + "epoch": 0.08003, + "grad_norm": 0.6366399526596069, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 8003 + }, + { + "epoch": 0.08004, + "grad_norm": 0.7067066431045532, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 8004 + }, + { + "epoch": 0.08005, + "grad_norm": 0.6715254783630371, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 8005 + }, + { + "epoch": 0.08006, + "grad_norm": 0.6775758266448975, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 8006 + }, + { + "epoch": 0.08007, + "grad_norm": 0.7099198698997498, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 8007 + }, + { + "epoch": 0.08008, + "grad_norm": 0.8098099827766418, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 8008 + }, + { + "epoch": 0.08009, + "grad_norm": 0.9449841976165771, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 8009 + }, + { + "epoch": 0.0801, + "grad_norm": 1.0425149202346802, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 8010 + }, + { + "epoch": 0.08011, + "grad_norm": 0.9982419013977051, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 8011 + }, + { + "epoch": 0.08012, + "grad_norm": 0.8389059901237488, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 8012 + }, + { + "epoch": 0.08013, + "grad_norm": 0.8514482378959656, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 8013 + }, + { + "epoch": 0.08014, + "grad_norm": 0.8291131258010864, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 8014 + }, + { + "epoch": 0.08015, + "grad_norm": 0.8924878835678101, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 8015 + }, + { + "epoch": 0.08016, + "grad_norm": 0.8806622624397278, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 8016 + }, + { + "epoch": 0.08017, + "grad_norm": 0.7676506638526917, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 8017 + }, + { + "epoch": 0.08018, + "grad_norm": 0.6425415873527527, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 8018 + }, + { + "epoch": 0.08019, + "grad_norm": 0.7521240711212158, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 8019 + }, + { + "epoch": 0.0802, + "grad_norm": 1.029622197151184, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 8020 + }, + { + "epoch": 0.08021, + "grad_norm": 1.132253885269165, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 8021 + }, + { + "epoch": 0.08022, + "grad_norm": 0.6701259016990662, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 8022 + }, + { + "epoch": 0.08023, + "grad_norm": 0.6120415925979614, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 8023 + }, + { + "epoch": 0.08024, + "grad_norm": 0.7347866892814636, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 8024 + }, + { + "epoch": 0.08025, + "grad_norm": 0.6548226475715637, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 8025 + }, + { + "epoch": 0.08026, + "grad_norm": 0.5551509857177734, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 8026 + }, + { + "epoch": 0.08027, + "grad_norm": 0.5772050023078918, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 8027 + }, + { + "epoch": 0.08028, + "grad_norm": 0.5994802117347717, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 8028 + }, + { + "epoch": 0.08029, + "grad_norm": 0.654687762260437, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 8029 + }, + { + "epoch": 0.0803, + "grad_norm": 0.8716540932655334, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 8030 + }, + { + "epoch": 0.08031, + "grad_norm": 1.1394952535629272, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 8031 + }, + { + "epoch": 0.08032, + "grad_norm": 0.8767929673194885, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 8032 + }, + { + "epoch": 0.08033, + "grad_norm": 0.6166559457778931, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 8033 + }, + { + "epoch": 0.08034, + "grad_norm": 0.5533866882324219, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 8034 + }, + { + "epoch": 0.08035, + "grad_norm": 0.7748246788978577, + "learning_rate": 0.003, + "loss": 4.06, + "step": 8035 + }, + { + "epoch": 0.08036, + "grad_norm": 0.8875789642333984, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 8036 + }, + { + "epoch": 0.08037, + "grad_norm": 0.893352210521698, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 8037 + }, + { + "epoch": 0.08038, + "grad_norm": 0.7569796442985535, + "learning_rate": 0.003, + "loss": 4.033, + "step": 8038 + }, + { + "epoch": 0.08039, + "grad_norm": 0.7512573003768921, + "learning_rate": 0.003, + "loss": 4.044, + "step": 8039 + }, + { + "epoch": 0.0804, + "grad_norm": 0.7568359971046448, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 8040 + }, + { + "epoch": 0.08041, + "grad_norm": 0.7010595202445984, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 8041 + }, + { + "epoch": 0.08042, + "grad_norm": 0.7021541595458984, + "learning_rate": 0.003, + "loss": 4.045, + "step": 8042 + }, + { + "epoch": 0.08043, + "grad_norm": 0.7211666703224182, + "learning_rate": 0.003, + "loss": 4.035, + "step": 8043 + }, + { + "epoch": 0.08044, + "grad_norm": 0.669906497001648, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 8044 + }, + { + "epoch": 0.08045, + "grad_norm": 0.6533012390136719, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 8045 + }, + { + "epoch": 0.08046, + "grad_norm": 0.5999561548233032, + "learning_rate": 0.003, + "loss": 4.035, + "step": 8046 + }, + { + "epoch": 0.08047, + "grad_norm": 0.7110080122947693, + "learning_rate": 0.003, + "loss": 4.036, + "step": 8047 + }, + { + "epoch": 0.08048, + "grad_norm": 0.7337324619293213, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 8048 + }, + { + "epoch": 0.08049, + "grad_norm": 0.7428856492042542, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 8049 + }, + { + "epoch": 0.0805, + "grad_norm": 0.683529794216156, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 8050 + }, + { + "epoch": 0.08051, + "grad_norm": 0.5760079622268677, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 8051 + }, + { + "epoch": 0.08052, + "grad_norm": 0.6353437304496765, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 8052 + }, + { + "epoch": 0.08053, + "grad_norm": 0.7857322096824646, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 8053 + }, + { + "epoch": 0.08054, + "grad_norm": 0.9808282852172852, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 8054 + }, + { + "epoch": 0.08055, + "grad_norm": 1.1264392137527466, + "learning_rate": 0.003, + "loss": 4.015, + "step": 8055 + }, + { + "epoch": 0.08056, + "grad_norm": 0.8243107795715332, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 8056 + }, + { + "epoch": 0.08057, + "grad_norm": 0.7461767196655273, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 8057 + }, + { + "epoch": 0.08058, + "grad_norm": 0.8168344497680664, + "learning_rate": 0.003, + "loss": 4.041, + "step": 8058 + }, + { + "epoch": 0.08059, + "grad_norm": 0.8754106760025024, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 8059 + }, + { + "epoch": 0.0806, + "grad_norm": 0.9352402687072754, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 8060 + }, + { + "epoch": 0.08061, + "grad_norm": 0.9238023161888123, + "learning_rate": 0.003, + "loss": 4.022, + "step": 8061 + }, + { + "epoch": 0.08062, + "grad_norm": 0.8781481385231018, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 8062 + }, + { + "epoch": 0.08063, + "grad_norm": 0.9294982552528381, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 8063 + }, + { + "epoch": 0.08064, + "grad_norm": 0.9283519983291626, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 8064 + }, + { + "epoch": 0.08065, + "grad_norm": 0.7961224913597107, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 8065 + }, + { + "epoch": 0.08066, + "grad_norm": 0.7206274271011353, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 8066 + }, + { + "epoch": 0.08067, + "grad_norm": 0.7221958041191101, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 8067 + }, + { + "epoch": 0.08068, + "grad_norm": 0.687168300151825, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 8068 + }, + { + "epoch": 0.08069, + "grad_norm": 0.6589773297309875, + "learning_rate": 0.003, + "loss": 4.057, + "step": 8069 + }, + { + "epoch": 0.0807, + "grad_norm": 0.7254470586776733, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 8070 + }, + { + "epoch": 0.08071, + "grad_norm": 0.7805096507072449, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 8071 + }, + { + "epoch": 0.08072, + "grad_norm": 0.7921481728553772, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 8072 + }, + { + "epoch": 0.08073, + "grad_norm": 0.7291877865791321, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 8073 + }, + { + "epoch": 0.08074, + "grad_norm": 0.6952309012413025, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 8074 + }, + { + "epoch": 0.08075, + "grad_norm": 0.6951455473899841, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 8075 + }, + { + "epoch": 0.08076, + "grad_norm": 0.7872135043144226, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 8076 + }, + { + "epoch": 0.08077, + "grad_norm": 0.8542959094047546, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 8077 + }, + { + "epoch": 0.08078, + "grad_norm": 1.0374029874801636, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 8078 + }, + { + "epoch": 0.08079, + "grad_norm": 1.025814175605774, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 8079 + }, + { + "epoch": 0.0808, + "grad_norm": 0.65848308801651, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 8080 + }, + { + "epoch": 0.08081, + "grad_norm": 0.5402346849441528, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 8081 + }, + { + "epoch": 0.08082, + "grad_norm": 0.6986470222473145, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 8082 + }, + { + "epoch": 0.08083, + "grad_norm": 0.8801277875900269, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 8083 + }, + { + "epoch": 0.08084, + "grad_norm": 0.9807546734809875, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 8084 + }, + { + "epoch": 0.08085, + "grad_norm": 0.9804517030715942, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 8085 + }, + { + "epoch": 0.08086, + "grad_norm": 0.8663859963417053, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 8086 + }, + { + "epoch": 0.08087, + "grad_norm": 0.8679655194282532, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 8087 + }, + { + "epoch": 0.08088, + "grad_norm": 0.8380336165428162, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 8088 + }, + { + "epoch": 0.08089, + "grad_norm": 0.8112128973007202, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 8089 + }, + { + "epoch": 0.0809, + "grad_norm": 0.8074503540992737, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 8090 + }, + { + "epoch": 0.08091, + "grad_norm": 0.8186045289039612, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 8091 + }, + { + "epoch": 0.08092, + "grad_norm": 0.777157723903656, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 8092 + }, + { + "epoch": 0.08093, + "grad_norm": 0.6288996338844299, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 8093 + }, + { + "epoch": 0.08094, + "grad_norm": 0.6444592475891113, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 8094 + }, + { + "epoch": 0.08095, + "grad_norm": 0.6421931385993958, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 8095 + }, + { + "epoch": 0.08096, + "grad_norm": 0.6471747756004333, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 8096 + }, + { + "epoch": 0.08097, + "grad_norm": 0.6687425971031189, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 8097 + }, + { + "epoch": 0.08098, + "grad_norm": 0.6552982330322266, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 8098 + }, + { + "epoch": 0.08099, + "grad_norm": 0.6883662939071655, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 8099 + }, + { + "epoch": 0.081, + "grad_norm": 0.7512494921684265, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 8100 + }, + { + "epoch": 0.08101, + "grad_norm": 0.7864393591880798, + "learning_rate": 0.003, + "loss": 4.068, + "step": 8101 + }, + { + "epoch": 0.08102, + "grad_norm": 0.7945188879966736, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 8102 + }, + { + "epoch": 0.08103, + "grad_norm": 0.8009957671165466, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 8103 + }, + { + "epoch": 0.08104, + "grad_norm": 0.7752313613891602, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 8104 + }, + { + "epoch": 0.08105, + "grad_norm": 0.7074246406555176, + "learning_rate": 0.003, + "loss": 4.055, + "step": 8105 + }, + { + "epoch": 0.08106, + "grad_norm": 0.6730649471282959, + "learning_rate": 0.003, + "loss": 4.044, + "step": 8106 + }, + { + "epoch": 0.08107, + "grad_norm": 0.597204327583313, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 8107 + }, + { + "epoch": 0.08108, + "grad_norm": 0.5866988301277161, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 8108 + }, + { + "epoch": 0.08109, + "grad_norm": 0.541022777557373, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 8109 + }, + { + "epoch": 0.0811, + "grad_norm": 0.5541371703147888, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 8110 + }, + { + "epoch": 0.08111, + "grad_norm": 0.607240617275238, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 8111 + }, + { + "epoch": 0.08112, + "grad_norm": 0.6710116267204285, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 8112 + }, + { + "epoch": 0.08113, + "grad_norm": 0.7832820415496826, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 8113 + }, + { + "epoch": 0.08114, + "grad_norm": 0.9054648876190186, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 8114 + }, + { + "epoch": 0.08115, + "grad_norm": 0.9504172205924988, + "learning_rate": 0.003, + "loss": 4.051, + "step": 8115 + }, + { + "epoch": 0.08116, + "grad_norm": 0.819379448890686, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 8116 + }, + { + "epoch": 0.08117, + "grad_norm": 0.5076575875282288, + "learning_rate": 0.003, + "loss": 4.051, + "step": 8117 + }, + { + "epoch": 0.08118, + "grad_norm": 0.5695910453796387, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 8118 + }, + { + "epoch": 0.08119, + "grad_norm": 0.6920680999755859, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 8119 + }, + { + "epoch": 0.0812, + "grad_norm": 0.841937780380249, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 8120 + }, + { + "epoch": 0.08121, + "grad_norm": 0.9552656412124634, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 8121 + }, + { + "epoch": 0.08122, + "grad_norm": 0.8833564519882202, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 8122 + }, + { + "epoch": 0.08123, + "grad_norm": 0.8226878046989441, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 8123 + }, + { + "epoch": 0.08124, + "grad_norm": 0.8926153779029846, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 8124 + }, + { + "epoch": 0.08125, + "grad_norm": 0.8978990316390991, + "learning_rate": 0.003, + "loss": 4.075, + "step": 8125 + }, + { + "epoch": 0.08126, + "grad_norm": 0.9611136317253113, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 8126 + }, + { + "epoch": 0.08127, + "grad_norm": 0.7953163981437683, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 8127 + }, + { + "epoch": 0.08128, + "grad_norm": 0.6702011823654175, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 8128 + }, + { + "epoch": 0.08129, + "grad_norm": 0.6014225482940674, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 8129 + }, + { + "epoch": 0.0813, + "grad_norm": 0.6430684328079224, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 8130 + }, + { + "epoch": 0.08131, + "grad_norm": 0.5604881048202515, + "learning_rate": 0.003, + "loss": 4.044, + "step": 8131 + }, + { + "epoch": 0.08132, + "grad_norm": 0.6205559372901917, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 8132 + }, + { + "epoch": 0.08133, + "grad_norm": 0.5798535943031311, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 8133 + }, + { + "epoch": 0.08134, + "grad_norm": 0.5685412287712097, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 8134 + }, + { + "epoch": 0.08135, + "grad_norm": 0.6919317841529846, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 8135 + }, + { + "epoch": 0.08136, + "grad_norm": 0.7103545069694519, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 8136 + }, + { + "epoch": 0.08137, + "grad_norm": 0.7415788769721985, + "learning_rate": 0.003, + "loss": 3.999, + "step": 8137 + }, + { + "epoch": 0.08138, + "grad_norm": 0.9382996559143066, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 8138 + }, + { + "epoch": 0.08139, + "grad_norm": 1.084199070930481, + "learning_rate": 0.003, + "loss": 4.046, + "step": 8139 + }, + { + "epoch": 0.0814, + "grad_norm": 0.9915632605552673, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 8140 + }, + { + "epoch": 0.08141, + "grad_norm": 0.8965795040130615, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 8141 + }, + { + "epoch": 0.08142, + "grad_norm": 0.8707264065742493, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 8142 + }, + { + "epoch": 0.08143, + "grad_norm": 0.8575231432914734, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 8143 + }, + { + "epoch": 0.08144, + "grad_norm": 0.7178121209144592, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 8144 + }, + { + "epoch": 0.08145, + "grad_norm": 0.7352826595306396, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 8145 + }, + { + "epoch": 0.08146, + "grad_norm": 0.806141197681427, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 8146 + }, + { + "epoch": 0.08147, + "grad_norm": 0.8929685354232788, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 8147 + }, + { + "epoch": 0.08148, + "grad_norm": 0.9094710350036621, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 8148 + }, + { + "epoch": 0.08149, + "grad_norm": 0.9839245676994324, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 8149 + }, + { + "epoch": 0.0815, + "grad_norm": 0.9399632811546326, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 8150 + }, + { + "epoch": 0.08151, + "grad_norm": 0.9006126523017883, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 8151 + }, + { + "epoch": 0.08152, + "grad_norm": 0.893135666847229, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 8152 + }, + { + "epoch": 0.08153, + "grad_norm": 0.8419569730758667, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 8153 + }, + { + "epoch": 0.08154, + "grad_norm": 0.675416886806488, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 8154 + }, + { + "epoch": 0.08155, + "grad_norm": 0.6207221746444702, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 8155 + }, + { + "epoch": 0.08156, + "grad_norm": 0.6687820553779602, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 8156 + }, + { + "epoch": 0.08157, + "grad_norm": 0.7092146277427673, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 8157 + }, + { + "epoch": 0.08158, + "grad_norm": 0.8258717060089111, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 8158 + }, + { + "epoch": 0.08159, + "grad_norm": 0.9466468095779419, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 8159 + }, + { + "epoch": 0.0816, + "grad_norm": 0.951337993144989, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 8160 + }, + { + "epoch": 0.08161, + "grad_norm": 0.813227653503418, + "learning_rate": 0.003, + "loss": 4.051, + "step": 8161 + }, + { + "epoch": 0.08162, + "grad_norm": 0.7988176941871643, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 8162 + }, + { + "epoch": 0.08163, + "grad_norm": 0.7750150561332703, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 8163 + }, + { + "epoch": 0.08164, + "grad_norm": 0.7018991112709045, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 8164 + }, + { + "epoch": 0.08165, + "grad_norm": 0.7162664532661438, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 8165 + }, + { + "epoch": 0.08166, + "grad_norm": 0.7163441777229309, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 8166 + }, + { + "epoch": 0.08167, + "grad_norm": 0.6863809823989868, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 8167 + }, + { + "epoch": 0.08168, + "grad_norm": 0.620455265045166, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 8168 + }, + { + "epoch": 0.08169, + "grad_norm": 0.6343228220939636, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 8169 + }, + { + "epoch": 0.0817, + "grad_norm": 0.6355931162834167, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 8170 + }, + { + "epoch": 0.08171, + "grad_norm": 0.6743160486221313, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 8171 + }, + { + "epoch": 0.08172, + "grad_norm": 0.7267938256263733, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 8172 + }, + { + "epoch": 0.08173, + "grad_norm": 0.8629770278930664, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 8173 + }, + { + "epoch": 0.08174, + "grad_norm": 0.9949284195899963, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 8174 + }, + { + "epoch": 0.08175, + "grad_norm": 1.0875071287155151, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 8175 + }, + { + "epoch": 0.08176, + "grad_norm": 0.7273131608963013, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 8176 + }, + { + "epoch": 0.08177, + "grad_norm": 0.5219792723655701, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 8177 + }, + { + "epoch": 0.08178, + "grad_norm": 0.5456799864768982, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 8178 + }, + { + "epoch": 0.08179, + "grad_norm": 0.5848241448402405, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 8179 + }, + { + "epoch": 0.0818, + "grad_norm": 0.5659416317939758, + "learning_rate": 0.003, + "loss": 4.03, + "step": 8180 + }, + { + "epoch": 0.08181, + "grad_norm": 0.5623244047164917, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 8181 + }, + { + "epoch": 0.08182, + "grad_norm": 0.5621029138565063, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 8182 + }, + { + "epoch": 0.08183, + "grad_norm": 0.6393095254898071, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 8183 + }, + { + "epoch": 0.08184, + "grad_norm": 0.776958167552948, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 8184 + }, + { + "epoch": 0.08185, + "grad_norm": 0.8038877248764038, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 8185 + }, + { + "epoch": 0.08186, + "grad_norm": 0.7159028649330139, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 8186 + }, + { + "epoch": 0.08187, + "grad_norm": 0.6306201219558716, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 8187 + }, + { + "epoch": 0.08188, + "grad_norm": 0.6069129109382629, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 8188 + }, + { + "epoch": 0.08189, + "grad_norm": 0.7105355262756348, + "learning_rate": 0.003, + "loss": 4.051, + "step": 8189 + }, + { + "epoch": 0.0819, + "grad_norm": 0.8592032790184021, + "learning_rate": 0.003, + "loss": 4.03, + "step": 8190 + }, + { + "epoch": 0.08191, + "grad_norm": 1.0450857877731323, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 8191 + }, + { + "epoch": 0.08192, + "grad_norm": 0.8729857802391052, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 8192 + }, + { + "epoch": 0.08193, + "grad_norm": 0.6815803050994873, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 8193 + }, + { + "epoch": 0.08194, + "grad_norm": 0.7523767352104187, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 8194 + }, + { + "epoch": 0.08195, + "grad_norm": 0.9143834710121155, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 8195 + }, + { + "epoch": 0.08196, + "grad_norm": 1.0875829458236694, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 8196 + }, + { + "epoch": 0.08197, + "grad_norm": 0.944886326789856, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 8197 + }, + { + "epoch": 0.08198, + "grad_norm": 0.9827715754508972, + "learning_rate": 0.003, + "loss": 4.045, + "step": 8198 + }, + { + "epoch": 0.08199, + "grad_norm": 1.0734338760375977, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 8199 + }, + { + "epoch": 0.082, + "grad_norm": 0.953004777431488, + "learning_rate": 0.003, + "loss": 4.033, + "step": 8200 + }, + { + "epoch": 0.08201, + "grad_norm": 0.9488988518714905, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 8201 + }, + { + "epoch": 0.08202, + "grad_norm": 0.9760498404502869, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 8202 + }, + { + "epoch": 0.08203, + "grad_norm": 1.027105689048767, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 8203 + }, + { + "epoch": 0.08204, + "grad_norm": 1.0658153295516968, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 8204 + }, + { + "epoch": 0.08205, + "grad_norm": 1.0252691507339478, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 8205 + }, + { + "epoch": 0.08206, + "grad_norm": 0.7699146866798401, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 8206 + }, + { + "epoch": 0.08207, + "grad_norm": 0.7636826634407043, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 8207 + }, + { + "epoch": 0.08208, + "grad_norm": 0.8411189913749695, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 8208 + }, + { + "epoch": 0.08209, + "grad_norm": 0.9034724831581116, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 8209 + }, + { + "epoch": 0.0821, + "grad_norm": 0.9367898106575012, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 8210 + }, + { + "epoch": 0.08211, + "grad_norm": 0.7969280481338501, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 8211 + }, + { + "epoch": 0.08212, + "grad_norm": 0.7595282196998596, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 8212 + }, + { + "epoch": 0.08213, + "grad_norm": 0.6862097382545471, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 8213 + }, + { + "epoch": 0.08214, + "grad_norm": 0.6680603623390198, + "learning_rate": 0.003, + "loss": 4.072, + "step": 8214 + }, + { + "epoch": 0.08215, + "grad_norm": 0.6800533533096313, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 8215 + }, + { + "epoch": 0.08216, + "grad_norm": 0.673785388469696, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 8216 + }, + { + "epoch": 0.08217, + "grad_norm": 0.5008993148803711, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 8217 + }, + { + "epoch": 0.08218, + "grad_norm": 0.562592089176178, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 8218 + }, + { + "epoch": 0.08219, + "grad_norm": 0.49677103757858276, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 8219 + }, + { + "epoch": 0.0822, + "grad_norm": 0.46024832129478455, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 8220 + }, + { + "epoch": 0.08221, + "grad_norm": 0.6039679050445557, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 8221 + }, + { + "epoch": 0.08222, + "grad_norm": 0.7332128286361694, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 8222 + }, + { + "epoch": 0.08223, + "grad_norm": 0.9698393940925598, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 8223 + }, + { + "epoch": 0.08224, + "grad_norm": 1.07571280002594, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 8224 + }, + { + "epoch": 0.08225, + "grad_norm": 0.7147070169448853, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 8225 + }, + { + "epoch": 0.08226, + "grad_norm": 0.6036560535430908, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 8226 + }, + { + "epoch": 0.08227, + "grad_norm": 0.6823557615280151, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 8227 + }, + { + "epoch": 0.08228, + "grad_norm": 0.6876879930496216, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 8228 + }, + { + "epoch": 0.08229, + "grad_norm": 0.6408465504646301, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 8229 + }, + { + "epoch": 0.0823, + "grad_norm": 0.5930390357971191, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 8230 + }, + { + "epoch": 0.08231, + "grad_norm": 0.6377511024475098, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 8231 + }, + { + "epoch": 0.08232, + "grad_norm": 0.6173083782196045, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 8232 + }, + { + "epoch": 0.08233, + "grad_norm": 0.6037210822105408, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 8233 + }, + { + "epoch": 0.08234, + "grad_norm": 0.6184289455413818, + "learning_rate": 0.003, + "loss": 4.044, + "step": 8234 + }, + { + "epoch": 0.08235, + "grad_norm": 0.6736436486244202, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 8235 + }, + { + "epoch": 0.08236, + "grad_norm": 0.7247467637062073, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 8236 + }, + { + "epoch": 0.08237, + "grad_norm": 0.7914360761642456, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 8237 + }, + { + "epoch": 0.08238, + "grad_norm": 0.8948922157287598, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 8238 + }, + { + "epoch": 0.08239, + "grad_norm": 0.8952855467796326, + "learning_rate": 0.003, + "loss": 4.02, + "step": 8239 + }, + { + "epoch": 0.0824, + "grad_norm": 0.710752546787262, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 8240 + }, + { + "epoch": 0.08241, + "grad_norm": 0.7061362266540527, + "learning_rate": 0.003, + "loss": 4.037, + "step": 8241 + }, + { + "epoch": 0.08242, + "grad_norm": 0.755864679813385, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 8242 + }, + { + "epoch": 0.08243, + "grad_norm": 0.7905144095420837, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 8243 + }, + { + "epoch": 0.08244, + "grad_norm": 0.743241548538208, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 8244 + }, + { + "epoch": 0.08245, + "grad_norm": 0.7604659199714661, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 8245 + }, + { + "epoch": 0.08246, + "grad_norm": 0.6285468935966492, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 8246 + }, + { + "epoch": 0.08247, + "grad_norm": 0.6213167905807495, + "learning_rate": 0.003, + "loss": 4.037, + "step": 8247 + }, + { + "epoch": 0.08248, + "grad_norm": 0.6580025553703308, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 8248 + }, + { + "epoch": 0.08249, + "grad_norm": 0.7325441241264343, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 8249 + }, + { + "epoch": 0.0825, + "grad_norm": 0.698697566986084, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 8250 + }, + { + "epoch": 0.08251, + "grad_norm": 0.7475115656852722, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 8251 + }, + { + "epoch": 0.08252, + "grad_norm": 0.888262152671814, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 8252 + }, + { + "epoch": 0.08253, + "grad_norm": 0.888322651386261, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 8253 + }, + { + "epoch": 0.08254, + "grad_norm": 0.9198958873748779, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 8254 + }, + { + "epoch": 0.08255, + "grad_norm": 0.8950197100639343, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 8255 + }, + { + "epoch": 0.08256, + "grad_norm": 0.9500605463981628, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 8256 + }, + { + "epoch": 0.08257, + "grad_norm": 0.9830167293548584, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 8257 + }, + { + "epoch": 0.08258, + "grad_norm": 0.9901215434074402, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 8258 + }, + { + "epoch": 0.08259, + "grad_norm": 0.8542518019676208, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 8259 + }, + { + "epoch": 0.0826, + "grad_norm": 0.7408918738365173, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 8260 + }, + { + "epoch": 0.08261, + "grad_norm": 0.760269045829773, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 8261 + }, + { + "epoch": 0.08262, + "grad_norm": 0.7092325687408447, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 8262 + }, + { + "epoch": 0.08263, + "grad_norm": 0.6450809240341187, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 8263 + }, + { + "epoch": 0.08264, + "grad_norm": 0.7137148976325989, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 8264 + }, + { + "epoch": 0.08265, + "grad_norm": 0.6923044323921204, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 8265 + }, + { + "epoch": 0.08266, + "grad_norm": 0.6888769865036011, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 8266 + }, + { + "epoch": 0.08267, + "grad_norm": 0.5442489981651306, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 8267 + }, + { + "epoch": 0.08268, + "grad_norm": 0.4939229190349579, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 8268 + }, + { + "epoch": 0.08269, + "grad_norm": 0.5606672167778015, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 8269 + }, + { + "epoch": 0.0827, + "grad_norm": 0.5966683030128479, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 8270 + }, + { + "epoch": 0.08271, + "grad_norm": 0.6150082349777222, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 8271 + }, + { + "epoch": 0.08272, + "grad_norm": 0.697607159614563, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 8272 + }, + { + "epoch": 0.08273, + "grad_norm": 0.7327117323875427, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 8273 + }, + { + "epoch": 0.08274, + "grad_norm": 0.6971558332443237, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 8274 + }, + { + "epoch": 0.08275, + "grad_norm": 0.7999626994132996, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 8275 + }, + { + "epoch": 0.08276, + "grad_norm": 0.9401102066040039, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 8276 + }, + { + "epoch": 0.08277, + "grad_norm": 0.8911102414131165, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 8277 + }, + { + "epoch": 0.08278, + "grad_norm": 0.9046621918678284, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 8278 + }, + { + "epoch": 0.08279, + "grad_norm": 1.0346699953079224, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 8279 + }, + { + "epoch": 0.0828, + "grad_norm": 1.2517218589782715, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 8280 + }, + { + "epoch": 0.08281, + "grad_norm": 0.8035468459129333, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 8281 + }, + { + "epoch": 0.08282, + "grad_norm": 0.7403268218040466, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 8282 + }, + { + "epoch": 0.08283, + "grad_norm": 0.7293475866317749, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 8283 + }, + { + "epoch": 0.08284, + "grad_norm": 0.7586977481842041, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 8284 + }, + { + "epoch": 0.08285, + "grad_norm": 0.8085505366325378, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 8285 + }, + { + "epoch": 0.08286, + "grad_norm": 0.8938177824020386, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 8286 + }, + { + "epoch": 0.08287, + "grad_norm": 0.9943509101867676, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 8287 + }, + { + "epoch": 0.08288, + "grad_norm": 0.9675386548042297, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 8288 + }, + { + "epoch": 0.08289, + "grad_norm": 0.9429525136947632, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 8289 + }, + { + "epoch": 0.0829, + "grad_norm": 0.9028465151786804, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 8290 + }, + { + "epoch": 0.08291, + "grad_norm": 0.8231788873672485, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 8291 + }, + { + "epoch": 0.08292, + "grad_norm": 0.8863356113433838, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 8292 + }, + { + "epoch": 0.08293, + "grad_norm": 0.7855710983276367, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 8293 + }, + { + "epoch": 0.08294, + "grad_norm": 0.7704231142997742, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 8294 + }, + { + "epoch": 0.08295, + "grad_norm": 0.799931526184082, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 8295 + }, + { + "epoch": 0.08296, + "grad_norm": 0.8029519319534302, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 8296 + }, + { + "epoch": 0.08297, + "grad_norm": 0.8360529541969299, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 8297 + }, + { + "epoch": 0.08298, + "grad_norm": 0.8334509134292603, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 8298 + }, + { + "epoch": 0.08299, + "grad_norm": 0.8397669792175293, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 8299 + }, + { + "epoch": 0.083, + "grad_norm": 0.6753296256065369, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 8300 + }, + { + "epoch": 0.08301, + "grad_norm": 0.598036527633667, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 8301 + }, + { + "epoch": 0.08302, + "grad_norm": 0.5940173864364624, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 8302 + }, + { + "epoch": 0.08303, + "grad_norm": 0.6859657168388367, + "learning_rate": 0.003, + "loss": 4.029, + "step": 8303 + }, + { + "epoch": 0.08304, + "grad_norm": 0.7473353147506714, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 8304 + }, + { + "epoch": 0.08305, + "grad_norm": 0.7625237703323364, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 8305 + }, + { + "epoch": 0.08306, + "grad_norm": 0.6629177331924438, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 8306 + }, + { + "epoch": 0.08307, + "grad_norm": 0.6138797402381897, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 8307 + }, + { + "epoch": 0.08308, + "grad_norm": 0.6568807363510132, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 8308 + }, + { + "epoch": 0.08309, + "grad_norm": 0.5892602801322937, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 8309 + }, + { + "epoch": 0.0831, + "grad_norm": 0.5024274587631226, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 8310 + }, + { + "epoch": 0.08311, + "grad_norm": 0.5848341584205627, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 8311 + }, + { + "epoch": 0.08312, + "grad_norm": 0.6992815732955933, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 8312 + }, + { + "epoch": 0.08313, + "grad_norm": 0.7933979630470276, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 8313 + }, + { + "epoch": 0.08314, + "grad_norm": 0.7513596415519714, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 8314 + }, + { + "epoch": 0.08315, + "grad_norm": 0.6562137007713318, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 8315 + }, + { + "epoch": 0.08316, + "grad_norm": 0.7130046486854553, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 8316 + }, + { + "epoch": 0.08317, + "grad_norm": 0.9234418869018555, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 8317 + }, + { + "epoch": 0.08318, + "grad_norm": 0.9351674318313599, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 8318 + }, + { + "epoch": 0.08319, + "grad_norm": 0.9117304682731628, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 8319 + }, + { + "epoch": 0.0832, + "grad_norm": 0.8410887122154236, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 8320 + }, + { + "epoch": 0.08321, + "grad_norm": 0.6563390493392944, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 8321 + }, + { + "epoch": 0.08322, + "grad_norm": 0.6313189268112183, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 8322 + }, + { + "epoch": 0.08323, + "grad_norm": 0.6625282168388367, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 8323 + }, + { + "epoch": 0.08324, + "grad_norm": 0.7624503970146179, + "learning_rate": 0.003, + "loss": 4.038, + "step": 8324 + }, + { + "epoch": 0.08325, + "grad_norm": 0.828724205493927, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 8325 + }, + { + "epoch": 0.08326, + "grad_norm": 0.9103044271469116, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 8326 + }, + { + "epoch": 0.08327, + "grad_norm": 0.7918485999107361, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 8327 + }, + { + "epoch": 0.08328, + "grad_norm": 0.8280613422393799, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 8328 + }, + { + "epoch": 0.08329, + "grad_norm": 0.8416042327880859, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 8329 + }, + { + "epoch": 0.0833, + "grad_norm": 0.8294980525970459, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 8330 + }, + { + "epoch": 0.08331, + "grad_norm": 0.9347625970840454, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 8331 + }, + { + "epoch": 0.08332, + "grad_norm": 1.0204062461853027, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 8332 + }, + { + "epoch": 0.08333, + "grad_norm": 1.0275133848190308, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 8333 + }, + { + "epoch": 0.08334, + "grad_norm": 0.8066427111625671, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 8334 + }, + { + "epoch": 0.08335, + "grad_norm": 0.6177697777748108, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 8335 + }, + { + "epoch": 0.08336, + "grad_norm": 0.5968998670578003, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 8336 + }, + { + "epoch": 0.08337, + "grad_norm": 0.6666345000267029, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 8337 + }, + { + "epoch": 0.08338, + "grad_norm": 0.7190283536911011, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 8338 + }, + { + "epoch": 0.08339, + "grad_norm": 0.7866709232330322, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 8339 + }, + { + "epoch": 0.0834, + "grad_norm": 0.8350096344947815, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 8340 + }, + { + "epoch": 0.08341, + "grad_norm": 0.8239350914955139, + "learning_rate": 0.003, + "loss": 4.067, + "step": 8341 + }, + { + "epoch": 0.08342, + "grad_norm": 0.8311486840248108, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 8342 + }, + { + "epoch": 0.08343, + "grad_norm": 0.7377791404724121, + "learning_rate": 0.003, + "loss": 4.02, + "step": 8343 + }, + { + "epoch": 0.08344, + "grad_norm": 0.8086702823638916, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 8344 + }, + { + "epoch": 0.08345, + "grad_norm": 0.9363798499107361, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 8345 + }, + { + "epoch": 0.08346, + "grad_norm": 0.959582507610321, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 8346 + }, + { + "epoch": 0.08347, + "grad_norm": 0.9427142143249512, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 8347 + }, + { + "epoch": 0.08348, + "grad_norm": 0.9667640924453735, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 8348 + }, + { + "epoch": 0.08349, + "grad_norm": 0.89124596118927, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 8349 + }, + { + "epoch": 0.0835, + "grad_norm": 0.7665905952453613, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 8350 + }, + { + "epoch": 0.08351, + "grad_norm": 0.7947695851325989, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 8351 + }, + { + "epoch": 0.08352, + "grad_norm": 0.7680903673171997, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 8352 + }, + { + "epoch": 0.08353, + "grad_norm": 0.7358306050300598, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 8353 + }, + { + "epoch": 0.08354, + "grad_norm": 0.7130650281906128, + "learning_rate": 0.003, + "loss": 4.053, + "step": 8354 + }, + { + "epoch": 0.08355, + "grad_norm": 0.6731503009796143, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 8355 + }, + { + "epoch": 0.08356, + "grad_norm": 0.711292028427124, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 8356 + }, + { + "epoch": 0.08357, + "grad_norm": 0.7001312375068665, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 8357 + }, + { + "epoch": 0.08358, + "grad_norm": 0.6773588061332703, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 8358 + }, + { + "epoch": 0.08359, + "grad_norm": 0.7479463815689087, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 8359 + }, + { + "epoch": 0.0836, + "grad_norm": 0.8306698203086853, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 8360 + }, + { + "epoch": 0.08361, + "grad_norm": 0.878010094165802, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 8361 + }, + { + "epoch": 0.08362, + "grad_norm": 1.0022238492965698, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 8362 + }, + { + "epoch": 0.08363, + "grad_norm": 1.0330588817596436, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 8363 + }, + { + "epoch": 0.08364, + "grad_norm": 0.718574047088623, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 8364 + }, + { + "epoch": 0.08365, + "grad_norm": 0.5936159491539001, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 8365 + }, + { + "epoch": 0.08366, + "grad_norm": 0.7422720193862915, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 8366 + }, + { + "epoch": 0.08367, + "grad_norm": 0.743090808391571, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 8367 + }, + { + "epoch": 0.08368, + "grad_norm": 0.7947729825973511, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 8368 + }, + { + "epoch": 0.08369, + "grad_norm": 0.837806224822998, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 8369 + }, + { + "epoch": 0.0837, + "grad_norm": 0.783185601234436, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 8370 + }, + { + "epoch": 0.08371, + "grad_norm": 0.642991304397583, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 8371 + }, + { + "epoch": 0.08372, + "grad_norm": 0.5675342082977295, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 8372 + }, + { + "epoch": 0.08373, + "grad_norm": 0.6473720073699951, + "learning_rate": 0.003, + "loss": 4.036, + "step": 8373 + }, + { + "epoch": 0.08374, + "grad_norm": 0.7581577897071838, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 8374 + }, + { + "epoch": 0.08375, + "grad_norm": 0.710021436214447, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 8375 + }, + { + "epoch": 0.08376, + "grad_norm": 0.6476665735244751, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 8376 + }, + { + "epoch": 0.08377, + "grad_norm": 0.5653111338615417, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 8377 + }, + { + "epoch": 0.08378, + "grad_norm": 0.5026406645774841, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 8378 + }, + { + "epoch": 0.08379, + "grad_norm": 0.5006636381149292, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 8379 + }, + { + "epoch": 0.0838, + "grad_norm": 0.5302611589431763, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 8380 + }, + { + "epoch": 0.08381, + "grad_norm": 0.49844783544540405, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 8381 + }, + { + "epoch": 0.08382, + "grad_norm": 0.5049042105674744, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 8382 + }, + { + "epoch": 0.08383, + "grad_norm": 0.5383378267288208, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 8383 + }, + { + "epoch": 0.08384, + "grad_norm": 0.5285283327102661, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 8384 + }, + { + "epoch": 0.08385, + "grad_norm": 0.6071128249168396, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 8385 + }, + { + "epoch": 0.08386, + "grad_norm": 0.7259279489517212, + "learning_rate": 0.003, + "loss": 4.024, + "step": 8386 + }, + { + "epoch": 0.08387, + "grad_norm": 0.9096173644065857, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 8387 + }, + { + "epoch": 0.08388, + "grad_norm": 1.0137066841125488, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 8388 + }, + { + "epoch": 0.08389, + "grad_norm": 0.9626190662384033, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 8389 + }, + { + "epoch": 0.0839, + "grad_norm": 0.7948998212814331, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 8390 + }, + { + "epoch": 0.08391, + "grad_norm": 0.8329476714134216, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 8391 + }, + { + "epoch": 0.08392, + "grad_norm": 0.9139432311058044, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 8392 + }, + { + "epoch": 0.08393, + "grad_norm": 0.7734724283218384, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 8393 + }, + { + "epoch": 0.08394, + "grad_norm": 0.7710996270179749, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 8394 + }, + { + "epoch": 0.08395, + "grad_norm": 0.8262189030647278, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 8395 + }, + { + "epoch": 0.08396, + "grad_norm": 0.9476003646850586, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 8396 + }, + { + "epoch": 0.08397, + "grad_norm": 1.077710509300232, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 8397 + }, + { + "epoch": 0.08398, + "grad_norm": 0.8334670066833496, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 8398 + }, + { + "epoch": 0.08399, + "grad_norm": 0.8015785217285156, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 8399 + }, + { + "epoch": 0.084, + "grad_norm": 0.78190016746521, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 8400 + }, + { + "epoch": 0.08401, + "grad_norm": 0.7719368934631348, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 8401 + }, + { + "epoch": 0.08402, + "grad_norm": 0.7329378128051758, + "learning_rate": 0.003, + "loss": 4.02, + "step": 8402 + }, + { + "epoch": 0.08403, + "grad_norm": 0.6700614094734192, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 8403 + }, + { + "epoch": 0.08404, + "grad_norm": 0.6247296333312988, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 8404 + }, + { + "epoch": 0.08405, + "grad_norm": 0.7085357904434204, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 8405 + }, + { + "epoch": 0.08406, + "grad_norm": 0.864506185054779, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 8406 + }, + { + "epoch": 0.08407, + "grad_norm": 0.9666787981987, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 8407 + }, + { + "epoch": 0.08408, + "grad_norm": 0.9957094788551331, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 8408 + }, + { + "epoch": 0.08409, + "grad_norm": 1.145738124847412, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 8409 + }, + { + "epoch": 0.0841, + "grad_norm": 0.8387814164161682, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 8410 + }, + { + "epoch": 0.08411, + "grad_norm": 0.8150733709335327, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 8411 + }, + { + "epoch": 0.08412, + "grad_norm": 0.8348231315612793, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 8412 + }, + { + "epoch": 0.08413, + "grad_norm": 0.8107559680938721, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 8413 + }, + { + "epoch": 0.08414, + "grad_norm": 0.7736573219299316, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 8414 + }, + { + "epoch": 0.08415, + "grad_norm": 0.7075899839401245, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 8415 + }, + { + "epoch": 0.08416, + "grad_norm": 0.7205349206924438, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 8416 + }, + { + "epoch": 0.08417, + "grad_norm": 0.6560434699058533, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 8417 + }, + { + "epoch": 0.08418, + "grad_norm": 0.5477179288864136, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 8418 + }, + { + "epoch": 0.08419, + "grad_norm": 0.549923300743103, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 8419 + }, + { + "epoch": 0.0842, + "grad_norm": 0.4624122679233551, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 8420 + }, + { + "epoch": 0.08421, + "grad_norm": 0.4707403779029846, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 8421 + }, + { + "epoch": 0.08422, + "grad_norm": 0.6087749004364014, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 8422 + }, + { + "epoch": 0.08423, + "grad_norm": 0.7090073227882385, + "learning_rate": 0.003, + "loss": 4.033, + "step": 8423 + }, + { + "epoch": 0.08424, + "grad_norm": 0.8543527126312256, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 8424 + }, + { + "epoch": 0.08425, + "grad_norm": 1.0288057327270508, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 8425 + }, + { + "epoch": 0.08426, + "grad_norm": 0.989617109298706, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 8426 + }, + { + "epoch": 0.08427, + "grad_norm": 0.8570108413696289, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 8427 + }, + { + "epoch": 0.08428, + "grad_norm": 0.6294158101081848, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 8428 + }, + { + "epoch": 0.08429, + "grad_norm": 0.6182265877723694, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 8429 + }, + { + "epoch": 0.0843, + "grad_norm": 0.8876853585243225, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 8430 + }, + { + "epoch": 0.08431, + "grad_norm": 0.9176228046417236, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 8431 + }, + { + "epoch": 0.08432, + "grad_norm": 0.8123117089271545, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 8432 + }, + { + "epoch": 0.08433, + "grad_norm": 0.8903416991233826, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 8433 + }, + { + "epoch": 0.08434, + "grad_norm": 0.9202600121498108, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 8434 + }, + { + "epoch": 0.08435, + "grad_norm": 0.8581541180610657, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 8435 + }, + { + "epoch": 0.08436, + "grad_norm": 0.8430383205413818, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 8436 + }, + { + "epoch": 0.08437, + "grad_norm": 0.73482346534729, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 8437 + }, + { + "epoch": 0.08438, + "grad_norm": 0.851243257522583, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 8438 + }, + { + "epoch": 0.08439, + "grad_norm": 0.9350183606147766, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 8439 + }, + { + "epoch": 0.0844, + "grad_norm": 0.8600278496742249, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 8440 + }, + { + "epoch": 0.08441, + "grad_norm": 0.9163636565208435, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 8441 + }, + { + "epoch": 0.08442, + "grad_norm": 0.9736191034317017, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 8442 + }, + { + "epoch": 0.08443, + "grad_norm": 0.9900960922241211, + "learning_rate": 0.003, + "loss": 4.053, + "step": 8443 + }, + { + "epoch": 0.08444, + "grad_norm": 0.9450505971908569, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 8444 + }, + { + "epoch": 0.08445, + "grad_norm": 0.8399696946144104, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 8445 + }, + { + "epoch": 0.08446, + "grad_norm": 0.715254008769989, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 8446 + }, + { + "epoch": 0.08447, + "grad_norm": 0.7299708127975464, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 8447 + }, + { + "epoch": 0.08448, + "grad_norm": 0.7352356314659119, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 8448 + }, + { + "epoch": 0.08449, + "grad_norm": 0.7627168893814087, + "learning_rate": 0.003, + "loss": 4.068, + "step": 8449 + }, + { + "epoch": 0.0845, + "grad_norm": 0.8209385871887207, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 8450 + }, + { + "epoch": 0.08451, + "grad_norm": 0.9701636433601379, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 8451 + }, + { + "epoch": 0.08452, + "grad_norm": 0.9062989950180054, + "learning_rate": 0.003, + "loss": 4.043, + "step": 8452 + }, + { + "epoch": 0.08453, + "grad_norm": 0.7730754613876343, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 8453 + }, + { + "epoch": 0.08454, + "grad_norm": 0.6160298585891724, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 8454 + }, + { + "epoch": 0.08455, + "grad_norm": 0.5016544461250305, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 8455 + }, + { + "epoch": 0.08456, + "grad_norm": 0.4829849302768707, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 8456 + }, + { + "epoch": 0.08457, + "grad_norm": 0.6312108039855957, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 8457 + }, + { + "epoch": 0.08458, + "grad_norm": 0.6965609788894653, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 8458 + }, + { + "epoch": 0.08459, + "grad_norm": 0.6963859796524048, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 8459 + }, + { + "epoch": 0.0846, + "grad_norm": 0.6759956479072571, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 8460 + }, + { + "epoch": 0.08461, + "grad_norm": 0.5745083093643188, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 8461 + }, + { + "epoch": 0.08462, + "grad_norm": 0.5628439784049988, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 8462 + }, + { + "epoch": 0.08463, + "grad_norm": 0.5284938216209412, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 8463 + }, + { + "epoch": 0.08464, + "grad_norm": 0.4526340961456299, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 8464 + }, + { + "epoch": 0.08465, + "grad_norm": 0.42585697770118713, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 8465 + }, + { + "epoch": 0.08466, + "grad_norm": 0.45660510659217834, + "learning_rate": 0.003, + "loss": 4.047, + "step": 8466 + }, + { + "epoch": 0.08467, + "grad_norm": 0.4530385434627533, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 8467 + }, + { + "epoch": 0.08468, + "grad_norm": 0.5353749394416809, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 8468 + }, + { + "epoch": 0.08469, + "grad_norm": 0.557474672794342, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 8469 + }, + { + "epoch": 0.0847, + "grad_norm": 0.5882935523986816, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 8470 + }, + { + "epoch": 0.08471, + "grad_norm": 0.7039132118225098, + "learning_rate": 0.003, + "loss": 4.007, + "step": 8471 + }, + { + "epoch": 0.08472, + "grad_norm": 0.9956209063529968, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 8472 + }, + { + "epoch": 0.08473, + "grad_norm": 1.4970457553863525, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 8473 + }, + { + "epoch": 0.08474, + "grad_norm": 0.5254650115966797, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 8474 + }, + { + "epoch": 0.08475, + "grad_norm": 0.8430723547935486, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 8475 + }, + { + "epoch": 0.08476, + "grad_norm": 0.8634896278381348, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 8476 + }, + { + "epoch": 0.08477, + "grad_norm": 0.8257098197937012, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 8477 + }, + { + "epoch": 0.08478, + "grad_norm": 0.8612629175186157, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 8478 + }, + { + "epoch": 0.08479, + "grad_norm": 0.9358077645301819, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 8479 + }, + { + "epoch": 0.0848, + "grad_norm": 0.8783276677131653, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 8480 + }, + { + "epoch": 0.08481, + "grad_norm": 0.9027069211006165, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 8481 + }, + { + "epoch": 0.08482, + "grad_norm": 0.7429633140563965, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 8482 + }, + { + "epoch": 0.08483, + "grad_norm": 0.749813437461853, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 8483 + }, + { + "epoch": 0.08484, + "grad_norm": 0.8067286014556885, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 8484 + }, + { + "epoch": 0.08485, + "grad_norm": 0.8065735697746277, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 8485 + }, + { + "epoch": 0.08486, + "grad_norm": 0.6747359037399292, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 8486 + }, + { + "epoch": 0.08487, + "grad_norm": 0.6390485763549805, + "learning_rate": 0.003, + "loss": 4.063, + "step": 8487 + }, + { + "epoch": 0.08488, + "grad_norm": 0.6853650808334351, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 8488 + }, + { + "epoch": 0.08489, + "grad_norm": 0.7082487940788269, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 8489 + }, + { + "epoch": 0.0849, + "grad_norm": 0.6889592409133911, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 8490 + }, + { + "epoch": 0.08491, + "grad_norm": 0.6803368926048279, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 8491 + }, + { + "epoch": 0.08492, + "grad_norm": 0.7069889307022095, + "learning_rate": 0.003, + "loss": 4.036, + "step": 8492 + }, + { + "epoch": 0.08493, + "grad_norm": 0.7506759166717529, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 8493 + }, + { + "epoch": 0.08494, + "grad_norm": 0.7356934547424316, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 8494 + }, + { + "epoch": 0.08495, + "grad_norm": 0.7744423747062683, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 8495 + }, + { + "epoch": 0.08496, + "grad_norm": 0.7731289863586426, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 8496 + }, + { + "epoch": 0.08497, + "grad_norm": 0.7522779107093811, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 8497 + }, + { + "epoch": 0.08498, + "grad_norm": 0.7073262333869934, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 8498 + }, + { + "epoch": 0.08499, + "grad_norm": 0.7390640377998352, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 8499 + }, + { + "epoch": 0.085, + "grad_norm": 0.7686527967453003, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 8500 + }, + { + "epoch": 0.08501, + "grad_norm": 0.723380446434021, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 8501 + }, + { + "epoch": 0.08502, + "grad_norm": 0.6123513579368591, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 8502 + }, + { + "epoch": 0.08503, + "grad_norm": 0.6763044595718384, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 8503 + }, + { + "epoch": 0.08504, + "grad_norm": 0.7245738506317139, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 8504 + }, + { + "epoch": 0.08505, + "grad_norm": 0.6983570456504822, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 8505 + }, + { + "epoch": 0.08506, + "grad_norm": 0.6399862170219421, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 8506 + }, + { + "epoch": 0.08507, + "grad_norm": 0.6284667253494263, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 8507 + }, + { + "epoch": 0.08508, + "grad_norm": 0.6627932786941528, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 8508 + }, + { + "epoch": 0.08509, + "grad_norm": 0.7929001450538635, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 8509 + }, + { + "epoch": 0.0851, + "grad_norm": 1.0402476787567139, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 8510 + }, + { + "epoch": 0.08511, + "grad_norm": 1.2262110710144043, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 8511 + }, + { + "epoch": 0.08512, + "grad_norm": 0.6842292547225952, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 8512 + }, + { + "epoch": 0.08513, + "grad_norm": 0.7122688293457031, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 8513 + }, + { + "epoch": 0.08514, + "grad_norm": 0.8808112144470215, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 8514 + }, + { + "epoch": 0.08515, + "grad_norm": 0.9627430438995361, + "learning_rate": 0.003, + "loss": 4.039, + "step": 8515 + }, + { + "epoch": 0.08516, + "grad_norm": 1.0835819244384766, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 8516 + }, + { + "epoch": 0.08517, + "grad_norm": 1.015756368637085, + "learning_rate": 0.003, + "loss": 4.041, + "step": 8517 + }, + { + "epoch": 0.08518, + "grad_norm": 0.8399810791015625, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 8518 + }, + { + "epoch": 0.08519, + "grad_norm": 0.8079990744590759, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 8519 + }, + { + "epoch": 0.0852, + "grad_norm": 0.8063580393791199, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 8520 + }, + { + "epoch": 0.08521, + "grad_norm": 0.7848697900772095, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 8521 + }, + { + "epoch": 0.08522, + "grad_norm": 0.7846394181251526, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 8522 + }, + { + "epoch": 0.08523, + "grad_norm": 0.8157345056533813, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 8523 + }, + { + "epoch": 0.08524, + "grad_norm": 0.949707567691803, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 8524 + }, + { + "epoch": 0.08525, + "grad_norm": 1.0673058032989502, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 8525 + }, + { + "epoch": 0.08526, + "grad_norm": 0.9210948944091797, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 8526 + }, + { + "epoch": 0.08527, + "grad_norm": 1.0052591562271118, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 8527 + }, + { + "epoch": 0.08528, + "grad_norm": 1.0522360801696777, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 8528 + }, + { + "epoch": 0.08529, + "grad_norm": 0.8486297130584717, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 8529 + }, + { + "epoch": 0.0853, + "grad_norm": 0.7309673428535461, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 8530 + }, + { + "epoch": 0.08531, + "grad_norm": 0.7416118383407593, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 8531 + }, + { + "epoch": 0.08532, + "grad_norm": 0.7339397668838501, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 8532 + }, + { + "epoch": 0.08533, + "grad_norm": 0.5952752232551575, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 8533 + }, + { + "epoch": 0.08534, + "grad_norm": 0.47096192836761475, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 8534 + }, + { + "epoch": 0.08535, + "grad_norm": 0.48998767137527466, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 8535 + }, + { + "epoch": 0.08536, + "grad_norm": 0.4841568171977997, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 8536 + }, + { + "epoch": 0.08537, + "grad_norm": 0.560813844203949, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 8537 + }, + { + "epoch": 0.08538, + "grad_norm": 0.6467888951301575, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 8538 + }, + { + "epoch": 0.08539, + "grad_norm": 0.7536951899528503, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 8539 + }, + { + "epoch": 0.0854, + "grad_norm": 0.8326345682144165, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 8540 + }, + { + "epoch": 0.08541, + "grad_norm": 0.7367427349090576, + "learning_rate": 0.003, + "loss": 4.015, + "step": 8541 + }, + { + "epoch": 0.08542, + "grad_norm": 0.5583395957946777, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 8542 + }, + { + "epoch": 0.08543, + "grad_norm": 0.48341140151023865, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 8543 + }, + { + "epoch": 0.08544, + "grad_norm": 0.5066325664520264, + "learning_rate": 0.003, + "loss": 4.015, + "step": 8544 + }, + { + "epoch": 0.08545, + "grad_norm": 0.4845677614212036, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 8545 + }, + { + "epoch": 0.08546, + "grad_norm": 0.4747871458530426, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 8546 + }, + { + "epoch": 0.08547, + "grad_norm": 0.4408625364303589, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 8547 + }, + { + "epoch": 0.08548, + "grad_norm": 0.5065740942955017, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 8548 + }, + { + "epoch": 0.08549, + "grad_norm": 0.5792751908302307, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 8549 + }, + { + "epoch": 0.0855, + "grad_norm": 0.6657066345214844, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 8550 + }, + { + "epoch": 0.08551, + "grad_norm": 0.7656866312026978, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 8551 + }, + { + "epoch": 0.08552, + "grad_norm": 0.9525842666625977, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 8552 + }, + { + "epoch": 0.08553, + "grad_norm": 0.986405611038208, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 8553 + }, + { + "epoch": 0.08554, + "grad_norm": 0.9743383526802063, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 8554 + }, + { + "epoch": 0.08555, + "grad_norm": 1.08023202419281, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 8555 + }, + { + "epoch": 0.08556, + "grad_norm": 0.8308687210083008, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 8556 + }, + { + "epoch": 0.08557, + "grad_norm": 0.8413991928100586, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 8557 + }, + { + "epoch": 0.08558, + "grad_norm": 0.8547917008399963, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 8558 + }, + { + "epoch": 0.08559, + "grad_norm": 0.8460670113563538, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 8559 + }, + { + "epoch": 0.0856, + "grad_norm": 0.8860325217247009, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 8560 + }, + { + "epoch": 0.08561, + "grad_norm": 0.8015673756599426, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 8561 + }, + { + "epoch": 0.08562, + "grad_norm": 0.9084799885749817, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 8562 + }, + { + "epoch": 0.08563, + "grad_norm": 0.9597362875938416, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 8563 + }, + { + "epoch": 0.08564, + "grad_norm": 0.8412767648696899, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 8564 + }, + { + "epoch": 0.08565, + "grad_norm": 0.8772968649864197, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 8565 + }, + { + "epoch": 0.08566, + "grad_norm": 0.8721387982368469, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 8566 + }, + { + "epoch": 0.08567, + "grad_norm": 0.9334637522697449, + "learning_rate": 0.003, + "loss": 4.069, + "step": 8567 + }, + { + "epoch": 0.08568, + "grad_norm": 1.1856471300125122, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 8568 + }, + { + "epoch": 0.08569, + "grad_norm": 1.0100841522216797, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 8569 + }, + { + "epoch": 0.0857, + "grad_norm": 0.9219196438789368, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 8570 + }, + { + "epoch": 0.08571, + "grad_norm": 1.1001014709472656, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 8571 + }, + { + "epoch": 0.08572, + "grad_norm": 1.129728078842163, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 8572 + }, + { + "epoch": 0.08573, + "grad_norm": 0.8418722748756409, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 8573 + }, + { + "epoch": 0.08574, + "grad_norm": 0.8911969065666199, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 8574 + }, + { + "epoch": 0.08575, + "grad_norm": 0.9890245199203491, + "learning_rate": 0.003, + "loss": 4.07, + "step": 8575 + }, + { + "epoch": 0.08576, + "grad_norm": 0.9305644631385803, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 8576 + }, + { + "epoch": 0.08577, + "grad_norm": 0.7585141062736511, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 8577 + }, + { + "epoch": 0.08578, + "grad_norm": 0.6668472290039062, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 8578 + }, + { + "epoch": 0.08579, + "grad_norm": 0.6890462040901184, + "learning_rate": 0.003, + "loss": 4.047, + "step": 8579 + }, + { + "epoch": 0.0858, + "grad_norm": 0.5876961946487427, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 8580 + }, + { + "epoch": 0.08581, + "grad_norm": 0.5115332007408142, + "learning_rate": 0.003, + "loss": 4.026, + "step": 8581 + }, + { + "epoch": 0.08582, + "grad_norm": 0.5275139212608337, + "learning_rate": 0.003, + "loss": 4.049, + "step": 8582 + }, + { + "epoch": 0.08583, + "grad_norm": 0.5889208912849426, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 8583 + }, + { + "epoch": 0.08584, + "grad_norm": 0.6184143424034119, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 8584 + }, + { + "epoch": 0.08585, + "grad_norm": 0.5801510810852051, + "learning_rate": 0.003, + "loss": 4.052, + "step": 8585 + }, + { + "epoch": 0.08586, + "grad_norm": 0.5840868353843689, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 8586 + }, + { + "epoch": 0.08587, + "grad_norm": 0.5485032796859741, + "learning_rate": 0.003, + "loss": 4.041, + "step": 8587 + }, + { + "epoch": 0.08588, + "grad_norm": 0.596748411655426, + "learning_rate": 0.003, + "loss": 4.02, + "step": 8588 + }, + { + "epoch": 0.08589, + "grad_norm": 0.5736215710639954, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 8589 + }, + { + "epoch": 0.0859, + "grad_norm": 0.5293859839439392, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 8590 + }, + { + "epoch": 0.08591, + "grad_norm": 0.5721410512924194, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 8591 + }, + { + "epoch": 0.08592, + "grad_norm": 0.7228186130523682, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 8592 + }, + { + "epoch": 0.08593, + "grad_norm": 0.9087763428688049, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 8593 + }, + { + "epoch": 0.08594, + "grad_norm": 1.1849374771118164, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 8594 + }, + { + "epoch": 0.08595, + "grad_norm": 0.8020411133766174, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 8595 + }, + { + "epoch": 0.08596, + "grad_norm": 0.6049209833145142, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 8596 + }, + { + "epoch": 0.08597, + "grad_norm": 0.5967742204666138, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 8597 + }, + { + "epoch": 0.08598, + "grad_norm": 0.6498646140098572, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 8598 + }, + { + "epoch": 0.08599, + "grad_norm": 0.6824044585227966, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 8599 + }, + { + "epoch": 0.086, + "grad_norm": 0.6595525741577148, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 8600 + }, + { + "epoch": 0.08601, + "grad_norm": 0.5955501198768616, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 8601 + }, + { + "epoch": 0.08602, + "grad_norm": 0.5499855279922485, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 8602 + }, + { + "epoch": 0.08603, + "grad_norm": 0.6001824140548706, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 8603 + }, + { + "epoch": 0.08604, + "grad_norm": 0.6126075983047485, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 8604 + }, + { + "epoch": 0.08605, + "grad_norm": 0.7098419666290283, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 8605 + }, + { + "epoch": 0.08606, + "grad_norm": 0.7325479388237, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 8606 + }, + { + "epoch": 0.08607, + "grad_norm": 0.7961235046386719, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 8607 + }, + { + "epoch": 0.08608, + "grad_norm": 0.7546827793121338, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 8608 + }, + { + "epoch": 0.08609, + "grad_norm": 0.715076744556427, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 8609 + }, + { + "epoch": 0.0861, + "grad_norm": 0.6888747215270996, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 8610 + }, + { + "epoch": 0.08611, + "grad_norm": 0.6887075304985046, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 8611 + }, + { + "epoch": 0.08612, + "grad_norm": 0.8714022040367126, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 8612 + }, + { + "epoch": 0.08613, + "grad_norm": 1.1010104417800903, + "learning_rate": 0.003, + "loss": 4.052, + "step": 8613 + }, + { + "epoch": 0.08614, + "grad_norm": 1.023213267326355, + "learning_rate": 0.003, + "loss": 4.033, + "step": 8614 + }, + { + "epoch": 0.08615, + "grad_norm": 0.9047683477401733, + "learning_rate": 0.003, + "loss": 4.036, + "step": 8615 + }, + { + "epoch": 0.08616, + "grad_norm": 0.9519088268280029, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 8616 + }, + { + "epoch": 0.08617, + "grad_norm": 0.8695728182792664, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 8617 + }, + { + "epoch": 0.08618, + "grad_norm": 0.8256757259368896, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 8618 + }, + { + "epoch": 0.08619, + "grad_norm": 0.8691619634628296, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 8619 + }, + { + "epoch": 0.0862, + "grad_norm": 0.9401669502258301, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 8620 + }, + { + "epoch": 0.08621, + "grad_norm": 1.0106126070022583, + "learning_rate": 0.003, + "loss": 4.048, + "step": 8621 + }, + { + "epoch": 0.08622, + "grad_norm": 1.0970956087112427, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 8622 + }, + { + "epoch": 0.08623, + "grad_norm": 0.9480116367340088, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 8623 + }, + { + "epoch": 0.08624, + "grad_norm": 0.8120977282524109, + "learning_rate": 0.003, + "loss": 4.046, + "step": 8624 + }, + { + "epoch": 0.08625, + "grad_norm": 0.7313287258148193, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 8625 + }, + { + "epoch": 0.08626, + "grad_norm": 0.7254028916358948, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 8626 + }, + { + "epoch": 0.08627, + "grad_norm": 0.8220468163490295, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 8627 + }, + { + "epoch": 0.08628, + "grad_norm": 0.8760457038879395, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 8628 + }, + { + "epoch": 0.08629, + "grad_norm": 0.6859437227249146, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 8629 + }, + { + "epoch": 0.0863, + "grad_norm": 0.7072209715843201, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 8630 + }, + { + "epoch": 0.08631, + "grad_norm": 0.54953932762146, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 8631 + }, + { + "epoch": 0.08632, + "grad_norm": 0.5660358667373657, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 8632 + }, + { + "epoch": 0.08633, + "grad_norm": 0.621670663356781, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 8633 + }, + { + "epoch": 0.08634, + "grad_norm": 0.6216848492622375, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 8634 + }, + { + "epoch": 0.08635, + "grad_norm": 0.573436439037323, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 8635 + }, + { + "epoch": 0.08636, + "grad_norm": 0.5630894899368286, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 8636 + }, + { + "epoch": 0.08637, + "grad_norm": 0.776887834072113, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 8637 + }, + { + "epoch": 0.08638, + "grad_norm": 0.9575220346450806, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 8638 + }, + { + "epoch": 0.08639, + "grad_norm": 1.1651415824890137, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 8639 + }, + { + "epoch": 0.0864, + "grad_norm": 0.8188918828964233, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 8640 + }, + { + "epoch": 0.08641, + "grad_norm": 0.49055084586143494, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 8641 + }, + { + "epoch": 0.08642, + "grad_norm": 0.5482150316238403, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 8642 + }, + { + "epoch": 0.08643, + "grad_norm": 0.6328015923500061, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 8643 + }, + { + "epoch": 0.08644, + "grad_norm": 0.6240196824073792, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 8644 + }, + { + "epoch": 0.08645, + "grad_norm": 0.5255370140075684, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 8645 + }, + { + "epoch": 0.08646, + "grad_norm": 0.5354486703872681, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 8646 + }, + { + "epoch": 0.08647, + "grad_norm": 0.5208635926246643, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 8647 + }, + { + "epoch": 0.08648, + "grad_norm": 0.4734020233154297, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 8648 + }, + { + "epoch": 0.08649, + "grad_norm": 0.49996834993362427, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 8649 + }, + { + "epoch": 0.0865, + "grad_norm": 0.6042665839195251, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 8650 + }, + { + "epoch": 0.08651, + "grad_norm": 0.647171139717102, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 8651 + }, + { + "epoch": 0.08652, + "grad_norm": 0.6678754687309265, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 8652 + }, + { + "epoch": 0.08653, + "grad_norm": 0.7537620663642883, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 8653 + }, + { + "epoch": 0.08654, + "grad_norm": 0.8044998049736023, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 8654 + }, + { + "epoch": 0.08655, + "grad_norm": 0.8797716498374939, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 8655 + }, + { + "epoch": 0.08656, + "grad_norm": 0.9785940051078796, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 8656 + }, + { + "epoch": 0.08657, + "grad_norm": 0.9777019023895264, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 8657 + }, + { + "epoch": 0.08658, + "grad_norm": 0.9151862263679504, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 8658 + }, + { + "epoch": 0.08659, + "grad_norm": 0.8125520348548889, + "learning_rate": 0.003, + "loss": 4.035, + "step": 8659 + }, + { + "epoch": 0.0866, + "grad_norm": 0.8756903409957886, + "learning_rate": 0.003, + "loss": 4.038, + "step": 8660 + }, + { + "epoch": 0.08661, + "grad_norm": 0.9547138810157776, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 8661 + }, + { + "epoch": 0.08662, + "grad_norm": 0.7263685464859009, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 8662 + }, + { + "epoch": 0.08663, + "grad_norm": 0.7474812269210815, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 8663 + }, + { + "epoch": 0.08664, + "grad_norm": 0.7189542055130005, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 8664 + }, + { + "epoch": 0.08665, + "grad_norm": 0.833405077457428, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 8665 + }, + { + "epoch": 0.08666, + "grad_norm": 1.123273491859436, + "learning_rate": 0.003, + "loss": 4.048, + "step": 8666 + }, + { + "epoch": 0.08667, + "grad_norm": 0.9802038073539734, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 8667 + }, + { + "epoch": 0.08668, + "grad_norm": 0.8807769417762756, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 8668 + }, + { + "epoch": 0.08669, + "grad_norm": 0.7354418635368347, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 8669 + }, + { + "epoch": 0.0867, + "grad_norm": 0.6019794940948486, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 8670 + }, + { + "epoch": 0.08671, + "grad_norm": 0.5857205986976624, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 8671 + }, + { + "epoch": 0.08672, + "grad_norm": 0.5742001533508301, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 8672 + }, + { + "epoch": 0.08673, + "grad_norm": 0.6621288657188416, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 8673 + }, + { + "epoch": 0.08674, + "grad_norm": 0.719313383102417, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 8674 + }, + { + "epoch": 0.08675, + "grad_norm": 0.8486753106117249, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 8675 + }, + { + "epoch": 0.08676, + "grad_norm": 0.7985334992408752, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 8676 + }, + { + "epoch": 0.08677, + "grad_norm": 0.7688699960708618, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 8677 + }, + { + "epoch": 0.08678, + "grad_norm": 0.801563024520874, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 8678 + }, + { + "epoch": 0.08679, + "grad_norm": 0.791769802570343, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 8679 + }, + { + "epoch": 0.0868, + "grad_norm": 0.8047665953636169, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 8680 + }, + { + "epoch": 0.08681, + "grad_norm": 0.9945483803749084, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 8681 + }, + { + "epoch": 0.08682, + "grad_norm": 0.9851046800613403, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 8682 + }, + { + "epoch": 0.08683, + "grad_norm": 1.0485612154006958, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 8683 + }, + { + "epoch": 0.08684, + "grad_norm": 1.0201568603515625, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 8684 + }, + { + "epoch": 0.08685, + "grad_norm": 1.0376875400543213, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 8685 + }, + { + "epoch": 0.08686, + "grad_norm": 0.805310845375061, + "learning_rate": 0.003, + "loss": 4.064, + "step": 8686 + }, + { + "epoch": 0.08687, + "grad_norm": 0.754871666431427, + "learning_rate": 0.003, + "loss": 4.067, + "step": 8687 + }, + { + "epoch": 0.08688, + "grad_norm": 0.8903035521507263, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 8688 + }, + { + "epoch": 0.08689, + "grad_norm": 1.157721996307373, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 8689 + }, + { + "epoch": 0.0869, + "grad_norm": 0.8955743312835693, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 8690 + }, + { + "epoch": 0.08691, + "grad_norm": 0.7679038047790527, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 8691 + }, + { + "epoch": 0.08692, + "grad_norm": 0.5723802447319031, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 8692 + }, + { + "epoch": 0.08693, + "grad_norm": 0.6420039534568787, + "learning_rate": 0.003, + "loss": 4.024, + "step": 8693 + }, + { + "epoch": 0.08694, + "grad_norm": 0.8482261896133423, + "learning_rate": 0.003, + "loss": 4.041, + "step": 8694 + }, + { + "epoch": 0.08695, + "grad_norm": 0.9729529619216919, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 8695 + }, + { + "epoch": 0.08696, + "grad_norm": 1.0107791423797607, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 8696 + }, + { + "epoch": 0.08697, + "grad_norm": 0.7928081154823303, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 8697 + }, + { + "epoch": 0.08698, + "grad_norm": 0.664397656917572, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 8698 + }, + { + "epoch": 0.08699, + "grad_norm": 0.6472232937812805, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 8699 + }, + { + "epoch": 0.087, + "grad_norm": 0.644534170627594, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 8700 + }, + { + "epoch": 0.08701, + "grad_norm": 0.7472826838493347, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 8701 + }, + { + "epoch": 0.08702, + "grad_norm": 0.7674583196640015, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 8702 + }, + { + "epoch": 0.08703, + "grad_norm": 0.7745077013969421, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 8703 + }, + { + "epoch": 0.08704, + "grad_norm": 0.695412814617157, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 8704 + }, + { + "epoch": 0.08705, + "grad_norm": 0.6802660822868347, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 8705 + }, + { + "epoch": 0.08706, + "grad_norm": 0.6403552889823914, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 8706 + }, + { + "epoch": 0.08707, + "grad_norm": 0.5612523555755615, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 8707 + }, + { + "epoch": 0.08708, + "grad_norm": 0.5290988683700562, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 8708 + }, + { + "epoch": 0.08709, + "grad_norm": 0.5752286911010742, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 8709 + }, + { + "epoch": 0.0871, + "grad_norm": 0.5760716199874878, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 8710 + }, + { + "epoch": 0.08711, + "grad_norm": 0.6479002237319946, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 8711 + }, + { + "epoch": 0.08712, + "grad_norm": 0.6316835880279541, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 8712 + }, + { + "epoch": 0.08713, + "grad_norm": 0.6163070797920227, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 8713 + }, + { + "epoch": 0.08714, + "grad_norm": 0.6544851064682007, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 8714 + }, + { + "epoch": 0.08715, + "grad_norm": 0.6510108113288879, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 8715 + }, + { + "epoch": 0.08716, + "grad_norm": 0.6509416103363037, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 8716 + }, + { + "epoch": 0.08717, + "grad_norm": 0.6482957601547241, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 8717 + }, + { + "epoch": 0.08718, + "grad_norm": 0.6902303099632263, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 8718 + }, + { + "epoch": 0.08719, + "grad_norm": 0.7940294146537781, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 8719 + }, + { + "epoch": 0.0872, + "grad_norm": 0.9621227979660034, + "learning_rate": 0.003, + "loss": 4.024, + "step": 8720 + }, + { + "epoch": 0.08721, + "grad_norm": 1.084385871887207, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 8721 + }, + { + "epoch": 0.08722, + "grad_norm": 0.8606046438217163, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 8722 + }, + { + "epoch": 0.08723, + "grad_norm": 0.7293562293052673, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 8723 + }, + { + "epoch": 0.08724, + "grad_norm": 0.8104085922241211, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 8724 + }, + { + "epoch": 0.08725, + "grad_norm": 1.134535789489746, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 8725 + }, + { + "epoch": 0.08726, + "grad_norm": 0.9137052893638611, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 8726 + }, + { + "epoch": 0.08727, + "grad_norm": 0.7882857918739319, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 8727 + }, + { + "epoch": 0.08728, + "grad_norm": 0.6920211911201477, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 8728 + }, + { + "epoch": 0.08729, + "grad_norm": 0.769177258014679, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 8729 + }, + { + "epoch": 0.0873, + "grad_norm": 0.7354542016983032, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 8730 + }, + { + "epoch": 0.08731, + "grad_norm": 0.6996873617172241, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 8731 + }, + { + "epoch": 0.08732, + "grad_norm": 0.6775871515274048, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 8732 + }, + { + "epoch": 0.08733, + "grad_norm": 0.6581737995147705, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 8733 + }, + { + "epoch": 0.08734, + "grad_norm": 0.6336266994476318, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 8734 + }, + { + "epoch": 0.08735, + "grad_norm": 0.6769811511039734, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 8735 + }, + { + "epoch": 0.08736, + "grad_norm": 0.6477938890457153, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 8736 + }, + { + "epoch": 0.08737, + "grad_norm": 0.5973692536354065, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 8737 + }, + { + "epoch": 0.08738, + "grad_norm": 0.6644995808601379, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 8738 + }, + { + "epoch": 0.08739, + "grad_norm": 0.8802757859230042, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 8739 + }, + { + "epoch": 0.0874, + "grad_norm": 1.0935834646224976, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 8740 + }, + { + "epoch": 0.08741, + "grad_norm": 0.9527744650840759, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 8741 + }, + { + "epoch": 0.08742, + "grad_norm": 0.7332203984260559, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 8742 + }, + { + "epoch": 0.08743, + "grad_norm": 0.7583456635475159, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 8743 + }, + { + "epoch": 0.08744, + "grad_norm": 0.7854973673820496, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 8744 + }, + { + "epoch": 0.08745, + "grad_norm": 0.721187949180603, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 8745 + }, + { + "epoch": 0.08746, + "grad_norm": 0.7170279026031494, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 8746 + }, + { + "epoch": 0.08747, + "grad_norm": 0.6679021120071411, + "learning_rate": 0.003, + "loss": 4.043, + "step": 8747 + }, + { + "epoch": 0.08748, + "grad_norm": 0.6513135433197021, + "learning_rate": 0.003, + "loss": 4.03, + "step": 8748 + }, + { + "epoch": 0.08749, + "grad_norm": 0.7048290371894836, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 8749 + }, + { + "epoch": 0.0875, + "grad_norm": 0.8222705721855164, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 8750 + }, + { + "epoch": 0.08751, + "grad_norm": 0.8615703582763672, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 8751 + }, + { + "epoch": 0.08752, + "grad_norm": 0.9447921514511108, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 8752 + }, + { + "epoch": 0.08753, + "grad_norm": 1.176733136177063, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 8753 + }, + { + "epoch": 0.08754, + "grad_norm": 0.6545034050941467, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 8754 + }, + { + "epoch": 0.08755, + "grad_norm": 0.681919276714325, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 8755 + }, + { + "epoch": 0.08756, + "grad_norm": 0.7190672159194946, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 8756 + }, + { + "epoch": 0.08757, + "grad_norm": 0.7491109371185303, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 8757 + }, + { + "epoch": 0.08758, + "grad_norm": 0.8422783017158508, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 8758 + }, + { + "epoch": 0.08759, + "grad_norm": 0.8749980330467224, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 8759 + }, + { + "epoch": 0.0876, + "grad_norm": 0.8432993292808533, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 8760 + }, + { + "epoch": 0.08761, + "grad_norm": 0.8244159817695618, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 8761 + }, + { + "epoch": 0.08762, + "grad_norm": 0.8379217982292175, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 8762 + }, + { + "epoch": 0.08763, + "grad_norm": 0.8017216920852661, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 8763 + }, + { + "epoch": 0.08764, + "grad_norm": 0.6845731139183044, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 8764 + }, + { + "epoch": 0.08765, + "grad_norm": 0.6732616424560547, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 8765 + }, + { + "epoch": 0.08766, + "grad_norm": 0.6401374340057373, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 8766 + }, + { + "epoch": 0.08767, + "grad_norm": 0.5827595591545105, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 8767 + }, + { + "epoch": 0.08768, + "grad_norm": 0.5755667686462402, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 8768 + }, + { + "epoch": 0.08769, + "grad_norm": 0.5682708024978638, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 8769 + }, + { + "epoch": 0.0877, + "grad_norm": 0.5289202332496643, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 8770 + }, + { + "epoch": 0.08771, + "grad_norm": 0.5895897746086121, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 8771 + }, + { + "epoch": 0.08772, + "grad_norm": 0.8289152383804321, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 8772 + }, + { + "epoch": 0.08773, + "grad_norm": 1.1563746929168701, + "learning_rate": 0.003, + "loss": 4.058, + "step": 8773 + }, + { + "epoch": 0.08774, + "grad_norm": 0.7927180528640747, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 8774 + }, + { + "epoch": 0.08775, + "grad_norm": 0.6851915717124939, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 8775 + }, + { + "epoch": 0.08776, + "grad_norm": 0.8875740766525269, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 8776 + }, + { + "epoch": 0.08777, + "grad_norm": 1.134511947631836, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 8777 + }, + { + "epoch": 0.08778, + "grad_norm": 0.9227170348167419, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 8778 + }, + { + "epoch": 0.08779, + "grad_norm": 0.8097333312034607, + "learning_rate": 0.003, + "loss": 4.058, + "step": 8779 + }, + { + "epoch": 0.0878, + "grad_norm": 0.6999717354774475, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 8780 + }, + { + "epoch": 0.08781, + "grad_norm": 0.6974648237228394, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 8781 + }, + { + "epoch": 0.08782, + "grad_norm": 0.7920845746994019, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 8782 + }, + { + "epoch": 0.08783, + "grad_norm": 0.9354571104049683, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 8783 + }, + { + "epoch": 0.08784, + "grad_norm": 0.8569172620773315, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 8784 + }, + { + "epoch": 0.08785, + "grad_norm": 0.8295965194702148, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 8785 + }, + { + "epoch": 0.08786, + "grad_norm": 0.7757327556610107, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 8786 + }, + { + "epoch": 0.08787, + "grad_norm": 0.7948103547096252, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 8787 + }, + { + "epoch": 0.08788, + "grad_norm": 0.9465476870536804, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 8788 + }, + { + "epoch": 0.08789, + "grad_norm": 1.0520426034927368, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 8789 + }, + { + "epoch": 0.0879, + "grad_norm": 0.8785009384155273, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 8790 + }, + { + "epoch": 0.08791, + "grad_norm": 0.7729325890541077, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 8791 + }, + { + "epoch": 0.08792, + "grad_norm": 0.7058838605880737, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 8792 + }, + { + "epoch": 0.08793, + "grad_norm": 0.6069271564483643, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 8793 + }, + { + "epoch": 0.08794, + "grad_norm": 0.6468567252159119, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 8794 + }, + { + "epoch": 0.08795, + "grad_norm": 0.6967955231666565, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 8795 + }, + { + "epoch": 0.08796, + "grad_norm": 0.6414638757705688, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 8796 + }, + { + "epoch": 0.08797, + "grad_norm": 0.5752788782119751, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 8797 + }, + { + "epoch": 0.08798, + "grad_norm": 0.5287693738937378, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 8798 + }, + { + "epoch": 0.08799, + "grad_norm": 0.5393201112747192, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 8799 + }, + { + "epoch": 0.088, + "grad_norm": 0.6019169688224792, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 8800 + }, + { + "epoch": 0.08801, + "grad_norm": 0.7389690279960632, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 8801 + }, + { + "epoch": 0.08802, + "grad_norm": 1.005857229232788, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 8802 + }, + { + "epoch": 0.08803, + "grad_norm": 1.0771352052688599, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 8803 + }, + { + "epoch": 0.08804, + "grad_norm": 0.7955772876739502, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 8804 + }, + { + "epoch": 0.08805, + "grad_norm": 0.7112957239151001, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 8805 + }, + { + "epoch": 0.08806, + "grad_norm": 0.7480246424674988, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 8806 + }, + { + "epoch": 0.08807, + "grad_norm": 0.6301780343055725, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 8807 + }, + { + "epoch": 0.08808, + "grad_norm": 0.7244052886962891, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 8808 + }, + { + "epoch": 0.08809, + "grad_norm": 0.6920201182365417, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 8809 + }, + { + "epoch": 0.0881, + "grad_norm": 0.6484432816505432, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 8810 + }, + { + "epoch": 0.08811, + "grad_norm": 0.6578159332275391, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 8811 + }, + { + "epoch": 0.08812, + "grad_norm": 0.7231255173683167, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 8812 + }, + { + "epoch": 0.08813, + "grad_norm": 0.7631604671478271, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 8813 + }, + { + "epoch": 0.08814, + "grad_norm": 0.7701826095581055, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 8814 + }, + { + "epoch": 0.08815, + "grad_norm": 0.8050733804702759, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 8815 + }, + { + "epoch": 0.08816, + "grad_norm": 0.8618783950805664, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 8816 + }, + { + "epoch": 0.08817, + "grad_norm": 0.9637402296066284, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 8817 + }, + { + "epoch": 0.08818, + "grad_norm": 1.014755129814148, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 8818 + }, + { + "epoch": 0.08819, + "grad_norm": 0.7961846590042114, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 8819 + }, + { + "epoch": 0.0882, + "grad_norm": 0.7335184812545776, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 8820 + }, + { + "epoch": 0.08821, + "grad_norm": 0.6655312180519104, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 8821 + }, + { + "epoch": 0.08822, + "grad_norm": 0.5640453696250916, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 8822 + }, + { + "epoch": 0.08823, + "grad_norm": 0.5673083066940308, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 8823 + }, + { + "epoch": 0.08824, + "grad_norm": 0.6634215712547302, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 8824 + }, + { + "epoch": 0.08825, + "grad_norm": 0.8428478837013245, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 8825 + }, + { + "epoch": 0.08826, + "grad_norm": 0.9593620896339417, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 8826 + }, + { + "epoch": 0.08827, + "grad_norm": 0.9794454574584961, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 8827 + }, + { + "epoch": 0.08828, + "grad_norm": 0.8954617381095886, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 8828 + }, + { + "epoch": 0.08829, + "grad_norm": 0.8380072712898254, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 8829 + }, + { + "epoch": 0.0883, + "grad_norm": 0.7732906341552734, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 8830 + }, + { + "epoch": 0.08831, + "grad_norm": 0.7492073774337769, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 8831 + }, + { + "epoch": 0.08832, + "grad_norm": 0.7155649065971375, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 8832 + }, + { + "epoch": 0.08833, + "grad_norm": 0.6484860181808472, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 8833 + }, + { + "epoch": 0.08834, + "grad_norm": 0.6914095878601074, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 8834 + }, + { + "epoch": 0.08835, + "grad_norm": 0.8174039721488953, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 8835 + }, + { + "epoch": 0.08836, + "grad_norm": 0.7501481771469116, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 8836 + }, + { + "epoch": 0.08837, + "grad_norm": 0.6577468514442444, + "learning_rate": 0.003, + "loss": 4.017, + "step": 8837 + }, + { + "epoch": 0.08838, + "grad_norm": 0.6486355066299438, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 8838 + }, + { + "epoch": 0.08839, + "grad_norm": 0.7385262250900269, + "learning_rate": 0.003, + "loss": 4.024, + "step": 8839 + }, + { + "epoch": 0.0884, + "grad_norm": 0.9628206491470337, + "learning_rate": 0.003, + "loss": 4.023, + "step": 8840 + }, + { + "epoch": 0.08841, + "grad_norm": 1.0862576961517334, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 8841 + }, + { + "epoch": 0.08842, + "grad_norm": 0.941209077835083, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 8842 + }, + { + "epoch": 0.08843, + "grad_norm": 0.7774216532707214, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 8843 + }, + { + "epoch": 0.08844, + "grad_norm": 0.6513327956199646, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 8844 + }, + { + "epoch": 0.08845, + "grad_norm": 0.7469035983085632, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 8845 + }, + { + "epoch": 0.08846, + "grad_norm": 0.770413875579834, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 8846 + }, + { + "epoch": 0.08847, + "grad_norm": 0.8194745182991028, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 8847 + }, + { + "epoch": 0.08848, + "grad_norm": 0.9441425800323486, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 8848 + }, + { + "epoch": 0.08849, + "grad_norm": 0.8423975110054016, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 8849 + }, + { + "epoch": 0.0885, + "grad_norm": 0.928989589214325, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 8850 + }, + { + "epoch": 0.08851, + "grad_norm": 0.954427182674408, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 8851 + }, + { + "epoch": 0.08852, + "grad_norm": 0.9665095210075378, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 8852 + }, + { + "epoch": 0.08853, + "grad_norm": 0.8344659209251404, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 8853 + }, + { + "epoch": 0.08854, + "grad_norm": 0.8386094570159912, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 8854 + }, + { + "epoch": 0.08855, + "grad_norm": 0.862771213054657, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 8855 + }, + { + "epoch": 0.08856, + "grad_norm": 0.779102623462677, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 8856 + }, + { + "epoch": 0.08857, + "grad_norm": 0.8765580654144287, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 8857 + }, + { + "epoch": 0.08858, + "grad_norm": 0.917957067489624, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 8858 + }, + { + "epoch": 0.08859, + "grad_norm": 0.855716347694397, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 8859 + }, + { + "epoch": 0.0886, + "grad_norm": 0.866002082824707, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 8860 + }, + { + "epoch": 0.08861, + "grad_norm": 0.7670868635177612, + "learning_rate": 0.003, + "loss": 4.035, + "step": 8861 + }, + { + "epoch": 0.08862, + "grad_norm": 0.7301899790763855, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 8862 + }, + { + "epoch": 0.08863, + "grad_norm": 0.728283166885376, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 8863 + }, + { + "epoch": 0.08864, + "grad_norm": 0.7372556924819946, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 8864 + }, + { + "epoch": 0.08865, + "grad_norm": 0.7424993515014648, + "learning_rate": 0.003, + "loss": 4.053, + "step": 8865 + }, + { + "epoch": 0.08866, + "grad_norm": 0.7967926263809204, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 8866 + }, + { + "epoch": 0.08867, + "grad_norm": 0.7535709738731384, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 8867 + }, + { + "epoch": 0.08868, + "grad_norm": 0.8728439211845398, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 8868 + }, + { + "epoch": 0.08869, + "grad_norm": 0.8164510130882263, + "learning_rate": 0.003, + "loss": 4.084, + "step": 8869 + }, + { + "epoch": 0.0887, + "grad_norm": 0.750055730342865, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 8870 + }, + { + "epoch": 0.08871, + "grad_norm": 0.7114226222038269, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 8871 + }, + { + "epoch": 0.08872, + "grad_norm": 0.6471983790397644, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 8872 + }, + { + "epoch": 0.08873, + "grad_norm": 0.5536442995071411, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 8873 + }, + { + "epoch": 0.08874, + "grad_norm": 0.5485543608665466, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 8874 + }, + { + "epoch": 0.08875, + "grad_norm": 0.5261621475219727, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 8875 + }, + { + "epoch": 0.08876, + "grad_norm": 0.486948162317276, + "learning_rate": 0.003, + "loss": 4.004, + "step": 8876 + }, + { + "epoch": 0.08877, + "grad_norm": 0.5024458765983582, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 8877 + }, + { + "epoch": 0.08878, + "grad_norm": 0.5072246193885803, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 8878 + }, + { + "epoch": 0.08879, + "grad_norm": 0.4610580503940582, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 8879 + }, + { + "epoch": 0.0888, + "grad_norm": 0.4206774830818176, + "learning_rate": 0.003, + "loss": 4.011, + "step": 8880 + }, + { + "epoch": 0.08881, + "grad_norm": 0.4502694010734558, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 8881 + }, + { + "epoch": 0.08882, + "grad_norm": 0.42741695046424866, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 8882 + }, + { + "epoch": 0.08883, + "grad_norm": 0.40149998664855957, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 8883 + }, + { + "epoch": 0.08884, + "grad_norm": 0.4078429043292999, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 8884 + }, + { + "epoch": 0.08885, + "grad_norm": 0.5221074819564819, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 8885 + }, + { + "epoch": 0.08886, + "grad_norm": 0.617674708366394, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 8886 + }, + { + "epoch": 0.08887, + "grad_norm": 0.6875243782997131, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 8887 + }, + { + "epoch": 0.08888, + "grad_norm": 0.8016972541809082, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 8888 + }, + { + "epoch": 0.08889, + "grad_norm": 1.1390345096588135, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 8889 + }, + { + "epoch": 0.0889, + "grad_norm": 1.3257744312286377, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 8890 + }, + { + "epoch": 0.08891, + "grad_norm": 0.6148163080215454, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 8891 + }, + { + "epoch": 0.08892, + "grad_norm": 0.7419478297233582, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 8892 + }, + { + "epoch": 0.08893, + "grad_norm": 0.9414969086647034, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 8893 + }, + { + "epoch": 0.08894, + "grad_norm": 0.9926276803016663, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 8894 + }, + { + "epoch": 0.08895, + "grad_norm": 1.1406117677688599, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 8895 + }, + { + "epoch": 0.08896, + "grad_norm": 0.8471540212631226, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 8896 + }, + { + "epoch": 0.08897, + "grad_norm": 0.7552980780601501, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 8897 + }, + { + "epoch": 0.08898, + "grad_norm": 0.8421490788459778, + "learning_rate": 0.003, + "loss": 4.042, + "step": 8898 + }, + { + "epoch": 0.08899, + "grad_norm": 0.9038390517234802, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 8899 + }, + { + "epoch": 0.089, + "grad_norm": 0.9617167115211487, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 8900 + }, + { + "epoch": 0.08901, + "grad_norm": 0.9217119812965393, + "learning_rate": 0.003, + "loss": 4.054, + "step": 8901 + }, + { + "epoch": 0.08902, + "grad_norm": 0.8489481210708618, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 8902 + }, + { + "epoch": 0.08903, + "grad_norm": 0.9893561005592346, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 8903 + }, + { + "epoch": 0.08904, + "grad_norm": 1.184971570968628, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 8904 + }, + { + "epoch": 0.08905, + "grad_norm": 0.829010546207428, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 8905 + }, + { + "epoch": 0.08906, + "grad_norm": 0.8626430630683899, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 8906 + }, + { + "epoch": 0.08907, + "grad_norm": 0.961496889591217, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 8907 + }, + { + "epoch": 0.08908, + "grad_norm": 1.0651192665100098, + "learning_rate": 0.003, + "loss": 4.073, + "step": 8908 + }, + { + "epoch": 0.08909, + "grad_norm": 0.9574206471443176, + "learning_rate": 0.003, + "loss": 4.066, + "step": 8909 + }, + { + "epoch": 0.0891, + "grad_norm": 0.8476295471191406, + "learning_rate": 0.003, + "loss": 4.046, + "step": 8910 + }, + { + "epoch": 0.08911, + "grad_norm": 0.8135413527488708, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 8911 + }, + { + "epoch": 0.08912, + "grad_norm": 0.8378600478172302, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 8912 + }, + { + "epoch": 0.08913, + "grad_norm": 0.7921456694602966, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 8913 + }, + { + "epoch": 0.08914, + "grad_norm": 0.8137087821960449, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 8914 + }, + { + "epoch": 0.08915, + "grad_norm": 0.7893593311309814, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 8915 + }, + { + "epoch": 0.08916, + "grad_norm": 0.7994639873504639, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 8916 + }, + { + "epoch": 0.08917, + "grad_norm": 0.7975410223007202, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 8917 + }, + { + "epoch": 0.08918, + "grad_norm": 0.6501461863517761, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 8918 + }, + { + "epoch": 0.08919, + "grad_norm": 0.5666654706001282, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 8919 + }, + { + "epoch": 0.0892, + "grad_norm": 0.5936571359634399, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 8920 + }, + { + "epoch": 0.08921, + "grad_norm": 0.5901350378990173, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 8921 + }, + { + "epoch": 0.08922, + "grad_norm": 0.6528334617614746, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 8922 + }, + { + "epoch": 0.08923, + "grad_norm": 0.7454528212547302, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 8923 + }, + { + "epoch": 0.08924, + "grad_norm": 0.7825729250907898, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 8924 + }, + { + "epoch": 0.08925, + "grad_norm": 0.7133893370628357, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 8925 + }, + { + "epoch": 0.08926, + "grad_norm": 0.7308877110481262, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 8926 + }, + { + "epoch": 0.08927, + "grad_norm": 0.7993197441101074, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 8927 + }, + { + "epoch": 0.08928, + "grad_norm": 0.8782841563224792, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 8928 + }, + { + "epoch": 0.08929, + "grad_norm": 0.9822357296943665, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 8929 + }, + { + "epoch": 0.0893, + "grad_norm": 0.9535257816314697, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 8930 + }, + { + "epoch": 0.08931, + "grad_norm": 0.830475926399231, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 8931 + }, + { + "epoch": 0.08932, + "grad_norm": 0.7698143124580383, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 8932 + }, + { + "epoch": 0.08933, + "grad_norm": 0.8662832379341125, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 8933 + }, + { + "epoch": 0.08934, + "grad_norm": 0.8279432654380798, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 8934 + }, + { + "epoch": 0.08935, + "grad_norm": 0.7569650411605835, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 8935 + }, + { + "epoch": 0.08936, + "grad_norm": 0.6220132112503052, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 8936 + }, + { + "epoch": 0.08937, + "grad_norm": 0.618433952331543, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 8937 + }, + { + "epoch": 0.08938, + "grad_norm": 0.7037031650543213, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 8938 + }, + { + "epoch": 0.08939, + "grad_norm": 0.7682626843452454, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 8939 + }, + { + "epoch": 0.0894, + "grad_norm": 0.8006645441055298, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 8940 + }, + { + "epoch": 0.08941, + "grad_norm": 0.7939675450325012, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 8941 + }, + { + "epoch": 0.08942, + "grad_norm": 0.7516507506370544, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 8942 + }, + { + "epoch": 0.08943, + "grad_norm": 0.657136082649231, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 8943 + }, + { + "epoch": 0.08944, + "grad_norm": 0.6682058572769165, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 8944 + }, + { + "epoch": 0.08945, + "grad_norm": 0.6049004793167114, + "learning_rate": 0.003, + "loss": 4.001, + "step": 8945 + }, + { + "epoch": 0.08946, + "grad_norm": 0.6015246510505676, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 8946 + }, + { + "epoch": 0.08947, + "grad_norm": 0.727924108505249, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 8947 + }, + { + "epoch": 0.08948, + "grad_norm": 0.7806482315063477, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 8948 + }, + { + "epoch": 0.08949, + "grad_norm": 0.836334764957428, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 8949 + }, + { + "epoch": 0.0895, + "grad_norm": 0.8896807432174683, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 8950 + }, + { + "epoch": 0.08951, + "grad_norm": 0.9908595681190491, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 8951 + }, + { + "epoch": 0.08952, + "grad_norm": 0.992614209651947, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 8952 + }, + { + "epoch": 0.08953, + "grad_norm": 0.997110903263092, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 8953 + }, + { + "epoch": 0.08954, + "grad_norm": 0.9277966618537903, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 8954 + }, + { + "epoch": 0.08955, + "grad_norm": 0.8257642388343811, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 8955 + }, + { + "epoch": 0.08956, + "grad_norm": 0.9073073267936707, + "learning_rate": 0.003, + "loss": 4.052, + "step": 8956 + }, + { + "epoch": 0.08957, + "grad_norm": 0.8745629191398621, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 8957 + }, + { + "epoch": 0.08958, + "grad_norm": 0.86850905418396, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 8958 + }, + { + "epoch": 0.08959, + "grad_norm": 0.8888902068138123, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 8959 + }, + { + "epoch": 0.0896, + "grad_norm": 0.9012948870658875, + "learning_rate": 0.003, + "loss": 4.016, + "step": 8960 + }, + { + "epoch": 0.08961, + "grad_norm": 0.8690242767333984, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 8961 + }, + { + "epoch": 0.08962, + "grad_norm": 0.863722562789917, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 8962 + }, + { + "epoch": 0.08963, + "grad_norm": 0.7099302411079407, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 8963 + }, + { + "epoch": 0.08964, + "grad_norm": 0.8618674278259277, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 8964 + }, + { + "epoch": 0.08965, + "grad_norm": 1.0636932849884033, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 8965 + }, + { + "epoch": 0.08966, + "grad_norm": 0.7878800630569458, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 8966 + }, + { + "epoch": 0.08967, + "grad_norm": 0.8019261956214905, + "learning_rate": 0.003, + "loss": 4.071, + "step": 8967 + }, + { + "epoch": 0.08968, + "grad_norm": 0.7324358224868774, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 8968 + }, + { + "epoch": 0.08969, + "grad_norm": 0.5871388912200928, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 8969 + }, + { + "epoch": 0.0897, + "grad_norm": 0.6406157612800598, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 8970 + }, + { + "epoch": 0.08971, + "grad_norm": 0.6205124855041504, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 8971 + }, + { + "epoch": 0.08972, + "grad_norm": 0.630068302154541, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 8972 + }, + { + "epoch": 0.08973, + "grad_norm": 0.6608004570007324, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 8973 + }, + { + "epoch": 0.08974, + "grad_norm": 0.675197184085846, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 8974 + }, + { + "epoch": 0.08975, + "grad_norm": 0.7026122808456421, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 8975 + }, + { + "epoch": 0.08976, + "grad_norm": 0.7652435898780823, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 8976 + }, + { + "epoch": 0.08977, + "grad_norm": 0.8026885390281677, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 8977 + }, + { + "epoch": 0.08978, + "grad_norm": 0.7495819926261902, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 8978 + }, + { + "epoch": 0.08979, + "grad_norm": 0.7030617594718933, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 8979 + }, + { + "epoch": 0.0898, + "grad_norm": 0.6318227052688599, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 8980 + }, + { + "epoch": 0.08981, + "grad_norm": 0.6919098496437073, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 8981 + }, + { + "epoch": 0.08982, + "grad_norm": 0.7900656461715698, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 8982 + }, + { + "epoch": 0.08983, + "grad_norm": 0.691038191318512, + "learning_rate": 0.003, + "loss": 4.008, + "step": 8983 + }, + { + "epoch": 0.08984, + "grad_norm": 0.6485254168510437, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 8984 + }, + { + "epoch": 0.08985, + "grad_norm": 0.5956553816795349, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 8985 + }, + { + "epoch": 0.08986, + "grad_norm": 0.6086512207984924, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 8986 + }, + { + "epoch": 0.08987, + "grad_norm": 0.6024112701416016, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 8987 + }, + { + "epoch": 0.08988, + "grad_norm": 0.6147379279136658, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 8988 + }, + { + "epoch": 0.08989, + "grad_norm": 0.7100562453269958, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 8989 + }, + { + "epoch": 0.0899, + "grad_norm": 0.7910034656524658, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 8990 + }, + { + "epoch": 0.08991, + "grad_norm": 0.9195103645324707, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 8991 + }, + { + "epoch": 0.08992, + "grad_norm": 0.8702215552330017, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 8992 + }, + { + "epoch": 0.08993, + "grad_norm": 0.7406909465789795, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 8993 + }, + { + "epoch": 0.08994, + "grad_norm": 0.6627404689788818, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 8994 + }, + { + "epoch": 0.08995, + "grad_norm": 0.6060569882392883, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 8995 + }, + { + "epoch": 0.08996, + "grad_norm": 0.48137080669403076, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 8996 + }, + { + "epoch": 0.08997, + "grad_norm": 0.5556135177612305, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 8997 + }, + { + "epoch": 0.08998, + "grad_norm": 0.6087490320205688, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 8998 + }, + { + "epoch": 0.08999, + "grad_norm": 0.600788950920105, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 8999 + }, + { + "epoch": 0.09, + "grad_norm": 0.6633055210113525, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 9000 + }, + { + "epoch": 0.09001, + "grad_norm": 0.7464064359664917, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 9001 + }, + { + "epoch": 0.09002, + "grad_norm": 0.8089091181755066, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 9002 + }, + { + "epoch": 0.09003, + "grad_norm": 0.8683592081069946, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 9003 + }, + { + "epoch": 0.09004, + "grad_norm": 0.9728555679321289, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 9004 + }, + { + "epoch": 0.09005, + "grad_norm": 1.038753867149353, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 9005 + }, + { + "epoch": 0.09006, + "grad_norm": 0.8108296394348145, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 9006 + }, + { + "epoch": 0.09007, + "grad_norm": 0.7779677510261536, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 9007 + }, + { + "epoch": 0.09008, + "grad_norm": 0.7952173948287964, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 9008 + }, + { + "epoch": 0.09009, + "grad_norm": 0.963347315788269, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 9009 + }, + { + "epoch": 0.0901, + "grad_norm": 1.054080605506897, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 9010 + }, + { + "epoch": 0.09011, + "grad_norm": 0.9431594014167786, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 9011 + }, + { + "epoch": 0.09012, + "grad_norm": 0.9199943542480469, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 9012 + }, + { + "epoch": 0.09013, + "grad_norm": 0.9955227971076965, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 9013 + }, + { + "epoch": 0.09014, + "grad_norm": 1.195615291595459, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 9014 + }, + { + "epoch": 0.09015, + "grad_norm": 0.9534950256347656, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 9015 + }, + { + "epoch": 0.09016, + "grad_norm": 0.8363763689994812, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 9016 + }, + { + "epoch": 0.09017, + "grad_norm": 0.8595864772796631, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 9017 + }, + { + "epoch": 0.09018, + "grad_norm": 0.8017337322235107, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 9018 + }, + { + "epoch": 0.09019, + "grad_norm": 0.7216896414756775, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 9019 + }, + { + "epoch": 0.0902, + "grad_norm": 0.7337148189544678, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 9020 + }, + { + "epoch": 0.09021, + "grad_norm": 0.7746081352233887, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 9021 + }, + { + "epoch": 0.09022, + "grad_norm": 0.7762857675552368, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 9022 + }, + { + "epoch": 0.09023, + "grad_norm": 0.6827755570411682, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 9023 + }, + { + "epoch": 0.09024, + "grad_norm": 0.6656430959701538, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 9024 + }, + { + "epoch": 0.09025, + "grad_norm": 0.5954445004463196, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 9025 + }, + { + "epoch": 0.09026, + "grad_norm": 0.6387386918067932, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 9026 + }, + { + "epoch": 0.09027, + "grad_norm": 0.6285505294799805, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 9027 + }, + { + "epoch": 0.09028, + "grad_norm": 0.6999052166938782, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 9028 + }, + { + "epoch": 0.09029, + "grad_norm": 0.9033819437026978, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 9029 + }, + { + "epoch": 0.0903, + "grad_norm": 1.188402533531189, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 9030 + }, + { + "epoch": 0.09031, + "grad_norm": 0.7485238313674927, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 9031 + }, + { + "epoch": 0.09032, + "grad_norm": 0.5360812544822693, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 9032 + }, + { + "epoch": 0.09033, + "grad_norm": 0.6563149690628052, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 9033 + }, + { + "epoch": 0.09034, + "grad_norm": 0.7612379193305969, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 9034 + }, + { + "epoch": 0.09035, + "grad_norm": 0.7884421348571777, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 9035 + }, + { + "epoch": 0.09036, + "grad_norm": 0.7918843626976013, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 9036 + }, + { + "epoch": 0.09037, + "grad_norm": 0.8294838070869446, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 9037 + }, + { + "epoch": 0.09038, + "grad_norm": 0.712574303150177, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 9038 + }, + { + "epoch": 0.09039, + "grad_norm": 0.6491422653198242, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 9039 + }, + { + "epoch": 0.0904, + "grad_norm": 0.6552371382713318, + "learning_rate": 0.003, + "loss": 4.012, + "step": 9040 + }, + { + "epoch": 0.09041, + "grad_norm": 0.5920805335044861, + "learning_rate": 0.003, + "loss": 4.008, + "step": 9041 + }, + { + "epoch": 0.09042, + "grad_norm": 0.5702484846115112, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 9042 + }, + { + "epoch": 0.09043, + "grad_norm": 0.5535892844200134, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 9043 + }, + { + "epoch": 0.09044, + "grad_norm": 0.610673189163208, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 9044 + }, + { + "epoch": 0.09045, + "grad_norm": 0.5978013873100281, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 9045 + }, + { + "epoch": 0.09046, + "grad_norm": 0.6004743576049805, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 9046 + }, + { + "epoch": 0.09047, + "grad_norm": 0.7006171345710754, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 9047 + }, + { + "epoch": 0.09048, + "grad_norm": 0.7652348279953003, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 9048 + }, + { + "epoch": 0.09049, + "grad_norm": 0.9703059792518616, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 9049 + }, + { + "epoch": 0.0905, + "grad_norm": 1.1826372146606445, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 9050 + }, + { + "epoch": 0.09051, + "grad_norm": 0.655387818813324, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 9051 + }, + { + "epoch": 0.09052, + "grad_norm": 0.5878477692604065, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 9052 + }, + { + "epoch": 0.09053, + "grad_norm": 0.5366464257240295, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 9053 + }, + { + "epoch": 0.09054, + "grad_norm": 0.5335559844970703, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 9054 + }, + { + "epoch": 0.09055, + "grad_norm": 0.673572301864624, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 9055 + }, + { + "epoch": 0.09056, + "grad_norm": 0.7973268032073975, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 9056 + }, + { + "epoch": 0.09057, + "grad_norm": 0.8581774234771729, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 9057 + }, + { + "epoch": 0.09058, + "grad_norm": 0.8413546681404114, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 9058 + }, + { + "epoch": 0.09059, + "grad_norm": 0.8598448634147644, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 9059 + }, + { + "epoch": 0.0906, + "grad_norm": 0.8993529677391052, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 9060 + }, + { + "epoch": 0.09061, + "grad_norm": 1.0165449380874634, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 9061 + }, + { + "epoch": 0.09062, + "grad_norm": 0.9853262305259705, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 9062 + }, + { + "epoch": 0.09063, + "grad_norm": 0.8807757496833801, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 9063 + }, + { + "epoch": 0.09064, + "grad_norm": 0.8548852801322937, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 9064 + }, + { + "epoch": 0.09065, + "grad_norm": 0.9431764483451843, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 9065 + }, + { + "epoch": 0.09066, + "grad_norm": 1.1532913446426392, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 9066 + }, + { + "epoch": 0.09067, + "grad_norm": 1.311736822128296, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 9067 + }, + { + "epoch": 0.09068, + "grad_norm": 0.796724796295166, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 9068 + }, + { + "epoch": 0.09069, + "grad_norm": 0.7498504519462585, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 9069 + }, + { + "epoch": 0.0907, + "grad_norm": 0.8895300030708313, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 9070 + }, + { + "epoch": 0.09071, + "grad_norm": 0.9477413296699524, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 9071 + }, + { + "epoch": 0.09072, + "grad_norm": 0.7607925534248352, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 9072 + }, + { + "epoch": 0.09073, + "grad_norm": 0.7120130658149719, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 9073 + }, + { + "epoch": 0.09074, + "grad_norm": 0.7336847186088562, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 9074 + }, + { + "epoch": 0.09075, + "grad_norm": 0.6090520024299622, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 9075 + }, + { + "epoch": 0.09076, + "grad_norm": 0.6007069945335388, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 9076 + }, + { + "epoch": 0.09077, + "grad_norm": 0.6927357316017151, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 9077 + }, + { + "epoch": 0.09078, + "grad_norm": 0.863045334815979, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 9078 + }, + { + "epoch": 0.09079, + "grad_norm": 1.1068956851959229, + "learning_rate": 0.003, + "loss": 4.055, + "step": 9079 + }, + { + "epoch": 0.0908, + "grad_norm": 0.8308747410774231, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 9080 + }, + { + "epoch": 0.09081, + "grad_norm": 0.6511629819869995, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 9081 + }, + { + "epoch": 0.09082, + "grad_norm": 0.582959771156311, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 9082 + }, + { + "epoch": 0.09083, + "grad_norm": 0.6266202330589294, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 9083 + }, + { + "epoch": 0.09084, + "grad_norm": 0.6483187079429626, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 9084 + }, + { + "epoch": 0.09085, + "grad_norm": 0.6330807209014893, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 9085 + }, + { + "epoch": 0.09086, + "grad_norm": 0.5214323401451111, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 9086 + }, + { + "epoch": 0.09087, + "grad_norm": 0.5562803745269775, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 9087 + }, + { + "epoch": 0.09088, + "grad_norm": 0.606450080871582, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 9088 + }, + { + "epoch": 0.09089, + "grad_norm": 0.5667656064033508, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 9089 + }, + { + "epoch": 0.0909, + "grad_norm": 0.5123067498207092, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 9090 + }, + { + "epoch": 0.09091, + "grad_norm": 0.4887959957122803, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 9091 + }, + { + "epoch": 0.09092, + "grad_norm": 0.4976811110973358, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 9092 + }, + { + "epoch": 0.09093, + "grad_norm": 0.4877670109272003, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 9093 + }, + { + "epoch": 0.09094, + "grad_norm": 0.48510006070137024, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 9094 + }, + { + "epoch": 0.09095, + "grad_norm": 0.5514529347419739, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 9095 + }, + { + "epoch": 0.09096, + "grad_norm": 0.6349421739578247, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 9096 + }, + { + "epoch": 0.09097, + "grad_norm": 0.9241090416908264, + "learning_rate": 0.003, + "loss": 4.028, + "step": 9097 + }, + { + "epoch": 0.09098, + "grad_norm": 1.3963031768798828, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 9098 + }, + { + "epoch": 0.09099, + "grad_norm": 0.6102014183998108, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 9099 + }, + { + "epoch": 0.091, + "grad_norm": 0.6364977359771729, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 9100 + }, + { + "epoch": 0.09101, + "grad_norm": 0.8607155680656433, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 9101 + }, + { + "epoch": 0.09102, + "grad_norm": 0.8906451463699341, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 9102 + }, + { + "epoch": 0.09103, + "grad_norm": 0.8266270756721497, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 9103 + }, + { + "epoch": 0.09104, + "grad_norm": 0.8795189261436462, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 9104 + }, + { + "epoch": 0.09105, + "grad_norm": 0.8761870265007019, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 9105 + }, + { + "epoch": 0.09106, + "grad_norm": 0.8591187596321106, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 9106 + }, + { + "epoch": 0.09107, + "grad_norm": 1.034685730934143, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 9107 + }, + { + "epoch": 0.09108, + "grad_norm": 1.0666166543960571, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 9108 + }, + { + "epoch": 0.09109, + "grad_norm": 0.9874948859214783, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 9109 + }, + { + "epoch": 0.0911, + "grad_norm": 0.9942079186439514, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 9110 + }, + { + "epoch": 0.09111, + "grad_norm": 1.0025235414505005, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 9111 + }, + { + "epoch": 0.09112, + "grad_norm": 1.0080291032791138, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 9112 + }, + { + "epoch": 0.09113, + "grad_norm": 0.8179364800453186, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 9113 + }, + { + "epoch": 0.09114, + "grad_norm": 0.6516106128692627, + "learning_rate": 0.003, + "loss": 4.034, + "step": 9114 + }, + { + "epoch": 0.09115, + "grad_norm": 0.6664100885391235, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 9115 + }, + { + "epoch": 0.09116, + "grad_norm": 0.7324456572532654, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 9116 + }, + { + "epoch": 0.09117, + "grad_norm": 0.7465541362762451, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 9117 + }, + { + "epoch": 0.09118, + "grad_norm": 0.7696415185928345, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 9118 + }, + { + "epoch": 0.09119, + "grad_norm": 0.6559749245643616, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 9119 + }, + { + "epoch": 0.0912, + "grad_norm": 0.7420315146446228, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 9120 + }, + { + "epoch": 0.09121, + "grad_norm": 0.8487631678581238, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 9121 + }, + { + "epoch": 0.09122, + "grad_norm": 1.019974946975708, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 9122 + }, + { + "epoch": 0.09123, + "grad_norm": 1.0010663270950317, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 9123 + }, + { + "epoch": 0.09124, + "grad_norm": 1.219331979751587, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 9124 + }, + { + "epoch": 0.09125, + "grad_norm": 1.1329954862594604, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 9125 + }, + { + "epoch": 0.09126, + "grad_norm": 0.9389865398406982, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 9126 + }, + { + "epoch": 0.09127, + "grad_norm": 0.782051682472229, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 9127 + }, + { + "epoch": 0.09128, + "grad_norm": 0.8009851574897766, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 9128 + }, + { + "epoch": 0.09129, + "grad_norm": 0.8631070256233215, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 9129 + }, + { + "epoch": 0.0913, + "grad_norm": 0.9466512799263, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 9130 + }, + { + "epoch": 0.09131, + "grad_norm": 0.840096652507782, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 9131 + }, + { + "epoch": 0.09132, + "grad_norm": 0.751104474067688, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 9132 + }, + { + "epoch": 0.09133, + "grad_norm": 0.7003132104873657, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 9133 + }, + { + "epoch": 0.09134, + "grad_norm": 0.6505602598190308, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 9134 + }, + { + "epoch": 0.09135, + "grad_norm": 0.6527884602546692, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 9135 + }, + { + "epoch": 0.09136, + "grad_norm": 0.7264319062232971, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 9136 + }, + { + "epoch": 0.09137, + "grad_norm": 0.8720904588699341, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 9137 + }, + { + "epoch": 0.09138, + "grad_norm": 0.9983429312705994, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 9138 + }, + { + "epoch": 0.09139, + "grad_norm": 0.9738618731498718, + "learning_rate": 0.003, + "loss": 4.041, + "step": 9139 + }, + { + "epoch": 0.0914, + "grad_norm": 0.7946332097053528, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 9140 + }, + { + "epoch": 0.09141, + "grad_norm": 0.5857697129249573, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 9141 + }, + { + "epoch": 0.09142, + "grad_norm": 0.5243735909461975, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 9142 + }, + { + "epoch": 0.09143, + "grad_norm": 0.501167893409729, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 9143 + }, + { + "epoch": 0.09144, + "grad_norm": 0.5033556818962097, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 9144 + }, + { + "epoch": 0.09145, + "grad_norm": 0.541192889213562, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 9145 + }, + { + "epoch": 0.09146, + "grad_norm": 0.7105612754821777, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 9146 + }, + { + "epoch": 0.09147, + "grad_norm": 0.8535759449005127, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 9147 + }, + { + "epoch": 0.09148, + "grad_norm": 0.7989993691444397, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 9148 + }, + { + "epoch": 0.09149, + "grad_norm": 0.6929615139961243, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 9149 + }, + { + "epoch": 0.0915, + "grad_norm": 0.58013516664505, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 9150 + }, + { + "epoch": 0.09151, + "grad_norm": 0.4813550114631653, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 9151 + }, + { + "epoch": 0.09152, + "grad_norm": 0.5421327948570251, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 9152 + }, + { + "epoch": 0.09153, + "grad_norm": 0.5891533493995667, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 9153 + }, + { + "epoch": 0.09154, + "grad_norm": 0.6357941031455994, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 9154 + }, + { + "epoch": 0.09155, + "grad_norm": 0.659079372882843, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 9155 + }, + { + "epoch": 0.09156, + "grad_norm": 0.657294511795044, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 9156 + }, + { + "epoch": 0.09157, + "grad_norm": 0.6846706867218018, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 9157 + }, + { + "epoch": 0.09158, + "grad_norm": 0.6189616918563843, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 9158 + }, + { + "epoch": 0.09159, + "grad_norm": 0.6369915008544922, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 9159 + }, + { + "epoch": 0.0916, + "grad_norm": 0.7117534875869751, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 9160 + }, + { + "epoch": 0.09161, + "grad_norm": 0.7279969453811646, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 9161 + }, + { + "epoch": 0.09162, + "grad_norm": 0.7147172689437866, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 9162 + }, + { + "epoch": 0.09163, + "grad_norm": 0.8350904583930969, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 9163 + }, + { + "epoch": 0.09164, + "grad_norm": 1.0859333276748657, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 9164 + }, + { + "epoch": 0.09165, + "grad_norm": 1.132325291633606, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 9165 + }, + { + "epoch": 0.09166, + "grad_norm": 0.6931542754173279, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 9166 + }, + { + "epoch": 0.09167, + "grad_norm": 0.5628350377082825, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 9167 + }, + { + "epoch": 0.09168, + "grad_norm": 0.691896378993988, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 9168 + }, + { + "epoch": 0.09169, + "grad_norm": 0.7813774347305298, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 9169 + }, + { + "epoch": 0.0917, + "grad_norm": 0.8740519285202026, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 9170 + }, + { + "epoch": 0.09171, + "grad_norm": 0.8779575824737549, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 9171 + }, + { + "epoch": 0.09172, + "grad_norm": 0.7493038773536682, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 9172 + }, + { + "epoch": 0.09173, + "grad_norm": 0.6481500864028931, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 9173 + }, + { + "epoch": 0.09174, + "grad_norm": 0.6999644637107849, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 9174 + }, + { + "epoch": 0.09175, + "grad_norm": 0.7405575513839722, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 9175 + }, + { + "epoch": 0.09176, + "grad_norm": 0.7775292992591858, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 9176 + }, + { + "epoch": 0.09177, + "grad_norm": 0.7605270743370056, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 9177 + }, + { + "epoch": 0.09178, + "grad_norm": 0.6888309121131897, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 9178 + }, + { + "epoch": 0.09179, + "grad_norm": 0.7338680624961853, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 9179 + }, + { + "epoch": 0.0918, + "grad_norm": 0.9125281572341919, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 9180 + }, + { + "epoch": 0.09181, + "grad_norm": 1.0294517278671265, + "learning_rate": 0.003, + "loss": 4.032, + "step": 9181 + }, + { + "epoch": 0.09182, + "grad_norm": 0.9629702568054199, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 9182 + }, + { + "epoch": 0.09183, + "grad_norm": 0.9055722951889038, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 9183 + }, + { + "epoch": 0.09184, + "grad_norm": 0.9649536609649658, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 9184 + }, + { + "epoch": 0.09185, + "grad_norm": 0.9440880417823792, + "learning_rate": 0.003, + "loss": 4.062, + "step": 9185 + }, + { + "epoch": 0.09186, + "grad_norm": 0.8451281189918518, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 9186 + }, + { + "epoch": 0.09187, + "grad_norm": 0.8534144759178162, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 9187 + }, + { + "epoch": 0.09188, + "grad_norm": 0.8250672817230225, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 9188 + }, + { + "epoch": 0.09189, + "grad_norm": 0.9062046408653259, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 9189 + }, + { + "epoch": 0.0919, + "grad_norm": 0.9313827157020569, + "learning_rate": 0.003, + "loss": 4.067, + "step": 9190 + }, + { + "epoch": 0.09191, + "grad_norm": 0.7684466242790222, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 9191 + }, + { + "epoch": 0.09192, + "grad_norm": 0.6470143795013428, + "learning_rate": 0.003, + "loss": 4.023, + "step": 9192 + }, + { + "epoch": 0.09193, + "grad_norm": 0.5500982403755188, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 9193 + }, + { + "epoch": 0.09194, + "grad_norm": 0.5531850457191467, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 9194 + }, + { + "epoch": 0.09195, + "grad_norm": 0.5240871906280518, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 9195 + }, + { + "epoch": 0.09196, + "grad_norm": 0.4800591468811035, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 9196 + }, + { + "epoch": 0.09197, + "grad_norm": 0.598760724067688, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 9197 + }, + { + "epoch": 0.09198, + "grad_norm": 0.8965339660644531, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 9198 + }, + { + "epoch": 0.09199, + "grad_norm": 1.3671212196350098, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 9199 + }, + { + "epoch": 0.092, + "grad_norm": 0.6588968634605408, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 9200 + }, + { + "epoch": 0.09201, + "grad_norm": 0.6256622672080994, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 9201 + }, + { + "epoch": 0.09202, + "grad_norm": 0.6034049987792969, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 9202 + }, + { + "epoch": 0.09203, + "grad_norm": 0.6372929811477661, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 9203 + }, + { + "epoch": 0.09204, + "grad_norm": 0.530689001083374, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 9204 + }, + { + "epoch": 0.09205, + "grad_norm": 0.5760714411735535, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 9205 + }, + { + "epoch": 0.09206, + "grad_norm": 0.6162959933280945, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 9206 + }, + { + "epoch": 0.09207, + "grad_norm": 0.8327687382698059, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 9207 + }, + { + "epoch": 0.09208, + "grad_norm": 1.0663541555404663, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 9208 + }, + { + "epoch": 0.09209, + "grad_norm": 0.99628084897995, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 9209 + }, + { + "epoch": 0.0921, + "grad_norm": 0.7654755711555481, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 9210 + }, + { + "epoch": 0.09211, + "grad_norm": 0.5291755795478821, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 9211 + }, + { + "epoch": 0.09212, + "grad_norm": 0.545681893825531, + "learning_rate": 0.003, + "loss": 4.037, + "step": 9212 + }, + { + "epoch": 0.09213, + "grad_norm": 0.7274799346923828, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 9213 + }, + { + "epoch": 0.09214, + "grad_norm": 0.8287873268127441, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 9214 + }, + { + "epoch": 0.09215, + "grad_norm": 0.8357604742050171, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 9215 + }, + { + "epoch": 0.09216, + "grad_norm": 0.9297974705696106, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 9216 + }, + { + "epoch": 0.09217, + "grad_norm": 1.03271484375, + "learning_rate": 0.003, + "loss": 4.004, + "step": 9217 + }, + { + "epoch": 0.09218, + "grad_norm": 0.8471750617027283, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 9218 + }, + { + "epoch": 0.09219, + "grad_norm": 0.7956122159957886, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 9219 + }, + { + "epoch": 0.0922, + "grad_norm": 0.7104371786117554, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 9220 + }, + { + "epoch": 0.09221, + "grad_norm": 0.8256221413612366, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 9221 + }, + { + "epoch": 0.09222, + "grad_norm": 0.914454460144043, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 9222 + }, + { + "epoch": 0.09223, + "grad_norm": 0.8593582510948181, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 9223 + }, + { + "epoch": 0.09224, + "grad_norm": 0.8074911236763, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 9224 + }, + { + "epoch": 0.09225, + "grad_norm": 0.7950632572174072, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 9225 + }, + { + "epoch": 0.09226, + "grad_norm": 0.7571146488189697, + "learning_rate": 0.003, + "loss": 4.021, + "step": 9226 + }, + { + "epoch": 0.09227, + "grad_norm": 0.7939788699150085, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 9227 + }, + { + "epoch": 0.09228, + "grad_norm": 0.999841034412384, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 9228 + }, + { + "epoch": 0.09229, + "grad_norm": 1.0757583379745483, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 9229 + }, + { + "epoch": 0.0923, + "grad_norm": 0.8179429769515991, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 9230 + }, + { + "epoch": 0.09231, + "grad_norm": 0.7525832056999207, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 9231 + }, + { + "epoch": 0.09232, + "grad_norm": 0.857976496219635, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 9232 + }, + { + "epoch": 0.09233, + "grad_norm": 1.0185387134552002, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 9233 + }, + { + "epoch": 0.09234, + "grad_norm": 0.9663023948669434, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 9234 + }, + { + "epoch": 0.09235, + "grad_norm": 1.0109349489212036, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 9235 + }, + { + "epoch": 0.09236, + "grad_norm": 1.1014132499694824, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 9236 + }, + { + "epoch": 0.09237, + "grad_norm": 0.919548511505127, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 9237 + }, + { + "epoch": 0.09238, + "grad_norm": 0.8530013561248779, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 9238 + }, + { + "epoch": 0.09239, + "grad_norm": 0.939599871635437, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 9239 + }, + { + "epoch": 0.0924, + "grad_norm": 1.015742540359497, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 9240 + }, + { + "epoch": 0.09241, + "grad_norm": 0.8002774119377136, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 9241 + }, + { + "epoch": 0.09242, + "grad_norm": 0.6762545704841614, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 9242 + }, + { + "epoch": 0.09243, + "grad_norm": 0.6825734376907349, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 9243 + }, + { + "epoch": 0.09244, + "grad_norm": 0.7166531682014465, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 9244 + }, + { + "epoch": 0.09245, + "grad_norm": 0.6903985738754272, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 9245 + }, + { + "epoch": 0.09246, + "grad_norm": 0.6596519351005554, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 9246 + }, + { + "epoch": 0.09247, + "grad_norm": 0.6328878998756409, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 9247 + }, + { + "epoch": 0.09248, + "grad_norm": 0.5870897769927979, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 9248 + }, + { + "epoch": 0.09249, + "grad_norm": 0.5670067071914673, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 9249 + }, + { + "epoch": 0.0925, + "grad_norm": 0.6085765957832336, + "learning_rate": 0.003, + "loss": 4.049, + "step": 9250 + }, + { + "epoch": 0.09251, + "grad_norm": 0.5557095408439636, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 9251 + }, + { + "epoch": 0.09252, + "grad_norm": 0.48302340507507324, + "learning_rate": 0.003, + "loss": 4.054, + "step": 9252 + }, + { + "epoch": 0.09253, + "grad_norm": 0.5740175843238831, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 9253 + }, + { + "epoch": 0.09254, + "grad_norm": 0.7579610347747803, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 9254 + }, + { + "epoch": 0.09255, + "grad_norm": 1.0486177206039429, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 9255 + }, + { + "epoch": 0.09256, + "grad_norm": 1.0068448781967163, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 9256 + }, + { + "epoch": 0.09257, + "grad_norm": 0.7558557987213135, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 9257 + }, + { + "epoch": 0.09258, + "grad_norm": 0.66719651222229, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 9258 + }, + { + "epoch": 0.09259, + "grad_norm": 0.7420986294746399, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 9259 + }, + { + "epoch": 0.0926, + "grad_norm": 0.8558784127235413, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 9260 + }, + { + "epoch": 0.09261, + "grad_norm": 0.9271590709686279, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 9261 + }, + { + "epoch": 0.09262, + "grad_norm": 0.8783436417579651, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 9262 + }, + { + "epoch": 0.09263, + "grad_norm": 0.7863196730613708, + "learning_rate": 0.003, + "loss": 4.041, + "step": 9263 + }, + { + "epoch": 0.09264, + "grad_norm": 0.6934552192687988, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 9264 + }, + { + "epoch": 0.09265, + "grad_norm": 0.6641402244567871, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 9265 + }, + { + "epoch": 0.09266, + "grad_norm": 0.6042680740356445, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 9266 + }, + { + "epoch": 0.09267, + "grad_norm": 0.6411666870117188, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 9267 + }, + { + "epoch": 0.09268, + "grad_norm": 0.5758857131004333, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 9268 + }, + { + "epoch": 0.09269, + "grad_norm": 0.5887449383735657, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 9269 + }, + { + "epoch": 0.0927, + "grad_norm": 0.6199323534965515, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 9270 + }, + { + "epoch": 0.09271, + "grad_norm": 0.6328174471855164, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 9271 + }, + { + "epoch": 0.09272, + "grad_norm": 0.6937214136123657, + "learning_rate": 0.003, + "loss": 4.024, + "step": 9272 + }, + { + "epoch": 0.09273, + "grad_norm": 0.7590581178665161, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 9273 + }, + { + "epoch": 0.09274, + "grad_norm": 0.775363564491272, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 9274 + }, + { + "epoch": 0.09275, + "grad_norm": 0.765939474105835, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 9275 + }, + { + "epoch": 0.09276, + "grad_norm": 0.6670629978179932, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 9276 + }, + { + "epoch": 0.09277, + "grad_norm": 0.555806577205658, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 9277 + }, + { + "epoch": 0.09278, + "grad_norm": 0.4805139899253845, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 9278 + }, + { + "epoch": 0.09279, + "grad_norm": 0.4400136172771454, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 9279 + }, + { + "epoch": 0.0928, + "grad_norm": 0.4695459008216858, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 9280 + }, + { + "epoch": 0.09281, + "grad_norm": 0.4340028762817383, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 9281 + }, + { + "epoch": 0.09282, + "grad_norm": 0.4828382730484009, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 9282 + }, + { + "epoch": 0.09283, + "grad_norm": 0.669910728931427, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 9283 + }, + { + "epoch": 0.09284, + "grad_norm": 0.787203848361969, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 9284 + }, + { + "epoch": 0.09285, + "grad_norm": 0.9136567115783691, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 9285 + }, + { + "epoch": 0.09286, + "grad_norm": 1.051871657371521, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 9286 + }, + { + "epoch": 0.09287, + "grad_norm": 1.0061372518539429, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 9287 + }, + { + "epoch": 0.09288, + "grad_norm": 0.91563880443573, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 9288 + }, + { + "epoch": 0.09289, + "grad_norm": 0.8254592418670654, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 9289 + }, + { + "epoch": 0.0929, + "grad_norm": 0.9599001407623291, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 9290 + }, + { + "epoch": 0.09291, + "grad_norm": 1.0785131454467773, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 9291 + }, + { + "epoch": 0.09292, + "grad_norm": 0.8951935172080994, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 9292 + }, + { + "epoch": 0.09293, + "grad_norm": 0.8868071436882019, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 9293 + }, + { + "epoch": 0.09294, + "grad_norm": 0.9149964451789856, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 9294 + }, + { + "epoch": 0.09295, + "grad_norm": 0.9641788601875305, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 9295 + }, + { + "epoch": 0.09296, + "grad_norm": 1.0511088371276855, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 9296 + }, + { + "epoch": 0.09297, + "grad_norm": 1.1237189769744873, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 9297 + }, + { + "epoch": 0.09298, + "grad_norm": 0.8537935614585876, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 9298 + }, + { + "epoch": 0.09299, + "grad_norm": 0.8743082880973816, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 9299 + }, + { + "epoch": 0.093, + "grad_norm": 0.8369987607002258, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 9300 + }, + { + "epoch": 0.09301, + "grad_norm": 0.7713850736618042, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 9301 + }, + { + "epoch": 0.09302, + "grad_norm": 0.7446997761726379, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 9302 + }, + { + "epoch": 0.09303, + "grad_norm": 0.7269630432128906, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 9303 + }, + { + "epoch": 0.09304, + "grad_norm": 0.6271345615386963, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 9304 + }, + { + "epoch": 0.09305, + "grad_norm": 0.6511696577072144, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 9305 + }, + { + "epoch": 0.09306, + "grad_norm": 0.7046614289283752, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 9306 + }, + { + "epoch": 0.09307, + "grad_norm": 0.6628211140632629, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 9307 + }, + { + "epoch": 0.09308, + "grad_norm": 0.6787553429603577, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 9308 + }, + { + "epoch": 0.09309, + "grad_norm": 0.7450267672538757, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 9309 + }, + { + "epoch": 0.0931, + "grad_norm": 0.8884741067886353, + "learning_rate": 0.003, + "loss": 4.057, + "step": 9310 + }, + { + "epoch": 0.09311, + "grad_norm": 1.3012408018112183, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 9311 + }, + { + "epoch": 0.09312, + "grad_norm": 0.8154418468475342, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 9312 + }, + { + "epoch": 0.09313, + "grad_norm": 0.7738723158836365, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 9313 + }, + { + "epoch": 0.09314, + "grad_norm": 0.7311467528343201, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 9314 + }, + { + "epoch": 0.09315, + "grad_norm": 0.6791749596595764, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 9315 + }, + { + "epoch": 0.09316, + "grad_norm": 0.7334395051002502, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 9316 + }, + { + "epoch": 0.09317, + "grad_norm": 0.7056670784950256, + "learning_rate": 0.003, + "loss": 4.058, + "step": 9317 + }, + { + "epoch": 0.09318, + "grad_norm": 0.7427524328231812, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 9318 + }, + { + "epoch": 0.09319, + "grad_norm": 0.773429274559021, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 9319 + }, + { + "epoch": 0.0932, + "grad_norm": 0.8293418884277344, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 9320 + }, + { + "epoch": 0.09321, + "grad_norm": 0.7913047671318054, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 9321 + }, + { + "epoch": 0.09322, + "grad_norm": 0.8516207337379456, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 9322 + }, + { + "epoch": 0.09323, + "grad_norm": 0.9764220714569092, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 9323 + }, + { + "epoch": 0.09324, + "grad_norm": 1.0407365560531616, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 9324 + }, + { + "epoch": 0.09325, + "grad_norm": 0.7859419584274292, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 9325 + }, + { + "epoch": 0.09326, + "grad_norm": 0.7470349073410034, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 9326 + }, + { + "epoch": 0.09327, + "grad_norm": 1.0527385473251343, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 9327 + }, + { + "epoch": 0.09328, + "grad_norm": 1.1210170984268188, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 9328 + }, + { + "epoch": 0.09329, + "grad_norm": 0.8028057217597961, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 9329 + }, + { + "epoch": 0.0933, + "grad_norm": 0.5863675475120544, + "learning_rate": 0.003, + "loss": 4.052, + "step": 9330 + }, + { + "epoch": 0.09331, + "grad_norm": 0.6289627552032471, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 9331 + }, + { + "epoch": 0.09332, + "grad_norm": 0.6613301634788513, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 9332 + }, + { + "epoch": 0.09333, + "grad_norm": 0.621925413608551, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 9333 + }, + { + "epoch": 0.09334, + "grad_norm": 0.6211190819740295, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 9334 + }, + { + "epoch": 0.09335, + "grad_norm": 0.6644201874732971, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 9335 + }, + { + "epoch": 0.09336, + "grad_norm": 0.6829946637153625, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 9336 + }, + { + "epoch": 0.09337, + "grad_norm": 0.5816702246665955, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 9337 + }, + { + "epoch": 0.09338, + "grad_norm": 0.5819098353385925, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 9338 + }, + { + "epoch": 0.09339, + "grad_norm": 0.5959669351577759, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 9339 + }, + { + "epoch": 0.0934, + "grad_norm": 0.5701149106025696, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 9340 + }, + { + "epoch": 0.09341, + "grad_norm": 0.43852511048316956, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 9341 + }, + { + "epoch": 0.09342, + "grad_norm": 0.48799461126327515, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 9342 + }, + { + "epoch": 0.09343, + "grad_norm": 0.5336135029792786, + "learning_rate": 0.003, + "loss": 4.022, + "step": 9343 + }, + { + "epoch": 0.09344, + "grad_norm": 0.5141116380691528, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 9344 + }, + { + "epoch": 0.09345, + "grad_norm": 0.47849857807159424, + "learning_rate": 0.003, + "loss": 4.041, + "step": 9345 + }, + { + "epoch": 0.09346, + "grad_norm": 0.49492859840393066, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 9346 + }, + { + "epoch": 0.09347, + "grad_norm": 0.6012299060821533, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 9347 + }, + { + "epoch": 0.09348, + "grad_norm": 0.8121467232704163, + "learning_rate": 0.003, + "loss": 4.04, + "step": 9348 + }, + { + "epoch": 0.09349, + "grad_norm": 1.0918225049972534, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 9349 + }, + { + "epoch": 0.0935, + "grad_norm": 1.1535876989364624, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 9350 + }, + { + "epoch": 0.09351, + "grad_norm": 0.6782242655754089, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 9351 + }, + { + "epoch": 0.09352, + "grad_norm": 0.6605706810951233, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 9352 + }, + { + "epoch": 0.09353, + "grad_norm": 0.8234308362007141, + "learning_rate": 0.003, + "loss": 4.006, + "step": 9353 + }, + { + "epoch": 0.09354, + "grad_norm": 0.7437732815742493, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 9354 + }, + { + "epoch": 0.09355, + "grad_norm": 0.7545703649520874, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 9355 + }, + { + "epoch": 0.09356, + "grad_norm": 0.9154151678085327, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 9356 + }, + { + "epoch": 0.09357, + "grad_norm": 0.9658426642417908, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 9357 + }, + { + "epoch": 0.09358, + "grad_norm": 0.9715717434883118, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 9358 + }, + { + "epoch": 0.09359, + "grad_norm": 0.8674546480178833, + "learning_rate": 0.003, + "loss": 4.055, + "step": 9359 + }, + { + "epoch": 0.0936, + "grad_norm": 0.8513766527175903, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 9360 + }, + { + "epoch": 0.09361, + "grad_norm": 0.9162145256996155, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 9361 + }, + { + "epoch": 0.09362, + "grad_norm": 1.0360206365585327, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 9362 + }, + { + "epoch": 0.09363, + "grad_norm": 1.0021082162857056, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 9363 + }, + { + "epoch": 0.09364, + "grad_norm": 0.9146813750267029, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 9364 + }, + { + "epoch": 0.09365, + "grad_norm": 0.8867185711860657, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 9365 + }, + { + "epoch": 0.09366, + "grad_norm": 0.9262175559997559, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 9366 + }, + { + "epoch": 0.09367, + "grad_norm": 0.9771784543991089, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 9367 + }, + { + "epoch": 0.09368, + "grad_norm": 1.0174373388290405, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 9368 + }, + { + "epoch": 0.09369, + "grad_norm": 0.972047746181488, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 9369 + }, + { + "epoch": 0.0937, + "grad_norm": 0.9600644111633301, + "learning_rate": 0.003, + "loss": 4.074, + "step": 9370 + }, + { + "epoch": 0.09371, + "grad_norm": 0.8365176916122437, + "learning_rate": 0.003, + "loss": 4.06, + "step": 9371 + }, + { + "epoch": 0.09372, + "grad_norm": 0.6783679127693176, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 9372 + }, + { + "epoch": 0.09373, + "grad_norm": 0.6796581149101257, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 9373 + }, + { + "epoch": 0.09374, + "grad_norm": 0.6576843857765198, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 9374 + }, + { + "epoch": 0.09375, + "grad_norm": 0.7115494608879089, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 9375 + }, + { + "epoch": 0.09376, + "grad_norm": 0.7252764105796814, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 9376 + }, + { + "epoch": 0.09377, + "grad_norm": 0.7984814047813416, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 9377 + }, + { + "epoch": 0.09378, + "grad_norm": 0.8357805609703064, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 9378 + }, + { + "epoch": 0.09379, + "grad_norm": 0.7846959233283997, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 9379 + }, + { + "epoch": 0.0938, + "grad_norm": 0.7526464462280273, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 9380 + }, + { + "epoch": 0.09381, + "grad_norm": 0.8302079439163208, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 9381 + }, + { + "epoch": 0.09382, + "grad_norm": 1.106316089630127, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 9382 + }, + { + "epoch": 0.09383, + "grad_norm": 1.034866452217102, + "learning_rate": 0.003, + "loss": 4.037, + "step": 9383 + }, + { + "epoch": 0.09384, + "grad_norm": 0.8781853914260864, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 9384 + }, + { + "epoch": 0.09385, + "grad_norm": 0.8163227438926697, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 9385 + }, + { + "epoch": 0.09386, + "grad_norm": 0.7359960675239563, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 9386 + }, + { + "epoch": 0.09387, + "grad_norm": 0.8685034513473511, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 9387 + }, + { + "epoch": 0.09388, + "grad_norm": 0.9002259969711304, + "learning_rate": 0.003, + "loss": 4.049, + "step": 9388 + }, + { + "epoch": 0.09389, + "grad_norm": 0.7710829973220825, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 9389 + }, + { + "epoch": 0.0939, + "grad_norm": 0.7765557169914246, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 9390 + }, + { + "epoch": 0.09391, + "grad_norm": 0.7747265100479126, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 9391 + }, + { + "epoch": 0.09392, + "grad_norm": 0.6100158095359802, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 9392 + }, + { + "epoch": 0.09393, + "grad_norm": 0.533496081829071, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 9393 + }, + { + "epoch": 0.09394, + "grad_norm": 0.5862325429916382, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 9394 + }, + { + "epoch": 0.09395, + "grad_norm": 0.5538999438285828, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 9395 + }, + { + "epoch": 0.09396, + "grad_norm": 0.5180746912956238, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 9396 + }, + { + "epoch": 0.09397, + "grad_norm": 0.5542228817939758, + "learning_rate": 0.003, + "loss": 4.013, + "step": 9397 + }, + { + "epoch": 0.09398, + "grad_norm": 0.5188986659049988, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 9398 + }, + { + "epoch": 0.09399, + "grad_norm": 0.5786181688308716, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 9399 + }, + { + "epoch": 0.094, + "grad_norm": 0.6772082448005676, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 9400 + }, + { + "epoch": 0.09401, + "grad_norm": 0.693780779838562, + "learning_rate": 0.003, + "loss": 4.041, + "step": 9401 + }, + { + "epoch": 0.09402, + "grad_norm": 0.6480074524879456, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 9402 + }, + { + "epoch": 0.09403, + "grad_norm": 0.599751353263855, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 9403 + }, + { + "epoch": 0.09404, + "grad_norm": 0.5232686996459961, + "learning_rate": 0.003, + "loss": 4.004, + "step": 9404 + }, + { + "epoch": 0.09405, + "grad_norm": 0.5285734534263611, + "learning_rate": 0.003, + "loss": 4.042, + "step": 9405 + }, + { + "epoch": 0.09406, + "grad_norm": 0.5814787149429321, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 9406 + }, + { + "epoch": 0.09407, + "grad_norm": 0.6249456405639648, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 9407 + }, + { + "epoch": 0.09408, + "grad_norm": 0.774571418762207, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 9408 + }, + { + "epoch": 0.09409, + "grad_norm": 0.8388532400131226, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 9409 + }, + { + "epoch": 0.0941, + "grad_norm": 0.783204197883606, + "learning_rate": 0.003, + "loss": 4.05, + "step": 9410 + }, + { + "epoch": 0.09411, + "grad_norm": 0.9117321372032166, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 9411 + }, + { + "epoch": 0.09412, + "grad_norm": 0.9661768078804016, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 9412 + }, + { + "epoch": 0.09413, + "grad_norm": 1.01078462600708, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 9413 + }, + { + "epoch": 0.09414, + "grad_norm": 1.0741658210754395, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 9414 + }, + { + "epoch": 0.09415, + "grad_norm": 0.9591096043586731, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 9415 + }, + { + "epoch": 0.09416, + "grad_norm": 0.8794155716896057, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 9416 + }, + { + "epoch": 0.09417, + "grad_norm": 0.841712236404419, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 9417 + }, + { + "epoch": 0.09418, + "grad_norm": 0.7917954921722412, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 9418 + }, + { + "epoch": 0.09419, + "grad_norm": 0.7964409589767456, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 9419 + }, + { + "epoch": 0.0942, + "grad_norm": 0.8282405734062195, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 9420 + }, + { + "epoch": 0.09421, + "grad_norm": 0.8015482425689697, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 9421 + }, + { + "epoch": 0.09422, + "grad_norm": 0.7137182950973511, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 9422 + }, + { + "epoch": 0.09423, + "grad_norm": 0.8045299053192139, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 9423 + }, + { + "epoch": 0.09424, + "grad_norm": 0.7985638380050659, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 9424 + }, + { + "epoch": 0.09425, + "grad_norm": 0.9419013261795044, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 9425 + }, + { + "epoch": 0.09426, + "grad_norm": 0.8842517137527466, + "learning_rate": 0.003, + "loss": 4.065, + "step": 9426 + }, + { + "epoch": 0.09427, + "grad_norm": 0.8842640519142151, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 9427 + }, + { + "epoch": 0.09428, + "grad_norm": 0.8510754108428955, + "learning_rate": 0.003, + "loss": 4.057, + "step": 9428 + }, + { + "epoch": 0.09429, + "grad_norm": 0.7526082396507263, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 9429 + }, + { + "epoch": 0.0943, + "grad_norm": 0.7133992314338684, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 9430 + }, + { + "epoch": 0.09431, + "grad_norm": 0.7032984495162964, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 9431 + }, + { + "epoch": 0.09432, + "grad_norm": 0.7109702229499817, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 9432 + }, + { + "epoch": 0.09433, + "grad_norm": 0.8378332257270813, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 9433 + }, + { + "epoch": 0.09434, + "grad_norm": 0.8901048302650452, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 9434 + }, + { + "epoch": 0.09435, + "grad_norm": 0.8476412296295166, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 9435 + }, + { + "epoch": 0.09436, + "grad_norm": 0.7854786515235901, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 9436 + }, + { + "epoch": 0.09437, + "grad_norm": 0.6704685091972351, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 9437 + }, + { + "epoch": 0.09438, + "grad_norm": 0.6411417722702026, + "learning_rate": 0.003, + "loss": 4.05, + "step": 9438 + }, + { + "epoch": 0.09439, + "grad_norm": 0.6393620371818542, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 9439 + }, + { + "epoch": 0.0944, + "grad_norm": 0.728570818901062, + "learning_rate": 0.003, + "loss": 4.024, + "step": 9440 + }, + { + "epoch": 0.09441, + "grad_norm": 0.9385724067687988, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 9441 + }, + { + "epoch": 0.09442, + "grad_norm": 0.9962633848190308, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 9442 + }, + { + "epoch": 0.09443, + "grad_norm": 0.8917745351791382, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 9443 + }, + { + "epoch": 0.09444, + "grad_norm": 0.7558419704437256, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 9444 + }, + { + "epoch": 0.09445, + "grad_norm": 0.6210421919822693, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 9445 + }, + { + "epoch": 0.09446, + "grad_norm": 0.6951500177383423, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 9446 + }, + { + "epoch": 0.09447, + "grad_norm": 0.8409909605979919, + "learning_rate": 0.003, + "loss": 4.05, + "step": 9447 + }, + { + "epoch": 0.09448, + "grad_norm": 0.9376384019851685, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 9448 + }, + { + "epoch": 0.09449, + "grad_norm": 1.001735806465149, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 9449 + }, + { + "epoch": 0.0945, + "grad_norm": 1.2646803855895996, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 9450 + }, + { + "epoch": 0.09451, + "grad_norm": 0.7255167365074158, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 9451 + }, + { + "epoch": 0.09452, + "grad_norm": 0.775261402130127, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 9452 + }, + { + "epoch": 0.09453, + "grad_norm": 0.7530305981636047, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 9453 + }, + { + "epoch": 0.09454, + "grad_norm": 0.8406985402107239, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 9454 + }, + { + "epoch": 0.09455, + "grad_norm": 0.8127760887145996, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 9455 + }, + { + "epoch": 0.09456, + "grad_norm": 0.679355800151825, + "learning_rate": 0.003, + "loss": 4.026, + "step": 9456 + }, + { + "epoch": 0.09457, + "grad_norm": 0.660081148147583, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 9457 + }, + { + "epoch": 0.09458, + "grad_norm": 0.5813243985176086, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 9458 + }, + { + "epoch": 0.09459, + "grad_norm": 0.4867233633995056, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 9459 + }, + { + "epoch": 0.0946, + "grad_norm": 0.5141322612762451, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 9460 + }, + { + "epoch": 0.09461, + "grad_norm": 0.6217049956321716, + "learning_rate": 0.003, + "loss": 4.06, + "step": 9461 + }, + { + "epoch": 0.09462, + "grad_norm": 0.7092083692550659, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 9462 + }, + { + "epoch": 0.09463, + "grad_norm": 0.8736013770103455, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 9463 + }, + { + "epoch": 0.09464, + "grad_norm": 0.9941328763961792, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 9464 + }, + { + "epoch": 0.09465, + "grad_norm": 1.0948302745819092, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 9465 + }, + { + "epoch": 0.09466, + "grad_norm": 0.8153526782989502, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 9466 + }, + { + "epoch": 0.09467, + "grad_norm": 0.6857858300209045, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 9467 + }, + { + "epoch": 0.09468, + "grad_norm": 0.7290007472038269, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 9468 + }, + { + "epoch": 0.09469, + "grad_norm": 0.7589170932769775, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 9469 + }, + { + "epoch": 0.0947, + "grad_norm": 0.7441050410270691, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 9470 + }, + { + "epoch": 0.09471, + "grad_norm": 0.699702262878418, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 9471 + }, + { + "epoch": 0.09472, + "grad_norm": 0.6244022250175476, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 9472 + }, + { + "epoch": 0.09473, + "grad_norm": 0.5642143487930298, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 9473 + }, + { + "epoch": 0.09474, + "grad_norm": 0.6282489895820618, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 9474 + }, + { + "epoch": 0.09475, + "grad_norm": 0.538658082485199, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 9475 + }, + { + "epoch": 0.09476, + "grad_norm": 0.5374001264572144, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 9476 + }, + { + "epoch": 0.09477, + "grad_norm": 0.509412944316864, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 9477 + }, + { + "epoch": 0.09478, + "grad_norm": 0.5749872326850891, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 9478 + }, + { + "epoch": 0.09479, + "grad_norm": 0.6383438110351562, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 9479 + }, + { + "epoch": 0.0948, + "grad_norm": 0.8217725157737732, + "learning_rate": 0.003, + "loss": 4.011, + "step": 9480 + }, + { + "epoch": 0.09481, + "grad_norm": 1.1610746383666992, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 9481 + }, + { + "epoch": 0.09482, + "grad_norm": 0.9759625196456909, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 9482 + }, + { + "epoch": 0.09483, + "grad_norm": 0.8113720417022705, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 9483 + }, + { + "epoch": 0.09484, + "grad_norm": 0.7075552344322205, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 9484 + }, + { + "epoch": 0.09485, + "grad_norm": 0.718109667301178, + "learning_rate": 0.003, + "loss": 4.028, + "step": 9485 + }, + { + "epoch": 0.09486, + "grad_norm": 0.7402860522270203, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 9486 + }, + { + "epoch": 0.09487, + "grad_norm": 0.8021457195281982, + "learning_rate": 0.003, + "loss": 4.032, + "step": 9487 + }, + { + "epoch": 0.09488, + "grad_norm": 0.9090721607208252, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 9488 + }, + { + "epoch": 0.09489, + "grad_norm": 1.036071538925171, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 9489 + }, + { + "epoch": 0.0949, + "grad_norm": 0.9327799677848816, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 9490 + }, + { + "epoch": 0.09491, + "grad_norm": 0.9281620383262634, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 9491 + }, + { + "epoch": 0.09492, + "grad_norm": 0.8979142904281616, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 9492 + }, + { + "epoch": 0.09493, + "grad_norm": 0.8455875515937805, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 9493 + }, + { + "epoch": 0.09494, + "grad_norm": 0.7073626518249512, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 9494 + }, + { + "epoch": 0.09495, + "grad_norm": 0.6978188753128052, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 9495 + }, + { + "epoch": 0.09496, + "grad_norm": 0.6406537294387817, + "learning_rate": 0.003, + "loss": 4.055, + "step": 9496 + }, + { + "epoch": 0.09497, + "grad_norm": 0.584082841873169, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 9497 + }, + { + "epoch": 0.09498, + "grad_norm": 0.650418758392334, + "learning_rate": 0.003, + "loss": 4.015, + "step": 9498 + }, + { + "epoch": 0.09499, + "grad_norm": 0.6627920269966125, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 9499 + }, + { + "epoch": 0.095, + "grad_norm": 0.7207992672920227, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 9500 + }, + { + "epoch": 0.09501, + "grad_norm": 0.8046587705612183, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 9501 + }, + { + "epoch": 0.09502, + "grad_norm": 0.9965620040893555, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 9502 + }, + { + "epoch": 0.09503, + "grad_norm": 1.2637056112289429, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 9503 + }, + { + "epoch": 0.09504, + "grad_norm": 0.7111979722976685, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 9504 + }, + { + "epoch": 0.09505, + "grad_norm": 0.5596445798873901, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 9505 + }, + { + "epoch": 0.09506, + "grad_norm": 0.6463555693626404, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 9506 + }, + { + "epoch": 0.09507, + "grad_norm": 0.8175804615020752, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 9507 + }, + { + "epoch": 0.09508, + "grad_norm": 0.995126485824585, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 9508 + }, + { + "epoch": 0.09509, + "grad_norm": 1.0844470262527466, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 9509 + }, + { + "epoch": 0.0951, + "grad_norm": 0.8322227001190186, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 9510 + }, + { + "epoch": 0.09511, + "grad_norm": 0.838636040687561, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 9511 + }, + { + "epoch": 0.09512, + "grad_norm": 0.8196890354156494, + "learning_rate": 0.003, + "loss": 4.03, + "step": 9512 + }, + { + "epoch": 0.09513, + "grad_norm": 0.7390263080596924, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 9513 + }, + { + "epoch": 0.09514, + "grad_norm": 0.5975653529167175, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 9514 + }, + { + "epoch": 0.09515, + "grad_norm": 0.544066846370697, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 9515 + }, + { + "epoch": 0.09516, + "grad_norm": 0.6771560907363892, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 9516 + }, + { + "epoch": 0.09517, + "grad_norm": 0.7807263135910034, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 9517 + }, + { + "epoch": 0.09518, + "grad_norm": 0.8398572206497192, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 9518 + }, + { + "epoch": 0.09519, + "grad_norm": 0.8214612007141113, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 9519 + }, + { + "epoch": 0.0952, + "grad_norm": 0.7340238690376282, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 9520 + }, + { + "epoch": 0.09521, + "grad_norm": 0.6527026295661926, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 9521 + }, + { + "epoch": 0.09522, + "grad_norm": 0.7640578746795654, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 9522 + }, + { + "epoch": 0.09523, + "grad_norm": 0.8175305724143982, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 9523 + }, + { + "epoch": 0.09524, + "grad_norm": 0.9655996561050415, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 9524 + }, + { + "epoch": 0.09525, + "grad_norm": 0.8379790186882019, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 9525 + }, + { + "epoch": 0.09526, + "grad_norm": 0.9293071627616882, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 9526 + }, + { + "epoch": 0.09527, + "grad_norm": 0.9869701266288757, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 9527 + }, + { + "epoch": 0.09528, + "grad_norm": 1.0038269758224487, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 9528 + }, + { + "epoch": 0.09529, + "grad_norm": 1.0940920114517212, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 9529 + }, + { + "epoch": 0.0953, + "grad_norm": 0.8704492449760437, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 9530 + }, + { + "epoch": 0.09531, + "grad_norm": 0.7913749814033508, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 9531 + }, + { + "epoch": 0.09532, + "grad_norm": 0.9196059703826904, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 9532 + }, + { + "epoch": 0.09533, + "grad_norm": 1.0615483522415161, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 9533 + }, + { + "epoch": 0.09534, + "grad_norm": 0.9020791053771973, + "learning_rate": 0.003, + "loss": 4.035, + "step": 9534 + }, + { + "epoch": 0.09535, + "grad_norm": 0.948823094367981, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 9535 + }, + { + "epoch": 0.09536, + "grad_norm": 1.080247163772583, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 9536 + }, + { + "epoch": 0.09537, + "grad_norm": 0.874646008014679, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 9537 + }, + { + "epoch": 0.09538, + "grad_norm": 0.9942668676376343, + "learning_rate": 0.003, + "loss": 4.05, + "step": 9538 + }, + { + "epoch": 0.09539, + "grad_norm": 1.0160984992980957, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 9539 + }, + { + "epoch": 0.0954, + "grad_norm": 0.9498040676116943, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 9540 + }, + { + "epoch": 0.09541, + "grad_norm": 0.7964853644371033, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 9541 + }, + { + "epoch": 0.09542, + "grad_norm": 0.6286479830741882, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 9542 + }, + { + "epoch": 0.09543, + "grad_norm": 0.5860509872436523, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 9543 + }, + { + "epoch": 0.09544, + "grad_norm": 0.6001367568969727, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 9544 + }, + { + "epoch": 0.09545, + "grad_norm": 0.5685854554176331, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 9545 + }, + { + "epoch": 0.09546, + "grad_norm": 0.6297983527183533, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 9546 + }, + { + "epoch": 0.09547, + "grad_norm": 0.594254195690155, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 9547 + }, + { + "epoch": 0.09548, + "grad_norm": 0.5334281921386719, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 9548 + }, + { + "epoch": 0.09549, + "grad_norm": 0.47183743119239807, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 9549 + }, + { + "epoch": 0.0955, + "grad_norm": 0.4987733066082001, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 9550 + }, + { + "epoch": 0.09551, + "grad_norm": 0.5213873386383057, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 9551 + }, + { + "epoch": 0.09552, + "grad_norm": 0.5530688166618347, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 9552 + }, + { + "epoch": 0.09553, + "grad_norm": 0.5738921761512756, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 9553 + }, + { + "epoch": 0.09554, + "grad_norm": 0.5256248116493225, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 9554 + }, + { + "epoch": 0.09555, + "grad_norm": 0.5314795970916748, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 9555 + }, + { + "epoch": 0.09556, + "grad_norm": 0.6204894781112671, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 9556 + }, + { + "epoch": 0.09557, + "grad_norm": 0.8138146996498108, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 9557 + }, + { + "epoch": 0.09558, + "grad_norm": 1.0319538116455078, + "learning_rate": 0.003, + "loss": 4.035, + "step": 9558 + }, + { + "epoch": 0.09559, + "grad_norm": 1.1002483367919922, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 9559 + }, + { + "epoch": 0.0956, + "grad_norm": 0.7374374270439148, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 9560 + }, + { + "epoch": 0.09561, + "grad_norm": 0.6492986083030701, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 9561 + }, + { + "epoch": 0.09562, + "grad_norm": 0.6662077903747559, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 9562 + }, + { + "epoch": 0.09563, + "grad_norm": 0.6137615442276001, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 9563 + }, + { + "epoch": 0.09564, + "grad_norm": 0.6091228127479553, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 9564 + }, + { + "epoch": 0.09565, + "grad_norm": 0.6556311845779419, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 9565 + }, + { + "epoch": 0.09566, + "grad_norm": 0.7306276559829712, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 9566 + }, + { + "epoch": 0.09567, + "grad_norm": 0.8696443438529968, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 9567 + }, + { + "epoch": 0.09568, + "grad_norm": 0.9888119101524353, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 9568 + }, + { + "epoch": 0.09569, + "grad_norm": 0.8965682983398438, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 9569 + }, + { + "epoch": 0.0957, + "grad_norm": 0.6938888430595398, + "learning_rate": 0.003, + "loss": 4.037, + "step": 9570 + }, + { + "epoch": 0.09571, + "grad_norm": 0.6471319794654846, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 9571 + }, + { + "epoch": 0.09572, + "grad_norm": 0.6941522359848022, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 9572 + }, + { + "epoch": 0.09573, + "grad_norm": 0.7525028586387634, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 9573 + }, + { + "epoch": 0.09574, + "grad_norm": 0.7359570860862732, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 9574 + }, + { + "epoch": 0.09575, + "grad_norm": 0.679182231426239, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 9575 + }, + { + "epoch": 0.09576, + "grad_norm": 0.634701669216156, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 9576 + }, + { + "epoch": 0.09577, + "grad_norm": 0.5520635843276978, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 9577 + }, + { + "epoch": 0.09578, + "grad_norm": 0.6691205501556396, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 9578 + }, + { + "epoch": 0.09579, + "grad_norm": 0.8371917605400085, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 9579 + }, + { + "epoch": 0.0958, + "grad_norm": 1.0776106119155884, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 9580 + }, + { + "epoch": 0.09581, + "grad_norm": 1.0245087146759033, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 9581 + }, + { + "epoch": 0.09582, + "grad_norm": 0.8801830410957336, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 9582 + }, + { + "epoch": 0.09583, + "grad_norm": 0.8247877359390259, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 9583 + }, + { + "epoch": 0.09584, + "grad_norm": 0.7499005198478699, + "learning_rate": 0.003, + "loss": 4.032, + "step": 9584 + }, + { + "epoch": 0.09585, + "grad_norm": 0.7511628866195679, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 9585 + }, + { + "epoch": 0.09586, + "grad_norm": 0.6854458451271057, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 9586 + }, + { + "epoch": 0.09587, + "grad_norm": 0.6980164051055908, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 9587 + }, + { + "epoch": 0.09588, + "grad_norm": 0.7041137218475342, + "learning_rate": 0.003, + "loss": 4.029, + "step": 9588 + }, + { + "epoch": 0.09589, + "grad_norm": 0.7521610856056213, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 9589 + }, + { + "epoch": 0.0959, + "grad_norm": 0.7036656737327576, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 9590 + }, + { + "epoch": 0.09591, + "grad_norm": 0.7238452434539795, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 9591 + }, + { + "epoch": 0.09592, + "grad_norm": 0.8542155027389526, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 9592 + }, + { + "epoch": 0.09593, + "grad_norm": 0.9843723177909851, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 9593 + }, + { + "epoch": 0.09594, + "grad_norm": 1.0649001598358154, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 9594 + }, + { + "epoch": 0.09595, + "grad_norm": 0.8816093802452087, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 9595 + }, + { + "epoch": 0.09596, + "grad_norm": 0.7636190056800842, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 9596 + }, + { + "epoch": 0.09597, + "grad_norm": 0.7464763522148132, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 9597 + }, + { + "epoch": 0.09598, + "grad_norm": 0.8498935699462891, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 9598 + }, + { + "epoch": 0.09599, + "grad_norm": 0.9224379062652588, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 9599 + }, + { + "epoch": 0.096, + "grad_norm": 0.8882635235786438, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 9600 + }, + { + "epoch": 0.09601, + "grad_norm": 0.8223801255226135, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 9601 + }, + { + "epoch": 0.09602, + "grad_norm": 0.8122618198394775, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 9602 + }, + { + "epoch": 0.09603, + "grad_norm": 0.7895182967185974, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 9603 + }, + { + "epoch": 0.09604, + "grad_norm": 0.6999941468238831, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 9604 + }, + { + "epoch": 0.09605, + "grad_norm": 0.7563785910606384, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 9605 + }, + { + "epoch": 0.09606, + "grad_norm": 0.8351266980171204, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 9606 + }, + { + "epoch": 0.09607, + "grad_norm": 0.9096396565437317, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 9607 + }, + { + "epoch": 0.09608, + "grad_norm": 0.9197260141372681, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 9608 + }, + { + "epoch": 0.09609, + "grad_norm": 0.7963103652000427, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 9609 + }, + { + "epoch": 0.0961, + "grad_norm": 0.6685717701911926, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 9610 + }, + { + "epoch": 0.09611, + "grad_norm": 0.7270292043685913, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 9611 + }, + { + "epoch": 0.09612, + "grad_norm": 0.7644274830818176, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 9612 + }, + { + "epoch": 0.09613, + "grad_norm": 0.790471613407135, + "learning_rate": 0.003, + "loss": 4.049, + "step": 9613 + }, + { + "epoch": 0.09614, + "grad_norm": 0.8529259562492371, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 9614 + }, + { + "epoch": 0.09615, + "grad_norm": 0.7713029384613037, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 9615 + }, + { + "epoch": 0.09616, + "grad_norm": 0.6528921127319336, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 9616 + }, + { + "epoch": 0.09617, + "grad_norm": 0.7347012758255005, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 9617 + }, + { + "epoch": 0.09618, + "grad_norm": 0.7659905552864075, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 9618 + }, + { + "epoch": 0.09619, + "grad_norm": 0.8106355667114258, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 9619 + }, + { + "epoch": 0.0962, + "grad_norm": 0.6627798080444336, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 9620 + }, + { + "epoch": 0.09621, + "grad_norm": 0.7969244122505188, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 9621 + }, + { + "epoch": 0.09622, + "grad_norm": 0.903766393661499, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 9622 + }, + { + "epoch": 0.09623, + "grad_norm": 1.0692620277404785, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 9623 + }, + { + "epoch": 0.09624, + "grad_norm": 0.8282185196876526, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 9624 + }, + { + "epoch": 0.09625, + "grad_norm": 0.713636040687561, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 9625 + }, + { + "epoch": 0.09626, + "grad_norm": 0.7236156463623047, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 9626 + }, + { + "epoch": 0.09627, + "grad_norm": 0.694229245185852, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 9627 + }, + { + "epoch": 0.09628, + "grad_norm": 0.6868857741355896, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 9628 + }, + { + "epoch": 0.09629, + "grad_norm": 0.7673381567001343, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 9629 + }, + { + "epoch": 0.0963, + "grad_norm": 0.8264285922050476, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 9630 + }, + { + "epoch": 0.09631, + "grad_norm": 0.8883211016654968, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 9631 + }, + { + "epoch": 0.09632, + "grad_norm": 0.8831930756568909, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 9632 + }, + { + "epoch": 0.09633, + "grad_norm": 0.8816713690757751, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 9633 + }, + { + "epoch": 0.09634, + "grad_norm": 0.8870523571968079, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 9634 + }, + { + "epoch": 0.09635, + "grad_norm": 0.9964356422424316, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 9635 + }, + { + "epoch": 0.09636, + "grad_norm": 0.8906300067901611, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 9636 + }, + { + "epoch": 0.09637, + "grad_norm": 0.8000807762145996, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 9637 + }, + { + "epoch": 0.09638, + "grad_norm": 0.7173836827278137, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 9638 + }, + { + "epoch": 0.09639, + "grad_norm": 0.67566978931427, + "learning_rate": 0.003, + "loss": 4.022, + "step": 9639 + }, + { + "epoch": 0.0964, + "grad_norm": 0.5778241753578186, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 9640 + }, + { + "epoch": 0.09641, + "grad_norm": 0.7081737518310547, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 9641 + }, + { + "epoch": 0.09642, + "grad_norm": 0.741016149520874, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 9642 + }, + { + "epoch": 0.09643, + "grad_norm": 0.778562605381012, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 9643 + }, + { + "epoch": 0.09644, + "grad_norm": 0.9048216342926025, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 9644 + }, + { + "epoch": 0.09645, + "grad_norm": 0.9883460402488708, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 9645 + }, + { + "epoch": 0.09646, + "grad_norm": 0.7610167860984802, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 9646 + }, + { + "epoch": 0.09647, + "grad_norm": 0.6392595767974854, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 9647 + }, + { + "epoch": 0.09648, + "grad_norm": 0.6436144709587097, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 9648 + }, + { + "epoch": 0.09649, + "grad_norm": 0.6952054500579834, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 9649 + }, + { + "epoch": 0.0965, + "grad_norm": 0.6287616491317749, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 9650 + }, + { + "epoch": 0.09651, + "grad_norm": 0.6473034024238586, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 9651 + }, + { + "epoch": 0.09652, + "grad_norm": 0.6925511360168457, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 9652 + }, + { + "epoch": 0.09653, + "grad_norm": 0.708454966545105, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 9653 + }, + { + "epoch": 0.09654, + "grad_norm": 0.7384510040283203, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 9654 + }, + { + "epoch": 0.09655, + "grad_norm": 0.6604750156402588, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 9655 + }, + { + "epoch": 0.09656, + "grad_norm": 0.5758141279220581, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 9656 + }, + { + "epoch": 0.09657, + "grad_norm": 0.6268450021743774, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 9657 + }, + { + "epoch": 0.09658, + "grad_norm": 0.7189260125160217, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 9658 + }, + { + "epoch": 0.09659, + "grad_norm": 0.7771549224853516, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 9659 + }, + { + "epoch": 0.0966, + "grad_norm": 0.8796934485435486, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 9660 + }, + { + "epoch": 0.09661, + "grad_norm": 0.896022617816925, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 9661 + }, + { + "epoch": 0.09662, + "grad_norm": 0.938194751739502, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 9662 + }, + { + "epoch": 0.09663, + "grad_norm": 0.8051214218139648, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 9663 + }, + { + "epoch": 0.09664, + "grad_norm": 0.6830205321311951, + "learning_rate": 0.003, + "loss": 4.044, + "step": 9664 + }, + { + "epoch": 0.09665, + "grad_norm": 0.793272078037262, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 9665 + }, + { + "epoch": 0.09666, + "grad_norm": 0.9261988401412964, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 9666 + }, + { + "epoch": 0.09667, + "grad_norm": 0.8233135938644409, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 9667 + }, + { + "epoch": 0.09668, + "grad_norm": 0.7708394527435303, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 9668 + }, + { + "epoch": 0.09669, + "grad_norm": 0.7485005855560303, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 9669 + }, + { + "epoch": 0.0967, + "grad_norm": 0.7645751237869263, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 9670 + }, + { + "epoch": 0.09671, + "grad_norm": 0.882091224193573, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 9671 + }, + { + "epoch": 0.09672, + "grad_norm": 1.0624443292617798, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 9672 + }, + { + "epoch": 0.09673, + "grad_norm": 0.900747537612915, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 9673 + }, + { + "epoch": 0.09674, + "grad_norm": 0.8541334867477417, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 9674 + }, + { + "epoch": 0.09675, + "grad_norm": 0.9348879456520081, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 9675 + }, + { + "epoch": 0.09676, + "grad_norm": 0.8951733112335205, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 9676 + }, + { + "epoch": 0.09677, + "grad_norm": 1.0163615942001343, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 9677 + }, + { + "epoch": 0.09678, + "grad_norm": 0.9605147242546082, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 9678 + }, + { + "epoch": 0.09679, + "grad_norm": 0.8225002288818359, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 9679 + }, + { + "epoch": 0.0968, + "grad_norm": 0.8135245442390442, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 9680 + }, + { + "epoch": 0.09681, + "grad_norm": 0.8713023066520691, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 9681 + }, + { + "epoch": 0.09682, + "grad_norm": 0.8497050404548645, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 9682 + }, + { + "epoch": 0.09683, + "grad_norm": 0.7430970072746277, + "learning_rate": 0.003, + "loss": 4.036, + "step": 9683 + }, + { + "epoch": 0.09684, + "grad_norm": 0.7872548699378967, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 9684 + }, + { + "epoch": 0.09685, + "grad_norm": 0.6804424524307251, + "learning_rate": 0.003, + "loss": 4.04, + "step": 9685 + }, + { + "epoch": 0.09686, + "grad_norm": 0.5914613008499146, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 9686 + }, + { + "epoch": 0.09687, + "grad_norm": 0.5871429443359375, + "learning_rate": 0.003, + "loss": 4.043, + "step": 9687 + }, + { + "epoch": 0.09688, + "grad_norm": 0.7036808133125305, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 9688 + }, + { + "epoch": 0.09689, + "grad_norm": 0.7072443962097168, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 9689 + }, + { + "epoch": 0.0969, + "grad_norm": 0.6006872653961182, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 9690 + }, + { + "epoch": 0.09691, + "grad_norm": 0.627549409866333, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 9691 + }, + { + "epoch": 0.09692, + "grad_norm": 0.6675216555595398, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 9692 + }, + { + "epoch": 0.09693, + "grad_norm": 0.8195521831512451, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 9693 + }, + { + "epoch": 0.09694, + "grad_norm": 1.020370602607727, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 9694 + }, + { + "epoch": 0.09695, + "grad_norm": 1.159254550933838, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 9695 + }, + { + "epoch": 0.09696, + "grad_norm": 0.7246578335762024, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 9696 + }, + { + "epoch": 0.09697, + "grad_norm": 0.7815423607826233, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 9697 + }, + { + "epoch": 0.09698, + "grad_norm": 0.9583694934844971, + "learning_rate": 0.003, + "loss": 4.053, + "step": 9698 + }, + { + "epoch": 0.09699, + "grad_norm": 0.948357343673706, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 9699 + }, + { + "epoch": 0.097, + "grad_norm": 0.8365939855575562, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 9700 + }, + { + "epoch": 0.09701, + "grad_norm": 0.7484097480773926, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 9701 + }, + { + "epoch": 0.09702, + "grad_norm": 0.8812928795814514, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 9702 + }, + { + "epoch": 0.09703, + "grad_norm": 0.8621808886528015, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 9703 + }, + { + "epoch": 0.09704, + "grad_norm": 0.7826769351959229, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 9704 + }, + { + "epoch": 0.09705, + "grad_norm": 0.6948601603507996, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 9705 + }, + { + "epoch": 0.09706, + "grad_norm": 0.7186203002929688, + "learning_rate": 0.003, + "loss": 4.042, + "step": 9706 + }, + { + "epoch": 0.09707, + "grad_norm": 0.8767298460006714, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 9707 + }, + { + "epoch": 0.09708, + "grad_norm": 1.0957658290863037, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 9708 + }, + { + "epoch": 0.09709, + "grad_norm": 0.939866304397583, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 9709 + }, + { + "epoch": 0.0971, + "grad_norm": 0.9766221046447754, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 9710 + }, + { + "epoch": 0.09711, + "grad_norm": 0.938903272151947, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 9711 + }, + { + "epoch": 0.09712, + "grad_norm": 0.9619114995002747, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 9712 + }, + { + "epoch": 0.09713, + "grad_norm": 0.8209119439125061, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 9713 + }, + { + "epoch": 0.09714, + "grad_norm": 0.5877271890640259, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 9714 + }, + { + "epoch": 0.09715, + "grad_norm": 0.6621226668357849, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 9715 + }, + { + "epoch": 0.09716, + "grad_norm": 0.7536833882331848, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 9716 + }, + { + "epoch": 0.09717, + "grad_norm": 0.8996556401252747, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 9717 + }, + { + "epoch": 0.09718, + "grad_norm": 0.8144115805625916, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 9718 + }, + { + "epoch": 0.09719, + "grad_norm": 0.7471117377281189, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 9719 + }, + { + "epoch": 0.0972, + "grad_norm": 0.7669509649276733, + "learning_rate": 0.003, + "loss": 4.017, + "step": 9720 + }, + { + "epoch": 0.09721, + "grad_norm": 0.7159215807914734, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 9721 + }, + { + "epoch": 0.09722, + "grad_norm": 0.6609331369400024, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 9722 + }, + { + "epoch": 0.09723, + "grad_norm": 0.5876916646957397, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 9723 + }, + { + "epoch": 0.09724, + "grad_norm": 0.48599714040756226, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 9724 + }, + { + "epoch": 0.09725, + "grad_norm": 0.5811905860900879, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 9725 + }, + { + "epoch": 0.09726, + "grad_norm": 0.7120689153671265, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 9726 + }, + { + "epoch": 0.09727, + "grad_norm": 0.8851310014724731, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 9727 + }, + { + "epoch": 0.09728, + "grad_norm": 0.842197060585022, + "learning_rate": 0.003, + "loss": 4.02, + "step": 9728 + }, + { + "epoch": 0.09729, + "grad_norm": 0.6467271447181702, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 9729 + }, + { + "epoch": 0.0973, + "grad_norm": 0.48625877499580383, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 9730 + }, + { + "epoch": 0.09731, + "grad_norm": 0.5645377039909363, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 9731 + }, + { + "epoch": 0.09732, + "grad_norm": 0.5628966093063354, + "learning_rate": 0.003, + "loss": 4.024, + "step": 9732 + }, + { + "epoch": 0.09733, + "grad_norm": 0.46445903182029724, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 9733 + }, + { + "epoch": 0.09734, + "grad_norm": 0.4646936357021332, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 9734 + }, + { + "epoch": 0.09735, + "grad_norm": 0.43636274337768555, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 9735 + }, + { + "epoch": 0.09736, + "grad_norm": 0.5397223830223083, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 9736 + }, + { + "epoch": 0.09737, + "grad_norm": 0.5813911557197571, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 9737 + }, + { + "epoch": 0.09738, + "grad_norm": 0.6568850874900818, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 9738 + }, + { + "epoch": 0.09739, + "grad_norm": 0.7499068379402161, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 9739 + }, + { + "epoch": 0.0974, + "grad_norm": 0.971897542476654, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 9740 + }, + { + "epoch": 0.09741, + "grad_norm": 1.23680579662323, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 9741 + }, + { + "epoch": 0.09742, + "grad_norm": 0.6952364444732666, + "learning_rate": 0.003, + "loss": 4.007, + "step": 9742 + }, + { + "epoch": 0.09743, + "grad_norm": 0.7512345314025879, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 9743 + }, + { + "epoch": 0.09744, + "grad_norm": 0.8438845872879028, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 9744 + }, + { + "epoch": 0.09745, + "grad_norm": 0.881648063659668, + "learning_rate": 0.003, + "loss": 4.06, + "step": 9745 + }, + { + "epoch": 0.09746, + "grad_norm": 0.8643108010292053, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 9746 + }, + { + "epoch": 0.09747, + "grad_norm": 1.1098822355270386, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 9747 + }, + { + "epoch": 0.09748, + "grad_norm": 0.9371985197067261, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 9748 + }, + { + "epoch": 0.09749, + "grad_norm": 0.9462610483169556, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 9749 + }, + { + "epoch": 0.0975, + "grad_norm": 0.8169135451316833, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 9750 + }, + { + "epoch": 0.09751, + "grad_norm": 0.7658066749572754, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 9751 + }, + { + "epoch": 0.09752, + "grad_norm": 0.7241183519363403, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 9752 + }, + { + "epoch": 0.09753, + "grad_norm": 0.7353487014770508, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 9753 + }, + { + "epoch": 0.09754, + "grad_norm": 0.7551058530807495, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 9754 + }, + { + "epoch": 0.09755, + "grad_norm": 0.6881881356239319, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 9755 + }, + { + "epoch": 0.09756, + "grad_norm": 0.6139550805091858, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 9756 + }, + { + "epoch": 0.09757, + "grad_norm": 0.7223386168479919, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 9757 + }, + { + "epoch": 0.09758, + "grad_norm": 0.8988397121429443, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 9758 + }, + { + "epoch": 0.09759, + "grad_norm": 1.005197286605835, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 9759 + }, + { + "epoch": 0.0976, + "grad_norm": 1.1311531066894531, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 9760 + }, + { + "epoch": 0.09761, + "grad_norm": 0.7937563061714172, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 9761 + }, + { + "epoch": 0.09762, + "grad_norm": 0.6707890033721924, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 9762 + }, + { + "epoch": 0.09763, + "grad_norm": 0.7106285691261292, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 9763 + }, + { + "epoch": 0.09764, + "grad_norm": 0.7116639614105225, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 9764 + }, + { + "epoch": 0.09765, + "grad_norm": 0.7842574119567871, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 9765 + }, + { + "epoch": 0.09766, + "grad_norm": 0.902124285697937, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 9766 + }, + { + "epoch": 0.09767, + "grad_norm": 1.0107566118240356, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 9767 + }, + { + "epoch": 0.09768, + "grad_norm": 0.9981918334960938, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 9768 + }, + { + "epoch": 0.09769, + "grad_norm": 0.7847188115119934, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 9769 + }, + { + "epoch": 0.0977, + "grad_norm": 0.8173283934593201, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 9770 + }, + { + "epoch": 0.09771, + "grad_norm": 0.8425480127334595, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 9771 + }, + { + "epoch": 0.09772, + "grad_norm": 0.8838803172111511, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 9772 + }, + { + "epoch": 0.09773, + "grad_norm": 0.8878778219223022, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 9773 + }, + { + "epoch": 0.09774, + "grad_norm": 0.9738554954528809, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 9774 + }, + { + "epoch": 0.09775, + "grad_norm": 0.969858705997467, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 9775 + }, + { + "epoch": 0.09776, + "grad_norm": 1.0172128677368164, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 9776 + }, + { + "epoch": 0.09777, + "grad_norm": 0.7204704284667969, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 9777 + }, + { + "epoch": 0.09778, + "grad_norm": 0.6760993003845215, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 9778 + }, + { + "epoch": 0.09779, + "grad_norm": 0.723325788974762, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 9779 + }, + { + "epoch": 0.0978, + "grad_norm": 0.7009559273719788, + "learning_rate": 0.003, + "loss": 4.026, + "step": 9780 + }, + { + "epoch": 0.09781, + "grad_norm": 0.7214966416358948, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 9781 + }, + { + "epoch": 0.09782, + "grad_norm": 0.7424318790435791, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 9782 + }, + { + "epoch": 0.09783, + "grad_norm": 0.6193498969078064, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 9783 + }, + { + "epoch": 0.09784, + "grad_norm": 0.6854250431060791, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 9784 + }, + { + "epoch": 0.09785, + "grad_norm": 0.7265077233314514, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 9785 + }, + { + "epoch": 0.09786, + "grad_norm": 0.7454569935798645, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 9786 + }, + { + "epoch": 0.09787, + "grad_norm": 0.8594154119491577, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 9787 + }, + { + "epoch": 0.09788, + "grad_norm": 1.0226812362670898, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 9788 + }, + { + "epoch": 0.09789, + "grad_norm": 1.1170358657836914, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 9789 + }, + { + "epoch": 0.0979, + "grad_norm": 0.7604365348815918, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 9790 + }, + { + "epoch": 0.09791, + "grad_norm": 0.6184024810791016, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 9791 + }, + { + "epoch": 0.09792, + "grad_norm": 0.7284930348396301, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 9792 + }, + { + "epoch": 0.09793, + "grad_norm": 0.6939471364021301, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 9793 + }, + { + "epoch": 0.09794, + "grad_norm": 0.7011038661003113, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 9794 + }, + { + "epoch": 0.09795, + "grad_norm": 0.6286042332649231, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 9795 + }, + { + "epoch": 0.09796, + "grad_norm": 0.5014172792434692, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 9796 + }, + { + "epoch": 0.09797, + "grad_norm": 0.5449438691139221, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 9797 + }, + { + "epoch": 0.09798, + "grad_norm": 0.6074647903442383, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 9798 + }, + { + "epoch": 0.09799, + "grad_norm": 0.6542043685913086, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 9799 + }, + { + "epoch": 0.098, + "grad_norm": 0.785072386264801, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 9800 + }, + { + "epoch": 0.09801, + "grad_norm": 0.7558362483978271, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 9801 + }, + { + "epoch": 0.09802, + "grad_norm": 0.7080808281898499, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 9802 + }, + { + "epoch": 0.09803, + "grad_norm": 0.9028113484382629, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 9803 + }, + { + "epoch": 0.09804, + "grad_norm": 1.1181584596633911, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 9804 + }, + { + "epoch": 0.09805, + "grad_norm": 1.098592758178711, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 9805 + }, + { + "epoch": 0.09806, + "grad_norm": 0.807129442691803, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 9806 + }, + { + "epoch": 0.09807, + "grad_norm": 0.7035567164421082, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 9807 + }, + { + "epoch": 0.09808, + "grad_norm": 0.7286992073059082, + "learning_rate": 0.003, + "loss": 4.06, + "step": 9808 + }, + { + "epoch": 0.09809, + "grad_norm": 0.7307597994804382, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 9809 + }, + { + "epoch": 0.0981, + "grad_norm": 0.721434473991394, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 9810 + }, + { + "epoch": 0.09811, + "grad_norm": 0.6393311023712158, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 9811 + }, + { + "epoch": 0.09812, + "grad_norm": 0.6354530453681946, + "learning_rate": 0.003, + "loss": 4.02, + "step": 9812 + }, + { + "epoch": 0.09813, + "grad_norm": 0.7901116609573364, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 9813 + }, + { + "epoch": 0.09814, + "grad_norm": 0.7675938010215759, + "learning_rate": 0.003, + "loss": 4.02, + "step": 9814 + }, + { + "epoch": 0.09815, + "grad_norm": 0.6112906336784363, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 9815 + }, + { + "epoch": 0.09816, + "grad_norm": 0.7082177400588989, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 9816 + }, + { + "epoch": 0.09817, + "grad_norm": 0.7757717370986938, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 9817 + }, + { + "epoch": 0.09818, + "grad_norm": 0.9383963346481323, + "learning_rate": 0.003, + "loss": 4.045, + "step": 9818 + }, + { + "epoch": 0.09819, + "grad_norm": 1.1323498487472534, + "learning_rate": 0.003, + "loss": 4.05, + "step": 9819 + }, + { + "epoch": 0.0982, + "grad_norm": 0.9455949068069458, + "learning_rate": 0.003, + "loss": 4.028, + "step": 9820 + }, + { + "epoch": 0.09821, + "grad_norm": 0.7970534563064575, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 9821 + }, + { + "epoch": 0.09822, + "grad_norm": 0.7729085683822632, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 9822 + }, + { + "epoch": 0.09823, + "grad_norm": 0.7592610716819763, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 9823 + }, + { + "epoch": 0.09824, + "grad_norm": 0.8357574343681335, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 9824 + }, + { + "epoch": 0.09825, + "grad_norm": 0.9557722210884094, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 9825 + }, + { + "epoch": 0.09826, + "grad_norm": 1.0662164688110352, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 9826 + }, + { + "epoch": 0.09827, + "grad_norm": 1.1117784976959229, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 9827 + }, + { + "epoch": 0.09828, + "grad_norm": 0.6650376319885254, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 9828 + }, + { + "epoch": 0.09829, + "grad_norm": 0.6073980331420898, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 9829 + }, + { + "epoch": 0.0983, + "grad_norm": 0.6785967350006104, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 9830 + }, + { + "epoch": 0.09831, + "grad_norm": 0.8099671006202698, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 9831 + }, + { + "epoch": 0.09832, + "grad_norm": 0.8946943283081055, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 9832 + }, + { + "epoch": 0.09833, + "grad_norm": 0.9274506568908691, + "learning_rate": 0.003, + "loss": 4.058, + "step": 9833 + }, + { + "epoch": 0.09834, + "grad_norm": 0.7214897274971008, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 9834 + }, + { + "epoch": 0.09835, + "grad_norm": 0.5876917839050293, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 9835 + }, + { + "epoch": 0.09836, + "grad_norm": 0.585267186164856, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 9836 + }, + { + "epoch": 0.09837, + "grad_norm": 0.6745166778564453, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 9837 + }, + { + "epoch": 0.09838, + "grad_norm": 0.7141098380088806, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 9838 + }, + { + "epoch": 0.09839, + "grad_norm": 0.7113993167877197, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 9839 + }, + { + "epoch": 0.0984, + "grad_norm": 0.7230168581008911, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 9840 + }, + { + "epoch": 0.09841, + "grad_norm": 0.6224821209907532, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 9841 + }, + { + "epoch": 0.09842, + "grad_norm": 0.6236535906791687, + "learning_rate": 0.003, + "loss": 4.015, + "step": 9842 + }, + { + "epoch": 0.09843, + "grad_norm": 0.5319640040397644, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 9843 + }, + { + "epoch": 0.09844, + "grad_norm": 0.5817659497261047, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 9844 + }, + { + "epoch": 0.09845, + "grad_norm": 0.6393766403198242, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 9845 + }, + { + "epoch": 0.09846, + "grad_norm": 0.6582833528518677, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 9846 + }, + { + "epoch": 0.09847, + "grad_norm": 0.6721013188362122, + "learning_rate": 0.003, + "loss": 4.029, + "step": 9847 + }, + { + "epoch": 0.09848, + "grad_norm": 0.6297866702079773, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 9848 + }, + { + "epoch": 0.09849, + "grad_norm": 0.7541521787643433, + "learning_rate": 0.003, + "loss": 4.018, + "step": 9849 + }, + { + "epoch": 0.0985, + "grad_norm": 0.9827596545219421, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 9850 + }, + { + "epoch": 0.09851, + "grad_norm": 1.3056297302246094, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 9851 + }, + { + "epoch": 0.09852, + "grad_norm": 0.595446765422821, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 9852 + }, + { + "epoch": 0.09853, + "grad_norm": 0.6285281181335449, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 9853 + }, + { + "epoch": 0.09854, + "grad_norm": 0.7268086671829224, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 9854 + }, + { + "epoch": 0.09855, + "grad_norm": 0.7416141033172607, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 9855 + }, + { + "epoch": 0.09856, + "grad_norm": 0.8398358225822449, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 9856 + }, + { + "epoch": 0.09857, + "grad_norm": 0.87022465467453, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 9857 + }, + { + "epoch": 0.09858, + "grad_norm": 0.9862394332885742, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 9858 + }, + { + "epoch": 0.09859, + "grad_norm": 1.0218671560287476, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 9859 + }, + { + "epoch": 0.0986, + "grad_norm": 0.9830066561698914, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 9860 + }, + { + "epoch": 0.09861, + "grad_norm": 0.9263790845870972, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 9861 + }, + { + "epoch": 0.09862, + "grad_norm": 0.9523604512214661, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 9862 + }, + { + "epoch": 0.09863, + "grad_norm": 0.992346465587616, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 9863 + }, + { + "epoch": 0.09864, + "grad_norm": 1.2398070096969604, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 9864 + }, + { + "epoch": 0.09865, + "grad_norm": 0.8953294157981873, + "learning_rate": 0.003, + "loss": 4.057, + "step": 9865 + }, + { + "epoch": 0.09866, + "grad_norm": 0.9769341349601746, + "learning_rate": 0.003, + "loss": 4.048, + "step": 9866 + }, + { + "epoch": 0.09867, + "grad_norm": 1.0070929527282715, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 9867 + }, + { + "epoch": 0.09868, + "grad_norm": 0.9202451705932617, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 9868 + }, + { + "epoch": 0.09869, + "grad_norm": 1.0466346740722656, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 9869 + }, + { + "epoch": 0.0987, + "grad_norm": 1.0406720638275146, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 9870 + }, + { + "epoch": 0.09871, + "grad_norm": 0.9774965047836304, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 9871 + }, + { + "epoch": 0.09872, + "grad_norm": 1.0072287321090698, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 9872 + }, + { + "epoch": 0.09873, + "grad_norm": 1.0737255811691284, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 9873 + }, + { + "epoch": 0.09874, + "grad_norm": 0.970801591873169, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 9874 + }, + { + "epoch": 0.09875, + "grad_norm": 1.119269847869873, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 9875 + }, + { + "epoch": 0.09876, + "grad_norm": 0.9882863759994507, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 9876 + }, + { + "epoch": 0.09877, + "grad_norm": 0.8598633408546448, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 9877 + }, + { + "epoch": 0.09878, + "grad_norm": 0.8976407647132874, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 9878 + }, + { + "epoch": 0.09879, + "grad_norm": 1.0029267072677612, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 9879 + }, + { + "epoch": 0.0988, + "grad_norm": 1.0229839086532593, + "learning_rate": 0.003, + "loss": 4.037, + "step": 9880 + }, + { + "epoch": 0.09881, + "grad_norm": 1.1409919261932373, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 9881 + }, + { + "epoch": 0.09882, + "grad_norm": 0.9878159165382385, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 9882 + }, + { + "epoch": 0.09883, + "grad_norm": 0.8589847087860107, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 9883 + }, + { + "epoch": 0.09884, + "grad_norm": 0.645190417766571, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 9884 + }, + { + "epoch": 0.09885, + "grad_norm": 0.5982037782669067, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 9885 + }, + { + "epoch": 0.09886, + "grad_norm": 0.5794589519500732, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 9886 + }, + { + "epoch": 0.09887, + "grad_norm": 0.622081995010376, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 9887 + }, + { + "epoch": 0.09888, + "grad_norm": 0.6355760097503662, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 9888 + }, + { + "epoch": 0.09889, + "grad_norm": 0.5906744003295898, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 9889 + }, + { + "epoch": 0.0989, + "grad_norm": 0.7298018336296082, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 9890 + }, + { + "epoch": 0.09891, + "grad_norm": 0.7977643609046936, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 9891 + }, + { + "epoch": 0.09892, + "grad_norm": 0.7684615254402161, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 9892 + }, + { + "epoch": 0.09893, + "grad_norm": 0.6070899963378906, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 9893 + }, + { + "epoch": 0.09894, + "grad_norm": 0.49417591094970703, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 9894 + }, + { + "epoch": 0.09895, + "grad_norm": 0.4215652644634247, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 9895 + }, + { + "epoch": 0.09896, + "grad_norm": 0.3874669373035431, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 9896 + }, + { + "epoch": 0.09897, + "grad_norm": 0.4470568895339966, + "learning_rate": 0.003, + "loss": 4.018, + "step": 9897 + }, + { + "epoch": 0.09898, + "grad_norm": 0.4585300087928772, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 9898 + }, + { + "epoch": 0.09899, + "grad_norm": 0.40234246850013733, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 9899 + }, + { + "epoch": 0.099, + "grad_norm": 0.4411037564277649, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 9900 + }, + { + "epoch": 0.09901, + "grad_norm": 0.4925287663936615, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 9901 + }, + { + "epoch": 0.09902, + "grad_norm": 0.5623918771743774, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 9902 + }, + { + "epoch": 0.09903, + "grad_norm": 0.6239189505577087, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 9903 + }, + { + "epoch": 0.09904, + "grad_norm": 0.6974774599075317, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 9904 + }, + { + "epoch": 0.09905, + "grad_norm": 0.847841739654541, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 9905 + }, + { + "epoch": 0.09906, + "grad_norm": 0.9315091967582703, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 9906 + }, + { + "epoch": 0.09907, + "grad_norm": 0.9649443626403809, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 9907 + }, + { + "epoch": 0.09908, + "grad_norm": 1.1252580881118774, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 9908 + }, + { + "epoch": 0.09909, + "grad_norm": 0.7856508493423462, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 9909 + }, + { + "epoch": 0.0991, + "grad_norm": 0.6007184386253357, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 9910 + }, + { + "epoch": 0.09911, + "grad_norm": 0.5230900645256042, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 9911 + }, + { + "epoch": 0.09912, + "grad_norm": 0.6029075384140015, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 9912 + }, + { + "epoch": 0.09913, + "grad_norm": 0.6787557601928711, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 9913 + }, + { + "epoch": 0.09914, + "grad_norm": 0.7057299017906189, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 9914 + }, + { + "epoch": 0.09915, + "grad_norm": 0.8509723544120789, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 9915 + }, + { + "epoch": 0.09916, + "grad_norm": 1.0318037271499634, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 9916 + }, + { + "epoch": 0.09917, + "grad_norm": 0.9847546219825745, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 9917 + }, + { + "epoch": 0.09918, + "grad_norm": 0.8432486653327942, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 9918 + }, + { + "epoch": 0.09919, + "grad_norm": 0.7586849331855774, + "learning_rate": 0.003, + "loss": 4.057, + "step": 9919 + }, + { + "epoch": 0.0992, + "grad_norm": 0.7176387906074524, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 9920 + }, + { + "epoch": 0.09921, + "grad_norm": 0.6509994864463806, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 9921 + }, + { + "epoch": 0.09922, + "grad_norm": 0.7130661606788635, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 9922 + }, + { + "epoch": 0.09923, + "grad_norm": 0.9413552284240723, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 9923 + }, + { + "epoch": 0.09924, + "grad_norm": 1.1508615016937256, + "learning_rate": 0.003, + "loss": 4.002, + "step": 9924 + }, + { + "epoch": 0.09925, + "grad_norm": 0.6644554138183594, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 9925 + }, + { + "epoch": 0.09926, + "grad_norm": 0.5794418454170227, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 9926 + }, + { + "epoch": 0.09927, + "grad_norm": 0.5938196778297424, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 9927 + }, + { + "epoch": 0.09928, + "grad_norm": 0.5763348937034607, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 9928 + }, + { + "epoch": 0.09929, + "grad_norm": 0.6028631329536438, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 9929 + }, + { + "epoch": 0.0993, + "grad_norm": 0.6171523928642273, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 9930 + }, + { + "epoch": 0.09931, + "grad_norm": 0.7361319065093994, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 9931 + }, + { + "epoch": 0.09932, + "grad_norm": 0.7786489725112915, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 9932 + }, + { + "epoch": 0.09933, + "grad_norm": 0.7424275279045105, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 9933 + }, + { + "epoch": 0.09934, + "grad_norm": 0.6433824300765991, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 9934 + }, + { + "epoch": 0.09935, + "grad_norm": 0.6526421308517456, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 9935 + }, + { + "epoch": 0.09936, + "grad_norm": 0.5947155952453613, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 9936 + }, + { + "epoch": 0.09937, + "grad_norm": 0.6477751731872559, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 9937 + }, + { + "epoch": 0.09938, + "grad_norm": 0.845565676689148, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 9938 + }, + { + "epoch": 0.09939, + "grad_norm": 1.120976448059082, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 9939 + }, + { + "epoch": 0.0994, + "grad_norm": 0.9667741656303406, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 9940 + }, + { + "epoch": 0.09941, + "grad_norm": 0.7664325833320618, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 9941 + }, + { + "epoch": 0.09942, + "grad_norm": 0.6430301666259766, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 9942 + }, + { + "epoch": 0.09943, + "grad_norm": 0.6047747135162354, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 9943 + }, + { + "epoch": 0.09944, + "grad_norm": 0.582067608833313, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 9944 + }, + { + "epoch": 0.09945, + "grad_norm": 0.8082442283630371, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 9945 + }, + { + "epoch": 0.09946, + "grad_norm": 1.1517560482025146, + "learning_rate": 0.003, + "loss": 4.05, + "step": 9946 + }, + { + "epoch": 0.09947, + "grad_norm": 0.9363013505935669, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 9947 + }, + { + "epoch": 0.09948, + "grad_norm": 0.7373224496841431, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 9948 + }, + { + "epoch": 0.09949, + "grad_norm": 0.6157240867614746, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 9949 + }, + { + "epoch": 0.0995, + "grad_norm": 0.6930040121078491, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 9950 + }, + { + "epoch": 0.09951, + "grad_norm": 0.7911564111709595, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 9951 + }, + { + "epoch": 0.09952, + "grad_norm": 0.8492519855499268, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 9952 + }, + { + "epoch": 0.09953, + "grad_norm": 0.8654104471206665, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 9953 + }, + { + "epoch": 0.09954, + "grad_norm": 0.9337427020072937, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 9954 + }, + { + "epoch": 0.09955, + "grad_norm": 1.0466641187667847, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 9955 + }, + { + "epoch": 0.09956, + "grad_norm": 0.9967005848884583, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 9956 + }, + { + "epoch": 0.09957, + "grad_norm": 1.014816164970398, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 9957 + }, + { + "epoch": 0.09958, + "grad_norm": 0.9045199155807495, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 9958 + }, + { + "epoch": 0.09959, + "grad_norm": 0.9327095746994019, + "learning_rate": 0.003, + "loss": 4.06, + "step": 9959 + }, + { + "epoch": 0.0996, + "grad_norm": 0.9663153886795044, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 9960 + }, + { + "epoch": 0.09961, + "grad_norm": 0.983515739440918, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 9961 + }, + { + "epoch": 0.09962, + "grad_norm": 1.0190386772155762, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 9962 + }, + { + "epoch": 0.09963, + "grad_norm": 0.8583796620368958, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 9963 + }, + { + "epoch": 0.09964, + "grad_norm": 0.7809299826622009, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 9964 + }, + { + "epoch": 0.09965, + "grad_norm": 0.8016514778137207, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 9965 + }, + { + "epoch": 0.09966, + "grad_norm": 0.743135929107666, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 9966 + }, + { + "epoch": 0.09967, + "grad_norm": 0.7582171559333801, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 9967 + }, + { + "epoch": 0.09968, + "grad_norm": 0.6683290004730225, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 9968 + }, + { + "epoch": 0.09969, + "grad_norm": 0.6653891205787659, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 9969 + }, + { + "epoch": 0.0997, + "grad_norm": 0.6902685761451721, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 9970 + }, + { + "epoch": 0.09971, + "grad_norm": 1.0166124105453491, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 9971 + }, + { + "epoch": 0.09972, + "grad_norm": 1.0039961338043213, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 9972 + }, + { + "epoch": 0.09973, + "grad_norm": 0.9374507665634155, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 9973 + }, + { + "epoch": 0.09974, + "grad_norm": 0.8050880432128906, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 9974 + }, + { + "epoch": 0.09975, + "grad_norm": 0.6863667964935303, + "learning_rate": 0.003, + "loss": 4.042, + "step": 9975 + }, + { + "epoch": 0.09976, + "grad_norm": 0.5954990983009338, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 9976 + }, + { + "epoch": 0.09977, + "grad_norm": 0.7258602976799011, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 9977 + }, + { + "epoch": 0.09978, + "grad_norm": 0.8950538039207458, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 9978 + }, + { + "epoch": 0.09979, + "grad_norm": 1.0293006896972656, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 9979 + }, + { + "epoch": 0.0998, + "grad_norm": 0.9026445746421814, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 9980 + }, + { + "epoch": 0.09981, + "grad_norm": 0.7071680426597595, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 9981 + }, + { + "epoch": 0.09982, + "grad_norm": 0.5843554735183716, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 9982 + }, + { + "epoch": 0.09983, + "grad_norm": 0.585106611251831, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 9983 + }, + { + "epoch": 0.09984, + "grad_norm": 0.600675106048584, + "learning_rate": 0.003, + "loss": 4.045, + "step": 9984 + }, + { + "epoch": 0.09985, + "grad_norm": 0.6129768490791321, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 9985 + }, + { + "epoch": 0.09986, + "grad_norm": 0.6755698323249817, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 9986 + }, + { + "epoch": 0.09987, + "grad_norm": 0.7202430367469788, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 9987 + }, + { + "epoch": 0.09988, + "grad_norm": 0.8828611969947815, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 9988 + }, + { + "epoch": 0.09989, + "grad_norm": 0.9760780930519104, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 9989 + }, + { + "epoch": 0.0999, + "grad_norm": 0.8992119431495667, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 9990 + }, + { + "epoch": 0.09991, + "grad_norm": 0.730566680431366, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 9991 + }, + { + "epoch": 0.09992, + "grad_norm": 0.6713852882385254, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 9992 + }, + { + "epoch": 0.09993, + "grad_norm": 0.6310202479362488, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 9993 + }, + { + "epoch": 0.09994, + "grad_norm": 0.5760572552680969, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 9994 + }, + { + "epoch": 0.09995, + "grad_norm": 0.5813117027282715, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 9995 + }, + { + "epoch": 0.09996, + "grad_norm": 0.6562214493751526, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 9996 + }, + { + "epoch": 0.09997, + "grad_norm": 0.6452088952064514, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 9997 + }, + { + "epoch": 0.09998, + "grad_norm": 0.6939143538475037, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 9998 + }, + { + "epoch": 0.09999, + "grad_norm": 0.8035266399383545, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 9999 + }, + { + "epoch": 0.1, + "grad_norm": 0.8756668567657471, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 10000 + }, + { + "epoch": 0.10001, + "grad_norm": 0.8760703802108765, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 10001 + }, + { + "epoch": 0.10002, + "grad_norm": 0.7529221177101135, + "learning_rate": 0.003, + "loss": 4.042, + "step": 10002 + }, + { + "epoch": 0.10003, + "grad_norm": 0.7163582444190979, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 10003 + }, + { + "epoch": 0.10004, + "grad_norm": 0.8936163783073425, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 10004 + }, + { + "epoch": 0.10005, + "grad_norm": 1.0073555707931519, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 10005 + }, + { + "epoch": 0.10006, + "grad_norm": 0.917222797870636, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 10006 + }, + { + "epoch": 0.10007, + "grad_norm": 0.8444756269454956, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 10007 + }, + { + "epoch": 0.10008, + "grad_norm": 0.7267651557922363, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 10008 + }, + { + "epoch": 0.10009, + "grad_norm": 0.7675696015357971, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 10009 + }, + { + "epoch": 0.1001, + "grad_norm": 0.7582693696022034, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 10010 + }, + { + "epoch": 0.10011, + "grad_norm": 0.6846451163291931, + "learning_rate": 0.003, + "loss": 4.028, + "step": 10011 + }, + { + "epoch": 0.10012, + "grad_norm": 0.729497492313385, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 10012 + }, + { + "epoch": 0.10013, + "grad_norm": 0.856055498123169, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 10013 + }, + { + "epoch": 0.10014, + "grad_norm": 0.9490603804588318, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 10014 + }, + { + "epoch": 0.10015, + "grad_norm": 0.8758609890937805, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 10015 + }, + { + "epoch": 0.10016, + "grad_norm": 0.8611948490142822, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 10016 + }, + { + "epoch": 0.10017, + "grad_norm": 0.7740161418914795, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 10017 + }, + { + "epoch": 0.10018, + "grad_norm": 0.6666895151138306, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 10018 + }, + { + "epoch": 0.10019, + "grad_norm": 0.6955499053001404, + "learning_rate": 0.003, + "loss": 4.035, + "step": 10019 + }, + { + "epoch": 0.1002, + "grad_norm": 0.6729419827461243, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 10020 + }, + { + "epoch": 0.10021, + "grad_norm": 0.7229282855987549, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 10021 + }, + { + "epoch": 0.10022, + "grad_norm": 0.8583958745002747, + "learning_rate": 0.003, + "loss": 4.026, + "step": 10022 + }, + { + "epoch": 0.10023, + "grad_norm": 0.8976644277572632, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 10023 + }, + { + "epoch": 0.10024, + "grad_norm": 0.8202301859855652, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 10024 + }, + { + "epoch": 0.10025, + "grad_norm": 0.7559810280799866, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 10025 + }, + { + "epoch": 0.10026, + "grad_norm": 0.7426246404647827, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 10026 + }, + { + "epoch": 0.10027, + "grad_norm": 0.7772095203399658, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 10027 + }, + { + "epoch": 0.10028, + "grad_norm": 0.7090165615081787, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 10028 + }, + { + "epoch": 0.10029, + "grad_norm": 0.751837432384491, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 10029 + }, + { + "epoch": 0.1003, + "grad_norm": 0.7565678358078003, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 10030 + }, + { + "epoch": 0.10031, + "grad_norm": 0.723259687423706, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 10031 + }, + { + "epoch": 0.10032, + "grad_norm": 0.6929975748062134, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 10032 + }, + { + "epoch": 0.10033, + "grad_norm": 0.7323466539382935, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 10033 + }, + { + "epoch": 0.10034, + "grad_norm": 0.7538563013076782, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 10034 + }, + { + "epoch": 0.10035, + "grad_norm": 0.7490745782852173, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 10035 + }, + { + "epoch": 0.10036, + "grad_norm": 0.7792319655418396, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 10036 + }, + { + "epoch": 0.10037, + "grad_norm": 0.82288658618927, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 10037 + }, + { + "epoch": 0.10038, + "grad_norm": 0.8214668035507202, + "learning_rate": 0.003, + "loss": 4.051, + "step": 10038 + }, + { + "epoch": 0.10039, + "grad_norm": 0.8127548098564148, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 10039 + }, + { + "epoch": 0.1004, + "grad_norm": 0.9639657139778137, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 10040 + }, + { + "epoch": 0.10041, + "grad_norm": 1.0831257104873657, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 10041 + }, + { + "epoch": 0.10042, + "grad_norm": 0.8846466541290283, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 10042 + }, + { + "epoch": 0.10043, + "grad_norm": 0.8289374709129333, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 10043 + }, + { + "epoch": 0.10044, + "grad_norm": 0.7353184223175049, + "learning_rate": 0.003, + "loss": 4.041, + "step": 10044 + }, + { + "epoch": 0.10045, + "grad_norm": 0.7047366499900818, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 10045 + }, + { + "epoch": 0.10046, + "grad_norm": 0.6355120539665222, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 10046 + }, + { + "epoch": 0.10047, + "grad_norm": 0.6405349969863892, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 10047 + }, + { + "epoch": 0.10048, + "grad_norm": 0.5647373795509338, + "learning_rate": 0.003, + "loss": 4.029, + "step": 10048 + }, + { + "epoch": 0.10049, + "grad_norm": 0.5304549336433411, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 10049 + }, + { + "epoch": 0.1005, + "grad_norm": 0.5612930059432983, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 10050 + }, + { + "epoch": 0.10051, + "grad_norm": 0.6478097438812256, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 10051 + }, + { + "epoch": 0.10052, + "grad_norm": 0.7477694153785706, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 10052 + }, + { + "epoch": 0.10053, + "grad_norm": 0.8523183465003967, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 10053 + }, + { + "epoch": 0.10054, + "grad_norm": 0.9100677371025085, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 10054 + }, + { + "epoch": 0.10055, + "grad_norm": 0.969436764717102, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 10055 + }, + { + "epoch": 0.10056, + "grad_norm": 0.8593199849128723, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 10056 + }, + { + "epoch": 0.10057, + "grad_norm": 0.7369247078895569, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 10057 + }, + { + "epoch": 0.10058, + "grad_norm": 0.6502119898796082, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 10058 + }, + { + "epoch": 0.10059, + "grad_norm": 0.6754664182662964, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 10059 + }, + { + "epoch": 0.1006, + "grad_norm": 0.7896672487258911, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 10060 + }, + { + "epoch": 0.10061, + "grad_norm": 0.7812275886535645, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 10061 + }, + { + "epoch": 0.10062, + "grad_norm": 0.7402509450912476, + "learning_rate": 0.003, + "loss": 4.006, + "step": 10062 + }, + { + "epoch": 0.10063, + "grad_norm": 0.7005326747894287, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 10063 + }, + { + "epoch": 0.10064, + "grad_norm": 0.6713504791259766, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 10064 + }, + { + "epoch": 0.10065, + "grad_norm": 0.6878002285957336, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 10065 + }, + { + "epoch": 0.10066, + "grad_norm": 0.7443453669548035, + "learning_rate": 0.003, + "loss": 4.018, + "step": 10066 + }, + { + "epoch": 0.10067, + "grad_norm": 0.8456590175628662, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 10067 + }, + { + "epoch": 0.10068, + "grad_norm": 1.0721101760864258, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 10068 + }, + { + "epoch": 0.10069, + "grad_norm": 0.9349250197410583, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 10069 + }, + { + "epoch": 0.1007, + "grad_norm": 0.7791523933410645, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 10070 + }, + { + "epoch": 0.10071, + "grad_norm": 0.643773078918457, + "learning_rate": 0.003, + "loss": 4.022, + "step": 10071 + }, + { + "epoch": 0.10072, + "grad_norm": 0.5889289379119873, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 10072 + }, + { + "epoch": 0.10073, + "grad_norm": 0.6941524147987366, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 10073 + }, + { + "epoch": 0.10074, + "grad_norm": 0.812353253364563, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 10074 + }, + { + "epoch": 0.10075, + "grad_norm": 0.848740816116333, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 10075 + }, + { + "epoch": 0.10076, + "grad_norm": 0.8380178213119507, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 10076 + }, + { + "epoch": 0.10077, + "grad_norm": 0.8449537754058838, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 10077 + }, + { + "epoch": 0.10078, + "grad_norm": 0.8786484599113464, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 10078 + }, + { + "epoch": 0.10079, + "grad_norm": 0.7593846321105957, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 10079 + }, + { + "epoch": 0.1008, + "grad_norm": 0.7023867964744568, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 10080 + }, + { + "epoch": 0.10081, + "grad_norm": 0.7081405520439148, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 10081 + }, + { + "epoch": 0.10082, + "grad_norm": 0.6782942414283752, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 10082 + }, + { + "epoch": 0.10083, + "grad_norm": 0.7835003137588501, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 10083 + }, + { + "epoch": 0.10084, + "grad_norm": 1.1654982566833496, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 10084 + }, + { + "epoch": 0.10085, + "grad_norm": 1.077035665512085, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 10085 + }, + { + "epoch": 0.10086, + "grad_norm": 0.7526758313179016, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 10086 + }, + { + "epoch": 0.10087, + "grad_norm": 0.572903037071228, + "learning_rate": 0.003, + "loss": 4.042, + "step": 10087 + }, + { + "epoch": 0.10088, + "grad_norm": 0.5392580032348633, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 10088 + }, + { + "epoch": 0.10089, + "grad_norm": 0.5970590710639954, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 10089 + }, + { + "epoch": 0.1009, + "grad_norm": 0.6526892185211182, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 10090 + }, + { + "epoch": 0.10091, + "grad_norm": 0.7266455888748169, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 10091 + }, + { + "epoch": 0.10092, + "grad_norm": 0.7619020342826843, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 10092 + }, + { + "epoch": 0.10093, + "grad_norm": 0.7967933416366577, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 10093 + }, + { + "epoch": 0.10094, + "grad_norm": 0.8304819464683533, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 10094 + }, + { + "epoch": 0.10095, + "grad_norm": 0.7637600898742676, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 10095 + }, + { + "epoch": 0.10096, + "grad_norm": 0.6716521382331848, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 10096 + }, + { + "epoch": 0.10097, + "grad_norm": 0.7301881909370422, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 10097 + }, + { + "epoch": 0.10098, + "grad_norm": 0.8763124942779541, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 10098 + }, + { + "epoch": 0.10099, + "grad_norm": 0.9246903657913208, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 10099 + }, + { + "epoch": 0.101, + "grad_norm": 0.852162778377533, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 10100 + }, + { + "epoch": 0.10101, + "grad_norm": 0.7550850510597229, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 10101 + }, + { + "epoch": 0.10102, + "grad_norm": 0.7212772369384766, + "learning_rate": 0.003, + "loss": 4.045, + "step": 10102 + }, + { + "epoch": 0.10103, + "grad_norm": 0.749382734298706, + "learning_rate": 0.003, + "loss": 4.006, + "step": 10103 + }, + { + "epoch": 0.10104, + "grad_norm": 0.8882433772087097, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 10104 + }, + { + "epoch": 0.10105, + "grad_norm": 1.2675986289978027, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 10105 + }, + { + "epoch": 0.10106, + "grad_norm": 0.8278036117553711, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 10106 + }, + { + "epoch": 0.10107, + "grad_norm": 0.7198725342750549, + "learning_rate": 0.003, + "loss": 4.032, + "step": 10107 + }, + { + "epoch": 0.10108, + "grad_norm": 0.6387250423431396, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 10108 + }, + { + "epoch": 0.10109, + "grad_norm": 0.5777958631515503, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 10109 + }, + { + "epoch": 0.1011, + "grad_norm": 0.6204786896705627, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 10110 + }, + { + "epoch": 0.10111, + "grad_norm": 0.5600338578224182, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 10111 + }, + { + "epoch": 0.10112, + "grad_norm": 0.6853824853897095, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 10112 + }, + { + "epoch": 0.10113, + "grad_norm": 0.8152238130569458, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 10113 + }, + { + "epoch": 0.10114, + "grad_norm": 0.9304043054580688, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 10114 + }, + { + "epoch": 0.10115, + "grad_norm": 1.061358094215393, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 10115 + }, + { + "epoch": 0.10116, + "grad_norm": 0.9930733442306519, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 10116 + }, + { + "epoch": 0.10117, + "grad_norm": 0.8263692855834961, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 10117 + }, + { + "epoch": 0.10118, + "grad_norm": 0.6010589599609375, + "learning_rate": 0.003, + "loss": 4.049, + "step": 10118 + }, + { + "epoch": 0.10119, + "grad_norm": 0.7774295210838318, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 10119 + }, + { + "epoch": 0.1012, + "grad_norm": 0.9463669061660767, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 10120 + }, + { + "epoch": 0.10121, + "grad_norm": 1.0663201808929443, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 10121 + }, + { + "epoch": 0.10122, + "grad_norm": 0.7437269687652588, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 10122 + }, + { + "epoch": 0.10123, + "grad_norm": 0.5977786779403687, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 10123 + }, + { + "epoch": 0.10124, + "grad_norm": 0.8191568851470947, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 10124 + }, + { + "epoch": 0.10125, + "grad_norm": 0.847891628742218, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 10125 + }, + { + "epoch": 0.10126, + "grad_norm": 0.9732335805892944, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 10126 + }, + { + "epoch": 0.10127, + "grad_norm": 1.0408259630203247, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 10127 + }, + { + "epoch": 0.10128, + "grad_norm": 0.8466975092887878, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 10128 + }, + { + "epoch": 0.10129, + "grad_norm": 0.8226129412651062, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 10129 + }, + { + "epoch": 0.1013, + "grad_norm": 0.7009112238883972, + "learning_rate": 0.003, + "loss": 4.002, + "step": 10130 + }, + { + "epoch": 0.10131, + "grad_norm": 0.6498497724533081, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 10131 + }, + { + "epoch": 0.10132, + "grad_norm": 0.718295693397522, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 10132 + }, + { + "epoch": 0.10133, + "grad_norm": 0.8029374480247498, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 10133 + }, + { + "epoch": 0.10134, + "grad_norm": 0.7989256381988525, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 10134 + }, + { + "epoch": 0.10135, + "grad_norm": 0.7786409854888916, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 10135 + }, + { + "epoch": 0.10136, + "grad_norm": 0.8297443985939026, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 10136 + }, + { + "epoch": 0.10137, + "grad_norm": 0.8420387506484985, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 10137 + }, + { + "epoch": 0.10138, + "grad_norm": 0.7902016043663025, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 10138 + }, + { + "epoch": 0.10139, + "grad_norm": 0.6987302899360657, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 10139 + }, + { + "epoch": 0.1014, + "grad_norm": 0.7255557179450989, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 10140 + }, + { + "epoch": 0.10141, + "grad_norm": 0.8124567866325378, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 10141 + }, + { + "epoch": 0.10142, + "grad_norm": 0.8950157761573792, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 10142 + }, + { + "epoch": 0.10143, + "grad_norm": 1.0962830781936646, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 10143 + }, + { + "epoch": 0.10144, + "grad_norm": 0.9343271255493164, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 10144 + }, + { + "epoch": 0.10145, + "grad_norm": 0.8869456052780151, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 10145 + }, + { + "epoch": 0.10146, + "grad_norm": 0.875791609287262, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 10146 + }, + { + "epoch": 0.10147, + "grad_norm": 0.7010773420333862, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 10147 + }, + { + "epoch": 0.10148, + "grad_norm": 0.6326318383216858, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 10148 + }, + { + "epoch": 0.10149, + "grad_norm": 0.6735743284225464, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 10149 + }, + { + "epoch": 0.1015, + "grad_norm": 0.6270667910575867, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 10150 + }, + { + "epoch": 0.10151, + "grad_norm": 0.7168855667114258, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 10151 + }, + { + "epoch": 0.10152, + "grad_norm": 0.7000669836997986, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 10152 + }, + { + "epoch": 0.10153, + "grad_norm": 0.7381427884101868, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 10153 + }, + { + "epoch": 0.10154, + "grad_norm": 0.7226046919822693, + "learning_rate": 0.003, + "loss": 4.038, + "step": 10154 + }, + { + "epoch": 0.10155, + "grad_norm": 0.653364896774292, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 10155 + }, + { + "epoch": 0.10156, + "grad_norm": 0.7581577301025391, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 10156 + }, + { + "epoch": 0.10157, + "grad_norm": 0.7567413449287415, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 10157 + }, + { + "epoch": 0.10158, + "grad_norm": 0.6570749282836914, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 10158 + }, + { + "epoch": 0.10159, + "grad_norm": 0.8515116572380066, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 10159 + }, + { + "epoch": 0.1016, + "grad_norm": 0.830341637134552, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 10160 + }, + { + "epoch": 0.10161, + "grad_norm": 0.6846141219139099, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 10161 + }, + { + "epoch": 0.10162, + "grad_norm": 0.7712964415550232, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 10162 + }, + { + "epoch": 0.10163, + "grad_norm": 0.8644002079963684, + "learning_rate": 0.003, + "loss": 4.03, + "step": 10163 + }, + { + "epoch": 0.10164, + "grad_norm": 1.1422466039657593, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 10164 + }, + { + "epoch": 0.10165, + "grad_norm": 0.9871522188186646, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 10165 + }, + { + "epoch": 0.10166, + "grad_norm": 1.0399481058120728, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 10166 + }, + { + "epoch": 0.10167, + "grad_norm": 0.9927687644958496, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 10167 + }, + { + "epoch": 0.10168, + "grad_norm": 0.9502347111701965, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 10168 + }, + { + "epoch": 0.10169, + "grad_norm": 0.784534215927124, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 10169 + }, + { + "epoch": 0.1017, + "grad_norm": 0.6386174559593201, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 10170 + }, + { + "epoch": 0.10171, + "grad_norm": 0.5796541571617126, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 10171 + }, + { + "epoch": 0.10172, + "grad_norm": 0.5448047518730164, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 10172 + }, + { + "epoch": 0.10173, + "grad_norm": 0.5247647166252136, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 10173 + }, + { + "epoch": 0.10174, + "grad_norm": 0.6259518265724182, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 10174 + }, + { + "epoch": 0.10175, + "grad_norm": 0.6033033132553101, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 10175 + }, + { + "epoch": 0.10176, + "grad_norm": 0.5712244510650635, + "learning_rate": 0.003, + "loss": 4.022, + "step": 10176 + }, + { + "epoch": 0.10177, + "grad_norm": 0.515041172504425, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 10177 + }, + { + "epoch": 0.10178, + "grad_norm": 0.45743808150291443, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 10178 + }, + { + "epoch": 0.10179, + "grad_norm": 0.42596837878227234, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 10179 + }, + { + "epoch": 0.1018, + "grad_norm": 0.47992241382598877, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 10180 + }, + { + "epoch": 0.10181, + "grad_norm": 0.5535048246383667, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 10181 + }, + { + "epoch": 0.10182, + "grad_norm": 0.7175001502037048, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 10182 + }, + { + "epoch": 0.10183, + "grad_norm": 0.9698479175567627, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 10183 + }, + { + "epoch": 0.10184, + "grad_norm": 1.2325149774551392, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 10184 + }, + { + "epoch": 0.10185, + "grad_norm": 0.6251413226127625, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 10185 + }, + { + "epoch": 0.10186, + "grad_norm": 0.6279639005661011, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 10186 + }, + { + "epoch": 0.10187, + "grad_norm": 0.7553200721740723, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 10187 + }, + { + "epoch": 0.10188, + "grad_norm": 0.9480890035629272, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 10188 + }, + { + "epoch": 0.10189, + "grad_norm": 1.0561100244522095, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 10189 + }, + { + "epoch": 0.1019, + "grad_norm": 0.9566508531570435, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 10190 + }, + { + "epoch": 0.10191, + "grad_norm": 0.9126545786857605, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 10191 + }, + { + "epoch": 0.10192, + "grad_norm": 0.8582307696342468, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 10192 + }, + { + "epoch": 0.10193, + "grad_norm": 0.859767735004425, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 10193 + }, + { + "epoch": 0.10194, + "grad_norm": 0.9116681218147278, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 10194 + }, + { + "epoch": 0.10195, + "grad_norm": 0.8132261037826538, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 10195 + }, + { + "epoch": 0.10196, + "grad_norm": 0.7543628215789795, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 10196 + }, + { + "epoch": 0.10197, + "grad_norm": 0.7262082099914551, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 10197 + }, + { + "epoch": 0.10198, + "grad_norm": 0.6579393744468689, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 10198 + }, + { + "epoch": 0.10199, + "grad_norm": 0.5839563608169556, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 10199 + }, + { + "epoch": 0.102, + "grad_norm": 0.6411651968955994, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10200 + }, + { + "epoch": 0.10201, + "grad_norm": 0.8397372364997864, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 10201 + }, + { + "epoch": 0.10202, + "grad_norm": 1.038834810256958, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 10202 + }, + { + "epoch": 0.10203, + "grad_norm": 1.2775077819824219, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 10203 + }, + { + "epoch": 0.10204, + "grad_norm": 0.778706431388855, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 10204 + }, + { + "epoch": 0.10205, + "grad_norm": 0.7379663586616516, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 10205 + }, + { + "epoch": 0.10206, + "grad_norm": 0.7955054640769958, + "learning_rate": 0.003, + "loss": 4.043, + "step": 10206 + }, + { + "epoch": 0.10207, + "grad_norm": 0.8948112726211548, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 10207 + }, + { + "epoch": 0.10208, + "grad_norm": 1.1121752262115479, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 10208 + }, + { + "epoch": 0.10209, + "grad_norm": 1.0435898303985596, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 10209 + }, + { + "epoch": 0.1021, + "grad_norm": 0.8086763024330139, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 10210 + }, + { + "epoch": 0.10211, + "grad_norm": 0.796384334564209, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 10211 + }, + { + "epoch": 0.10212, + "grad_norm": 0.8457871079444885, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 10212 + }, + { + "epoch": 0.10213, + "grad_norm": 0.8126521110534668, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 10213 + }, + { + "epoch": 0.10214, + "grad_norm": 0.7107754945755005, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 10214 + }, + { + "epoch": 0.10215, + "grad_norm": 0.6939681172370911, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 10215 + }, + { + "epoch": 0.10216, + "grad_norm": 0.7338607907295227, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 10216 + }, + { + "epoch": 0.10217, + "grad_norm": 0.8102380037307739, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 10217 + }, + { + "epoch": 0.10218, + "grad_norm": 0.8121943473815918, + "learning_rate": 0.003, + "loss": 4.059, + "step": 10218 + }, + { + "epoch": 0.10219, + "grad_norm": 0.7863835692405701, + "learning_rate": 0.003, + "loss": 4.001, + "step": 10219 + }, + { + "epoch": 0.1022, + "grad_norm": 0.6928033232688904, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 10220 + }, + { + "epoch": 0.10221, + "grad_norm": 0.675073504447937, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 10221 + }, + { + "epoch": 0.10222, + "grad_norm": 0.5772913098335266, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 10222 + }, + { + "epoch": 0.10223, + "grad_norm": 0.5414987802505493, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 10223 + }, + { + "epoch": 0.10224, + "grad_norm": 0.5829790830612183, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 10224 + }, + { + "epoch": 0.10225, + "grad_norm": 0.638757586479187, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 10225 + }, + { + "epoch": 0.10226, + "grad_norm": 0.6315435767173767, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 10226 + }, + { + "epoch": 0.10227, + "grad_norm": 0.7391036152839661, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 10227 + }, + { + "epoch": 0.10228, + "grad_norm": 0.8362815976142883, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 10228 + }, + { + "epoch": 0.10229, + "grad_norm": 1.0286784172058105, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 10229 + }, + { + "epoch": 0.1023, + "grad_norm": 1.009440541267395, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 10230 + }, + { + "epoch": 0.10231, + "grad_norm": 0.8905285000801086, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 10231 + }, + { + "epoch": 0.10232, + "grad_norm": 0.921527624130249, + "learning_rate": 0.003, + "loss": 4.046, + "step": 10232 + }, + { + "epoch": 0.10233, + "grad_norm": 0.8988191485404968, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 10233 + }, + { + "epoch": 0.10234, + "grad_norm": 0.7774051427841187, + "learning_rate": 0.003, + "loss": 4.05, + "step": 10234 + }, + { + "epoch": 0.10235, + "grad_norm": 0.7421836256980896, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 10235 + }, + { + "epoch": 0.10236, + "grad_norm": 0.7218093872070312, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 10236 + }, + { + "epoch": 0.10237, + "grad_norm": 0.6867312788963318, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 10237 + }, + { + "epoch": 0.10238, + "grad_norm": 0.6688364148139954, + "learning_rate": 0.003, + "loss": 4.019, + "step": 10238 + }, + { + "epoch": 0.10239, + "grad_norm": 0.6108981370925903, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 10239 + }, + { + "epoch": 0.1024, + "grad_norm": 0.6995348930358887, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 10240 + }, + { + "epoch": 0.10241, + "grad_norm": 0.8642280101776123, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 10241 + }, + { + "epoch": 0.10242, + "grad_norm": 1.1544699668884277, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 10242 + }, + { + "epoch": 0.10243, + "grad_norm": 0.9920125007629395, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 10243 + }, + { + "epoch": 0.10244, + "grad_norm": 0.9115767478942871, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 10244 + }, + { + "epoch": 0.10245, + "grad_norm": 0.7791694402694702, + "learning_rate": 0.003, + "loss": 4.011, + "step": 10245 + }, + { + "epoch": 0.10246, + "grad_norm": 0.7047367691993713, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 10246 + }, + { + "epoch": 0.10247, + "grad_norm": 0.6797937154769897, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 10247 + }, + { + "epoch": 0.10248, + "grad_norm": 0.7237415313720703, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 10248 + }, + { + "epoch": 0.10249, + "grad_norm": 0.8145940899848938, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 10249 + }, + { + "epoch": 0.1025, + "grad_norm": 0.7724632024765015, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 10250 + }, + { + "epoch": 0.10251, + "grad_norm": 0.7524970769882202, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 10251 + }, + { + "epoch": 0.10252, + "grad_norm": 0.8881193399429321, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 10252 + }, + { + "epoch": 0.10253, + "grad_norm": 1.0016326904296875, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 10253 + }, + { + "epoch": 0.10254, + "grad_norm": 1.0284184217453003, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 10254 + }, + { + "epoch": 0.10255, + "grad_norm": 0.7833191156387329, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 10255 + }, + { + "epoch": 0.10256, + "grad_norm": 0.6149386167526245, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 10256 + }, + { + "epoch": 0.10257, + "grad_norm": 0.5318538546562195, + "learning_rate": 0.003, + "loss": 4.01, + "step": 10257 + }, + { + "epoch": 0.10258, + "grad_norm": 0.6054723262786865, + "learning_rate": 0.003, + "loss": 4.02, + "step": 10258 + }, + { + "epoch": 0.10259, + "grad_norm": 0.690203845500946, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 10259 + }, + { + "epoch": 0.1026, + "grad_norm": 0.6764129400253296, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 10260 + }, + { + "epoch": 0.10261, + "grad_norm": 0.6417571902275085, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 10261 + }, + { + "epoch": 0.10262, + "grad_norm": 0.6889247894287109, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10262 + }, + { + "epoch": 0.10263, + "grad_norm": 0.8735169172286987, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 10263 + }, + { + "epoch": 0.10264, + "grad_norm": 1.0683711767196655, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 10264 + }, + { + "epoch": 0.10265, + "grad_norm": 1.0963785648345947, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 10265 + }, + { + "epoch": 0.10266, + "grad_norm": 0.9746417999267578, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 10266 + }, + { + "epoch": 0.10267, + "grad_norm": 0.9337226152420044, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 10267 + }, + { + "epoch": 0.10268, + "grad_norm": 0.8803682327270508, + "learning_rate": 0.003, + "loss": 4.018, + "step": 10268 + }, + { + "epoch": 0.10269, + "grad_norm": 0.8062136769294739, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 10269 + }, + { + "epoch": 0.1027, + "grad_norm": 0.6083574891090393, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 10270 + }, + { + "epoch": 0.10271, + "grad_norm": 0.5812376141548157, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 10271 + }, + { + "epoch": 0.10272, + "grad_norm": 0.5521930456161499, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 10272 + }, + { + "epoch": 0.10273, + "grad_norm": 0.528708815574646, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 10273 + }, + { + "epoch": 0.10274, + "grad_norm": 0.5105339288711548, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 10274 + }, + { + "epoch": 0.10275, + "grad_norm": 0.465678870677948, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 10275 + }, + { + "epoch": 0.10276, + "grad_norm": 0.49017584323883057, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 10276 + }, + { + "epoch": 0.10277, + "grad_norm": 0.5855206251144409, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 10277 + }, + { + "epoch": 0.10278, + "grad_norm": 0.5988587141036987, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 10278 + }, + { + "epoch": 0.10279, + "grad_norm": 0.625211238861084, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 10279 + }, + { + "epoch": 0.1028, + "grad_norm": 0.803665041923523, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 10280 + }, + { + "epoch": 0.10281, + "grad_norm": 0.9801792502403259, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 10281 + }, + { + "epoch": 0.10282, + "grad_norm": 1.018873691558838, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 10282 + }, + { + "epoch": 0.10283, + "grad_norm": 0.8730368614196777, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 10283 + }, + { + "epoch": 0.10284, + "grad_norm": 0.7729176878929138, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 10284 + }, + { + "epoch": 0.10285, + "grad_norm": 0.7743112444877625, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 10285 + }, + { + "epoch": 0.10286, + "grad_norm": 0.7332449555397034, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 10286 + }, + { + "epoch": 0.10287, + "grad_norm": 0.9000879526138306, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 10287 + }, + { + "epoch": 0.10288, + "grad_norm": 0.9760184288024902, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 10288 + }, + { + "epoch": 0.10289, + "grad_norm": 1.027477741241455, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 10289 + }, + { + "epoch": 0.1029, + "grad_norm": 0.8924638032913208, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 10290 + }, + { + "epoch": 0.10291, + "grad_norm": 0.7968407869338989, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 10291 + }, + { + "epoch": 0.10292, + "grad_norm": 0.8795375227928162, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 10292 + }, + { + "epoch": 0.10293, + "grad_norm": 0.9991517663002014, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 10293 + }, + { + "epoch": 0.10294, + "grad_norm": 0.9769951701164246, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 10294 + }, + { + "epoch": 0.10295, + "grad_norm": 0.9654301404953003, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 10295 + }, + { + "epoch": 0.10296, + "grad_norm": 1.0473653078079224, + "learning_rate": 0.003, + "loss": 4.023, + "step": 10296 + }, + { + "epoch": 0.10297, + "grad_norm": 0.9287710785865784, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 10297 + }, + { + "epoch": 0.10298, + "grad_norm": 0.8924084305763245, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 10298 + }, + { + "epoch": 0.10299, + "grad_norm": 0.9555318355560303, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 10299 + }, + { + "epoch": 0.103, + "grad_norm": 1.0484460592269897, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 10300 + }, + { + "epoch": 0.10301, + "grad_norm": 0.8988922238349915, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 10301 + }, + { + "epoch": 0.10302, + "grad_norm": 0.7982579469680786, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 10302 + }, + { + "epoch": 0.10303, + "grad_norm": 0.8110583424568176, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 10303 + }, + { + "epoch": 0.10304, + "grad_norm": 0.8419620394706726, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 10304 + }, + { + "epoch": 0.10305, + "grad_norm": 0.9462392926216125, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 10305 + }, + { + "epoch": 0.10306, + "grad_norm": 1.0094115734100342, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 10306 + }, + { + "epoch": 0.10307, + "grad_norm": 0.9687174558639526, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 10307 + }, + { + "epoch": 0.10308, + "grad_norm": 0.9006624221801758, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 10308 + }, + { + "epoch": 0.10309, + "grad_norm": 0.8222994208335876, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10309 + }, + { + "epoch": 0.1031, + "grad_norm": 0.8331215381622314, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 10310 + }, + { + "epoch": 0.10311, + "grad_norm": 0.6884284019470215, + "learning_rate": 0.003, + "loss": 4.034, + "step": 10311 + }, + { + "epoch": 0.10312, + "grad_norm": 0.6938837766647339, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 10312 + }, + { + "epoch": 0.10313, + "grad_norm": 0.659064769744873, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 10313 + }, + { + "epoch": 0.10314, + "grad_norm": 0.6628589630126953, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 10314 + }, + { + "epoch": 0.10315, + "grad_norm": 0.7294396758079529, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 10315 + }, + { + "epoch": 0.10316, + "grad_norm": 0.6614591479301453, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 10316 + }, + { + "epoch": 0.10317, + "grad_norm": 0.684831440448761, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 10317 + }, + { + "epoch": 0.10318, + "grad_norm": 0.7045228481292725, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 10318 + }, + { + "epoch": 0.10319, + "grad_norm": 0.7830377817153931, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 10319 + }, + { + "epoch": 0.1032, + "grad_norm": 0.9162354469299316, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 10320 + }, + { + "epoch": 0.10321, + "grad_norm": 0.9490805864334106, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 10321 + }, + { + "epoch": 0.10322, + "grad_norm": 0.6477469205856323, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 10322 + }, + { + "epoch": 0.10323, + "grad_norm": 0.5870311856269836, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 10323 + }, + { + "epoch": 0.10324, + "grad_norm": 0.5911048054695129, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 10324 + }, + { + "epoch": 0.10325, + "grad_norm": 0.6378191709518433, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 10325 + }, + { + "epoch": 0.10326, + "grad_norm": 0.6543301343917847, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10326 + }, + { + "epoch": 0.10327, + "grad_norm": 0.6948620080947876, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 10327 + }, + { + "epoch": 0.10328, + "grad_norm": 0.7429428100585938, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 10328 + }, + { + "epoch": 0.10329, + "grad_norm": 0.7913004159927368, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 10329 + }, + { + "epoch": 0.1033, + "grad_norm": 0.8257701992988586, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 10330 + }, + { + "epoch": 0.10331, + "grad_norm": 0.8367582559585571, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 10331 + }, + { + "epoch": 0.10332, + "grad_norm": 0.7799370884895325, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 10332 + }, + { + "epoch": 0.10333, + "grad_norm": 0.7020469307899475, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 10333 + }, + { + "epoch": 0.10334, + "grad_norm": 0.6871790289878845, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 10334 + }, + { + "epoch": 0.10335, + "grad_norm": 0.583629846572876, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 10335 + }, + { + "epoch": 0.10336, + "grad_norm": 0.6028733849525452, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 10336 + }, + { + "epoch": 0.10337, + "grad_norm": 0.7084194421768188, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 10337 + }, + { + "epoch": 0.10338, + "grad_norm": 0.7569435834884644, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 10338 + }, + { + "epoch": 0.10339, + "grad_norm": 0.6454418301582336, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 10339 + }, + { + "epoch": 0.1034, + "grad_norm": 0.6388258337974548, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 10340 + }, + { + "epoch": 0.10341, + "grad_norm": 0.6123440861701965, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 10341 + }, + { + "epoch": 0.10342, + "grad_norm": 0.600523054599762, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 10342 + }, + { + "epoch": 0.10343, + "grad_norm": 0.7250362634658813, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 10343 + }, + { + "epoch": 0.10344, + "grad_norm": 0.8328801393508911, + "learning_rate": 0.003, + "loss": 4.035, + "step": 10344 + }, + { + "epoch": 0.10345, + "grad_norm": 0.9125838875770569, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 10345 + }, + { + "epoch": 0.10346, + "grad_norm": 0.9645897746086121, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 10346 + }, + { + "epoch": 0.10347, + "grad_norm": 0.9420467019081116, + "learning_rate": 0.003, + "loss": 4.03, + "step": 10347 + }, + { + "epoch": 0.10348, + "grad_norm": 0.9969478249549866, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 10348 + }, + { + "epoch": 0.10349, + "grad_norm": 1.1243056058883667, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 10349 + }, + { + "epoch": 0.1035, + "grad_norm": 1.056504726409912, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 10350 + }, + { + "epoch": 0.10351, + "grad_norm": 1.0015921592712402, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 10351 + }, + { + "epoch": 0.10352, + "grad_norm": 1.054189682006836, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 10352 + }, + { + "epoch": 0.10353, + "grad_norm": 0.9281718730926514, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 10353 + }, + { + "epoch": 0.10354, + "grad_norm": 0.8637033700942993, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 10354 + }, + { + "epoch": 0.10355, + "grad_norm": 0.6888679265975952, + "learning_rate": 0.003, + "loss": 4.026, + "step": 10355 + }, + { + "epoch": 0.10356, + "grad_norm": 0.5851486921310425, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 10356 + }, + { + "epoch": 0.10357, + "grad_norm": 0.5784017443656921, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 10357 + }, + { + "epoch": 0.10358, + "grad_norm": 0.569531261920929, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 10358 + }, + { + "epoch": 0.10359, + "grad_norm": 0.5688461065292358, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 10359 + }, + { + "epoch": 0.1036, + "grad_norm": 0.6449005007743835, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 10360 + }, + { + "epoch": 0.10361, + "grad_norm": 0.7200685739517212, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 10361 + }, + { + "epoch": 0.10362, + "grad_norm": 0.7981273531913757, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 10362 + }, + { + "epoch": 0.10363, + "grad_norm": 0.8022664785385132, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 10363 + }, + { + "epoch": 0.10364, + "grad_norm": 0.8001551628112793, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 10364 + }, + { + "epoch": 0.10365, + "grad_norm": 0.8692436814308167, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 10365 + }, + { + "epoch": 0.10366, + "grad_norm": 0.8658043146133423, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 10366 + }, + { + "epoch": 0.10367, + "grad_norm": 0.8140656352043152, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 10367 + }, + { + "epoch": 0.10368, + "grad_norm": 0.9589678049087524, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 10368 + }, + { + "epoch": 0.10369, + "grad_norm": 0.9657092094421387, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 10369 + }, + { + "epoch": 0.1037, + "grad_norm": 0.9821603298187256, + "learning_rate": 0.003, + "loss": 4.029, + "step": 10370 + }, + { + "epoch": 0.10371, + "grad_norm": 1.0179591178894043, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 10371 + }, + { + "epoch": 0.10372, + "grad_norm": 1.0082340240478516, + "learning_rate": 0.003, + "loss": 4.026, + "step": 10372 + }, + { + "epoch": 0.10373, + "grad_norm": 1.0298233032226562, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 10373 + }, + { + "epoch": 0.10374, + "grad_norm": 0.8384741544723511, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 10374 + }, + { + "epoch": 0.10375, + "grad_norm": 0.7988961338996887, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 10375 + }, + { + "epoch": 0.10376, + "grad_norm": 0.9472295641899109, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 10376 + }, + { + "epoch": 0.10377, + "grad_norm": 1.0347899198532104, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 10377 + }, + { + "epoch": 0.10378, + "grad_norm": 1.1437108516693115, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 10378 + }, + { + "epoch": 0.10379, + "grad_norm": 0.7097597718238831, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 10379 + }, + { + "epoch": 0.1038, + "grad_norm": 0.6190682649612427, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 10380 + }, + { + "epoch": 0.10381, + "grad_norm": 0.6848435401916504, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 10381 + }, + { + "epoch": 0.10382, + "grad_norm": 0.8380157947540283, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 10382 + }, + { + "epoch": 0.10383, + "grad_norm": 0.9357398748397827, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 10383 + }, + { + "epoch": 0.10384, + "grad_norm": 0.8331278562545776, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 10384 + }, + { + "epoch": 0.10385, + "grad_norm": 0.7992599010467529, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 10385 + }, + { + "epoch": 0.10386, + "grad_norm": 0.7941585779190063, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 10386 + }, + { + "epoch": 0.10387, + "grad_norm": 0.7473763227462769, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 10387 + }, + { + "epoch": 0.10388, + "grad_norm": 0.6912757158279419, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 10388 + }, + { + "epoch": 0.10389, + "grad_norm": 0.6043686866760254, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 10389 + }, + { + "epoch": 0.1039, + "grad_norm": 0.593113124370575, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 10390 + }, + { + "epoch": 0.10391, + "grad_norm": 0.5677477717399597, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 10391 + }, + { + "epoch": 0.10392, + "grad_norm": 0.5792213082313538, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 10392 + }, + { + "epoch": 0.10393, + "grad_norm": 0.6402269005775452, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 10393 + }, + { + "epoch": 0.10394, + "grad_norm": 0.6701036691665649, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10394 + }, + { + "epoch": 0.10395, + "grad_norm": 0.7115254402160645, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 10395 + }, + { + "epoch": 0.10396, + "grad_norm": 0.7768694162368774, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 10396 + }, + { + "epoch": 0.10397, + "grad_norm": 0.7983322739601135, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 10397 + }, + { + "epoch": 0.10398, + "grad_norm": 0.8996577262878418, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 10398 + }, + { + "epoch": 0.10399, + "grad_norm": 0.9261710047721863, + "learning_rate": 0.003, + "loss": 4.038, + "step": 10399 + }, + { + "epoch": 0.104, + "grad_norm": 0.8924296498298645, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 10400 + }, + { + "epoch": 0.10401, + "grad_norm": 0.6983252763748169, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 10401 + }, + { + "epoch": 0.10402, + "grad_norm": 0.6650741100311279, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 10402 + }, + { + "epoch": 0.10403, + "grad_norm": 0.7275328636169434, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 10403 + }, + { + "epoch": 0.10404, + "grad_norm": 0.7682779431343079, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 10404 + }, + { + "epoch": 0.10405, + "grad_norm": 0.6787878274917603, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 10405 + }, + { + "epoch": 0.10406, + "grad_norm": 0.8612582087516785, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 10406 + }, + { + "epoch": 0.10407, + "grad_norm": 0.9241533279418945, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 10407 + }, + { + "epoch": 0.10408, + "grad_norm": 0.9687588810920715, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 10408 + }, + { + "epoch": 0.10409, + "grad_norm": 1.0829697847366333, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 10409 + }, + { + "epoch": 0.1041, + "grad_norm": 0.7547574043273926, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 10410 + }, + { + "epoch": 0.10411, + "grad_norm": 0.6516535878181458, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 10411 + }, + { + "epoch": 0.10412, + "grad_norm": 0.732597291469574, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 10412 + }, + { + "epoch": 0.10413, + "grad_norm": 0.6244125366210938, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 10413 + }, + { + "epoch": 0.10414, + "grad_norm": 0.6561394929885864, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 10414 + }, + { + "epoch": 0.10415, + "grad_norm": 0.7480183839797974, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 10415 + }, + { + "epoch": 0.10416, + "grad_norm": 0.7821226716041565, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 10416 + }, + { + "epoch": 0.10417, + "grad_norm": 0.8230247497558594, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 10417 + }, + { + "epoch": 0.10418, + "grad_norm": 0.9134896993637085, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 10418 + }, + { + "epoch": 0.10419, + "grad_norm": 0.976138710975647, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 10419 + }, + { + "epoch": 0.1042, + "grad_norm": 0.9060264229774475, + "learning_rate": 0.003, + "loss": 4.052, + "step": 10420 + }, + { + "epoch": 0.10421, + "grad_norm": 0.8544201850891113, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 10421 + }, + { + "epoch": 0.10422, + "grad_norm": 0.7737321257591248, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 10422 + }, + { + "epoch": 0.10423, + "grad_norm": 0.7796673774719238, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 10423 + }, + { + "epoch": 0.10424, + "grad_norm": 0.9727042317390442, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 10424 + }, + { + "epoch": 0.10425, + "grad_norm": 1.0700771808624268, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10425 + }, + { + "epoch": 0.10426, + "grad_norm": 1.0908294916152954, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 10426 + }, + { + "epoch": 0.10427, + "grad_norm": 0.800527811050415, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 10427 + }, + { + "epoch": 0.10428, + "grad_norm": 0.6922642588615417, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 10428 + }, + { + "epoch": 0.10429, + "grad_norm": 0.7603942155838013, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 10429 + }, + { + "epoch": 0.1043, + "grad_norm": 0.9487042427062988, + "learning_rate": 0.003, + "loss": 4.016, + "step": 10430 + }, + { + "epoch": 0.10431, + "grad_norm": 0.9603556394577026, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 10431 + }, + { + "epoch": 0.10432, + "grad_norm": 0.7391037344932556, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 10432 + }, + { + "epoch": 0.10433, + "grad_norm": 0.5578556060791016, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 10433 + }, + { + "epoch": 0.10434, + "grad_norm": 0.7265412211418152, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 10434 + }, + { + "epoch": 0.10435, + "grad_norm": 0.8804828524589539, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 10435 + }, + { + "epoch": 0.10436, + "grad_norm": 0.7199718952178955, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 10436 + }, + { + "epoch": 0.10437, + "grad_norm": 0.6212375164031982, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 10437 + }, + { + "epoch": 0.10438, + "grad_norm": 0.621557354927063, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 10438 + }, + { + "epoch": 0.10439, + "grad_norm": 0.6311930418014526, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 10439 + }, + { + "epoch": 0.1044, + "grad_norm": 0.6799618601799011, + "learning_rate": 0.003, + "loss": 4.033, + "step": 10440 + }, + { + "epoch": 0.10441, + "grad_norm": 0.6392710208892822, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 10441 + }, + { + "epoch": 0.10442, + "grad_norm": 0.5309514403343201, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 10442 + }, + { + "epoch": 0.10443, + "grad_norm": 0.46626293659210205, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 10443 + }, + { + "epoch": 0.10444, + "grad_norm": 0.5864386558532715, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 10444 + }, + { + "epoch": 0.10445, + "grad_norm": 0.6950222849845886, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 10445 + }, + { + "epoch": 0.10446, + "grad_norm": 0.7721315622329712, + "learning_rate": 0.003, + "loss": 4.003, + "step": 10446 + }, + { + "epoch": 0.10447, + "grad_norm": 0.8094618320465088, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 10447 + }, + { + "epoch": 0.10448, + "grad_norm": 0.7723525762557983, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 10448 + }, + { + "epoch": 0.10449, + "grad_norm": 0.6535624265670776, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 10449 + }, + { + "epoch": 0.1045, + "grad_norm": 0.615446925163269, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 10450 + }, + { + "epoch": 0.10451, + "grad_norm": 0.8183608055114746, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 10451 + }, + { + "epoch": 0.10452, + "grad_norm": 0.9758654832839966, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 10452 + }, + { + "epoch": 0.10453, + "grad_norm": 1.1344112157821655, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 10453 + }, + { + "epoch": 0.10454, + "grad_norm": 0.7159709930419922, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 10454 + }, + { + "epoch": 0.10455, + "grad_norm": 0.6631528735160828, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 10455 + }, + { + "epoch": 0.10456, + "grad_norm": 0.747515857219696, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 10456 + }, + { + "epoch": 0.10457, + "grad_norm": 0.7781107425689697, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 10457 + }, + { + "epoch": 0.10458, + "grad_norm": 0.9523149728775024, + "learning_rate": 0.003, + "loss": 4.047, + "step": 10458 + }, + { + "epoch": 0.10459, + "grad_norm": 1.1178628206253052, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 10459 + }, + { + "epoch": 0.1046, + "grad_norm": 0.6914238929748535, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 10460 + }, + { + "epoch": 0.10461, + "grad_norm": 0.611783504486084, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 10461 + }, + { + "epoch": 0.10462, + "grad_norm": 0.7709699869155884, + "learning_rate": 0.003, + "loss": 4.035, + "step": 10462 + }, + { + "epoch": 0.10463, + "grad_norm": 0.6816902160644531, + "learning_rate": 0.003, + "loss": 4.006, + "step": 10463 + }, + { + "epoch": 0.10464, + "grad_norm": 0.6006267070770264, + "learning_rate": 0.003, + "loss": 4.035, + "step": 10464 + }, + { + "epoch": 0.10465, + "grad_norm": 0.5709377527236938, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 10465 + }, + { + "epoch": 0.10466, + "grad_norm": 0.5002264976501465, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 10466 + }, + { + "epoch": 0.10467, + "grad_norm": 0.5063186287879944, + "learning_rate": 0.003, + "loss": 4.016, + "step": 10467 + }, + { + "epoch": 0.10468, + "grad_norm": 0.594136118888855, + "learning_rate": 0.003, + "loss": 4.0, + "step": 10468 + }, + { + "epoch": 0.10469, + "grad_norm": 0.5744364261627197, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 10469 + }, + { + "epoch": 0.1047, + "grad_norm": 0.6634460687637329, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 10470 + }, + { + "epoch": 0.10471, + "grad_norm": 0.7383639812469482, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 10471 + }, + { + "epoch": 0.10472, + "grad_norm": 0.9079957604408264, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 10472 + }, + { + "epoch": 0.10473, + "grad_norm": 1.1386297941207886, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 10473 + }, + { + "epoch": 0.10474, + "grad_norm": 0.7866982817649841, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 10474 + }, + { + "epoch": 0.10475, + "grad_norm": 0.7195823192596436, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 10475 + }, + { + "epoch": 0.10476, + "grad_norm": 0.7635900974273682, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 10476 + }, + { + "epoch": 0.10477, + "grad_norm": 0.7521525621414185, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 10477 + }, + { + "epoch": 0.10478, + "grad_norm": 0.8314422369003296, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 10478 + }, + { + "epoch": 0.10479, + "grad_norm": 0.9330824613571167, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 10479 + }, + { + "epoch": 0.1048, + "grad_norm": 0.9943487644195557, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 10480 + }, + { + "epoch": 0.10481, + "grad_norm": 1.000788927078247, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 10481 + }, + { + "epoch": 0.10482, + "grad_norm": 0.9992671608924866, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 10482 + }, + { + "epoch": 0.10483, + "grad_norm": 0.9347127676010132, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 10483 + }, + { + "epoch": 0.10484, + "grad_norm": 0.9052039384841919, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 10484 + }, + { + "epoch": 0.10485, + "grad_norm": 0.9239107966423035, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 10485 + }, + { + "epoch": 0.10486, + "grad_norm": 0.996957004070282, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 10486 + }, + { + "epoch": 0.10487, + "grad_norm": 0.9323636889457703, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 10487 + }, + { + "epoch": 0.10488, + "grad_norm": 0.9457101821899414, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 10488 + }, + { + "epoch": 0.10489, + "grad_norm": 0.8397849202156067, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 10489 + }, + { + "epoch": 0.1049, + "grad_norm": 0.7614753842353821, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 10490 + }, + { + "epoch": 0.10491, + "grad_norm": 0.7990318536758423, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 10491 + }, + { + "epoch": 0.10492, + "grad_norm": 0.7685605883598328, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 10492 + }, + { + "epoch": 0.10493, + "grad_norm": 0.8393089771270752, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 10493 + }, + { + "epoch": 0.10494, + "grad_norm": 0.7939583659172058, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 10494 + }, + { + "epoch": 0.10495, + "grad_norm": 0.6594287753105164, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 10495 + }, + { + "epoch": 0.10496, + "grad_norm": 0.6255642175674438, + "learning_rate": 0.003, + "loss": 4.037, + "step": 10496 + }, + { + "epoch": 0.10497, + "grad_norm": 0.5651243925094604, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 10497 + }, + { + "epoch": 0.10498, + "grad_norm": 0.4910907447338104, + "learning_rate": 0.003, + "loss": 4.04, + "step": 10498 + }, + { + "epoch": 0.10499, + "grad_norm": 0.5701528191566467, + "learning_rate": 0.003, + "loss": 4.007, + "step": 10499 + }, + { + "epoch": 0.105, + "grad_norm": 0.6363165378570557, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 10500 + }, + { + "epoch": 0.10501, + "grad_norm": 0.687296450138092, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 10501 + }, + { + "epoch": 0.10502, + "grad_norm": 0.9270140528678894, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 10502 + }, + { + "epoch": 0.10503, + "grad_norm": 1.2313417196273804, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 10503 + }, + { + "epoch": 0.10504, + "grad_norm": 0.6528260707855225, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 10504 + }, + { + "epoch": 0.10505, + "grad_norm": 0.5283305644989014, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 10505 + }, + { + "epoch": 0.10506, + "grad_norm": 0.5796220898628235, + "learning_rate": 0.003, + "loss": 4.022, + "step": 10506 + }, + { + "epoch": 0.10507, + "grad_norm": 0.623137891292572, + "learning_rate": 0.003, + "loss": 4.023, + "step": 10507 + }, + { + "epoch": 0.10508, + "grad_norm": 0.6215649843215942, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 10508 + }, + { + "epoch": 0.10509, + "grad_norm": 0.6104245781898499, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 10509 + }, + { + "epoch": 0.1051, + "grad_norm": 0.6872464418411255, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 10510 + }, + { + "epoch": 0.10511, + "grad_norm": 0.7551093101501465, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 10511 + }, + { + "epoch": 0.10512, + "grad_norm": 0.8498075008392334, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 10512 + }, + { + "epoch": 0.10513, + "grad_norm": 0.832697331905365, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 10513 + }, + { + "epoch": 0.10514, + "grad_norm": 0.7095122337341309, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 10514 + }, + { + "epoch": 0.10515, + "grad_norm": 0.6831356883049011, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 10515 + }, + { + "epoch": 0.10516, + "grad_norm": 0.7304404377937317, + "learning_rate": 0.003, + "loss": 3.99, + "step": 10516 + }, + { + "epoch": 0.10517, + "grad_norm": 0.7606459856033325, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 10517 + }, + { + "epoch": 0.10518, + "grad_norm": 0.7579407095909119, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 10518 + }, + { + "epoch": 0.10519, + "grad_norm": 0.7492026090621948, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 10519 + }, + { + "epoch": 0.1052, + "grad_norm": 0.7874739766120911, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 10520 + }, + { + "epoch": 0.10521, + "grad_norm": 0.7780115008354187, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 10521 + }, + { + "epoch": 0.10522, + "grad_norm": 0.774425745010376, + "learning_rate": 0.003, + "loss": 4.035, + "step": 10522 + }, + { + "epoch": 0.10523, + "grad_norm": 0.8482930064201355, + "learning_rate": 0.003, + "loss": 4.051, + "step": 10523 + }, + { + "epoch": 0.10524, + "grad_norm": 1.0661581754684448, + "learning_rate": 0.003, + "loss": 4.026, + "step": 10524 + }, + { + "epoch": 0.10525, + "grad_norm": 1.2727166414260864, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 10525 + }, + { + "epoch": 0.10526, + "grad_norm": 0.8416107296943665, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 10526 + }, + { + "epoch": 0.10527, + "grad_norm": 0.7589594721794128, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 10527 + }, + { + "epoch": 0.10528, + "grad_norm": 0.7027614116668701, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 10528 + }, + { + "epoch": 0.10529, + "grad_norm": 0.7037528157234192, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 10529 + }, + { + "epoch": 0.1053, + "grad_norm": 0.6253616213798523, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 10530 + }, + { + "epoch": 0.10531, + "grad_norm": 0.7041178941726685, + "learning_rate": 0.003, + "loss": 4.021, + "step": 10531 + }, + { + "epoch": 0.10532, + "grad_norm": 0.7070190906524658, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 10532 + }, + { + "epoch": 0.10533, + "grad_norm": 0.8403984308242798, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 10533 + }, + { + "epoch": 0.10534, + "grad_norm": 0.8914600014686584, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 10534 + }, + { + "epoch": 0.10535, + "grad_norm": 0.8591518998146057, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 10535 + }, + { + "epoch": 0.10536, + "grad_norm": 0.8553505539894104, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 10536 + }, + { + "epoch": 0.10537, + "grad_norm": 0.9386612772941589, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 10537 + }, + { + "epoch": 0.10538, + "grad_norm": 1.1708186864852905, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 10538 + }, + { + "epoch": 0.10539, + "grad_norm": 0.7604055404663086, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 10539 + }, + { + "epoch": 0.1054, + "grad_norm": 0.6964765787124634, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 10540 + }, + { + "epoch": 0.10541, + "grad_norm": 0.7417082190513611, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 10541 + }, + { + "epoch": 0.10542, + "grad_norm": 0.783746600151062, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 10542 + }, + { + "epoch": 0.10543, + "grad_norm": 0.8471524119377136, + "learning_rate": 0.003, + "loss": 4.03, + "step": 10543 + }, + { + "epoch": 0.10544, + "grad_norm": 0.8709812760353088, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 10544 + }, + { + "epoch": 0.10545, + "grad_norm": 0.8089157938957214, + "learning_rate": 0.003, + "loss": 4.042, + "step": 10545 + }, + { + "epoch": 0.10546, + "grad_norm": 0.8024004101753235, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 10546 + }, + { + "epoch": 0.10547, + "grad_norm": 0.771805465221405, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 10547 + }, + { + "epoch": 0.10548, + "grad_norm": 0.7884362936019897, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 10548 + }, + { + "epoch": 0.10549, + "grad_norm": 0.8107030987739563, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 10549 + }, + { + "epoch": 0.1055, + "grad_norm": 0.7356095314025879, + "learning_rate": 0.003, + "loss": 4.023, + "step": 10550 + }, + { + "epoch": 0.10551, + "grad_norm": 0.7435442805290222, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 10551 + }, + { + "epoch": 0.10552, + "grad_norm": 0.6796212792396545, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 10552 + }, + { + "epoch": 0.10553, + "grad_norm": 0.6590840816497803, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 10553 + }, + { + "epoch": 0.10554, + "grad_norm": 0.6578130722045898, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 10554 + }, + { + "epoch": 0.10555, + "grad_norm": 0.854569673538208, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 10555 + }, + { + "epoch": 0.10556, + "grad_norm": 1.234406590461731, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 10556 + }, + { + "epoch": 0.10557, + "grad_norm": 0.9004664421081543, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 10557 + }, + { + "epoch": 0.10558, + "grad_norm": 0.8317961692810059, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 10558 + }, + { + "epoch": 0.10559, + "grad_norm": 0.7906206250190735, + "learning_rate": 0.003, + "loss": 4.011, + "step": 10559 + }, + { + "epoch": 0.1056, + "grad_norm": 0.7413365244865417, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 10560 + }, + { + "epoch": 0.10561, + "grad_norm": 0.6539586782455444, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 10561 + }, + { + "epoch": 0.10562, + "grad_norm": 0.5383809804916382, + "learning_rate": 0.003, + "loss": 3.997, + "step": 10562 + }, + { + "epoch": 0.10563, + "grad_norm": 0.529626190662384, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 10563 + }, + { + "epoch": 0.10564, + "grad_norm": 0.557378351688385, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 10564 + }, + { + "epoch": 0.10565, + "grad_norm": 0.5710329413414001, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 10565 + }, + { + "epoch": 0.10566, + "grad_norm": 0.7581164240837097, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 10566 + }, + { + "epoch": 0.10567, + "grad_norm": 0.9188153147697449, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 10567 + }, + { + "epoch": 0.10568, + "grad_norm": 1.2158697843551636, + "learning_rate": 0.003, + "loss": 4.054, + "step": 10568 + }, + { + "epoch": 0.10569, + "grad_norm": 0.799419105052948, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 10569 + }, + { + "epoch": 0.1057, + "grad_norm": 0.8842980265617371, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 10570 + }, + { + "epoch": 0.10571, + "grad_norm": 1.089672565460205, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 10571 + }, + { + "epoch": 0.10572, + "grad_norm": 0.9158157110214233, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 10572 + }, + { + "epoch": 0.10573, + "grad_norm": 1.0048152208328247, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 10573 + }, + { + "epoch": 0.10574, + "grad_norm": 0.9467569589614868, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 10574 + }, + { + "epoch": 0.10575, + "grad_norm": 0.9308494329452515, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 10575 + }, + { + "epoch": 0.10576, + "grad_norm": 0.8847599029541016, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 10576 + }, + { + "epoch": 0.10577, + "grad_norm": 0.8620772957801819, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 10577 + }, + { + "epoch": 0.10578, + "grad_norm": 0.9054552316665649, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 10578 + }, + { + "epoch": 0.10579, + "grad_norm": 0.9128457903862, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 10579 + }, + { + "epoch": 0.1058, + "grad_norm": 0.734600305557251, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 10580 + }, + { + "epoch": 0.10581, + "grad_norm": 0.7643128037452698, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 10581 + }, + { + "epoch": 0.10582, + "grad_norm": 0.6332104802131653, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 10582 + }, + { + "epoch": 0.10583, + "grad_norm": 0.5396142601966858, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 10583 + }, + { + "epoch": 0.10584, + "grad_norm": 0.5539082884788513, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 10584 + }, + { + "epoch": 0.10585, + "grad_norm": 0.5713188648223877, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 10585 + }, + { + "epoch": 0.10586, + "grad_norm": 0.6109192371368408, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 10586 + }, + { + "epoch": 0.10587, + "grad_norm": 0.7975706458091736, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 10587 + }, + { + "epoch": 0.10588, + "grad_norm": 1.0050781965255737, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 10588 + }, + { + "epoch": 0.10589, + "grad_norm": 1.3392750024795532, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 10589 + }, + { + "epoch": 0.1059, + "grad_norm": 0.63167804479599, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 10590 + }, + { + "epoch": 0.10591, + "grad_norm": 0.6843758821487427, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 10591 + }, + { + "epoch": 0.10592, + "grad_norm": 0.7719295620918274, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 10592 + }, + { + "epoch": 0.10593, + "grad_norm": 0.9197074174880981, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 10593 + }, + { + "epoch": 0.10594, + "grad_norm": 1.031716227531433, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 10594 + }, + { + "epoch": 0.10595, + "grad_norm": 0.7711551785469055, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 10595 + }, + { + "epoch": 0.10596, + "grad_norm": 0.6337047815322876, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 10596 + }, + { + "epoch": 0.10597, + "grad_norm": 0.5785505175590515, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 10597 + }, + { + "epoch": 0.10598, + "grad_norm": 0.5775535702705383, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 10598 + }, + { + "epoch": 0.10599, + "grad_norm": 0.5415910482406616, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 10599 + }, + { + "epoch": 0.106, + "grad_norm": 0.5284204483032227, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 10600 + }, + { + "epoch": 0.10601, + "grad_norm": 0.485622763633728, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 10601 + }, + { + "epoch": 0.10602, + "grad_norm": 0.5149875283241272, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 10602 + }, + { + "epoch": 0.10603, + "grad_norm": 0.6656335592269897, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 10603 + }, + { + "epoch": 0.10604, + "grad_norm": 0.8385307788848877, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 10604 + }, + { + "epoch": 0.10605, + "grad_norm": 1.0076050758361816, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 10605 + }, + { + "epoch": 0.10606, + "grad_norm": 1.0600577592849731, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 10606 + }, + { + "epoch": 0.10607, + "grad_norm": 0.7597001791000366, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 10607 + }, + { + "epoch": 0.10608, + "grad_norm": 0.6262304186820984, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 10608 + }, + { + "epoch": 0.10609, + "grad_norm": 0.6129266619682312, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 10609 + }, + { + "epoch": 0.1061, + "grad_norm": 0.6446408033370972, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 10610 + }, + { + "epoch": 0.10611, + "grad_norm": 0.5875228047370911, + "learning_rate": 0.003, + "loss": 3.984, + "step": 10611 + }, + { + "epoch": 0.10612, + "grad_norm": 0.5728915929794312, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 10612 + }, + { + "epoch": 0.10613, + "grad_norm": 0.6378050446510315, + "learning_rate": 0.003, + "loss": 4.008, + "step": 10613 + }, + { + "epoch": 0.10614, + "grad_norm": 0.6682537198066711, + "learning_rate": 0.003, + "loss": 4.03, + "step": 10614 + }, + { + "epoch": 0.10615, + "grad_norm": 0.6920739412307739, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 10615 + }, + { + "epoch": 0.10616, + "grad_norm": 0.7054862380027771, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 10616 + }, + { + "epoch": 0.10617, + "grad_norm": 0.8225484490394592, + "learning_rate": 0.003, + "loss": 4.042, + "step": 10617 + }, + { + "epoch": 0.10618, + "grad_norm": 0.8290918469429016, + "learning_rate": 0.003, + "loss": 4.022, + "step": 10618 + }, + { + "epoch": 0.10619, + "grad_norm": 0.9801556468009949, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 10619 + }, + { + "epoch": 0.1062, + "grad_norm": 1.2758433818817139, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 10620 + }, + { + "epoch": 0.10621, + "grad_norm": 0.7299548387527466, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 10621 + }, + { + "epoch": 0.10622, + "grad_norm": 0.5834721326828003, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 10622 + }, + { + "epoch": 0.10623, + "grad_norm": 0.5982388257980347, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 10623 + }, + { + "epoch": 0.10624, + "grad_norm": 0.5697194933891296, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 10624 + }, + { + "epoch": 0.10625, + "grad_norm": 0.5805581212043762, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 10625 + }, + { + "epoch": 0.10626, + "grad_norm": 0.6096617579460144, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 10626 + }, + { + "epoch": 0.10627, + "grad_norm": 0.7148505449295044, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 10627 + }, + { + "epoch": 0.10628, + "grad_norm": 0.9222334623336792, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 10628 + }, + { + "epoch": 0.10629, + "grad_norm": 1.1694058179855347, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 10629 + }, + { + "epoch": 0.1063, + "grad_norm": 1.0497586727142334, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 10630 + }, + { + "epoch": 0.10631, + "grad_norm": 0.8866661190986633, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 10631 + }, + { + "epoch": 0.10632, + "grad_norm": 0.7329339981079102, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 10632 + }, + { + "epoch": 0.10633, + "grad_norm": 0.7718518972396851, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 10633 + }, + { + "epoch": 0.10634, + "grad_norm": 0.7611184120178223, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 10634 + }, + { + "epoch": 0.10635, + "grad_norm": 0.7974138855934143, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 10635 + }, + { + "epoch": 0.10636, + "grad_norm": 0.6915683150291443, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 10636 + }, + { + "epoch": 0.10637, + "grad_norm": 0.6084825396537781, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 10637 + }, + { + "epoch": 0.10638, + "grad_norm": 0.611084520816803, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 10638 + }, + { + "epoch": 0.10639, + "grad_norm": 0.6165266633033752, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 10639 + }, + { + "epoch": 0.1064, + "grad_norm": 0.6598978042602539, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 10640 + }, + { + "epoch": 0.10641, + "grad_norm": 0.7608402967453003, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 10641 + }, + { + "epoch": 0.10642, + "grad_norm": 0.7880569100379944, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 10642 + }, + { + "epoch": 0.10643, + "grad_norm": 0.8869519829750061, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 10643 + }, + { + "epoch": 0.10644, + "grad_norm": 1.1587555408477783, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 10644 + }, + { + "epoch": 0.10645, + "grad_norm": 0.9315388202667236, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 10645 + }, + { + "epoch": 0.10646, + "grad_norm": 0.7919732928276062, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 10646 + }, + { + "epoch": 0.10647, + "grad_norm": 0.812877357006073, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 10647 + }, + { + "epoch": 0.10648, + "grad_norm": 0.9086081981658936, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 10648 + }, + { + "epoch": 0.10649, + "grad_norm": 0.9742897152900696, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 10649 + }, + { + "epoch": 0.1065, + "grad_norm": 0.9627885818481445, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 10650 + }, + { + "epoch": 0.10651, + "grad_norm": 0.99714595079422, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 10651 + }, + { + "epoch": 0.10652, + "grad_norm": 0.9961947202682495, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 10652 + }, + { + "epoch": 0.10653, + "grad_norm": 0.9269364476203918, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 10653 + }, + { + "epoch": 0.10654, + "grad_norm": 0.8528988361358643, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 10654 + }, + { + "epoch": 0.10655, + "grad_norm": 0.9138254523277283, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 10655 + }, + { + "epoch": 0.10656, + "grad_norm": 1.1768444776535034, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 10656 + }, + { + "epoch": 0.10657, + "grad_norm": 0.8890489935874939, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 10657 + }, + { + "epoch": 0.10658, + "grad_norm": 0.961041271686554, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 10658 + }, + { + "epoch": 0.10659, + "grad_norm": 1.1152033805847168, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 10659 + }, + { + "epoch": 0.1066, + "grad_norm": 0.9879071712493896, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 10660 + }, + { + "epoch": 0.10661, + "grad_norm": 0.9049113392829895, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 10661 + }, + { + "epoch": 0.10662, + "grad_norm": 0.9326181411743164, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 10662 + }, + { + "epoch": 0.10663, + "grad_norm": 0.8698120713233948, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 10663 + }, + { + "epoch": 0.10664, + "grad_norm": 0.8739762902259827, + "learning_rate": 0.003, + "loss": 4.029, + "step": 10664 + }, + { + "epoch": 0.10665, + "grad_norm": 0.9346017241477966, + "learning_rate": 0.003, + "loss": 4.038, + "step": 10665 + }, + { + "epoch": 0.10666, + "grad_norm": 0.8549253940582275, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 10666 + }, + { + "epoch": 0.10667, + "grad_norm": 0.8709663152694702, + "learning_rate": 0.003, + "loss": 4.004, + "step": 10667 + }, + { + "epoch": 0.10668, + "grad_norm": 1.018907070159912, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 10668 + }, + { + "epoch": 0.10669, + "grad_norm": 1.186867594718933, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 10669 + }, + { + "epoch": 0.1067, + "grad_norm": 0.7143198251724243, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 10670 + }, + { + "epoch": 0.10671, + "grad_norm": 0.7373873591423035, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 10671 + }, + { + "epoch": 0.10672, + "grad_norm": 0.8877958059310913, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 10672 + }, + { + "epoch": 0.10673, + "grad_norm": 0.8976678848266602, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 10673 + }, + { + "epoch": 0.10674, + "grad_norm": 0.809621274471283, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 10674 + }, + { + "epoch": 0.10675, + "grad_norm": 0.6401908993721008, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 10675 + }, + { + "epoch": 0.10676, + "grad_norm": 0.6899451017379761, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 10676 + }, + { + "epoch": 0.10677, + "grad_norm": 0.47715431451797485, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 10677 + }, + { + "epoch": 0.10678, + "grad_norm": 0.5788038372993469, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 10678 + }, + { + "epoch": 0.10679, + "grad_norm": 0.5781792998313904, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 10679 + }, + { + "epoch": 0.1068, + "grad_norm": 0.5460559725761414, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 10680 + }, + { + "epoch": 0.10681, + "grad_norm": 0.531968891620636, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 10681 + }, + { + "epoch": 0.10682, + "grad_norm": 0.5341914892196655, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 10682 + }, + { + "epoch": 0.10683, + "grad_norm": 0.6349906921386719, + "learning_rate": 0.003, + "loss": 4.026, + "step": 10683 + }, + { + "epoch": 0.10684, + "grad_norm": 0.6948190927505493, + "learning_rate": 0.003, + "loss": 4.028, + "step": 10684 + }, + { + "epoch": 0.10685, + "grad_norm": 0.7251952886581421, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 10685 + }, + { + "epoch": 0.10686, + "grad_norm": 0.7874016761779785, + "learning_rate": 0.003, + "loss": 4.032, + "step": 10686 + }, + { + "epoch": 0.10687, + "grad_norm": 0.8624654412269592, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 10687 + }, + { + "epoch": 0.10688, + "grad_norm": 0.8007801175117493, + "learning_rate": 0.003, + "loss": 4.039, + "step": 10688 + }, + { + "epoch": 0.10689, + "grad_norm": 0.8008077144622803, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 10689 + }, + { + "epoch": 0.1069, + "grad_norm": 0.7206010818481445, + "learning_rate": 0.003, + "loss": 4.031, + "step": 10690 + }, + { + "epoch": 0.10691, + "grad_norm": 0.8642178177833557, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 10691 + }, + { + "epoch": 0.10692, + "grad_norm": 1.0821588039398193, + "learning_rate": 0.003, + "loss": 4.038, + "step": 10692 + }, + { + "epoch": 0.10693, + "grad_norm": 1.2005447149276733, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 10693 + }, + { + "epoch": 0.10694, + "grad_norm": 0.9520975351333618, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 10694 + }, + { + "epoch": 0.10695, + "grad_norm": 0.7902743220329285, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 10695 + }, + { + "epoch": 0.10696, + "grad_norm": 0.6577290892601013, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 10696 + }, + { + "epoch": 0.10697, + "grad_norm": 0.677960991859436, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 10697 + }, + { + "epoch": 0.10698, + "grad_norm": 0.6639751195907593, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 10698 + }, + { + "epoch": 0.10699, + "grad_norm": 0.6529691815376282, + "learning_rate": 0.003, + "loss": 4.056, + "step": 10699 + }, + { + "epoch": 0.107, + "grad_norm": 0.6431300640106201, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 10700 + }, + { + "epoch": 0.10701, + "grad_norm": 0.5310377478599548, + "learning_rate": 0.003, + "loss": 4.01, + "step": 10701 + }, + { + "epoch": 0.10702, + "grad_norm": 0.6834575533866882, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 10702 + }, + { + "epoch": 0.10703, + "grad_norm": 0.8570904731750488, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 10703 + }, + { + "epoch": 0.10704, + "grad_norm": 0.9088355898857117, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 10704 + }, + { + "epoch": 0.10705, + "grad_norm": 0.9915702939033508, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 10705 + }, + { + "epoch": 0.10706, + "grad_norm": 0.8690432906150818, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 10706 + }, + { + "epoch": 0.10707, + "grad_norm": 0.7811831831932068, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 10707 + }, + { + "epoch": 0.10708, + "grad_norm": 0.7660472989082336, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 10708 + }, + { + "epoch": 0.10709, + "grad_norm": 0.7577682733535767, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 10709 + }, + { + "epoch": 0.1071, + "grad_norm": 0.688403308391571, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 10710 + }, + { + "epoch": 0.10711, + "grad_norm": 0.5480352640151978, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 10711 + }, + { + "epoch": 0.10712, + "grad_norm": 0.6282423734664917, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 10712 + }, + { + "epoch": 0.10713, + "grad_norm": 0.6788033246994019, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 10713 + }, + { + "epoch": 0.10714, + "grad_norm": 0.713354229927063, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 10714 + }, + { + "epoch": 0.10715, + "grad_norm": 0.7679527997970581, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 10715 + }, + { + "epoch": 0.10716, + "grad_norm": 0.9965485334396362, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 10716 + }, + { + "epoch": 0.10717, + "grad_norm": 1.2599714994430542, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 10717 + }, + { + "epoch": 0.10718, + "grad_norm": 0.6802042126655579, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 10718 + }, + { + "epoch": 0.10719, + "grad_norm": 0.5995518565177917, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 10719 + }, + { + "epoch": 0.1072, + "grad_norm": 0.6362910270690918, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 10720 + }, + { + "epoch": 0.10721, + "grad_norm": 0.6041771173477173, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 10721 + }, + { + "epoch": 0.10722, + "grad_norm": 0.7603263258934021, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 10722 + }, + { + "epoch": 0.10723, + "grad_norm": 0.8332676291465759, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 10723 + }, + { + "epoch": 0.10724, + "grad_norm": 0.9279446601867676, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 10724 + }, + { + "epoch": 0.10725, + "grad_norm": 1.125430703163147, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 10725 + }, + { + "epoch": 0.10726, + "grad_norm": 1.0059179067611694, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 10726 + }, + { + "epoch": 0.10727, + "grad_norm": 1.1123380661010742, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 10727 + }, + { + "epoch": 0.10728, + "grad_norm": 0.9447655081748962, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 10728 + }, + { + "epoch": 0.10729, + "grad_norm": 0.9554311633110046, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 10729 + }, + { + "epoch": 0.1073, + "grad_norm": 1.0092507600784302, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 10730 + }, + { + "epoch": 0.10731, + "grad_norm": 1.0152878761291504, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 10731 + }, + { + "epoch": 0.10732, + "grad_norm": 0.9422562718391418, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 10732 + }, + { + "epoch": 0.10733, + "grad_norm": 0.9458036422729492, + "learning_rate": 0.003, + "loss": 4.053, + "step": 10733 + }, + { + "epoch": 0.10734, + "grad_norm": 0.9883454442024231, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 10734 + }, + { + "epoch": 0.10735, + "grad_norm": 0.8949477672576904, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 10735 + }, + { + "epoch": 0.10736, + "grad_norm": 0.7392562627792358, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 10736 + }, + { + "epoch": 0.10737, + "grad_norm": 0.7315913438796997, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 10737 + }, + { + "epoch": 0.10738, + "grad_norm": 0.7146841883659363, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 10738 + }, + { + "epoch": 0.10739, + "grad_norm": 0.7426468729972839, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 10739 + }, + { + "epoch": 0.1074, + "grad_norm": 0.5836867690086365, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 10740 + }, + { + "epoch": 0.10741, + "grad_norm": 0.5937772989273071, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 10741 + }, + { + "epoch": 0.10742, + "grad_norm": 0.694475531578064, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 10742 + }, + { + "epoch": 0.10743, + "grad_norm": 0.6357514262199402, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 10743 + }, + { + "epoch": 0.10744, + "grad_norm": 0.5116270184516907, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 10744 + }, + { + "epoch": 0.10745, + "grad_norm": 0.5230345726013184, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 10745 + }, + { + "epoch": 0.10746, + "grad_norm": 0.5541887283325195, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 10746 + }, + { + "epoch": 0.10747, + "grad_norm": 0.5771496891975403, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 10747 + }, + { + "epoch": 0.10748, + "grad_norm": 0.6870571970939636, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 10748 + }, + { + "epoch": 0.10749, + "grad_norm": 0.8241485357284546, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 10749 + }, + { + "epoch": 0.1075, + "grad_norm": 0.9226048588752747, + "learning_rate": 0.003, + "loss": 4.037, + "step": 10750 + }, + { + "epoch": 0.10751, + "grad_norm": 0.9833307266235352, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 10751 + }, + { + "epoch": 0.10752, + "grad_norm": 1.0102068185806274, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 10752 + }, + { + "epoch": 0.10753, + "grad_norm": 0.9321718811988831, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 10753 + }, + { + "epoch": 0.10754, + "grad_norm": 0.8044393062591553, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 10754 + }, + { + "epoch": 0.10755, + "grad_norm": 0.8157967925071716, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 10755 + }, + { + "epoch": 0.10756, + "grad_norm": 0.9599362015724182, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 10756 + }, + { + "epoch": 0.10757, + "grad_norm": 0.8057109117507935, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 10757 + }, + { + "epoch": 0.10758, + "grad_norm": 0.84975266456604, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 10758 + }, + { + "epoch": 0.10759, + "grad_norm": 0.7976199388504028, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 10759 + }, + { + "epoch": 0.1076, + "grad_norm": 0.8996418714523315, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 10760 + }, + { + "epoch": 0.10761, + "grad_norm": 0.952308714389801, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 10761 + }, + { + "epoch": 0.10762, + "grad_norm": 1.00287663936615, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 10762 + }, + { + "epoch": 0.10763, + "grad_norm": 0.9806409478187561, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 10763 + }, + { + "epoch": 0.10764, + "grad_norm": 0.7794227004051208, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 10764 + }, + { + "epoch": 0.10765, + "grad_norm": 0.7748934030532837, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 10765 + }, + { + "epoch": 0.10766, + "grad_norm": 0.8416560888290405, + "learning_rate": 0.003, + "loss": 4.021, + "step": 10766 + }, + { + "epoch": 0.10767, + "grad_norm": 0.8533070087432861, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 10767 + }, + { + "epoch": 0.10768, + "grad_norm": 0.8723945617675781, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 10768 + }, + { + "epoch": 0.10769, + "grad_norm": 0.8659939765930176, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 10769 + }, + { + "epoch": 0.1077, + "grad_norm": 0.8741641640663147, + "learning_rate": 0.003, + "loss": 4.036, + "step": 10770 + }, + { + "epoch": 0.10771, + "grad_norm": 0.8021601438522339, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 10771 + }, + { + "epoch": 0.10772, + "grad_norm": 0.6761582493782043, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 10772 + }, + { + "epoch": 0.10773, + "grad_norm": 0.584344744682312, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 10773 + }, + { + "epoch": 0.10774, + "grad_norm": 0.6257133483886719, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 10774 + }, + { + "epoch": 0.10775, + "grad_norm": 0.6287021040916443, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 10775 + }, + { + "epoch": 0.10776, + "grad_norm": 0.6459553837776184, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 10776 + }, + { + "epoch": 0.10777, + "grad_norm": 0.6836524605751038, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 10777 + }, + { + "epoch": 0.10778, + "grad_norm": 0.7361724972724915, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 10778 + }, + { + "epoch": 0.10779, + "grad_norm": 0.8304905295372009, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 10779 + }, + { + "epoch": 0.1078, + "grad_norm": 0.856549084186554, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 10780 + }, + { + "epoch": 0.10781, + "grad_norm": 0.942664623260498, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 10781 + }, + { + "epoch": 0.10782, + "grad_norm": 0.9167718887329102, + "learning_rate": 0.003, + "loss": 4.024, + "step": 10782 + }, + { + "epoch": 0.10783, + "grad_norm": 0.7978794574737549, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10783 + }, + { + "epoch": 0.10784, + "grad_norm": 0.7039570212364197, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 10784 + }, + { + "epoch": 0.10785, + "grad_norm": 0.6894760727882385, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 10785 + }, + { + "epoch": 0.10786, + "grad_norm": 0.7555667757987976, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 10786 + }, + { + "epoch": 0.10787, + "grad_norm": 0.7514839768409729, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 10787 + }, + { + "epoch": 0.10788, + "grad_norm": 0.6158230900764465, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 10788 + }, + { + "epoch": 0.10789, + "grad_norm": 0.5220729112625122, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 10789 + }, + { + "epoch": 0.1079, + "grad_norm": 0.4973139762878418, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 10790 + }, + { + "epoch": 0.10791, + "grad_norm": 0.5057622790336609, + "learning_rate": 0.003, + "loss": 4.035, + "step": 10791 + }, + { + "epoch": 0.10792, + "grad_norm": 0.459989458322525, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 10792 + }, + { + "epoch": 0.10793, + "grad_norm": 0.46567296981811523, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 10793 + }, + { + "epoch": 0.10794, + "grad_norm": 0.5016858577728271, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 10794 + }, + { + "epoch": 0.10795, + "grad_norm": 0.5655898451805115, + "learning_rate": 0.003, + "loss": 3.982, + "step": 10795 + }, + { + "epoch": 0.10796, + "grad_norm": 0.7223862409591675, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 10796 + }, + { + "epoch": 0.10797, + "grad_norm": 0.9427323937416077, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 10797 + }, + { + "epoch": 0.10798, + "grad_norm": 1.170189619064331, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 10798 + }, + { + "epoch": 0.10799, + "grad_norm": 0.9903934597969055, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 10799 + }, + { + "epoch": 0.108, + "grad_norm": 0.9310336709022522, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 10800 + }, + { + "epoch": 0.10801, + "grad_norm": 0.8072385787963867, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 10801 + }, + { + "epoch": 0.10802, + "grad_norm": 0.802480936050415, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 10802 + }, + { + "epoch": 0.10803, + "grad_norm": 0.8250213861465454, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 10803 + }, + { + "epoch": 0.10804, + "grad_norm": 0.9733414649963379, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 10804 + }, + { + "epoch": 0.10805, + "grad_norm": 0.8959097266197205, + "learning_rate": 0.003, + "loss": 4.033, + "step": 10805 + }, + { + "epoch": 0.10806, + "grad_norm": 0.8214556574821472, + "learning_rate": 0.003, + "loss": 4.012, + "step": 10806 + }, + { + "epoch": 0.10807, + "grad_norm": 0.8318617343902588, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 10807 + }, + { + "epoch": 0.10808, + "grad_norm": 0.8079638481140137, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 10808 + }, + { + "epoch": 0.10809, + "grad_norm": 0.7618487477302551, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 10809 + }, + { + "epoch": 0.1081, + "grad_norm": 0.8443053960800171, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 10810 + }, + { + "epoch": 0.10811, + "grad_norm": 0.8693640232086182, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 10811 + }, + { + "epoch": 0.10812, + "grad_norm": 0.802323579788208, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 10812 + }, + { + "epoch": 0.10813, + "grad_norm": 0.8824979662895203, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 10813 + }, + { + "epoch": 0.10814, + "grad_norm": 1.0556713342666626, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 10814 + }, + { + "epoch": 0.10815, + "grad_norm": 1.0019479990005493, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 10815 + }, + { + "epoch": 0.10816, + "grad_norm": 1.1382358074188232, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 10816 + }, + { + "epoch": 0.10817, + "grad_norm": 0.7958524227142334, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 10817 + }, + { + "epoch": 0.10818, + "grad_norm": 0.6067748665809631, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 10818 + }, + { + "epoch": 0.10819, + "grad_norm": 0.6149799227714539, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 10819 + }, + { + "epoch": 0.1082, + "grad_norm": 0.6444442272186279, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 10820 + }, + { + "epoch": 0.10821, + "grad_norm": 0.7835311889648438, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 10821 + }, + { + "epoch": 0.10822, + "grad_norm": 0.9068350195884705, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 10822 + }, + { + "epoch": 0.10823, + "grad_norm": 0.9789934754371643, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 10823 + }, + { + "epoch": 0.10824, + "grad_norm": 0.9751971364021301, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 10824 + }, + { + "epoch": 0.10825, + "grad_norm": 0.7828039526939392, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 10825 + }, + { + "epoch": 0.10826, + "grad_norm": 0.6331886649131775, + "learning_rate": 0.003, + "loss": 3.99, + "step": 10826 + }, + { + "epoch": 0.10827, + "grad_norm": 0.6150903105735779, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 10827 + }, + { + "epoch": 0.10828, + "grad_norm": 0.5984993577003479, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 10828 + }, + { + "epoch": 0.10829, + "grad_norm": 0.7864992022514343, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 10829 + }, + { + "epoch": 0.1083, + "grad_norm": 0.9425749778747559, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 10830 + }, + { + "epoch": 0.10831, + "grad_norm": 1.058080792427063, + "learning_rate": 0.003, + "loss": 4.021, + "step": 10831 + }, + { + "epoch": 0.10832, + "grad_norm": 0.7845868468284607, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 10832 + }, + { + "epoch": 0.10833, + "grad_norm": 0.6163697242736816, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 10833 + }, + { + "epoch": 0.10834, + "grad_norm": 0.6821464896202087, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 10834 + }, + { + "epoch": 0.10835, + "grad_norm": 0.6944262981414795, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 10835 + }, + { + "epoch": 0.10836, + "grad_norm": 0.6780229806900024, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 10836 + }, + { + "epoch": 0.10837, + "grad_norm": 0.703480064868927, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 10837 + }, + { + "epoch": 0.10838, + "grad_norm": 0.8821530938148499, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 10838 + }, + { + "epoch": 0.10839, + "grad_norm": 1.2261230945587158, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 10839 + }, + { + "epoch": 0.1084, + "grad_norm": 0.6746938228607178, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 10840 + }, + { + "epoch": 0.10841, + "grad_norm": 0.5455398559570312, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 10841 + }, + { + "epoch": 0.10842, + "grad_norm": 0.5697435736656189, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 10842 + }, + { + "epoch": 0.10843, + "grad_norm": 0.557073175907135, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 10843 + }, + { + "epoch": 0.10844, + "grad_norm": 0.5535928606987, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 10844 + }, + { + "epoch": 0.10845, + "grad_norm": 0.6609616875648499, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 10845 + }, + { + "epoch": 0.10846, + "grad_norm": 0.7913950085639954, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 10846 + }, + { + "epoch": 0.10847, + "grad_norm": 0.8395732045173645, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 10847 + }, + { + "epoch": 0.10848, + "grad_norm": 0.7630530595779419, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 10848 + }, + { + "epoch": 0.10849, + "grad_norm": 0.7382761836051941, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 10849 + }, + { + "epoch": 0.1085, + "grad_norm": 0.7676743865013123, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 10850 + }, + { + "epoch": 0.10851, + "grad_norm": 0.7738227248191833, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 10851 + }, + { + "epoch": 0.10852, + "grad_norm": 0.727118968963623, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 10852 + }, + { + "epoch": 0.10853, + "grad_norm": 0.5983321070671082, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 10853 + }, + { + "epoch": 0.10854, + "grad_norm": 0.5814149379730225, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 10854 + }, + { + "epoch": 0.10855, + "grad_norm": 0.7128687500953674, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 10855 + }, + { + "epoch": 0.10856, + "grad_norm": 0.8423973321914673, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 10856 + }, + { + "epoch": 0.10857, + "grad_norm": 1.0030944347381592, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 10857 + }, + { + "epoch": 0.10858, + "grad_norm": 1.2414257526397705, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 10858 + }, + { + "epoch": 0.10859, + "grad_norm": 0.8019863963127136, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 10859 + }, + { + "epoch": 0.1086, + "grad_norm": 0.7122812867164612, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 10860 + }, + { + "epoch": 0.10861, + "grad_norm": 0.6793990135192871, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 10861 + }, + { + "epoch": 0.10862, + "grad_norm": 0.7667950987815857, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 10862 + }, + { + "epoch": 0.10863, + "grad_norm": 0.7574916481971741, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 10863 + }, + { + "epoch": 0.10864, + "grad_norm": 0.7576637864112854, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 10864 + }, + { + "epoch": 0.10865, + "grad_norm": 0.7982226014137268, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 10865 + }, + { + "epoch": 0.10866, + "grad_norm": 0.831477165222168, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 10866 + }, + { + "epoch": 0.10867, + "grad_norm": 0.8126018643379211, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 10867 + }, + { + "epoch": 0.10868, + "grad_norm": 0.7804986834526062, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 10868 + }, + { + "epoch": 0.10869, + "grad_norm": 0.8123776912689209, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 10869 + }, + { + "epoch": 0.1087, + "grad_norm": 0.9683001041412354, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 10870 + }, + { + "epoch": 0.10871, + "grad_norm": 1.090463399887085, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 10871 + }, + { + "epoch": 0.10872, + "grad_norm": 0.8968812823295593, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 10872 + }, + { + "epoch": 0.10873, + "grad_norm": 0.8717519640922546, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 10873 + }, + { + "epoch": 0.10874, + "grad_norm": 0.7509618997573853, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 10874 + }, + { + "epoch": 0.10875, + "grad_norm": 0.6578442454338074, + "learning_rate": 0.003, + "loss": 4.043, + "step": 10875 + }, + { + "epoch": 0.10876, + "grad_norm": 0.6749911308288574, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 10876 + }, + { + "epoch": 0.10877, + "grad_norm": 0.7552892565727234, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 10877 + }, + { + "epoch": 0.10878, + "grad_norm": 0.9866586923599243, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 10878 + }, + { + "epoch": 0.10879, + "grad_norm": 1.1038225889205933, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 10879 + }, + { + "epoch": 0.1088, + "grad_norm": 0.9424269795417786, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 10880 + }, + { + "epoch": 0.10881, + "grad_norm": 0.931976318359375, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 10881 + }, + { + "epoch": 0.10882, + "grad_norm": 0.9206244349479675, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 10882 + }, + { + "epoch": 0.10883, + "grad_norm": 0.8365570306777954, + "learning_rate": 0.003, + "loss": 4.05, + "step": 10883 + }, + { + "epoch": 0.10884, + "grad_norm": 0.8806115388870239, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 10884 + }, + { + "epoch": 0.10885, + "grad_norm": 0.71012282371521, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10885 + }, + { + "epoch": 0.10886, + "grad_norm": 0.6836910247802734, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 10886 + }, + { + "epoch": 0.10887, + "grad_norm": 0.590783417224884, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 10887 + }, + { + "epoch": 0.10888, + "grad_norm": 0.6044482588768005, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 10888 + }, + { + "epoch": 0.10889, + "grad_norm": 0.5061213970184326, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 10889 + }, + { + "epoch": 0.1089, + "grad_norm": 0.4831678867340088, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 10890 + }, + { + "epoch": 0.10891, + "grad_norm": 0.5567334294319153, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 10891 + }, + { + "epoch": 0.10892, + "grad_norm": 0.7139768004417419, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 10892 + }, + { + "epoch": 0.10893, + "grad_norm": 0.939973771572113, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 10893 + }, + { + "epoch": 0.10894, + "grad_norm": 1.1920320987701416, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 10894 + }, + { + "epoch": 0.10895, + "grad_norm": 0.6929920315742493, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 10895 + }, + { + "epoch": 0.10896, + "grad_norm": 0.6842626333236694, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 10896 + }, + { + "epoch": 0.10897, + "grad_norm": 0.7854790091514587, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 10897 + }, + { + "epoch": 0.10898, + "grad_norm": 0.9311167597770691, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 10898 + }, + { + "epoch": 0.10899, + "grad_norm": 1.1679928302764893, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 10899 + }, + { + "epoch": 0.109, + "grad_norm": 0.8089187145233154, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 10900 + }, + { + "epoch": 0.10901, + "grad_norm": 0.6758765578269958, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 10901 + }, + { + "epoch": 0.10902, + "grad_norm": 0.7646622061729431, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 10902 + }, + { + "epoch": 0.10903, + "grad_norm": 0.7202122807502747, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 10903 + }, + { + "epoch": 0.10904, + "grad_norm": 0.7561991810798645, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 10904 + }, + { + "epoch": 0.10905, + "grad_norm": 0.5963756442070007, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 10905 + }, + { + "epoch": 0.10906, + "grad_norm": 0.5371752977371216, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 10906 + }, + { + "epoch": 0.10907, + "grad_norm": 0.5870004892349243, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 10907 + }, + { + "epoch": 0.10908, + "grad_norm": 0.6057379841804504, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 10908 + }, + { + "epoch": 0.10909, + "grad_norm": 0.593961238861084, + "learning_rate": 0.003, + "loss": 4.006, + "step": 10909 + }, + { + "epoch": 0.1091, + "grad_norm": 0.6798774600028992, + "learning_rate": 0.003, + "loss": 4.015, + "step": 10910 + }, + { + "epoch": 0.10911, + "grad_norm": 0.8016900420188904, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 10911 + }, + { + "epoch": 0.10912, + "grad_norm": 0.8332940340042114, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 10912 + }, + { + "epoch": 0.10913, + "grad_norm": 0.9600593447685242, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 10913 + }, + { + "epoch": 0.10914, + "grad_norm": 1.1089653968811035, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 10914 + }, + { + "epoch": 0.10915, + "grad_norm": 0.9480727314949036, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 10915 + }, + { + "epoch": 0.10916, + "grad_norm": 0.894866406917572, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 10916 + }, + { + "epoch": 0.10917, + "grad_norm": 0.8441586494445801, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 10917 + }, + { + "epoch": 0.10918, + "grad_norm": 0.9097299575805664, + "learning_rate": 0.003, + "loss": 4.036, + "step": 10918 + }, + { + "epoch": 0.10919, + "grad_norm": 0.9293150305747986, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 10919 + }, + { + "epoch": 0.1092, + "grad_norm": 1.0343148708343506, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 10920 + }, + { + "epoch": 0.10921, + "grad_norm": 1.063411831855774, + "learning_rate": 0.003, + "loss": 4.043, + "step": 10921 + }, + { + "epoch": 0.10922, + "grad_norm": 0.8918190598487854, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 10922 + }, + { + "epoch": 0.10923, + "grad_norm": 0.7885593771934509, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 10923 + }, + { + "epoch": 0.10924, + "grad_norm": 0.6758106350898743, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 10924 + }, + { + "epoch": 0.10925, + "grad_norm": 0.73740553855896, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 10925 + }, + { + "epoch": 0.10926, + "grad_norm": 0.7766514420509338, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 10926 + }, + { + "epoch": 0.10927, + "grad_norm": 0.8272937536239624, + "learning_rate": 0.003, + "loss": 4.038, + "step": 10927 + }, + { + "epoch": 0.10928, + "grad_norm": 0.8774792551994324, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 10928 + }, + { + "epoch": 0.10929, + "grad_norm": 1.0164790153503418, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 10929 + }, + { + "epoch": 0.1093, + "grad_norm": 1.0911238193511963, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 10930 + }, + { + "epoch": 0.10931, + "grad_norm": 0.8725080490112305, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 10931 + }, + { + "epoch": 0.10932, + "grad_norm": 0.6738268733024597, + "learning_rate": 0.003, + "loss": 4.029, + "step": 10932 + }, + { + "epoch": 0.10933, + "grad_norm": 0.6225630640983582, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 10933 + }, + { + "epoch": 0.10934, + "grad_norm": 0.7357386946678162, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 10934 + }, + { + "epoch": 0.10935, + "grad_norm": 0.7856181859970093, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 10935 + }, + { + "epoch": 0.10936, + "grad_norm": 0.6953492164611816, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 10936 + }, + { + "epoch": 0.10937, + "grad_norm": 0.6342230439186096, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 10937 + }, + { + "epoch": 0.10938, + "grad_norm": 0.8185663223266602, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 10938 + }, + { + "epoch": 0.10939, + "grad_norm": 0.9318892955780029, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 10939 + }, + { + "epoch": 0.1094, + "grad_norm": 0.8709768056869507, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 10940 + }, + { + "epoch": 0.10941, + "grad_norm": 1.0035088062286377, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 10941 + }, + { + "epoch": 0.10942, + "grad_norm": 0.9826143383979797, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 10942 + }, + { + "epoch": 0.10943, + "grad_norm": 0.8794414401054382, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 10943 + }, + { + "epoch": 0.10944, + "grad_norm": 0.8242260217666626, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 10944 + }, + { + "epoch": 0.10945, + "grad_norm": 0.7647247314453125, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 10945 + }, + { + "epoch": 0.10946, + "grad_norm": 0.7977977991104126, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 10946 + }, + { + "epoch": 0.10947, + "grad_norm": 0.7217475175857544, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 10947 + }, + { + "epoch": 0.10948, + "grad_norm": 0.68278568983078, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 10948 + }, + { + "epoch": 0.10949, + "grad_norm": 0.6589925289154053, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 10949 + }, + { + "epoch": 0.1095, + "grad_norm": 0.733100414276123, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 10950 + }, + { + "epoch": 0.10951, + "grad_norm": 0.9500106573104858, + "learning_rate": 0.003, + "loss": 4.01, + "step": 10951 + }, + { + "epoch": 0.10952, + "grad_norm": 1.2109297513961792, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 10952 + }, + { + "epoch": 0.10953, + "grad_norm": 0.8722065091133118, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 10953 + }, + { + "epoch": 0.10954, + "grad_norm": 0.7249215841293335, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 10954 + }, + { + "epoch": 0.10955, + "grad_norm": 0.6058104038238525, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 10955 + }, + { + "epoch": 0.10956, + "grad_norm": 0.6156740784645081, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 10956 + }, + { + "epoch": 0.10957, + "grad_norm": 0.6245015859603882, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 10957 + }, + { + "epoch": 0.10958, + "grad_norm": 0.740548849105835, + "learning_rate": 0.003, + "loss": 4.019, + "step": 10958 + }, + { + "epoch": 0.10959, + "grad_norm": 1.0325299501419067, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 10959 + }, + { + "epoch": 0.1096, + "grad_norm": 1.0633299350738525, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 10960 + }, + { + "epoch": 0.10961, + "grad_norm": 0.7050127983093262, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 10961 + }, + { + "epoch": 0.10962, + "grad_norm": 0.6579815149307251, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 10962 + }, + { + "epoch": 0.10963, + "grad_norm": 0.8424527645111084, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 10963 + }, + { + "epoch": 0.10964, + "grad_norm": 0.9569191336631775, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 10964 + }, + { + "epoch": 0.10965, + "grad_norm": 1.1351628303527832, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 10965 + }, + { + "epoch": 0.10966, + "grad_norm": 0.7953414916992188, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 10966 + }, + { + "epoch": 0.10967, + "grad_norm": 0.6877783536911011, + "learning_rate": 0.003, + "loss": 4.041, + "step": 10967 + }, + { + "epoch": 0.10968, + "grad_norm": 0.7406366467475891, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 10968 + }, + { + "epoch": 0.10969, + "grad_norm": 0.7145115733146667, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 10969 + }, + { + "epoch": 0.1097, + "grad_norm": 0.6158110499382019, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 10970 + }, + { + "epoch": 0.10971, + "grad_norm": 0.5443043112754822, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 10971 + }, + { + "epoch": 0.10972, + "grad_norm": 0.542644202709198, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 10972 + }, + { + "epoch": 0.10973, + "grad_norm": 0.6166226863861084, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 10973 + }, + { + "epoch": 0.10974, + "grad_norm": 0.6657989025115967, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 10974 + }, + { + "epoch": 0.10975, + "grad_norm": 0.7187350988388062, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 10975 + }, + { + "epoch": 0.10976, + "grad_norm": 0.7222906351089478, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 10976 + }, + { + "epoch": 0.10977, + "grad_norm": 0.8421234488487244, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 10977 + }, + { + "epoch": 0.10978, + "grad_norm": 1.1006399393081665, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 10978 + }, + { + "epoch": 0.10979, + "grad_norm": 0.8633862733840942, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 10979 + }, + { + "epoch": 0.1098, + "grad_norm": 0.6699802875518799, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 10980 + }, + { + "epoch": 0.10981, + "grad_norm": 0.6813558340072632, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 10981 + }, + { + "epoch": 0.10982, + "grad_norm": 0.9310502409934998, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 10982 + }, + { + "epoch": 0.10983, + "grad_norm": 1.1476507186889648, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 10983 + }, + { + "epoch": 0.10984, + "grad_norm": 0.706564724445343, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 10984 + }, + { + "epoch": 0.10985, + "grad_norm": 0.584486722946167, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 10985 + }, + { + "epoch": 0.10986, + "grad_norm": 0.6373727917671204, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 10986 + }, + { + "epoch": 0.10987, + "grad_norm": 0.715704083442688, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 10987 + }, + { + "epoch": 0.10988, + "grad_norm": 0.8571540713310242, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 10988 + }, + { + "epoch": 0.10989, + "grad_norm": 0.8616876602172852, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 10989 + }, + { + "epoch": 0.1099, + "grad_norm": 0.7467997670173645, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 10990 + }, + { + "epoch": 0.10991, + "grad_norm": 0.7104257941246033, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 10991 + }, + { + "epoch": 0.10992, + "grad_norm": 0.8819006085395813, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 10992 + }, + { + "epoch": 0.10993, + "grad_norm": 1.1179277896881104, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 10993 + }, + { + "epoch": 0.10994, + "grad_norm": 0.7531784772872925, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 10994 + }, + { + "epoch": 0.10995, + "grad_norm": 0.6037430763244629, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 10995 + }, + { + "epoch": 0.10996, + "grad_norm": 0.5394481420516968, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 10996 + }, + { + "epoch": 0.10997, + "grad_norm": 0.4965377449989319, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 10997 + }, + { + "epoch": 0.10998, + "grad_norm": 0.4652479290962219, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 10998 + }, + { + "epoch": 0.10999, + "grad_norm": 0.47271811962127686, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 10999 + }, + { + "epoch": 0.11, + "grad_norm": 0.5026476383209229, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 11000 + }, + { + "epoch": 0.11001, + "grad_norm": 0.600214958190918, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 11001 + }, + { + "epoch": 0.11002, + "grad_norm": 0.7847859859466553, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 11002 + }, + { + "epoch": 0.11003, + "grad_norm": 1.0072622299194336, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 11003 + }, + { + "epoch": 0.11004, + "grad_norm": 1.0832382440567017, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 11004 + }, + { + "epoch": 0.11005, + "grad_norm": 0.8995561003684998, + "learning_rate": 0.003, + "loss": 4.002, + "step": 11005 + }, + { + "epoch": 0.11006, + "grad_norm": 0.940889298915863, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 11006 + }, + { + "epoch": 0.11007, + "grad_norm": 0.9138043522834778, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 11007 + }, + { + "epoch": 0.11008, + "grad_norm": 0.8481074571609497, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 11008 + }, + { + "epoch": 0.11009, + "grad_norm": 0.9567734599113464, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 11009 + }, + { + "epoch": 0.1101, + "grad_norm": 1.285965919494629, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 11010 + }, + { + "epoch": 0.11011, + "grad_norm": 0.8564926981925964, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 11011 + }, + { + "epoch": 0.11012, + "grad_norm": 0.8530118465423584, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 11012 + }, + { + "epoch": 0.11013, + "grad_norm": 0.8939397931098938, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 11013 + }, + { + "epoch": 0.11014, + "grad_norm": 0.8238596320152283, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 11014 + }, + { + "epoch": 0.11015, + "grad_norm": 0.8376443386077881, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 11015 + }, + { + "epoch": 0.11016, + "grad_norm": 0.7546934485435486, + "learning_rate": 0.003, + "loss": 4.033, + "step": 11016 + }, + { + "epoch": 0.11017, + "grad_norm": 0.7620981931686401, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 11017 + }, + { + "epoch": 0.11018, + "grad_norm": 0.7299874424934387, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 11018 + }, + { + "epoch": 0.11019, + "grad_norm": 0.6225110292434692, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 11019 + }, + { + "epoch": 0.1102, + "grad_norm": 0.6360118389129639, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 11020 + }, + { + "epoch": 0.11021, + "grad_norm": 0.6085506081581116, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 11021 + }, + { + "epoch": 0.11022, + "grad_norm": 0.6367476582527161, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 11022 + }, + { + "epoch": 0.11023, + "grad_norm": 0.7030669450759888, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 11023 + }, + { + "epoch": 0.11024, + "grad_norm": 0.7498410940170288, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 11024 + }, + { + "epoch": 0.11025, + "grad_norm": 0.8611958622932434, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 11025 + }, + { + "epoch": 0.11026, + "grad_norm": 0.8228955268859863, + "learning_rate": 0.003, + "loss": 4.021, + "step": 11026 + }, + { + "epoch": 0.11027, + "grad_norm": 0.6890997290611267, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 11027 + }, + { + "epoch": 0.11028, + "grad_norm": 0.7558037638664246, + "learning_rate": 0.003, + "loss": 4.016, + "step": 11028 + }, + { + "epoch": 0.11029, + "grad_norm": 0.8738684058189392, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 11029 + }, + { + "epoch": 0.1103, + "grad_norm": 0.8827449083328247, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 11030 + }, + { + "epoch": 0.11031, + "grad_norm": 0.7657271027565002, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 11031 + }, + { + "epoch": 0.11032, + "grad_norm": 0.7138214111328125, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 11032 + }, + { + "epoch": 0.11033, + "grad_norm": 0.7493338584899902, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 11033 + }, + { + "epoch": 0.11034, + "grad_norm": 0.7448801398277283, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 11034 + }, + { + "epoch": 0.11035, + "grad_norm": 0.66679447889328, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 11035 + }, + { + "epoch": 0.11036, + "grad_norm": 0.6494250297546387, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 11036 + }, + { + "epoch": 0.11037, + "grad_norm": 0.700835108757019, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 11037 + }, + { + "epoch": 0.11038, + "grad_norm": 0.7485676407814026, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 11038 + }, + { + "epoch": 0.11039, + "grad_norm": 0.6627346277236938, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 11039 + }, + { + "epoch": 0.1104, + "grad_norm": 0.6443310379981995, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 11040 + }, + { + "epoch": 0.11041, + "grad_norm": 0.6941895484924316, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 11041 + }, + { + "epoch": 0.11042, + "grad_norm": 0.7193214893341064, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 11042 + }, + { + "epoch": 0.11043, + "grad_norm": 0.8162452578544617, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 11043 + }, + { + "epoch": 0.11044, + "grad_norm": 0.9481807351112366, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 11044 + }, + { + "epoch": 0.11045, + "grad_norm": 0.7185466885566711, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 11045 + }, + { + "epoch": 0.11046, + "grad_norm": 0.7170248031616211, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 11046 + }, + { + "epoch": 0.11047, + "grad_norm": 0.8825567364692688, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 11047 + }, + { + "epoch": 0.11048, + "grad_norm": 1.2538373470306396, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 11048 + }, + { + "epoch": 0.11049, + "grad_norm": 0.856818675994873, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 11049 + }, + { + "epoch": 0.1105, + "grad_norm": 0.7342973351478577, + "learning_rate": 0.003, + "loss": 4.02, + "step": 11050 + }, + { + "epoch": 0.11051, + "grad_norm": 0.8246771693229675, + "learning_rate": 0.003, + "loss": 4.037, + "step": 11051 + }, + { + "epoch": 0.11052, + "grad_norm": 1.0061192512512207, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 11052 + }, + { + "epoch": 0.11053, + "grad_norm": 1.0631734132766724, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 11053 + }, + { + "epoch": 0.11054, + "grad_norm": 1.0199459791183472, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 11054 + }, + { + "epoch": 0.11055, + "grad_norm": 0.9731594920158386, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 11055 + }, + { + "epoch": 0.11056, + "grad_norm": 0.8760973215103149, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 11056 + }, + { + "epoch": 0.11057, + "grad_norm": 0.789323627948761, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 11057 + }, + { + "epoch": 0.11058, + "grad_norm": 0.8253927826881409, + "learning_rate": 0.003, + "loss": 4.051, + "step": 11058 + }, + { + "epoch": 0.11059, + "grad_norm": 0.766977071762085, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 11059 + }, + { + "epoch": 0.1106, + "grad_norm": 0.7662436366081238, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 11060 + }, + { + "epoch": 0.11061, + "grad_norm": 0.8208341002464294, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 11061 + }, + { + "epoch": 0.11062, + "grad_norm": 0.8970367908477783, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 11062 + }, + { + "epoch": 0.11063, + "grad_norm": 1.0099105834960938, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 11063 + }, + { + "epoch": 0.11064, + "grad_norm": 0.9661181569099426, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 11064 + }, + { + "epoch": 0.11065, + "grad_norm": 1.0960055589675903, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 11065 + }, + { + "epoch": 0.11066, + "grad_norm": 0.9327487945556641, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 11066 + }, + { + "epoch": 0.11067, + "grad_norm": 0.7413168549537659, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 11067 + }, + { + "epoch": 0.11068, + "grad_norm": 0.6693038940429688, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 11068 + }, + { + "epoch": 0.11069, + "grad_norm": 0.6442914605140686, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 11069 + }, + { + "epoch": 0.1107, + "grad_norm": 0.6646853089332581, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 11070 + }, + { + "epoch": 0.11071, + "grad_norm": 0.5942994356155396, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 11071 + }, + { + "epoch": 0.11072, + "grad_norm": 0.547498345375061, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 11072 + }, + { + "epoch": 0.11073, + "grad_norm": 0.5513029098510742, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 11073 + }, + { + "epoch": 0.11074, + "grad_norm": 0.6343658566474915, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 11074 + }, + { + "epoch": 0.11075, + "grad_norm": 0.7820315957069397, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 11075 + }, + { + "epoch": 0.11076, + "grad_norm": 0.9613561630249023, + "learning_rate": 0.003, + "loss": 4.02, + "step": 11076 + }, + { + "epoch": 0.11077, + "grad_norm": 1.0932581424713135, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 11077 + }, + { + "epoch": 0.11078, + "grad_norm": 0.9842900633811951, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 11078 + }, + { + "epoch": 0.11079, + "grad_norm": 0.8640671968460083, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 11079 + }, + { + "epoch": 0.1108, + "grad_norm": 0.7365666627883911, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 11080 + }, + { + "epoch": 0.11081, + "grad_norm": 0.7927987575531006, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 11081 + }, + { + "epoch": 0.11082, + "grad_norm": 0.8462502360343933, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 11082 + }, + { + "epoch": 0.11083, + "grad_norm": 0.8290241360664368, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 11083 + }, + { + "epoch": 0.11084, + "grad_norm": 0.7888922095298767, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 11084 + }, + { + "epoch": 0.11085, + "grad_norm": 0.6967290639877319, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 11085 + }, + { + "epoch": 0.11086, + "grad_norm": 0.7021198868751526, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 11086 + }, + { + "epoch": 0.11087, + "grad_norm": 0.7120110988616943, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 11087 + }, + { + "epoch": 0.11088, + "grad_norm": 0.6755147576332092, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 11088 + }, + { + "epoch": 0.11089, + "grad_norm": 0.8151680827140808, + "learning_rate": 0.003, + "loss": 4.018, + "step": 11089 + }, + { + "epoch": 0.1109, + "grad_norm": 1.1533037424087524, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 11090 + }, + { + "epoch": 0.11091, + "grad_norm": 0.9690134525299072, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 11091 + }, + { + "epoch": 0.11092, + "grad_norm": 0.8287317752838135, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 11092 + }, + { + "epoch": 0.11093, + "grad_norm": 0.6802783608436584, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 11093 + }, + { + "epoch": 0.11094, + "grad_norm": 0.5430587530136108, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 11094 + }, + { + "epoch": 0.11095, + "grad_norm": 0.4955750107765198, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 11095 + }, + { + "epoch": 0.11096, + "grad_norm": 0.5845931768417358, + "learning_rate": 0.003, + "loss": 4.025, + "step": 11096 + }, + { + "epoch": 0.11097, + "grad_norm": 0.7685054540634155, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 11097 + }, + { + "epoch": 0.11098, + "grad_norm": 0.9369683265686035, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 11098 + }, + { + "epoch": 0.11099, + "grad_norm": 1.0887501239776611, + "learning_rate": 0.003, + "loss": 4.052, + "step": 11099 + }, + { + "epoch": 0.111, + "grad_norm": 0.7608708143234253, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 11100 + }, + { + "epoch": 0.11101, + "grad_norm": 0.6959680318832397, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 11101 + }, + { + "epoch": 0.11102, + "grad_norm": 1.0319355726242065, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 11102 + }, + { + "epoch": 0.11103, + "grad_norm": 1.1587380170822144, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 11103 + }, + { + "epoch": 0.11104, + "grad_norm": 0.6529046297073364, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 11104 + }, + { + "epoch": 0.11105, + "grad_norm": 0.6380882263183594, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 11105 + }, + { + "epoch": 0.11106, + "grad_norm": 0.801237940788269, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 11106 + }, + { + "epoch": 0.11107, + "grad_norm": 1.0300419330596924, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 11107 + }, + { + "epoch": 0.11108, + "grad_norm": 0.8984727263450623, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 11108 + }, + { + "epoch": 0.11109, + "grad_norm": 0.7662938237190247, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 11109 + }, + { + "epoch": 0.1111, + "grad_norm": 0.7453399300575256, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 11110 + }, + { + "epoch": 0.11111, + "grad_norm": 0.815127968788147, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 11111 + }, + { + "epoch": 0.11112, + "grad_norm": 0.7698167562484741, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 11112 + }, + { + "epoch": 0.11113, + "grad_norm": 0.6643550992012024, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 11113 + }, + { + "epoch": 0.11114, + "grad_norm": 0.5844383239746094, + "learning_rate": 0.003, + "loss": 4.056, + "step": 11114 + }, + { + "epoch": 0.11115, + "grad_norm": 0.6911113262176514, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 11115 + }, + { + "epoch": 0.11116, + "grad_norm": 0.7149133086204529, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 11116 + }, + { + "epoch": 0.11117, + "grad_norm": 0.6734727025032043, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 11117 + }, + { + "epoch": 0.11118, + "grad_norm": 0.6846103668212891, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 11118 + }, + { + "epoch": 0.11119, + "grad_norm": 0.7427072525024414, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 11119 + }, + { + "epoch": 0.1112, + "grad_norm": 0.8775559067726135, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 11120 + }, + { + "epoch": 0.11121, + "grad_norm": 1.1779645681381226, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 11121 + }, + { + "epoch": 0.11122, + "grad_norm": 0.8917052149772644, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 11122 + }, + { + "epoch": 0.11123, + "grad_norm": 0.8060603141784668, + "learning_rate": 0.003, + "loss": 4.046, + "step": 11123 + }, + { + "epoch": 0.11124, + "grad_norm": 0.8351340293884277, + "learning_rate": 0.003, + "loss": 4.021, + "step": 11124 + }, + { + "epoch": 0.11125, + "grad_norm": 0.7737427353858948, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 11125 + }, + { + "epoch": 0.11126, + "grad_norm": 0.7786157131195068, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 11126 + }, + { + "epoch": 0.11127, + "grad_norm": 0.8442794680595398, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 11127 + }, + { + "epoch": 0.11128, + "grad_norm": 0.9874286651611328, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 11128 + }, + { + "epoch": 0.11129, + "grad_norm": 1.0912935733795166, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 11129 + }, + { + "epoch": 0.1113, + "grad_norm": 0.9720292091369629, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 11130 + }, + { + "epoch": 0.11131, + "grad_norm": 0.866678774356842, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 11131 + }, + { + "epoch": 0.11132, + "grad_norm": 0.9276267886161804, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 11132 + }, + { + "epoch": 0.11133, + "grad_norm": 1.0052250623703003, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 11133 + }, + { + "epoch": 0.11134, + "grad_norm": 0.9750459790229797, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 11134 + }, + { + "epoch": 0.11135, + "grad_norm": 1.0204553604125977, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 11135 + }, + { + "epoch": 0.11136, + "grad_norm": 1.1887991428375244, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 11136 + }, + { + "epoch": 0.11137, + "grad_norm": 0.8008458018302917, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 11137 + }, + { + "epoch": 0.11138, + "grad_norm": 0.6591132879257202, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 11138 + }, + { + "epoch": 0.11139, + "grad_norm": 0.5977212190628052, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 11139 + }, + { + "epoch": 0.1114, + "grad_norm": 0.5454531311988831, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 11140 + }, + { + "epoch": 0.11141, + "grad_norm": 0.535681962966919, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 11141 + }, + { + "epoch": 0.11142, + "grad_norm": 0.5493536591529846, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 11142 + }, + { + "epoch": 0.11143, + "grad_norm": 0.6497412323951721, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 11143 + }, + { + "epoch": 0.11144, + "grad_norm": 0.7990849614143372, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 11144 + }, + { + "epoch": 0.11145, + "grad_norm": 0.7726136445999146, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 11145 + }, + { + "epoch": 0.11146, + "grad_norm": 0.6456999182701111, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 11146 + }, + { + "epoch": 0.11147, + "grad_norm": 0.6681917905807495, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 11147 + }, + { + "epoch": 0.11148, + "grad_norm": 0.6563037633895874, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 11148 + }, + { + "epoch": 0.11149, + "grad_norm": 0.6741673350334167, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 11149 + }, + { + "epoch": 0.1115, + "grad_norm": 0.8483068943023682, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 11150 + }, + { + "epoch": 0.11151, + "grad_norm": 1.0420079231262207, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 11151 + }, + { + "epoch": 0.11152, + "grad_norm": 1.0746169090270996, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 11152 + }, + { + "epoch": 0.11153, + "grad_norm": 0.8414022922515869, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 11153 + }, + { + "epoch": 0.11154, + "grad_norm": 0.7460391521453857, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 11154 + }, + { + "epoch": 0.11155, + "grad_norm": 0.8425396084785461, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 11155 + }, + { + "epoch": 0.11156, + "grad_norm": 0.7851555347442627, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 11156 + }, + { + "epoch": 0.11157, + "grad_norm": 0.7829062938690186, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 11157 + }, + { + "epoch": 0.11158, + "grad_norm": 0.8758364915847778, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 11158 + }, + { + "epoch": 0.11159, + "grad_norm": 0.8003804683685303, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 11159 + }, + { + "epoch": 0.1116, + "grad_norm": 0.7548210620880127, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 11160 + }, + { + "epoch": 0.11161, + "grad_norm": 0.7185114026069641, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 11161 + }, + { + "epoch": 0.11162, + "grad_norm": 0.6627770066261292, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 11162 + }, + { + "epoch": 0.11163, + "grad_norm": 0.6781023144721985, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 11163 + }, + { + "epoch": 0.11164, + "grad_norm": 0.7373916506767273, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 11164 + }, + { + "epoch": 0.11165, + "grad_norm": 0.8808543086051941, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 11165 + }, + { + "epoch": 0.11166, + "grad_norm": 0.934785783290863, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 11166 + }, + { + "epoch": 0.11167, + "grad_norm": 0.8570555448532104, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 11167 + }, + { + "epoch": 0.11168, + "grad_norm": 0.8547791242599487, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 11168 + }, + { + "epoch": 0.11169, + "grad_norm": 0.9320052266120911, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 11169 + }, + { + "epoch": 0.1117, + "grad_norm": 1.164999008178711, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 11170 + }, + { + "epoch": 0.11171, + "grad_norm": 0.9276113510131836, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 11171 + }, + { + "epoch": 0.11172, + "grad_norm": 0.8860996961593628, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 11172 + }, + { + "epoch": 0.11173, + "grad_norm": 0.9282903671264648, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 11173 + }, + { + "epoch": 0.11174, + "grad_norm": 0.8367762565612793, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 11174 + }, + { + "epoch": 0.11175, + "grad_norm": 0.7053303122520447, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 11175 + }, + { + "epoch": 0.11176, + "grad_norm": 0.6328437328338623, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 11176 + }, + { + "epoch": 0.11177, + "grad_norm": 0.6176173686981201, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 11177 + }, + { + "epoch": 0.11178, + "grad_norm": 0.6604607105255127, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 11178 + }, + { + "epoch": 0.11179, + "grad_norm": 0.6398141980171204, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 11179 + }, + { + "epoch": 0.1118, + "grad_norm": 0.5881091952323914, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 11180 + }, + { + "epoch": 0.11181, + "grad_norm": 0.5375813841819763, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 11181 + }, + { + "epoch": 0.11182, + "grad_norm": 0.5335777997970581, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 11182 + }, + { + "epoch": 0.11183, + "grad_norm": 0.4691711366176605, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 11183 + }, + { + "epoch": 0.11184, + "grad_norm": 0.524629533290863, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 11184 + }, + { + "epoch": 0.11185, + "grad_norm": 0.6009156703948975, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 11185 + }, + { + "epoch": 0.11186, + "grad_norm": 0.8225870132446289, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 11186 + }, + { + "epoch": 0.11187, + "grad_norm": 1.3781869411468506, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 11187 + }, + { + "epoch": 0.11188, + "grad_norm": 0.8902261257171631, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 11188 + }, + { + "epoch": 0.11189, + "grad_norm": 0.6796999573707581, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 11189 + }, + { + "epoch": 0.1119, + "grad_norm": 0.5967490673065186, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 11190 + }, + { + "epoch": 0.11191, + "grad_norm": 0.7360259890556335, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 11191 + }, + { + "epoch": 0.11192, + "grad_norm": 0.8131880164146423, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 11192 + }, + { + "epoch": 0.11193, + "grad_norm": 0.7584455609321594, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 11193 + }, + { + "epoch": 0.11194, + "grad_norm": 0.6299670934677124, + "learning_rate": 0.003, + "loss": 4.05, + "step": 11194 + }, + { + "epoch": 0.11195, + "grad_norm": 0.6245558857917786, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 11195 + }, + { + "epoch": 0.11196, + "grad_norm": 0.7212068438529968, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 11196 + }, + { + "epoch": 0.11197, + "grad_norm": 0.7425583600997925, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 11197 + }, + { + "epoch": 0.11198, + "grad_norm": 0.8863121271133423, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 11198 + }, + { + "epoch": 0.11199, + "grad_norm": 0.9826465845108032, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 11199 + }, + { + "epoch": 0.112, + "grad_norm": 0.9267696738243103, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 11200 + }, + { + "epoch": 0.11201, + "grad_norm": 0.9320189356803894, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 11201 + }, + { + "epoch": 0.11202, + "grad_norm": 0.9510216116905212, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 11202 + }, + { + "epoch": 0.11203, + "grad_norm": 0.9528765678405762, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 11203 + }, + { + "epoch": 0.11204, + "grad_norm": 1.0288702249526978, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 11204 + }, + { + "epoch": 0.11205, + "grad_norm": 1.0346468687057495, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 11205 + }, + { + "epoch": 0.11206, + "grad_norm": 0.92815101146698, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 11206 + }, + { + "epoch": 0.11207, + "grad_norm": 0.8600759506225586, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 11207 + }, + { + "epoch": 0.11208, + "grad_norm": 0.7563555836677551, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 11208 + }, + { + "epoch": 0.11209, + "grad_norm": 0.7098404169082642, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 11209 + }, + { + "epoch": 0.1121, + "grad_norm": 0.8382050395011902, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 11210 + }, + { + "epoch": 0.11211, + "grad_norm": 1.0424740314483643, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 11211 + }, + { + "epoch": 0.11212, + "grad_norm": 1.058436393737793, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 11212 + }, + { + "epoch": 0.11213, + "grad_norm": 0.9474709630012512, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 11213 + }, + { + "epoch": 0.11214, + "grad_norm": 0.9382772445678711, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 11214 + }, + { + "epoch": 0.11215, + "grad_norm": 1.0095430612564087, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 11215 + }, + { + "epoch": 0.11216, + "grad_norm": 0.99806809425354, + "learning_rate": 0.003, + "loss": 4.028, + "step": 11216 + }, + { + "epoch": 0.11217, + "grad_norm": 0.874643087387085, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 11217 + }, + { + "epoch": 0.11218, + "grad_norm": 0.8845378160476685, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 11218 + }, + { + "epoch": 0.11219, + "grad_norm": 1.0587880611419678, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 11219 + }, + { + "epoch": 0.1122, + "grad_norm": 0.9981693029403687, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 11220 + }, + { + "epoch": 0.11221, + "grad_norm": 0.8588231205940247, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 11221 + }, + { + "epoch": 0.11222, + "grad_norm": 0.67852783203125, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 11222 + }, + { + "epoch": 0.11223, + "grad_norm": 0.6346800923347473, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 11223 + }, + { + "epoch": 0.11224, + "grad_norm": 0.5292540192604065, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 11224 + }, + { + "epoch": 0.11225, + "grad_norm": 0.5228301882743835, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 11225 + }, + { + "epoch": 0.11226, + "grad_norm": 0.48205044865608215, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 11226 + }, + { + "epoch": 0.11227, + "grad_norm": 0.47511619329452515, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 11227 + }, + { + "epoch": 0.11228, + "grad_norm": 0.5459652543067932, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 11228 + }, + { + "epoch": 0.11229, + "grad_norm": 0.6039158701896667, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 11229 + }, + { + "epoch": 0.1123, + "grad_norm": 0.5797991156578064, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 11230 + }, + { + "epoch": 0.11231, + "grad_norm": 0.677551805973053, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 11231 + }, + { + "epoch": 0.11232, + "grad_norm": 0.6512748003005981, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 11232 + }, + { + "epoch": 0.11233, + "grad_norm": 0.6309075355529785, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 11233 + }, + { + "epoch": 0.11234, + "grad_norm": 0.5747607350349426, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 11234 + }, + { + "epoch": 0.11235, + "grad_norm": 0.6133723855018616, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 11235 + }, + { + "epoch": 0.11236, + "grad_norm": 0.9459915161132812, + "learning_rate": 0.003, + "loss": 4.039, + "step": 11236 + }, + { + "epoch": 0.11237, + "grad_norm": 1.2948063611984253, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 11237 + }, + { + "epoch": 0.11238, + "grad_norm": 0.9041596055030823, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 11238 + }, + { + "epoch": 0.11239, + "grad_norm": 0.8141499757766724, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 11239 + }, + { + "epoch": 0.1124, + "grad_norm": 0.8682344555854797, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 11240 + }, + { + "epoch": 0.11241, + "grad_norm": 0.7102754712104797, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 11241 + }, + { + "epoch": 0.11242, + "grad_norm": 0.5986071228981018, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 11242 + }, + { + "epoch": 0.11243, + "grad_norm": 0.6330735087394714, + "learning_rate": 0.003, + "loss": 4.022, + "step": 11243 + }, + { + "epoch": 0.11244, + "grad_norm": 0.7382640838623047, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 11244 + }, + { + "epoch": 0.11245, + "grad_norm": 0.7702611088752747, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 11245 + }, + { + "epoch": 0.11246, + "grad_norm": 0.9086140394210815, + "learning_rate": 0.003, + "loss": 4.011, + "step": 11246 + }, + { + "epoch": 0.11247, + "grad_norm": 1.0867910385131836, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 11247 + }, + { + "epoch": 0.11248, + "grad_norm": 0.8973993062973022, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 11248 + }, + { + "epoch": 0.11249, + "grad_norm": 0.741752028465271, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 11249 + }, + { + "epoch": 0.1125, + "grad_norm": 0.5408896803855896, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 11250 + }, + { + "epoch": 0.11251, + "grad_norm": 0.577225923538208, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 11251 + }, + { + "epoch": 0.11252, + "grad_norm": 0.6334212422370911, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 11252 + }, + { + "epoch": 0.11253, + "grad_norm": 0.6426181793212891, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 11253 + }, + { + "epoch": 0.11254, + "grad_norm": 0.7714022994041443, + "learning_rate": 0.003, + "loss": 4.018, + "step": 11254 + }, + { + "epoch": 0.11255, + "grad_norm": 0.8345813751220703, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 11255 + }, + { + "epoch": 0.11256, + "grad_norm": 0.8170779943466187, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 11256 + }, + { + "epoch": 0.11257, + "grad_norm": 0.8695367574691772, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 11257 + }, + { + "epoch": 0.11258, + "grad_norm": 0.9470844864845276, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 11258 + }, + { + "epoch": 0.11259, + "grad_norm": 1.0294305086135864, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 11259 + }, + { + "epoch": 0.1126, + "grad_norm": 0.9117608666419983, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 11260 + }, + { + "epoch": 0.11261, + "grad_norm": 0.7772311568260193, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 11261 + }, + { + "epoch": 0.11262, + "grad_norm": 0.7702427506446838, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 11262 + }, + { + "epoch": 0.11263, + "grad_norm": 0.7167341709136963, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 11263 + }, + { + "epoch": 0.11264, + "grad_norm": 0.7869431972503662, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 11264 + }, + { + "epoch": 0.11265, + "grad_norm": 0.9567169547080994, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 11265 + }, + { + "epoch": 0.11266, + "grad_norm": 1.15801203250885, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 11266 + }, + { + "epoch": 0.11267, + "grad_norm": 0.9887841939926147, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 11267 + }, + { + "epoch": 0.11268, + "grad_norm": 0.907789945602417, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 11268 + }, + { + "epoch": 0.11269, + "grad_norm": 0.7598255276679993, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 11269 + }, + { + "epoch": 0.1127, + "grad_norm": 0.7950443029403687, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 11270 + }, + { + "epoch": 0.11271, + "grad_norm": 0.9572848081588745, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 11271 + }, + { + "epoch": 0.11272, + "grad_norm": 0.9056628942489624, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 11272 + }, + { + "epoch": 0.11273, + "grad_norm": 1.0475438833236694, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 11273 + }, + { + "epoch": 0.11274, + "grad_norm": 1.0598156452178955, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 11274 + }, + { + "epoch": 0.11275, + "grad_norm": 0.7640100121498108, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 11275 + }, + { + "epoch": 0.11276, + "grad_norm": 0.666291356086731, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 11276 + }, + { + "epoch": 0.11277, + "grad_norm": 0.7809427380561829, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 11277 + }, + { + "epoch": 0.11278, + "grad_norm": 0.7285363078117371, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 11278 + }, + { + "epoch": 0.11279, + "grad_norm": 0.8132727146148682, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 11279 + }, + { + "epoch": 0.1128, + "grad_norm": 0.9549084901809692, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 11280 + }, + { + "epoch": 0.11281, + "grad_norm": 0.9229263067245483, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 11281 + }, + { + "epoch": 0.11282, + "grad_norm": 0.7626646161079407, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 11282 + }, + { + "epoch": 0.11283, + "grad_norm": 0.6786330342292786, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 11283 + }, + { + "epoch": 0.11284, + "grad_norm": 0.6982824802398682, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 11284 + }, + { + "epoch": 0.11285, + "grad_norm": 0.6288905739784241, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 11285 + }, + { + "epoch": 0.11286, + "grad_norm": 0.5336101651191711, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 11286 + }, + { + "epoch": 0.11287, + "grad_norm": 0.5692437291145325, + "learning_rate": 0.003, + "loss": 4.005, + "step": 11287 + }, + { + "epoch": 0.11288, + "grad_norm": 0.6209651827812195, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 11288 + }, + { + "epoch": 0.11289, + "grad_norm": 0.7525631785392761, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 11289 + }, + { + "epoch": 0.1129, + "grad_norm": 0.9396098852157593, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 11290 + }, + { + "epoch": 0.11291, + "grad_norm": 1.063955545425415, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 11291 + }, + { + "epoch": 0.11292, + "grad_norm": 0.863498866558075, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 11292 + }, + { + "epoch": 0.11293, + "grad_norm": 0.6577820181846619, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 11293 + }, + { + "epoch": 0.11294, + "grad_norm": 0.5650443434715271, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 11294 + }, + { + "epoch": 0.11295, + "grad_norm": 0.6451413631439209, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 11295 + }, + { + "epoch": 0.11296, + "grad_norm": 0.6615285873413086, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 11296 + }, + { + "epoch": 0.11297, + "grad_norm": 0.6570398807525635, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 11297 + }, + { + "epoch": 0.11298, + "grad_norm": 0.6592798233032227, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 11298 + }, + { + "epoch": 0.11299, + "grad_norm": 0.847843587398529, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 11299 + }, + { + "epoch": 0.113, + "grad_norm": 0.9787973165512085, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 11300 + }, + { + "epoch": 0.11301, + "grad_norm": 1.1148426532745361, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 11301 + }, + { + "epoch": 0.11302, + "grad_norm": 0.8683514595031738, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 11302 + }, + { + "epoch": 0.11303, + "grad_norm": 0.7785764932632446, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 11303 + }, + { + "epoch": 0.11304, + "grad_norm": 0.7863451242446899, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 11304 + }, + { + "epoch": 0.11305, + "grad_norm": 0.8932132720947266, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 11305 + }, + { + "epoch": 0.11306, + "grad_norm": 1.0904179811477661, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 11306 + }, + { + "epoch": 0.11307, + "grad_norm": 0.9348009824752808, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 11307 + }, + { + "epoch": 0.11308, + "grad_norm": 0.8359718918800354, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 11308 + }, + { + "epoch": 0.11309, + "grad_norm": 0.7915058732032776, + "learning_rate": 0.003, + "loss": 4.035, + "step": 11309 + }, + { + "epoch": 0.1131, + "grad_norm": 0.731500506401062, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 11310 + }, + { + "epoch": 0.11311, + "grad_norm": 0.6767420768737793, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 11311 + }, + { + "epoch": 0.11312, + "grad_norm": 0.7071685194969177, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 11312 + }, + { + "epoch": 0.11313, + "grad_norm": 0.8206322193145752, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 11313 + }, + { + "epoch": 0.11314, + "grad_norm": 0.995917558670044, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 11314 + }, + { + "epoch": 0.11315, + "grad_norm": 1.007768988609314, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 11315 + }, + { + "epoch": 0.11316, + "grad_norm": 0.917212188243866, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 11316 + }, + { + "epoch": 0.11317, + "grad_norm": 0.8176689147949219, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 11317 + }, + { + "epoch": 0.11318, + "grad_norm": 0.790153443813324, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 11318 + }, + { + "epoch": 0.11319, + "grad_norm": 0.77534019947052, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 11319 + }, + { + "epoch": 0.1132, + "grad_norm": 0.7776956558227539, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 11320 + }, + { + "epoch": 0.11321, + "grad_norm": 0.7186883687973022, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 11321 + }, + { + "epoch": 0.11322, + "grad_norm": 0.7190566062927246, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 11322 + }, + { + "epoch": 0.11323, + "grad_norm": 0.7112944722175598, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 11323 + }, + { + "epoch": 0.11324, + "grad_norm": 0.6839632391929626, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 11324 + }, + { + "epoch": 0.11325, + "grad_norm": 0.7430686354637146, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 11325 + }, + { + "epoch": 0.11326, + "grad_norm": 0.8290679454803467, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 11326 + }, + { + "epoch": 0.11327, + "grad_norm": 0.7175843119621277, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 11327 + }, + { + "epoch": 0.11328, + "grad_norm": 0.6538087725639343, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 11328 + }, + { + "epoch": 0.11329, + "grad_norm": 0.5562537908554077, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 11329 + }, + { + "epoch": 0.1133, + "grad_norm": 0.6857356429100037, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 11330 + }, + { + "epoch": 0.11331, + "grad_norm": 0.9902355074882507, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 11331 + }, + { + "epoch": 0.11332, + "grad_norm": 1.1442843675613403, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 11332 + }, + { + "epoch": 0.11333, + "grad_norm": 0.8154362440109253, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 11333 + }, + { + "epoch": 0.11334, + "grad_norm": 0.7913908362388611, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 11334 + }, + { + "epoch": 0.11335, + "grad_norm": 0.9230514764785767, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 11335 + }, + { + "epoch": 0.11336, + "grad_norm": 1.0062015056610107, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 11336 + }, + { + "epoch": 0.11337, + "grad_norm": 0.9774962663650513, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 11337 + }, + { + "epoch": 0.11338, + "grad_norm": 0.8366467356681824, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 11338 + }, + { + "epoch": 0.11339, + "grad_norm": 0.7167420983314514, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 11339 + }, + { + "epoch": 0.1134, + "grad_norm": 0.7209339141845703, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 11340 + }, + { + "epoch": 0.11341, + "grad_norm": 0.7862409353256226, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 11341 + }, + { + "epoch": 0.11342, + "grad_norm": 0.7756770849227905, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 11342 + }, + { + "epoch": 0.11343, + "grad_norm": 0.7953722476959229, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 11343 + }, + { + "epoch": 0.11344, + "grad_norm": 0.8854662179946899, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 11344 + }, + { + "epoch": 0.11345, + "grad_norm": 0.9794262051582336, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 11345 + }, + { + "epoch": 0.11346, + "grad_norm": 0.919201135635376, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 11346 + }, + { + "epoch": 0.11347, + "grad_norm": 0.7543264031410217, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 11347 + }, + { + "epoch": 0.11348, + "grad_norm": 0.7020623087882996, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 11348 + }, + { + "epoch": 0.11349, + "grad_norm": 0.7879665493965149, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 11349 + }, + { + "epoch": 0.1135, + "grad_norm": 0.862865686416626, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 11350 + }, + { + "epoch": 0.11351, + "grad_norm": 0.8149678111076355, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 11351 + }, + { + "epoch": 0.11352, + "grad_norm": 0.7770326137542725, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 11352 + }, + { + "epoch": 0.11353, + "grad_norm": 0.6807413101196289, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 11353 + }, + { + "epoch": 0.11354, + "grad_norm": 0.6837721467018127, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 11354 + }, + { + "epoch": 0.11355, + "grad_norm": 0.6766520142555237, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 11355 + }, + { + "epoch": 0.11356, + "grad_norm": 0.6636706590652466, + "learning_rate": 0.003, + "loss": 4.023, + "step": 11356 + }, + { + "epoch": 0.11357, + "grad_norm": 0.6780890226364136, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 11357 + }, + { + "epoch": 0.11358, + "grad_norm": 0.6905491948127747, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 11358 + }, + { + "epoch": 0.11359, + "grad_norm": 0.8020058274269104, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 11359 + }, + { + "epoch": 0.1136, + "grad_norm": 0.8286008238792419, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 11360 + }, + { + "epoch": 0.11361, + "grad_norm": 0.890533447265625, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 11361 + }, + { + "epoch": 0.11362, + "grad_norm": 1.0137696266174316, + "learning_rate": 0.003, + "loss": 4.03, + "step": 11362 + }, + { + "epoch": 0.11363, + "grad_norm": 1.0358784198760986, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 11363 + }, + { + "epoch": 0.11364, + "grad_norm": 0.8371180295944214, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 11364 + }, + { + "epoch": 0.11365, + "grad_norm": 0.8734986782073975, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 11365 + }, + { + "epoch": 0.11366, + "grad_norm": 0.9228023886680603, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 11366 + }, + { + "epoch": 0.11367, + "grad_norm": 0.9581348896026611, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 11367 + }, + { + "epoch": 0.11368, + "grad_norm": 1.0258618593215942, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 11368 + }, + { + "epoch": 0.11369, + "grad_norm": 1.1487102508544922, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 11369 + }, + { + "epoch": 0.1137, + "grad_norm": 0.7709944844245911, + "learning_rate": 0.003, + "loss": 4.051, + "step": 11370 + }, + { + "epoch": 0.11371, + "grad_norm": 0.6586733460426331, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 11371 + }, + { + "epoch": 0.11372, + "grad_norm": 0.6725128889083862, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 11372 + }, + { + "epoch": 0.11373, + "grad_norm": 0.7969658970832825, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 11373 + }, + { + "epoch": 0.11374, + "grad_norm": 0.8550029397010803, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 11374 + }, + { + "epoch": 0.11375, + "grad_norm": 0.9954065084457397, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 11375 + }, + { + "epoch": 0.11376, + "grad_norm": 0.9369992017745972, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 11376 + }, + { + "epoch": 0.11377, + "grad_norm": 0.780281662940979, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 11377 + }, + { + "epoch": 0.11378, + "grad_norm": 0.6823753118515015, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 11378 + }, + { + "epoch": 0.11379, + "grad_norm": 0.6865916848182678, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 11379 + }, + { + "epoch": 0.1138, + "grad_norm": 0.6688868403434753, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 11380 + }, + { + "epoch": 0.11381, + "grad_norm": 0.6455453634262085, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 11381 + }, + { + "epoch": 0.11382, + "grad_norm": 0.5758188962936401, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 11382 + }, + { + "epoch": 0.11383, + "grad_norm": 0.4809495508670807, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 11383 + }, + { + "epoch": 0.11384, + "grad_norm": 0.5226956009864807, + "learning_rate": 0.003, + "loss": 4.015, + "step": 11384 + }, + { + "epoch": 0.11385, + "grad_norm": 0.5507903695106506, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 11385 + }, + { + "epoch": 0.11386, + "grad_norm": 0.703472375869751, + "learning_rate": 0.003, + "loss": 4.032, + "step": 11386 + }, + { + "epoch": 0.11387, + "grad_norm": 0.8473612666130066, + "learning_rate": 0.003, + "loss": 4.01, + "step": 11387 + }, + { + "epoch": 0.11388, + "grad_norm": 1.0691862106323242, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 11388 + }, + { + "epoch": 0.11389, + "grad_norm": 1.0493425130844116, + "learning_rate": 0.003, + "loss": 4.025, + "step": 11389 + }, + { + "epoch": 0.1139, + "grad_norm": 0.8289149403572083, + "learning_rate": 0.003, + "loss": 4.009, + "step": 11390 + }, + { + "epoch": 0.11391, + "grad_norm": 0.6496636867523193, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 11391 + }, + { + "epoch": 0.11392, + "grad_norm": 0.6588830947875977, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 11392 + }, + { + "epoch": 0.11393, + "grad_norm": 0.8284120559692383, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 11393 + }, + { + "epoch": 0.11394, + "grad_norm": 0.8982833027839661, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 11394 + }, + { + "epoch": 0.11395, + "grad_norm": 0.8549256920814514, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 11395 + }, + { + "epoch": 0.11396, + "grad_norm": 0.7516512870788574, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 11396 + }, + { + "epoch": 0.11397, + "grad_norm": 0.6641618609428406, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 11397 + }, + { + "epoch": 0.11398, + "grad_norm": 0.7321159243583679, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 11398 + }, + { + "epoch": 0.11399, + "grad_norm": 0.7968395352363586, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 11399 + }, + { + "epoch": 0.114, + "grad_norm": 0.7829858660697937, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 11400 + }, + { + "epoch": 0.11401, + "grad_norm": 0.7570399641990662, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 11401 + }, + { + "epoch": 0.11402, + "grad_norm": 0.7252801656723022, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 11402 + }, + { + "epoch": 0.11403, + "grad_norm": 0.6975102424621582, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 11403 + }, + { + "epoch": 0.11404, + "grad_norm": 0.6238744854927063, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 11404 + }, + { + "epoch": 0.11405, + "grad_norm": 0.4995003938674927, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 11405 + }, + { + "epoch": 0.11406, + "grad_norm": 0.5959094762802124, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 11406 + }, + { + "epoch": 0.11407, + "grad_norm": 0.5805252194404602, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 11407 + }, + { + "epoch": 0.11408, + "grad_norm": 0.563025176525116, + "learning_rate": 0.003, + "loss": 4.025, + "step": 11408 + }, + { + "epoch": 0.11409, + "grad_norm": 0.6606787443161011, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 11409 + }, + { + "epoch": 0.1141, + "grad_norm": 0.8641210198402405, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 11410 + }, + { + "epoch": 0.11411, + "grad_norm": 1.1100224256515503, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 11411 + }, + { + "epoch": 0.11412, + "grad_norm": 0.9309477210044861, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 11412 + }, + { + "epoch": 0.11413, + "grad_norm": 0.7454750537872314, + "learning_rate": 0.003, + "loss": 4.008, + "step": 11413 + }, + { + "epoch": 0.11414, + "grad_norm": 0.7080267667770386, + "learning_rate": 0.003, + "loss": 4.026, + "step": 11414 + }, + { + "epoch": 0.11415, + "grad_norm": 0.7802665829658508, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 11415 + }, + { + "epoch": 0.11416, + "grad_norm": 0.8674970269203186, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 11416 + }, + { + "epoch": 0.11417, + "grad_norm": 0.9167194962501526, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 11417 + }, + { + "epoch": 0.11418, + "grad_norm": 0.8333175182342529, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 11418 + }, + { + "epoch": 0.11419, + "grad_norm": 0.7349749207496643, + "learning_rate": 0.003, + "loss": 4.002, + "step": 11419 + }, + { + "epoch": 0.1142, + "grad_norm": 0.7354753017425537, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 11420 + }, + { + "epoch": 0.11421, + "grad_norm": 0.7623195648193359, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 11421 + }, + { + "epoch": 0.11422, + "grad_norm": 0.8163384199142456, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 11422 + }, + { + "epoch": 0.11423, + "grad_norm": 0.9785248637199402, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 11423 + }, + { + "epoch": 0.11424, + "grad_norm": 1.2677289247512817, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 11424 + }, + { + "epoch": 0.11425, + "grad_norm": 0.9398715496063232, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 11425 + }, + { + "epoch": 0.11426, + "grad_norm": 0.9864420890808105, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 11426 + }, + { + "epoch": 0.11427, + "grad_norm": 0.9389703869819641, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 11427 + }, + { + "epoch": 0.11428, + "grad_norm": 0.884861171245575, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 11428 + }, + { + "epoch": 0.11429, + "grad_norm": 0.678497850894928, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 11429 + }, + { + "epoch": 0.1143, + "grad_norm": 0.9918597340583801, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 11430 + }, + { + "epoch": 0.11431, + "grad_norm": 1.2051234245300293, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 11431 + }, + { + "epoch": 0.11432, + "grad_norm": 0.9006490707397461, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 11432 + }, + { + "epoch": 0.11433, + "grad_norm": 0.862767219543457, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 11433 + }, + { + "epoch": 0.11434, + "grad_norm": 0.7508940100669861, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 11434 + }, + { + "epoch": 0.11435, + "grad_norm": 0.6505155563354492, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 11435 + }, + { + "epoch": 0.11436, + "grad_norm": 0.5951887965202332, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 11436 + }, + { + "epoch": 0.11437, + "grad_norm": 0.6015822291374207, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 11437 + }, + { + "epoch": 0.11438, + "grad_norm": 0.6409400105476379, + "learning_rate": 0.003, + "loss": 4.026, + "step": 11438 + }, + { + "epoch": 0.11439, + "grad_norm": 0.6430752277374268, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 11439 + }, + { + "epoch": 0.1144, + "grad_norm": 0.6330476999282837, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 11440 + }, + { + "epoch": 0.11441, + "grad_norm": 0.8735347390174866, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 11441 + }, + { + "epoch": 0.11442, + "grad_norm": 1.0421018600463867, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 11442 + }, + { + "epoch": 0.11443, + "grad_norm": 1.0504906177520752, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 11443 + }, + { + "epoch": 0.11444, + "grad_norm": 0.9561898112297058, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 11444 + }, + { + "epoch": 0.11445, + "grad_norm": 0.893307089805603, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 11445 + }, + { + "epoch": 0.11446, + "grad_norm": 0.8496833443641663, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 11446 + }, + { + "epoch": 0.11447, + "grad_norm": 0.9446256160736084, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 11447 + }, + { + "epoch": 0.11448, + "grad_norm": 1.1525483131408691, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 11448 + }, + { + "epoch": 0.11449, + "grad_norm": 0.8027827739715576, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 11449 + }, + { + "epoch": 0.1145, + "grad_norm": 0.7700136303901672, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 11450 + }, + { + "epoch": 0.11451, + "grad_norm": 0.6967355608940125, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 11451 + }, + { + "epoch": 0.11452, + "grad_norm": 0.6830270886421204, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 11452 + }, + { + "epoch": 0.11453, + "grad_norm": 0.811431348323822, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 11453 + }, + { + "epoch": 0.11454, + "grad_norm": 0.9994477033615112, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 11454 + }, + { + "epoch": 0.11455, + "grad_norm": 1.0607717037200928, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 11455 + }, + { + "epoch": 0.11456, + "grad_norm": 0.7838580012321472, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 11456 + }, + { + "epoch": 0.11457, + "grad_norm": 0.6020787954330444, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 11457 + }, + { + "epoch": 0.11458, + "grad_norm": 0.6495767831802368, + "learning_rate": 0.003, + "loss": 4.036, + "step": 11458 + }, + { + "epoch": 0.11459, + "grad_norm": 0.6524631381034851, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 11459 + }, + { + "epoch": 0.1146, + "grad_norm": 0.6308627724647522, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 11460 + }, + { + "epoch": 0.11461, + "grad_norm": 0.6008997559547424, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 11461 + }, + { + "epoch": 0.11462, + "grad_norm": 0.5560154914855957, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 11462 + }, + { + "epoch": 0.11463, + "grad_norm": 0.5246428847312927, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 11463 + }, + { + "epoch": 0.11464, + "grad_norm": 0.5368339419364929, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 11464 + }, + { + "epoch": 0.11465, + "grad_norm": 0.6533960103988647, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 11465 + }, + { + "epoch": 0.11466, + "grad_norm": 0.7779512405395508, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 11466 + }, + { + "epoch": 0.11467, + "grad_norm": 0.7873508334159851, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 11467 + }, + { + "epoch": 0.11468, + "grad_norm": 0.7781346440315247, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 11468 + }, + { + "epoch": 0.11469, + "grad_norm": 0.8823782801628113, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 11469 + }, + { + "epoch": 0.1147, + "grad_norm": 1.0907844305038452, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 11470 + }, + { + "epoch": 0.11471, + "grad_norm": 1.0541037321090698, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 11471 + }, + { + "epoch": 0.11472, + "grad_norm": 0.9238687753677368, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 11472 + }, + { + "epoch": 0.11473, + "grad_norm": 0.7540009617805481, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 11473 + }, + { + "epoch": 0.11474, + "grad_norm": 0.6382348537445068, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 11474 + }, + { + "epoch": 0.11475, + "grad_norm": 0.7358950972557068, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 11475 + }, + { + "epoch": 0.11476, + "grad_norm": 0.8675046563148499, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 11476 + }, + { + "epoch": 0.11477, + "grad_norm": 1.04617440700531, + "learning_rate": 0.003, + "loss": 4.056, + "step": 11477 + }, + { + "epoch": 0.11478, + "grad_norm": 1.0342538356781006, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 11478 + }, + { + "epoch": 0.11479, + "grad_norm": 0.8536825776100159, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 11479 + }, + { + "epoch": 0.1148, + "grad_norm": 0.8890334367752075, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 11480 + }, + { + "epoch": 0.11481, + "grad_norm": 0.8013483285903931, + "learning_rate": 0.003, + "loss": 4.011, + "step": 11481 + }, + { + "epoch": 0.11482, + "grad_norm": 0.7241466641426086, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 11482 + }, + { + "epoch": 0.11483, + "grad_norm": 0.671504020690918, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 11483 + }, + { + "epoch": 0.11484, + "grad_norm": 0.6110320687294006, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 11484 + }, + { + "epoch": 0.11485, + "grad_norm": 0.6042850017547607, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 11485 + }, + { + "epoch": 0.11486, + "grad_norm": 0.6030364036560059, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 11486 + }, + { + "epoch": 0.11487, + "grad_norm": 0.641995370388031, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 11487 + }, + { + "epoch": 0.11488, + "grad_norm": 0.787706196308136, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 11488 + }, + { + "epoch": 0.11489, + "grad_norm": 0.9720715284347534, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 11489 + }, + { + "epoch": 0.1149, + "grad_norm": 0.9054409861564636, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 11490 + }, + { + "epoch": 0.11491, + "grad_norm": 0.8776920437812805, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 11491 + }, + { + "epoch": 0.11492, + "grad_norm": 0.90711909532547, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 11492 + }, + { + "epoch": 0.11493, + "grad_norm": 0.8923183083534241, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 11493 + }, + { + "epoch": 0.11494, + "grad_norm": 0.7126554250717163, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 11494 + }, + { + "epoch": 0.11495, + "grad_norm": 0.6723785996437073, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 11495 + }, + { + "epoch": 0.11496, + "grad_norm": 0.7934660911560059, + "learning_rate": 0.003, + "loss": 4.007, + "step": 11496 + }, + { + "epoch": 0.11497, + "grad_norm": 0.8057811856269836, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 11497 + }, + { + "epoch": 0.11498, + "grad_norm": 0.7440505027770996, + "learning_rate": 0.003, + "loss": 4.023, + "step": 11498 + }, + { + "epoch": 0.11499, + "grad_norm": 0.7898901700973511, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 11499 + }, + { + "epoch": 0.115, + "grad_norm": 0.8212448358535767, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 11500 + }, + { + "epoch": 0.11501, + "grad_norm": 0.8043070435523987, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 11501 + }, + { + "epoch": 0.11502, + "grad_norm": 0.69261634349823, + "learning_rate": 0.003, + "loss": 4.038, + "step": 11502 + }, + { + "epoch": 0.11503, + "grad_norm": 0.5780364871025085, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 11503 + }, + { + "epoch": 0.11504, + "grad_norm": 0.6269531846046448, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 11504 + }, + { + "epoch": 0.11505, + "grad_norm": 0.7450412511825562, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 11505 + }, + { + "epoch": 0.11506, + "grad_norm": 0.8044024705886841, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 11506 + }, + { + "epoch": 0.11507, + "grad_norm": 0.8625448942184448, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 11507 + }, + { + "epoch": 0.11508, + "grad_norm": 0.9397217035293579, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 11508 + }, + { + "epoch": 0.11509, + "grad_norm": 1.151915431022644, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 11509 + }, + { + "epoch": 0.1151, + "grad_norm": 0.7171595692634583, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 11510 + }, + { + "epoch": 0.11511, + "grad_norm": 0.6709728240966797, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 11511 + }, + { + "epoch": 0.11512, + "grad_norm": 0.7095353007316589, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 11512 + }, + { + "epoch": 0.11513, + "grad_norm": 0.6780837774276733, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 11513 + }, + { + "epoch": 0.11514, + "grad_norm": 0.5921359062194824, + "learning_rate": 0.003, + "loss": 3.998, + "step": 11514 + }, + { + "epoch": 0.11515, + "grad_norm": 0.5906557440757751, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 11515 + }, + { + "epoch": 0.11516, + "grad_norm": 0.6455663442611694, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 11516 + }, + { + "epoch": 0.11517, + "grad_norm": 0.7563275098800659, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 11517 + }, + { + "epoch": 0.11518, + "grad_norm": 0.8928882479667664, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 11518 + }, + { + "epoch": 0.11519, + "grad_norm": 1.008765459060669, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 11519 + }, + { + "epoch": 0.1152, + "grad_norm": 1.1093732118606567, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 11520 + }, + { + "epoch": 0.11521, + "grad_norm": 0.9981747269630432, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 11521 + }, + { + "epoch": 0.11522, + "grad_norm": 0.9125166535377502, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 11522 + }, + { + "epoch": 0.11523, + "grad_norm": 0.7462885975837708, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 11523 + }, + { + "epoch": 0.11524, + "grad_norm": 0.7454661726951599, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 11524 + }, + { + "epoch": 0.11525, + "grad_norm": 0.7887188196182251, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 11525 + }, + { + "epoch": 0.11526, + "grad_norm": 1.0900375843048096, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 11526 + }, + { + "epoch": 0.11527, + "grad_norm": 1.1068854331970215, + "learning_rate": 0.003, + "loss": 4.045, + "step": 11527 + }, + { + "epoch": 0.11528, + "grad_norm": 0.9299308657646179, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 11528 + }, + { + "epoch": 0.11529, + "grad_norm": 0.8719900846481323, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 11529 + }, + { + "epoch": 0.1153, + "grad_norm": 0.9658557176589966, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 11530 + }, + { + "epoch": 0.11531, + "grad_norm": 0.979651689529419, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 11531 + }, + { + "epoch": 0.11532, + "grad_norm": 0.8919920325279236, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 11532 + }, + { + "epoch": 0.11533, + "grad_norm": 0.7383463382720947, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 11533 + }, + { + "epoch": 0.11534, + "grad_norm": 0.7078273892402649, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 11534 + }, + { + "epoch": 0.11535, + "grad_norm": 0.6356802582740784, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 11535 + }, + { + "epoch": 0.11536, + "grad_norm": 0.6901099681854248, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 11536 + }, + { + "epoch": 0.11537, + "grad_norm": 0.7354924082756042, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 11537 + }, + { + "epoch": 0.11538, + "grad_norm": 0.9668107032775879, + "learning_rate": 0.003, + "loss": 4.051, + "step": 11538 + }, + { + "epoch": 0.11539, + "grad_norm": 1.2265743017196655, + "learning_rate": 0.003, + "loss": 4.017, + "step": 11539 + }, + { + "epoch": 0.1154, + "grad_norm": 0.7557204961776733, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 11540 + }, + { + "epoch": 0.11541, + "grad_norm": 0.7328994870185852, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 11541 + }, + { + "epoch": 0.11542, + "grad_norm": 0.6833067536354065, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 11542 + }, + { + "epoch": 0.11543, + "grad_norm": 0.7724437117576599, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 11543 + }, + { + "epoch": 0.11544, + "grad_norm": 0.8594021201133728, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 11544 + }, + { + "epoch": 0.11545, + "grad_norm": 1.020458698272705, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 11545 + }, + { + "epoch": 0.11546, + "grad_norm": 1.0542057752609253, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 11546 + }, + { + "epoch": 0.11547, + "grad_norm": 0.8527884483337402, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 11547 + }, + { + "epoch": 0.11548, + "grad_norm": 0.7760061025619507, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 11548 + }, + { + "epoch": 0.11549, + "grad_norm": 0.8009873032569885, + "learning_rate": 0.003, + "loss": 4.034, + "step": 11549 + }, + { + "epoch": 0.1155, + "grad_norm": 0.7634437680244446, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 11550 + }, + { + "epoch": 0.11551, + "grad_norm": 0.7709644436836243, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 11551 + }, + { + "epoch": 0.11552, + "grad_norm": 0.7237449288368225, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 11552 + }, + { + "epoch": 0.11553, + "grad_norm": 0.6726607084274292, + "learning_rate": 0.003, + "loss": 4.003, + "step": 11553 + }, + { + "epoch": 0.11554, + "grad_norm": 0.5839000344276428, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 11554 + }, + { + "epoch": 0.11555, + "grad_norm": 0.5689242482185364, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 11555 + }, + { + "epoch": 0.11556, + "grad_norm": 0.6661663055419922, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 11556 + }, + { + "epoch": 0.11557, + "grad_norm": 0.7585426568984985, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 11557 + }, + { + "epoch": 0.11558, + "grad_norm": 0.8989922404289246, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 11558 + }, + { + "epoch": 0.11559, + "grad_norm": 0.972117006778717, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 11559 + }, + { + "epoch": 0.1156, + "grad_norm": 0.9820306301116943, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 11560 + }, + { + "epoch": 0.11561, + "grad_norm": 0.9749352931976318, + "learning_rate": 0.003, + "loss": 4.029, + "step": 11561 + }, + { + "epoch": 0.11562, + "grad_norm": 0.8870478868484497, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 11562 + }, + { + "epoch": 0.11563, + "grad_norm": 0.7449541687965393, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 11563 + }, + { + "epoch": 0.11564, + "grad_norm": 0.6444079875946045, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 11564 + }, + { + "epoch": 0.11565, + "grad_norm": 0.7940042018890381, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 11565 + }, + { + "epoch": 0.11566, + "grad_norm": 0.9880294799804688, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 11566 + }, + { + "epoch": 0.11567, + "grad_norm": 1.0673390626907349, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 11567 + }, + { + "epoch": 0.11568, + "grad_norm": 0.9975600242614746, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 11568 + }, + { + "epoch": 0.11569, + "grad_norm": 0.9561159610748291, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 11569 + }, + { + "epoch": 0.1157, + "grad_norm": 0.7889679074287415, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 11570 + }, + { + "epoch": 0.11571, + "grad_norm": 0.7783418893814087, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 11571 + }, + { + "epoch": 0.11572, + "grad_norm": 0.6229084730148315, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 11572 + }, + { + "epoch": 0.11573, + "grad_norm": 0.6400849223136902, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 11573 + }, + { + "epoch": 0.11574, + "grad_norm": 0.6409059762954712, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 11574 + }, + { + "epoch": 0.11575, + "grad_norm": 0.6521430611610413, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 11575 + }, + { + "epoch": 0.11576, + "grad_norm": 0.7270999550819397, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 11576 + }, + { + "epoch": 0.11577, + "grad_norm": 0.6581321954727173, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 11577 + }, + { + "epoch": 0.11578, + "grad_norm": 0.7205742597579956, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 11578 + }, + { + "epoch": 0.11579, + "grad_norm": 0.7195239067077637, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 11579 + }, + { + "epoch": 0.1158, + "grad_norm": 0.6457316875457764, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 11580 + }, + { + "epoch": 0.11581, + "grad_norm": 0.6417819857597351, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 11581 + }, + { + "epoch": 0.11582, + "grad_norm": 0.7540315389633179, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 11582 + }, + { + "epoch": 0.11583, + "grad_norm": 0.9404506087303162, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 11583 + }, + { + "epoch": 0.11584, + "grad_norm": 1.077072262763977, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 11584 + }, + { + "epoch": 0.11585, + "grad_norm": 0.9085767269134521, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 11585 + }, + { + "epoch": 0.11586, + "grad_norm": 0.8725526928901672, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 11586 + }, + { + "epoch": 0.11587, + "grad_norm": 0.8863524198532104, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 11587 + }, + { + "epoch": 0.11588, + "grad_norm": 0.8565869331359863, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 11588 + }, + { + "epoch": 0.11589, + "grad_norm": 0.8511981964111328, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 11589 + }, + { + "epoch": 0.1159, + "grad_norm": 0.8466473817825317, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 11590 + }, + { + "epoch": 0.11591, + "grad_norm": 0.8959945440292358, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 11591 + }, + { + "epoch": 0.11592, + "grad_norm": 0.8851884007453918, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 11592 + }, + { + "epoch": 0.11593, + "grad_norm": 0.7232153415679932, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 11593 + }, + { + "epoch": 0.11594, + "grad_norm": 0.7993742227554321, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 11594 + }, + { + "epoch": 0.11595, + "grad_norm": 0.8647189736366272, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 11595 + }, + { + "epoch": 0.11596, + "grad_norm": 0.8297176957130432, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 11596 + }, + { + "epoch": 0.11597, + "grad_norm": 0.9545539617538452, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 11597 + }, + { + "epoch": 0.11598, + "grad_norm": 1.1064233779907227, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 11598 + }, + { + "epoch": 0.11599, + "grad_norm": 0.842133641242981, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 11599 + }, + { + "epoch": 0.116, + "grad_norm": 0.7919009327888489, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 11600 + }, + { + "epoch": 0.11601, + "grad_norm": 0.8049101829528809, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 11601 + }, + { + "epoch": 0.11602, + "grad_norm": 0.9507090449333191, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 11602 + }, + { + "epoch": 0.11603, + "grad_norm": 0.9661722183227539, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 11603 + }, + { + "epoch": 0.11604, + "grad_norm": 0.9851440787315369, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 11604 + }, + { + "epoch": 0.11605, + "grad_norm": 0.95931077003479, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 11605 + }, + { + "epoch": 0.11606, + "grad_norm": 0.9839902520179749, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 11606 + }, + { + "epoch": 0.11607, + "grad_norm": 0.8679165840148926, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 11607 + }, + { + "epoch": 0.11608, + "grad_norm": 0.7510985136032104, + "learning_rate": 0.003, + "loss": 4.048, + "step": 11608 + }, + { + "epoch": 0.11609, + "grad_norm": 0.6729006767272949, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 11609 + }, + { + "epoch": 0.1161, + "grad_norm": 0.7698171734809875, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 11610 + }, + { + "epoch": 0.11611, + "grad_norm": 0.866439938545227, + "learning_rate": 0.003, + "loss": 4.041, + "step": 11611 + }, + { + "epoch": 0.11612, + "grad_norm": 0.9573925733566284, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 11612 + }, + { + "epoch": 0.11613, + "grad_norm": 1.0064102411270142, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 11613 + }, + { + "epoch": 0.11614, + "grad_norm": 0.951464056968689, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 11614 + }, + { + "epoch": 0.11615, + "grad_norm": 0.7083039283752441, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 11615 + }, + { + "epoch": 0.11616, + "grad_norm": 0.5653154253959656, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 11616 + }, + { + "epoch": 0.11617, + "grad_norm": 0.5044902563095093, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 11617 + }, + { + "epoch": 0.11618, + "grad_norm": 0.5124688744544983, + "learning_rate": 0.003, + "loss": 4.013, + "step": 11618 + }, + { + "epoch": 0.11619, + "grad_norm": 0.439615398645401, + "learning_rate": 0.003, + "loss": 4.006, + "step": 11619 + }, + { + "epoch": 0.1162, + "grad_norm": 0.48477035760879517, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 11620 + }, + { + "epoch": 0.11621, + "grad_norm": 0.4753328561782837, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 11621 + }, + { + "epoch": 0.11622, + "grad_norm": 0.5537976026535034, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 11622 + }, + { + "epoch": 0.11623, + "grad_norm": 0.5809506177902222, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 11623 + }, + { + "epoch": 0.11624, + "grad_norm": 0.6150876879692078, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 11624 + }, + { + "epoch": 0.11625, + "grad_norm": 0.714521050453186, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 11625 + }, + { + "epoch": 0.11626, + "grad_norm": 0.7555716633796692, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 11626 + }, + { + "epoch": 0.11627, + "grad_norm": 0.7596738934516907, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 11627 + }, + { + "epoch": 0.11628, + "grad_norm": 0.6994544863700867, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 11628 + }, + { + "epoch": 0.11629, + "grad_norm": 0.6706871390342712, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 11629 + }, + { + "epoch": 0.1163, + "grad_norm": 0.7496283054351807, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 11630 + }, + { + "epoch": 0.11631, + "grad_norm": 0.7881671786308289, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 11631 + }, + { + "epoch": 0.11632, + "grad_norm": 0.9565471410751343, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 11632 + }, + { + "epoch": 0.11633, + "grad_norm": 1.3986477851867676, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 11633 + }, + { + "epoch": 0.11634, + "grad_norm": 0.5135916471481323, + "learning_rate": 0.003, + "loss": 4.028, + "step": 11634 + }, + { + "epoch": 0.11635, + "grad_norm": 0.6713153719902039, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 11635 + }, + { + "epoch": 0.11636, + "grad_norm": 0.801653265953064, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 11636 + }, + { + "epoch": 0.11637, + "grad_norm": 0.8092045187950134, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 11637 + }, + { + "epoch": 0.11638, + "grad_norm": 0.8648455739021301, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 11638 + }, + { + "epoch": 0.11639, + "grad_norm": 0.7091639041900635, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 11639 + }, + { + "epoch": 0.1164, + "grad_norm": 0.6421645879745483, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 11640 + }, + { + "epoch": 0.11641, + "grad_norm": 0.6985124349594116, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 11641 + }, + { + "epoch": 0.11642, + "grad_norm": 0.8126601576805115, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 11642 + }, + { + "epoch": 0.11643, + "grad_norm": 0.8762948513031006, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 11643 + }, + { + "epoch": 0.11644, + "grad_norm": 0.9369477033615112, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 11644 + }, + { + "epoch": 0.11645, + "grad_norm": 0.9334920048713684, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 11645 + }, + { + "epoch": 0.11646, + "grad_norm": 1.0934345722198486, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 11646 + }, + { + "epoch": 0.11647, + "grad_norm": 0.9287505149841309, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 11647 + }, + { + "epoch": 0.11648, + "grad_norm": 0.9370683431625366, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 11648 + }, + { + "epoch": 0.11649, + "grad_norm": 0.9412454962730408, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 11649 + }, + { + "epoch": 0.1165, + "grad_norm": 0.8130035996437073, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 11650 + }, + { + "epoch": 0.11651, + "grad_norm": 0.7839321494102478, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 11651 + }, + { + "epoch": 0.11652, + "grad_norm": 0.8696938157081604, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 11652 + }, + { + "epoch": 0.11653, + "grad_norm": 0.9900596141815186, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 11653 + }, + { + "epoch": 0.11654, + "grad_norm": 1.0957406759262085, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 11654 + }, + { + "epoch": 0.11655, + "grad_norm": 1.00582754611969, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 11655 + }, + { + "epoch": 0.11656, + "grad_norm": 1.2449061870574951, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 11656 + }, + { + "epoch": 0.11657, + "grad_norm": 0.8895262479782104, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 11657 + }, + { + "epoch": 0.11658, + "grad_norm": 0.7397792339324951, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 11658 + }, + { + "epoch": 0.11659, + "grad_norm": 0.7503185868263245, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 11659 + }, + { + "epoch": 0.1166, + "grad_norm": 0.8214109539985657, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 11660 + }, + { + "epoch": 0.11661, + "grad_norm": 0.7903082966804504, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 11661 + }, + { + "epoch": 0.11662, + "grad_norm": 0.7165495157241821, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 11662 + }, + { + "epoch": 0.11663, + "grad_norm": 0.7824479937553406, + "learning_rate": 0.003, + "loss": 4.012, + "step": 11663 + }, + { + "epoch": 0.11664, + "grad_norm": 0.810527503490448, + "learning_rate": 0.003, + "loss": 4.053, + "step": 11664 + }, + { + "epoch": 0.11665, + "grad_norm": 0.817573606967926, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 11665 + }, + { + "epoch": 0.11666, + "grad_norm": 0.8660758137702942, + "learning_rate": 0.003, + "loss": 4.027, + "step": 11666 + }, + { + "epoch": 0.11667, + "grad_norm": 0.8072959780693054, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 11667 + }, + { + "epoch": 0.11668, + "grad_norm": 0.9078159332275391, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 11668 + }, + { + "epoch": 0.11669, + "grad_norm": 1.052850365638733, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 11669 + }, + { + "epoch": 0.1167, + "grad_norm": 0.9063073992729187, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 11670 + }, + { + "epoch": 0.11671, + "grad_norm": 0.7276980876922607, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 11671 + }, + { + "epoch": 0.11672, + "grad_norm": 0.6072031259536743, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 11672 + }, + { + "epoch": 0.11673, + "grad_norm": 0.6613339781761169, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 11673 + }, + { + "epoch": 0.11674, + "grad_norm": 0.6048155426979065, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 11674 + }, + { + "epoch": 0.11675, + "grad_norm": 0.6490438580513, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 11675 + }, + { + "epoch": 0.11676, + "grad_norm": 0.6653977632522583, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 11676 + }, + { + "epoch": 0.11677, + "grad_norm": 0.6288249492645264, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 11677 + }, + { + "epoch": 0.11678, + "grad_norm": 0.6480962038040161, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 11678 + }, + { + "epoch": 0.11679, + "grad_norm": 0.6132774353027344, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 11679 + }, + { + "epoch": 0.1168, + "grad_norm": 0.5827869772911072, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 11680 + }, + { + "epoch": 0.11681, + "grad_norm": 0.6305233836174011, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 11681 + }, + { + "epoch": 0.11682, + "grad_norm": 0.7440044283866882, + "learning_rate": 0.003, + "loss": 4.018, + "step": 11682 + }, + { + "epoch": 0.11683, + "grad_norm": 0.8142518997192383, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 11683 + }, + { + "epoch": 0.11684, + "grad_norm": 0.8391857147216797, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 11684 + }, + { + "epoch": 0.11685, + "grad_norm": 0.788082480430603, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 11685 + }, + { + "epoch": 0.11686, + "grad_norm": 0.8251965641975403, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 11686 + }, + { + "epoch": 0.11687, + "grad_norm": 0.9477576613426208, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 11687 + }, + { + "epoch": 0.11688, + "grad_norm": 0.9171106219291687, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 11688 + }, + { + "epoch": 0.11689, + "grad_norm": 0.7135186791419983, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 11689 + }, + { + "epoch": 0.1169, + "grad_norm": 0.6629257798194885, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 11690 + }, + { + "epoch": 0.11691, + "grad_norm": 0.6546911001205444, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 11691 + }, + { + "epoch": 0.11692, + "grad_norm": 0.6957620978355408, + "learning_rate": 0.003, + "loss": 4.01, + "step": 11692 + }, + { + "epoch": 0.11693, + "grad_norm": 0.7230929136276245, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 11693 + }, + { + "epoch": 0.11694, + "grad_norm": 0.7952154874801636, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 11694 + }, + { + "epoch": 0.11695, + "grad_norm": 0.866857647895813, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 11695 + }, + { + "epoch": 0.11696, + "grad_norm": 0.9634397625923157, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 11696 + }, + { + "epoch": 0.11697, + "grad_norm": 1.0934638977050781, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 11697 + }, + { + "epoch": 0.11698, + "grad_norm": 0.9745580554008484, + "learning_rate": 0.003, + "loss": 4.019, + "step": 11698 + }, + { + "epoch": 0.11699, + "grad_norm": 1.0540590286254883, + "learning_rate": 0.003, + "loss": 4.021, + "step": 11699 + }, + { + "epoch": 0.117, + "grad_norm": 0.8467910289764404, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 11700 + }, + { + "epoch": 0.11701, + "grad_norm": 0.6315522193908691, + "learning_rate": 0.003, + "loss": 4.022, + "step": 11701 + }, + { + "epoch": 0.11702, + "grad_norm": 0.6518235206604004, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 11702 + }, + { + "epoch": 0.11703, + "grad_norm": 0.7548579573631287, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 11703 + }, + { + "epoch": 0.11704, + "grad_norm": 0.7728471159934998, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 11704 + }, + { + "epoch": 0.11705, + "grad_norm": 0.7047388553619385, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 11705 + }, + { + "epoch": 0.11706, + "grad_norm": 0.6782577633857727, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 11706 + }, + { + "epoch": 0.11707, + "grad_norm": 0.7736063003540039, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 11707 + }, + { + "epoch": 0.11708, + "grad_norm": 0.7953140139579773, + "learning_rate": 0.003, + "loss": 4.009, + "step": 11708 + }, + { + "epoch": 0.11709, + "grad_norm": 0.7375701069831848, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 11709 + }, + { + "epoch": 0.1171, + "grad_norm": 0.7914718389511108, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 11710 + }, + { + "epoch": 0.11711, + "grad_norm": 0.7727771401405334, + "learning_rate": 0.003, + "loss": 3.995, + "step": 11711 + }, + { + "epoch": 0.11712, + "grad_norm": 0.8406480550765991, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 11712 + }, + { + "epoch": 0.11713, + "grad_norm": 0.9383452534675598, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 11713 + }, + { + "epoch": 0.11714, + "grad_norm": 1.0819647312164307, + "learning_rate": 0.003, + "loss": 4.02, + "step": 11714 + }, + { + "epoch": 0.11715, + "grad_norm": 0.9252278208732605, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 11715 + }, + { + "epoch": 0.11716, + "grad_norm": 0.7834035158157349, + "learning_rate": 0.003, + "loss": 3.991, + "step": 11716 + }, + { + "epoch": 0.11717, + "grad_norm": 0.8063236474990845, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 11717 + }, + { + "epoch": 0.11718, + "grad_norm": 0.8123244047164917, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 11718 + }, + { + "epoch": 0.11719, + "grad_norm": 0.767584502696991, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 11719 + }, + { + "epoch": 0.1172, + "grad_norm": 0.6858091354370117, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 11720 + }, + { + "epoch": 0.11721, + "grad_norm": 0.6874958276748657, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 11721 + }, + { + "epoch": 0.11722, + "grad_norm": 0.7214937210083008, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 11722 + }, + { + "epoch": 0.11723, + "grad_norm": 0.9050902724266052, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 11723 + }, + { + "epoch": 0.11724, + "grad_norm": 1.386954665184021, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 11724 + }, + { + "epoch": 0.11725, + "grad_norm": 0.7460734844207764, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 11725 + }, + { + "epoch": 0.11726, + "grad_norm": 0.6396922469139099, + "learning_rate": 0.003, + "loss": 4.024, + "step": 11726 + }, + { + "epoch": 0.11727, + "grad_norm": 0.6917034983634949, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 11727 + }, + { + "epoch": 0.11728, + "grad_norm": 0.7560396194458008, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 11728 + }, + { + "epoch": 0.11729, + "grad_norm": 0.8665305376052856, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 11729 + }, + { + "epoch": 0.1173, + "grad_norm": 0.772159993648529, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 11730 + }, + { + "epoch": 0.11731, + "grad_norm": 0.7486879825592041, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 11731 + }, + { + "epoch": 0.11732, + "grad_norm": 0.7846516966819763, + "learning_rate": 0.003, + "loss": 4.018, + "step": 11732 + }, + { + "epoch": 0.11733, + "grad_norm": 0.879023015499115, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 11733 + }, + { + "epoch": 0.11734, + "grad_norm": 1.1099352836608887, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 11734 + }, + { + "epoch": 0.11735, + "grad_norm": 0.7834532260894775, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 11735 + }, + { + "epoch": 0.11736, + "grad_norm": 0.5790771842002869, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 11736 + }, + { + "epoch": 0.11737, + "grad_norm": 0.5680363774299622, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 11737 + }, + { + "epoch": 0.11738, + "grad_norm": 0.6008859276771545, + "learning_rate": 0.003, + "loss": 4.033, + "step": 11738 + }, + { + "epoch": 0.11739, + "grad_norm": 0.7392567992210388, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 11739 + }, + { + "epoch": 0.1174, + "grad_norm": 0.916904091835022, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 11740 + }, + { + "epoch": 0.11741, + "grad_norm": 1.0528932809829712, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 11741 + }, + { + "epoch": 0.11742, + "grad_norm": 0.7817772626876831, + "learning_rate": 0.003, + "loss": 4.037, + "step": 11742 + }, + { + "epoch": 0.11743, + "grad_norm": 0.5557342171669006, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 11743 + }, + { + "epoch": 0.11744, + "grad_norm": 0.5775336623191833, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 11744 + }, + { + "epoch": 0.11745, + "grad_norm": 0.7112092971801758, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 11745 + }, + { + "epoch": 0.11746, + "grad_norm": 0.9123482704162598, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 11746 + }, + { + "epoch": 0.11747, + "grad_norm": 0.9527279734611511, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 11747 + }, + { + "epoch": 0.11748, + "grad_norm": 0.9975042343139648, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 11748 + }, + { + "epoch": 0.11749, + "grad_norm": 0.9033437371253967, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 11749 + }, + { + "epoch": 0.1175, + "grad_norm": 0.9851463437080383, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 11750 + }, + { + "epoch": 0.11751, + "grad_norm": 0.9741025567054749, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 11751 + }, + { + "epoch": 0.11752, + "grad_norm": 0.8289722800254822, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 11752 + }, + { + "epoch": 0.11753, + "grad_norm": 0.8503996729850769, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 11753 + }, + { + "epoch": 0.11754, + "grad_norm": 0.7909584045410156, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 11754 + }, + { + "epoch": 0.11755, + "grad_norm": 0.7145279049873352, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 11755 + }, + { + "epoch": 0.11756, + "grad_norm": 0.7444949150085449, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 11756 + }, + { + "epoch": 0.11757, + "grad_norm": 0.8783793449401855, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 11757 + }, + { + "epoch": 0.11758, + "grad_norm": 0.8041375875473022, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 11758 + }, + { + "epoch": 0.11759, + "grad_norm": 0.9612789750099182, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 11759 + }, + { + "epoch": 0.1176, + "grad_norm": 1.2083760499954224, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 11760 + }, + { + "epoch": 0.11761, + "grad_norm": 0.807439386844635, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 11761 + }, + { + "epoch": 0.11762, + "grad_norm": 0.7512252330780029, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 11762 + }, + { + "epoch": 0.11763, + "grad_norm": 0.7043628692626953, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 11763 + }, + { + "epoch": 0.11764, + "grad_norm": 0.6556148529052734, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 11764 + }, + { + "epoch": 0.11765, + "grad_norm": 0.6054413914680481, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 11765 + }, + { + "epoch": 0.11766, + "grad_norm": 0.5417591333389282, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 11766 + }, + { + "epoch": 0.11767, + "grad_norm": 0.5872727036476135, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 11767 + }, + { + "epoch": 0.11768, + "grad_norm": 0.6032535433769226, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 11768 + }, + { + "epoch": 0.11769, + "grad_norm": 0.6734449863433838, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 11769 + }, + { + "epoch": 0.1177, + "grad_norm": 0.7936435341835022, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 11770 + }, + { + "epoch": 0.11771, + "grad_norm": 1.079922080039978, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 11771 + }, + { + "epoch": 0.11772, + "grad_norm": 1.234328031539917, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 11772 + }, + { + "epoch": 0.11773, + "grad_norm": 0.5922883152961731, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 11773 + }, + { + "epoch": 0.11774, + "grad_norm": 0.8037434816360474, + "learning_rate": 0.003, + "loss": 4.021, + "step": 11774 + }, + { + "epoch": 0.11775, + "grad_norm": 1.0557984113693237, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 11775 + }, + { + "epoch": 0.11776, + "grad_norm": 0.8010382652282715, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 11776 + }, + { + "epoch": 0.11777, + "grad_norm": 0.7044901251792908, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 11777 + }, + { + "epoch": 0.11778, + "grad_norm": 0.7547171711921692, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 11778 + }, + { + "epoch": 0.11779, + "grad_norm": 0.8296960592269897, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 11779 + }, + { + "epoch": 0.1178, + "grad_norm": 0.7630957365036011, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 11780 + }, + { + "epoch": 0.11781, + "grad_norm": 0.7284735441207886, + "learning_rate": 0.003, + "loss": 4.007, + "step": 11781 + }, + { + "epoch": 0.11782, + "grad_norm": 0.6988702416419983, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 11782 + }, + { + "epoch": 0.11783, + "grad_norm": 0.7704843878746033, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 11783 + }, + { + "epoch": 0.11784, + "grad_norm": 0.797591507434845, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 11784 + }, + { + "epoch": 0.11785, + "grad_norm": 0.7599050998687744, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 11785 + }, + { + "epoch": 0.11786, + "grad_norm": 0.7562331557273865, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 11786 + }, + { + "epoch": 0.11787, + "grad_norm": 0.7886567115783691, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 11787 + }, + { + "epoch": 0.11788, + "grad_norm": 0.7673211693763733, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 11788 + }, + { + "epoch": 0.11789, + "grad_norm": 0.7491581439971924, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 11789 + }, + { + "epoch": 0.1179, + "grad_norm": 0.747179388999939, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 11790 + }, + { + "epoch": 0.11791, + "grad_norm": 0.8889058828353882, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 11791 + }, + { + "epoch": 0.11792, + "grad_norm": 0.9870392084121704, + "learning_rate": 0.003, + "loss": 4.039, + "step": 11792 + }, + { + "epoch": 0.11793, + "grad_norm": 1.0791133642196655, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 11793 + }, + { + "epoch": 0.11794, + "grad_norm": 1.0078823566436768, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 11794 + }, + { + "epoch": 0.11795, + "grad_norm": 0.903905987739563, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 11795 + }, + { + "epoch": 0.11796, + "grad_norm": 0.7996202111244202, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 11796 + }, + { + "epoch": 0.11797, + "grad_norm": 0.7034594416618347, + "learning_rate": 0.003, + "loss": 4.005, + "step": 11797 + }, + { + "epoch": 0.11798, + "grad_norm": 0.8433344960212708, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 11798 + }, + { + "epoch": 0.11799, + "grad_norm": 0.9259787201881409, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 11799 + }, + { + "epoch": 0.118, + "grad_norm": 0.9990931749343872, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 11800 + }, + { + "epoch": 0.11801, + "grad_norm": 1.0543739795684814, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 11801 + }, + { + "epoch": 0.11802, + "grad_norm": 1.006338357925415, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 11802 + }, + { + "epoch": 0.11803, + "grad_norm": 0.9883202910423279, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 11803 + }, + { + "epoch": 0.11804, + "grad_norm": 0.9899671077728271, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 11804 + }, + { + "epoch": 0.11805, + "grad_norm": 0.8588717579841614, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 11805 + }, + { + "epoch": 0.11806, + "grad_norm": 0.7318597435951233, + "learning_rate": 0.003, + "loss": 4.027, + "step": 11806 + }, + { + "epoch": 0.11807, + "grad_norm": 0.689117968082428, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 11807 + }, + { + "epoch": 0.11808, + "grad_norm": 0.6362882256507874, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 11808 + }, + { + "epoch": 0.11809, + "grad_norm": 0.641882061958313, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 11809 + }, + { + "epoch": 0.1181, + "grad_norm": 0.8091892004013062, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 11810 + }, + { + "epoch": 0.11811, + "grad_norm": 1.0445493459701538, + "learning_rate": 0.003, + "loss": 4.041, + "step": 11811 + }, + { + "epoch": 0.11812, + "grad_norm": 0.9836093187332153, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 11812 + }, + { + "epoch": 0.11813, + "grad_norm": 0.8166530728340149, + "learning_rate": 0.003, + "loss": 4.024, + "step": 11813 + }, + { + "epoch": 0.11814, + "grad_norm": 0.7115487456321716, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 11814 + }, + { + "epoch": 0.11815, + "grad_norm": 0.6834825277328491, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 11815 + }, + { + "epoch": 0.11816, + "grad_norm": 0.8012105226516724, + "learning_rate": 0.003, + "loss": 4.002, + "step": 11816 + }, + { + "epoch": 0.11817, + "grad_norm": 0.9633815288543701, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 11817 + }, + { + "epoch": 0.11818, + "grad_norm": 1.0355150699615479, + "learning_rate": 0.003, + "loss": 4.032, + "step": 11818 + }, + { + "epoch": 0.11819, + "grad_norm": 0.7960835695266724, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 11819 + }, + { + "epoch": 0.1182, + "grad_norm": 0.6536712646484375, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 11820 + }, + { + "epoch": 0.11821, + "grad_norm": 0.6484993100166321, + "learning_rate": 0.003, + "loss": 3.984, + "step": 11821 + }, + { + "epoch": 0.11822, + "grad_norm": 0.7504442930221558, + "learning_rate": 0.003, + "loss": 4.027, + "step": 11822 + }, + { + "epoch": 0.11823, + "grad_norm": 0.7606385350227356, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 11823 + }, + { + "epoch": 0.11824, + "grad_norm": 0.7401397824287415, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 11824 + }, + { + "epoch": 0.11825, + "grad_norm": 0.6549736857414246, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 11825 + }, + { + "epoch": 0.11826, + "grad_norm": 0.7295933365821838, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 11826 + }, + { + "epoch": 0.11827, + "grad_norm": 0.8805853724479675, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 11827 + }, + { + "epoch": 0.11828, + "grad_norm": 0.7789543271064758, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 11828 + }, + { + "epoch": 0.11829, + "grad_norm": 0.6787858009338379, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 11829 + }, + { + "epoch": 0.1183, + "grad_norm": 0.5975125432014465, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 11830 + }, + { + "epoch": 0.11831, + "grad_norm": 0.6243595480918884, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 11831 + }, + { + "epoch": 0.11832, + "grad_norm": 0.5720900893211365, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 11832 + }, + { + "epoch": 0.11833, + "grad_norm": 0.502043604850769, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 11833 + }, + { + "epoch": 0.11834, + "grad_norm": 0.4862186312675476, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 11834 + }, + { + "epoch": 0.11835, + "grad_norm": 0.4785534143447876, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 11835 + }, + { + "epoch": 0.11836, + "grad_norm": 0.4916263222694397, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 11836 + }, + { + "epoch": 0.11837, + "grad_norm": 0.5785505771636963, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 11837 + }, + { + "epoch": 0.11838, + "grad_norm": 0.7762029767036438, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 11838 + }, + { + "epoch": 0.11839, + "grad_norm": 1.147498369216919, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 11839 + }, + { + "epoch": 0.1184, + "grad_norm": 1.1743519306182861, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 11840 + }, + { + "epoch": 0.11841, + "grad_norm": 0.7633300423622131, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 11841 + }, + { + "epoch": 0.11842, + "grad_norm": 0.7302260398864746, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 11842 + }, + { + "epoch": 0.11843, + "grad_norm": 0.9037927389144897, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 11843 + }, + { + "epoch": 0.11844, + "grad_norm": 1.0036314725875854, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 11844 + }, + { + "epoch": 0.11845, + "grad_norm": 0.981715202331543, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 11845 + }, + { + "epoch": 0.11846, + "grad_norm": 0.9306796789169312, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 11846 + }, + { + "epoch": 0.11847, + "grad_norm": 0.9010899662971497, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 11847 + }, + { + "epoch": 0.11848, + "grad_norm": 0.9531976580619812, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 11848 + }, + { + "epoch": 0.11849, + "grad_norm": 0.9927566647529602, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 11849 + }, + { + "epoch": 0.1185, + "grad_norm": 1.0491796731948853, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 11850 + }, + { + "epoch": 0.11851, + "grad_norm": 0.8954125046730042, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 11851 + }, + { + "epoch": 0.11852, + "grad_norm": 0.9096914529800415, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 11852 + }, + { + "epoch": 0.11853, + "grad_norm": 0.7658776044845581, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 11853 + }, + { + "epoch": 0.11854, + "grad_norm": 0.724291980266571, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 11854 + }, + { + "epoch": 0.11855, + "grad_norm": 0.6638486981391907, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 11855 + }, + { + "epoch": 0.11856, + "grad_norm": 0.7768213152885437, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 11856 + }, + { + "epoch": 0.11857, + "grad_norm": 0.8802120685577393, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 11857 + }, + { + "epoch": 0.11858, + "grad_norm": 1.0737560987472534, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 11858 + }, + { + "epoch": 0.11859, + "grad_norm": 1.0319401025772095, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 11859 + }, + { + "epoch": 0.1186, + "grad_norm": 1.0573298931121826, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 11860 + }, + { + "epoch": 0.11861, + "grad_norm": 0.9240105152130127, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 11861 + }, + { + "epoch": 0.11862, + "grad_norm": 0.8155292272567749, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 11862 + }, + { + "epoch": 0.11863, + "grad_norm": 0.7273171544075012, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 11863 + }, + { + "epoch": 0.11864, + "grad_norm": 0.7226408123970032, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 11864 + }, + { + "epoch": 0.11865, + "grad_norm": 0.7888292670249939, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 11865 + }, + { + "epoch": 0.11866, + "grad_norm": 0.9124221205711365, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 11866 + }, + { + "epoch": 0.11867, + "grad_norm": 0.9692350029945374, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 11867 + }, + { + "epoch": 0.11868, + "grad_norm": 0.9330461621284485, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 11868 + }, + { + "epoch": 0.11869, + "grad_norm": 0.8661285042762756, + "learning_rate": 0.003, + "loss": 4.032, + "step": 11869 + }, + { + "epoch": 0.1187, + "grad_norm": 0.852867603302002, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 11870 + }, + { + "epoch": 0.11871, + "grad_norm": 0.8410532474517822, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 11871 + }, + { + "epoch": 0.11872, + "grad_norm": 0.7677245736122131, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 11872 + }, + { + "epoch": 0.11873, + "grad_norm": 0.7099470496177673, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 11873 + }, + { + "epoch": 0.11874, + "grad_norm": 0.5636304616928101, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 11874 + }, + { + "epoch": 0.11875, + "grad_norm": 0.6461973190307617, + "learning_rate": 0.003, + "loss": 4.027, + "step": 11875 + }, + { + "epoch": 0.11876, + "grad_norm": 0.6658841967582703, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 11876 + }, + { + "epoch": 0.11877, + "grad_norm": 0.655237078666687, + "learning_rate": 0.003, + "loss": 4.044, + "step": 11877 + }, + { + "epoch": 0.11878, + "grad_norm": 0.7163628935813904, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 11878 + }, + { + "epoch": 0.11879, + "grad_norm": 0.5543665885925293, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 11879 + }, + { + "epoch": 0.1188, + "grad_norm": 0.5670758485794067, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 11880 + }, + { + "epoch": 0.11881, + "grad_norm": 0.511709451675415, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 11881 + }, + { + "epoch": 0.11882, + "grad_norm": 0.4493480920791626, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 11882 + }, + { + "epoch": 0.11883, + "grad_norm": 0.5147427320480347, + "learning_rate": 0.003, + "loss": 4.032, + "step": 11883 + }, + { + "epoch": 0.11884, + "grad_norm": 0.6153466105461121, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 11884 + }, + { + "epoch": 0.11885, + "grad_norm": 0.8834508061408997, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 11885 + }, + { + "epoch": 0.11886, + "grad_norm": 1.244941234588623, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 11886 + }, + { + "epoch": 0.11887, + "grad_norm": 0.7770438194274902, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 11887 + }, + { + "epoch": 0.11888, + "grad_norm": 0.6514125466346741, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 11888 + }, + { + "epoch": 0.11889, + "grad_norm": 0.6172106862068176, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 11889 + }, + { + "epoch": 0.1189, + "grad_norm": 0.607948899269104, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 11890 + }, + { + "epoch": 0.11891, + "grad_norm": 0.7253835201263428, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 11891 + }, + { + "epoch": 0.11892, + "grad_norm": 0.6802498698234558, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 11892 + }, + { + "epoch": 0.11893, + "grad_norm": 0.7029039859771729, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 11893 + }, + { + "epoch": 0.11894, + "grad_norm": 0.7636929154396057, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 11894 + }, + { + "epoch": 0.11895, + "grad_norm": 0.7230116724967957, + "learning_rate": 0.003, + "loss": 4.003, + "step": 11895 + }, + { + "epoch": 0.11896, + "grad_norm": 0.6269598603248596, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 11896 + }, + { + "epoch": 0.11897, + "grad_norm": 0.5971459746360779, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 11897 + }, + { + "epoch": 0.11898, + "grad_norm": 0.7045280337333679, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 11898 + }, + { + "epoch": 0.11899, + "grad_norm": 0.7602986097335815, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 11899 + }, + { + "epoch": 0.119, + "grad_norm": 0.762942910194397, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 11900 + }, + { + "epoch": 0.11901, + "grad_norm": 0.8937726020812988, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 11901 + }, + { + "epoch": 0.11902, + "grad_norm": 1.0988658666610718, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 11902 + }, + { + "epoch": 0.11903, + "grad_norm": 1.0117043256759644, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 11903 + }, + { + "epoch": 0.11904, + "grad_norm": 1.114853858947754, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 11904 + }, + { + "epoch": 0.11905, + "grad_norm": 0.8835121393203735, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 11905 + }, + { + "epoch": 0.11906, + "grad_norm": 0.7715875506401062, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 11906 + }, + { + "epoch": 0.11907, + "grad_norm": 0.7026994824409485, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 11907 + }, + { + "epoch": 0.11908, + "grad_norm": 0.7263714671134949, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 11908 + }, + { + "epoch": 0.11909, + "grad_norm": 0.8062036037445068, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 11909 + }, + { + "epoch": 0.1191, + "grad_norm": 0.862729549407959, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 11910 + }, + { + "epoch": 0.11911, + "grad_norm": 0.8009734749794006, + "learning_rate": 0.003, + "loss": 4.024, + "step": 11911 + }, + { + "epoch": 0.11912, + "grad_norm": 0.9418559670448303, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 11912 + }, + { + "epoch": 0.11913, + "grad_norm": 1.1915781497955322, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 11913 + }, + { + "epoch": 0.11914, + "grad_norm": 0.8546242713928223, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 11914 + }, + { + "epoch": 0.11915, + "grad_norm": 0.697476327419281, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 11915 + }, + { + "epoch": 0.11916, + "grad_norm": 0.7126041650772095, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 11916 + }, + { + "epoch": 0.11917, + "grad_norm": 0.7683954238891602, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 11917 + }, + { + "epoch": 0.11918, + "grad_norm": 0.7003222703933716, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 11918 + }, + { + "epoch": 0.11919, + "grad_norm": 0.6144813299179077, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 11919 + }, + { + "epoch": 0.1192, + "grad_norm": 0.6127336025238037, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 11920 + }, + { + "epoch": 0.11921, + "grad_norm": 0.6253604292869568, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 11921 + }, + { + "epoch": 0.11922, + "grad_norm": 0.716572642326355, + "learning_rate": 0.003, + "loss": 4.04, + "step": 11922 + }, + { + "epoch": 0.11923, + "grad_norm": 0.8776218295097351, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 11923 + }, + { + "epoch": 0.11924, + "grad_norm": 0.9347663521766663, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 11924 + }, + { + "epoch": 0.11925, + "grad_norm": 1.0252727270126343, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 11925 + }, + { + "epoch": 0.11926, + "grad_norm": 1.1369107961654663, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 11926 + }, + { + "epoch": 0.11927, + "grad_norm": 0.8379679918289185, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 11927 + }, + { + "epoch": 0.11928, + "grad_norm": 0.6795943379402161, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 11928 + }, + { + "epoch": 0.11929, + "grad_norm": 0.6492370367050171, + "learning_rate": 0.003, + "loss": 4.019, + "step": 11929 + }, + { + "epoch": 0.1193, + "grad_norm": 0.7252062559127808, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 11930 + }, + { + "epoch": 0.11931, + "grad_norm": 0.6780815720558167, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 11931 + }, + { + "epoch": 0.11932, + "grad_norm": 0.678745448589325, + "learning_rate": 0.003, + "loss": 4.025, + "step": 11932 + }, + { + "epoch": 0.11933, + "grad_norm": 0.6801624298095703, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 11933 + }, + { + "epoch": 0.11934, + "grad_norm": 0.7659531235694885, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 11934 + }, + { + "epoch": 0.11935, + "grad_norm": 0.8605554699897766, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 11935 + }, + { + "epoch": 0.11936, + "grad_norm": 0.9928908944129944, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 11936 + }, + { + "epoch": 0.11937, + "grad_norm": 0.9970818161964417, + "learning_rate": 0.003, + "loss": 4.033, + "step": 11937 + }, + { + "epoch": 0.11938, + "grad_norm": 0.8282353281974792, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 11938 + }, + { + "epoch": 0.11939, + "grad_norm": 0.802817702293396, + "learning_rate": 0.003, + "loss": 4.01, + "step": 11939 + }, + { + "epoch": 0.1194, + "grad_norm": 0.7577781081199646, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 11940 + }, + { + "epoch": 0.11941, + "grad_norm": 0.7385069727897644, + "learning_rate": 0.003, + "loss": 4.029, + "step": 11941 + }, + { + "epoch": 0.11942, + "grad_norm": 0.7899826765060425, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 11942 + }, + { + "epoch": 0.11943, + "grad_norm": 0.7291315793991089, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 11943 + }, + { + "epoch": 0.11944, + "grad_norm": 0.6764593720436096, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 11944 + }, + { + "epoch": 0.11945, + "grad_norm": 0.7137207388877869, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 11945 + }, + { + "epoch": 0.11946, + "grad_norm": 0.9325737953186035, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 11946 + }, + { + "epoch": 0.11947, + "grad_norm": 1.2605060338974, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 11947 + }, + { + "epoch": 0.11948, + "grad_norm": 0.7151952981948853, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 11948 + }, + { + "epoch": 0.11949, + "grad_norm": 0.6542052030563354, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 11949 + }, + { + "epoch": 0.1195, + "grad_norm": 0.5991142988204956, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 11950 + }, + { + "epoch": 0.11951, + "grad_norm": 0.6021481156349182, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 11951 + }, + { + "epoch": 0.11952, + "grad_norm": 0.6924704313278198, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 11952 + }, + { + "epoch": 0.11953, + "grad_norm": 0.609024703502655, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 11953 + }, + { + "epoch": 0.11954, + "grad_norm": 0.6090964674949646, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 11954 + }, + { + "epoch": 0.11955, + "grad_norm": 0.6386412382125854, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 11955 + }, + { + "epoch": 0.11956, + "grad_norm": 0.7164059281349182, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 11956 + }, + { + "epoch": 0.11957, + "grad_norm": 0.7793961763381958, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 11957 + }, + { + "epoch": 0.11958, + "grad_norm": 0.6898953914642334, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 11958 + }, + { + "epoch": 0.11959, + "grad_norm": 0.7248098254203796, + "learning_rate": 0.003, + "loss": 3.99, + "step": 11959 + }, + { + "epoch": 0.1196, + "grad_norm": 0.9642515778541565, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 11960 + }, + { + "epoch": 0.11961, + "grad_norm": 1.08201265335083, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 11961 + }, + { + "epoch": 0.11962, + "grad_norm": 0.8699434995651245, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 11962 + }, + { + "epoch": 0.11963, + "grad_norm": 0.8050743341445923, + "learning_rate": 0.003, + "loss": 4.015, + "step": 11963 + }, + { + "epoch": 0.11964, + "grad_norm": 0.7765284776687622, + "learning_rate": 0.003, + "loss": 4.012, + "step": 11964 + }, + { + "epoch": 0.11965, + "grad_norm": 0.894541323184967, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 11965 + }, + { + "epoch": 0.11966, + "grad_norm": 1.0280059576034546, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 11966 + }, + { + "epoch": 0.11967, + "grad_norm": 1.1794424057006836, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 11967 + }, + { + "epoch": 0.11968, + "grad_norm": 0.886949896812439, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 11968 + }, + { + "epoch": 0.11969, + "grad_norm": 0.7620879411697388, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 11969 + }, + { + "epoch": 0.1197, + "grad_norm": 0.7351470589637756, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 11970 + }, + { + "epoch": 0.11971, + "grad_norm": 0.7624823451042175, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 11971 + }, + { + "epoch": 0.11972, + "grad_norm": 0.6433542966842651, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 11972 + }, + { + "epoch": 0.11973, + "grad_norm": 0.6336454153060913, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 11973 + }, + { + "epoch": 0.11974, + "grad_norm": 0.6942681670188904, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 11974 + }, + { + "epoch": 0.11975, + "grad_norm": 0.7114654183387756, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 11975 + }, + { + "epoch": 0.11976, + "grad_norm": 0.7263957858085632, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 11976 + }, + { + "epoch": 0.11977, + "grad_norm": 0.8133060932159424, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 11977 + }, + { + "epoch": 0.11978, + "grad_norm": 1.0520018339157104, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 11978 + }, + { + "epoch": 0.11979, + "grad_norm": 1.1112744808197021, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 11979 + }, + { + "epoch": 0.1198, + "grad_norm": 0.9063125848770142, + "learning_rate": 0.003, + "loss": 4.057, + "step": 11980 + }, + { + "epoch": 0.11981, + "grad_norm": 0.9346314668655396, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 11981 + }, + { + "epoch": 0.11982, + "grad_norm": 0.9792442321777344, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 11982 + }, + { + "epoch": 0.11983, + "grad_norm": 0.9704103469848633, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 11983 + }, + { + "epoch": 0.11984, + "grad_norm": 0.849890410900116, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 11984 + }, + { + "epoch": 0.11985, + "grad_norm": 0.9243372678756714, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 11985 + }, + { + "epoch": 0.11986, + "grad_norm": 1.137408971786499, + "learning_rate": 0.003, + "loss": 4.037, + "step": 11986 + }, + { + "epoch": 0.11987, + "grad_norm": 0.8184463381767273, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 11987 + }, + { + "epoch": 0.11988, + "grad_norm": 0.7495303153991699, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 11988 + }, + { + "epoch": 0.11989, + "grad_norm": 0.700872004032135, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 11989 + }, + { + "epoch": 0.1199, + "grad_norm": 0.7107323408126831, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 11990 + }, + { + "epoch": 0.11991, + "grad_norm": 0.8026617765426636, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 11991 + }, + { + "epoch": 0.11992, + "grad_norm": 0.8296177983283997, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 11992 + }, + { + "epoch": 0.11993, + "grad_norm": 0.8199352622032166, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 11993 + }, + { + "epoch": 0.11994, + "grad_norm": 0.711927592754364, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 11994 + }, + { + "epoch": 0.11995, + "grad_norm": 0.7301411628723145, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 11995 + }, + { + "epoch": 0.11996, + "grad_norm": 0.7096173763275146, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 11996 + }, + { + "epoch": 0.11997, + "grad_norm": 0.6520859003067017, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 11997 + }, + { + "epoch": 0.11998, + "grad_norm": 0.6798012852668762, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 11998 + }, + { + "epoch": 0.11999, + "grad_norm": 0.8672653436660767, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 11999 + }, + { + "epoch": 0.12, + "grad_norm": 1.0057185888290405, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 12000 + }, + { + "epoch": 0.12001, + "grad_norm": 0.9552724361419678, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 12001 + }, + { + "epoch": 0.12002, + "grad_norm": 1.129225492477417, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 12002 + }, + { + "epoch": 0.12003, + "grad_norm": 0.8080599904060364, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 12003 + }, + { + "epoch": 0.12004, + "grad_norm": 0.7223402857780457, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 12004 + }, + { + "epoch": 0.12005, + "grad_norm": 0.6406126022338867, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 12005 + }, + { + "epoch": 0.12006, + "grad_norm": 0.6637592315673828, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 12006 + }, + { + "epoch": 0.12007, + "grad_norm": 0.8314640522003174, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 12007 + }, + { + "epoch": 0.12008, + "grad_norm": 0.9080745577812195, + "learning_rate": 0.003, + "loss": 4.05, + "step": 12008 + }, + { + "epoch": 0.12009, + "grad_norm": 0.8800250887870789, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 12009 + }, + { + "epoch": 0.1201, + "grad_norm": 0.8031384348869324, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 12010 + }, + { + "epoch": 0.12011, + "grad_norm": 0.9114692807197571, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 12011 + }, + { + "epoch": 0.12012, + "grad_norm": 0.9691035747528076, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 12012 + }, + { + "epoch": 0.12013, + "grad_norm": 1.0572527647018433, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 12013 + }, + { + "epoch": 0.12014, + "grad_norm": 0.914512574672699, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 12014 + }, + { + "epoch": 0.12015, + "grad_norm": 0.7966519594192505, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 12015 + }, + { + "epoch": 0.12016, + "grad_norm": 0.5781070590019226, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 12016 + }, + { + "epoch": 0.12017, + "grad_norm": 0.5625675916671753, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 12017 + }, + { + "epoch": 0.12018, + "grad_norm": 0.7984493374824524, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 12018 + }, + { + "epoch": 0.12019, + "grad_norm": 0.8596245050430298, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 12019 + }, + { + "epoch": 0.1202, + "grad_norm": 0.8960752487182617, + "learning_rate": 0.003, + "loss": 4.03, + "step": 12020 + }, + { + "epoch": 0.12021, + "grad_norm": 0.8911660313606262, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 12021 + }, + { + "epoch": 0.12022, + "grad_norm": 0.7461158633232117, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 12022 + }, + { + "epoch": 0.12023, + "grad_norm": 0.7623320817947388, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 12023 + }, + { + "epoch": 0.12024, + "grad_norm": 0.781243622303009, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 12024 + }, + { + "epoch": 0.12025, + "grad_norm": 0.843302845954895, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 12025 + }, + { + "epoch": 0.12026, + "grad_norm": 0.902095377445221, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 12026 + }, + { + "epoch": 0.12027, + "grad_norm": 1.1767667531967163, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 12027 + }, + { + "epoch": 0.12028, + "grad_norm": 0.8134157061576843, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 12028 + }, + { + "epoch": 0.12029, + "grad_norm": 0.7799186706542969, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 12029 + }, + { + "epoch": 0.1203, + "grad_norm": 0.8834793567657471, + "learning_rate": 0.003, + "loss": 4.016, + "step": 12030 + }, + { + "epoch": 0.12031, + "grad_norm": 1.112605333328247, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 12031 + }, + { + "epoch": 0.12032, + "grad_norm": 1.2384040355682373, + "learning_rate": 0.003, + "loss": 4.026, + "step": 12032 + }, + { + "epoch": 0.12033, + "grad_norm": 0.752450168132782, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 12033 + }, + { + "epoch": 0.12034, + "grad_norm": 0.6615104079246521, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 12034 + }, + { + "epoch": 0.12035, + "grad_norm": 0.7580855488777161, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 12035 + }, + { + "epoch": 0.12036, + "grad_norm": 0.6694372892379761, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 12036 + }, + { + "epoch": 0.12037, + "grad_norm": 0.7598423361778259, + "learning_rate": 0.003, + "loss": 4.029, + "step": 12037 + }, + { + "epoch": 0.12038, + "grad_norm": 0.7612495422363281, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 12038 + }, + { + "epoch": 0.12039, + "grad_norm": 0.7796201705932617, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 12039 + }, + { + "epoch": 0.1204, + "grad_norm": 0.7708131670951843, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 12040 + }, + { + "epoch": 0.12041, + "grad_norm": 0.707992672920227, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 12041 + }, + { + "epoch": 0.12042, + "grad_norm": 0.6602921485900879, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 12042 + }, + { + "epoch": 0.12043, + "grad_norm": 0.7230878472328186, + "learning_rate": 0.003, + "loss": 4.038, + "step": 12043 + }, + { + "epoch": 0.12044, + "grad_norm": 0.7043576240539551, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 12044 + }, + { + "epoch": 0.12045, + "grad_norm": 0.7773805260658264, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 12045 + }, + { + "epoch": 0.12046, + "grad_norm": 0.9323230981826782, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 12046 + }, + { + "epoch": 0.12047, + "grad_norm": 0.9924766421318054, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 12047 + }, + { + "epoch": 0.12048, + "grad_norm": 0.9833324551582336, + "learning_rate": 0.003, + "loss": 4.022, + "step": 12048 + }, + { + "epoch": 0.12049, + "grad_norm": 0.8768059015274048, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 12049 + }, + { + "epoch": 0.1205, + "grad_norm": 0.6571690440177917, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 12050 + }, + { + "epoch": 0.12051, + "grad_norm": 0.6308192610740662, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 12051 + }, + { + "epoch": 0.12052, + "grad_norm": 0.7342305183410645, + "learning_rate": 0.003, + "loss": 4.028, + "step": 12052 + }, + { + "epoch": 0.12053, + "grad_norm": 0.9399965405464172, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 12053 + }, + { + "epoch": 0.12054, + "grad_norm": 1.0758345127105713, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 12054 + }, + { + "epoch": 0.12055, + "grad_norm": 0.7563015818595886, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 12055 + }, + { + "epoch": 0.12056, + "grad_norm": 0.6503219604492188, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 12056 + }, + { + "epoch": 0.12057, + "grad_norm": 0.6826764345169067, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 12057 + }, + { + "epoch": 0.12058, + "grad_norm": 0.7406718134880066, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 12058 + }, + { + "epoch": 0.12059, + "grad_norm": 0.7053825855255127, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 12059 + }, + { + "epoch": 0.1206, + "grad_norm": 0.7572804093360901, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 12060 + }, + { + "epoch": 0.12061, + "grad_norm": 0.9737385511398315, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 12061 + }, + { + "epoch": 0.12062, + "grad_norm": 1.229773998260498, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 12062 + }, + { + "epoch": 0.12063, + "grad_norm": 0.9908783435821533, + "learning_rate": 0.003, + "loss": 4.034, + "step": 12063 + }, + { + "epoch": 0.12064, + "grad_norm": 0.8750876188278198, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 12064 + }, + { + "epoch": 0.12065, + "grad_norm": 0.8608133792877197, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 12065 + }, + { + "epoch": 0.12066, + "grad_norm": 0.8134159445762634, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 12066 + }, + { + "epoch": 0.12067, + "grad_norm": 0.6643102169036865, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 12067 + }, + { + "epoch": 0.12068, + "grad_norm": 0.7150804400444031, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 12068 + }, + { + "epoch": 0.12069, + "grad_norm": 0.8066950440406799, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 12069 + }, + { + "epoch": 0.1207, + "grad_norm": 0.8921740055084229, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 12070 + }, + { + "epoch": 0.12071, + "grad_norm": 1.0176774263381958, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 12071 + }, + { + "epoch": 0.12072, + "grad_norm": 0.9523666501045227, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 12072 + }, + { + "epoch": 0.12073, + "grad_norm": 0.9739864468574524, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 12073 + }, + { + "epoch": 0.12074, + "grad_norm": 0.8286561369895935, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 12074 + }, + { + "epoch": 0.12075, + "grad_norm": 0.7403897643089294, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 12075 + }, + { + "epoch": 0.12076, + "grad_norm": 0.6872183084487915, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 12076 + }, + { + "epoch": 0.12077, + "grad_norm": 0.6335912942886353, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 12077 + }, + { + "epoch": 0.12078, + "grad_norm": 0.6662702560424805, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 12078 + }, + { + "epoch": 0.12079, + "grad_norm": 0.6359442472457886, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 12079 + }, + { + "epoch": 0.1208, + "grad_norm": 0.5857674479484558, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 12080 + }, + { + "epoch": 0.12081, + "grad_norm": 0.6044425964355469, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 12081 + }, + { + "epoch": 0.12082, + "grad_norm": 0.6441171169281006, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 12082 + }, + { + "epoch": 0.12083, + "grad_norm": 0.6959605813026428, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 12083 + }, + { + "epoch": 0.12084, + "grad_norm": 0.7471359968185425, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 12084 + }, + { + "epoch": 0.12085, + "grad_norm": 0.92557293176651, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 12085 + }, + { + "epoch": 0.12086, + "grad_norm": 1.1548380851745605, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 12086 + }, + { + "epoch": 0.12087, + "grad_norm": 0.8555243015289307, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 12087 + }, + { + "epoch": 0.12088, + "grad_norm": 0.6919422149658203, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 12088 + }, + { + "epoch": 0.12089, + "grad_norm": 0.654999315738678, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 12089 + }, + { + "epoch": 0.1209, + "grad_norm": 0.7054887413978577, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 12090 + }, + { + "epoch": 0.12091, + "grad_norm": 0.6020074486732483, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 12091 + }, + { + "epoch": 0.12092, + "grad_norm": 0.6386114954948425, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 12092 + }, + { + "epoch": 0.12093, + "grad_norm": 0.7278828620910645, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 12093 + }, + { + "epoch": 0.12094, + "grad_norm": 0.8216177821159363, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 12094 + }, + { + "epoch": 0.12095, + "grad_norm": 0.8740129470825195, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 12095 + }, + { + "epoch": 0.12096, + "grad_norm": 0.9309924244880676, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 12096 + }, + { + "epoch": 0.12097, + "grad_norm": 0.9449071884155273, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 12097 + }, + { + "epoch": 0.12098, + "grad_norm": 0.806206226348877, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 12098 + }, + { + "epoch": 0.12099, + "grad_norm": 0.6736234426498413, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 12099 + }, + { + "epoch": 0.121, + "grad_norm": 0.8320266008377075, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 12100 + }, + { + "epoch": 0.12101, + "grad_norm": 0.9483769536018372, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 12101 + }, + { + "epoch": 0.12102, + "grad_norm": 1.183903455734253, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 12102 + }, + { + "epoch": 0.12103, + "grad_norm": 0.9592525362968445, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 12103 + }, + { + "epoch": 0.12104, + "grad_norm": 0.8384464383125305, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 12104 + }, + { + "epoch": 0.12105, + "grad_norm": 0.7661356925964355, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 12105 + }, + { + "epoch": 0.12106, + "grad_norm": 0.8419826626777649, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 12106 + }, + { + "epoch": 0.12107, + "grad_norm": 0.9251189231872559, + "learning_rate": 0.003, + "loss": 4.048, + "step": 12107 + }, + { + "epoch": 0.12108, + "grad_norm": 1.063491702079773, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 12108 + }, + { + "epoch": 0.12109, + "grad_norm": 0.8947046995162964, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 12109 + }, + { + "epoch": 0.1211, + "grad_norm": 0.8604028820991516, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 12110 + }, + { + "epoch": 0.12111, + "grad_norm": 0.7970651984214783, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 12111 + }, + { + "epoch": 0.12112, + "grad_norm": 0.7963676452636719, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 12112 + }, + { + "epoch": 0.12113, + "grad_norm": 0.7746478319168091, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 12113 + }, + { + "epoch": 0.12114, + "grad_norm": 0.7988983988761902, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 12114 + }, + { + "epoch": 0.12115, + "grad_norm": 0.8888752460479736, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 12115 + }, + { + "epoch": 0.12116, + "grad_norm": 0.8377873301506042, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 12116 + }, + { + "epoch": 0.12117, + "grad_norm": 0.6950471997261047, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 12117 + }, + { + "epoch": 0.12118, + "grad_norm": 0.7134587168693542, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 12118 + }, + { + "epoch": 0.12119, + "grad_norm": 0.6574320197105408, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 12119 + }, + { + "epoch": 0.1212, + "grad_norm": 0.5444393754005432, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 12120 + }, + { + "epoch": 0.12121, + "grad_norm": 0.5818464159965515, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 12121 + }, + { + "epoch": 0.12122, + "grad_norm": 0.6494410037994385, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 12122 + }, + { + "epoch": 0.12123, + "grad_norm": 0.7310759425163269, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 12123 + }, + { + "epoch": 0.12124, + "grad_norm": 0.7354447841644287, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 12124 + }, + { + "epoch": 0.12125, + "grad_norm": 0.7390363216400146, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 12125 + }, + { + "epoch": 0.12126, + "grad_norm": 0.7380009293556213, + "learning_rate": 0.003, + "loss": 4.02, + "step": 12126 + }, + { + "epoch": 0.12127, + "grad_norm": 0.7050486207008362, + "learning_rate": 0.003, + "loss": 4.012, + "step": 12127 + }, + { + "epoch": 0.12128, + "grad_norm": 0.8125148415565491, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 12128 + }, + { + "epoch": 0.12129, + "grad_norm": 0.8589991927146912, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 12129 + }, + { + "epoch": 0.1213, + "grad_norm": 0.847008466720581, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 12130 + }, + { + "epoch": 0.12131, + "grad_norm": 0.9250040054321289, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 12131 + }, + { + "epoch": 0.12132, + "grad_norm": 1.12769615650177, + "learning_rate": 0.003, + "loss": 4.033, + "step": 12132 + }, + { + "epoch": 0.12133, + "grad_norm": 1.0908305644989014, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 12133 + }, + { + "epoch": 0.12134, + "grad_norm": 0.8170710802078247, + "learning_rate": 0.003, + "loss": 4.036, + "step": 12134 + }, + { + "epoch": 0.12135, + "grad_norm": 0.7351126074790955, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 12135 + }, + { + "epoch": 0.12136, + "grad_norm": 0.7365264892578125, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 12136 + }, + { + "epoch": 0.12137, + "grad_norm": 0.7570641040802002, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 12137 + }, + { + "epoch": 0.12138, + "grad_norm": 0.7057920098304749, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 12138 + }, + { + "epoch": 0.12139, + "grad_norm": 0.7471189498901367, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 12139 + }, + { + "epoch": 0.1214, + "grad_norm": 0.8883476257324219, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 12140 + }, + { + "epoch": 0.12141, + "grad_norm": 0.884615421295166, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 12141 + }, + { + "epoch": 0.12142, + "grad_norm": 0.7976682782173157, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 12142 + }, + { + "epoch": 0.12143, + "grad_norm": 0.8299143314361572, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 12143 + }, + { + "epoch": 0.12144, + "grad_norm": 0.9504224061965942, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 12144 + }, + { + "epoch": 0.12145, + "grad_norm": 0.9362991452217102, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 12145 + }, + { + "epoch": 0.12146, + "grad_norm": 0.9021977782249451, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 12146 + }, + { + "epoch": 0.12147, + "grad_norm": 0.95110684633255, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 12147 + }, + { + "epoch": 0.12148, + "grad_norm": 0.8893001675605774, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 12148 + }, + { + "epoch": 0.12149, + "grad_norm": 0.7543485164642334, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 12149 + }, + { + "epoch": 0.1215, + "grad_norm": 0.6045880317687988, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 12150 + }, + { + "epoch": 0.12151, + "grad_norm": 0.6619433760643005, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 12151 + }, + { + "epoch": 0.12152, + "grad_norm": 0.7473676800727844, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 12152 + }, + { + "epoch": 0.12153, + "grad_norm": 0.6784195899963379, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 12153 + }, + { + "epoch": 0.12154, + "grad_norm": 0.6124919056892395, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 12154 + }, + { + "epoch": 0.12155, + "grad_norm": 0.653683066368103, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 12155 + }, + { + "epoch": 0.12156, + "grad_norm": 0.8463802933692932, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 12156 + }, + { + "epoch": 0.12157, + "grad_norm": 0.9578853249549866, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 12157 + }, + { + "epoch": 0.12158, + "grad_norm": 1.1375261545181274, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 12158 + }, + { + "epoch": 0.12159, + "grad_norm": 0.8238728046417236, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 12159 + }, + { + "epoch": 0.1216, + "grad_norm": 0.7417867183685303, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 12160 + }, + { + "epoch": 0.12161, + "grad_norm": 0.8725582361221313, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 12161 + }, + { + "epoch": 0.12162, + "grad_norm": 0.9587368965148926, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 12162 + }, + { + "epoch": 0.12163, + "grad_norm": 0.9570406079292297, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 12163 + }, + { + "epoch": 0.12164, + "grad_norm": 0.7927216291427612, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 12164 + }, + { + "epoch": 0.12165, + "grad_norm": 0.8281391859054565, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 12165 + }, + { + "epoch": 0.12166, + "grad_norm": 0.8732867240905762, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 12166 + }, + { + "epoch": 0.12167, + "grad_norm": 1.0064791440963745, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 12167 + }, + { + "epoch": 0.12168, + "grad_norm": 1.2000722885131836, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 12168 + }, + { + "epoch": 0.12169, + "grad_norm": 0.7036688327789307, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 12169 + }, + { + "epoch": 0.1217, + "grad_norm": 0.6756927371025085, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 12170 + }, + { + "epoch": 0.12171, + "grad_norm": 0.6766944527626038, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 12171 + }, + { + "epoch": 0.12172, + "grad_norm": 0.6412097215652466, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 12172 + }, + { + "epoch": 0.12173, + "grad_norm": 0.6343686580657959, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 12173 + }, + { + "epoch": 0.12174, + "grad_norm": 0.728533923625946, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 12174 + }, + { + "epoch": 0.12175, + "grad_norm": 0.7192274928092957, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 12175 + }, + { + "epoch": 0.12176, + "grad_norm": 0.8144643902778625, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 12176 + }, + { + "epoch": 0.12177, + "grad_norm": 0.9071642160415649, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 12177 + }, + { + "epoch": 0.12178, + "grad_norm": 1.0799925327301025, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 12178 + }, + { + "epoch": 0.12179, + "grad_norm": 1.0543906688690186, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 12179 + }, + { + "epoch": 0.1218, + "grad_norm": 0.8628104329109192, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 12180 + }, + { + "epoch": 0.12181, + "grad_norm": 0.6865615248680115, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 12181 + }, + { + "epoch": 0.12182, + "grad_norm": 0.7942445278167725, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 12182 + }, + { + "epoch": 0.12183, + "grad_norm": 0.9711741805076599, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 12183 + }, + { + "epoch": 0.12184, + "grad_norm": 1.001904010772705, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 12184 + }, + { + "epoch": 0.12185, + "grad_norm": 0.8621499538421631, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 12185 + }, + { + "epoch": 0.12186, + "grad_norm": 0.7740522027015686, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 12186 + }, + { + "epoch": 0.12187, + "grad_norm": 0.7193096876144409, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 12187 + }, + { + "epoch": 0.12188, + "grad_norm": 0.7161280512809753, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 12188 + }, + { + "epoch": 0.12189, + "grad_norm": 0.6895385384559631, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 12189 + }, + { + "epoch": 0.1219, + "grad_norm": 0.6350817680358887, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 12190 + }, + { + "epoch": 0.12191, + "grad_norm": 0.5773914456367493, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 12191 + }, + { + "epoch": 0.12192, + "grad_norm": 0.6090174317359924, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 12192 + }, + { + "epoch": 0.12193, + "grad_norm": 0.662630021572113, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 12193 + }, + { + "epoch": 0.12194, + "grad_norm": 0.6350228786468506, + "learning_rate": 0.003, + "loss": 4.006, + "step": 12194 + }, + { + "epoch": 0.12195, + "grad_norm": 0.6361150741577148, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 12195 + }, + { + "epoch": 0.12196, + "grad_norm": 0.799187421798706, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 12196 + }, + { + "epoch": 0.12197, + "grad_norm": 0.8982497453689575, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 12197 + }, + { + "epoch": 0.12198, + "grad_norm": 1.0497699975967407, + "learning_rate": 0.003, + "loss": 4.023, + "step": 12198 + }, + { + "epoch": 0.12199, + "grad_norm": 1.0019128322601318, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 12199 + }, + { + "epoch": 0.122, + "grad_norm": 0.9094564914703369, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 12200 + }, + { + "epoch": 0.12201, + "grad_norm": 0.9067058563232422, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 12201 + }, + { + "epoch": 0.12202, + "grad_norm": 0.9713603258132935, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 12202 + }, + { + "epoch": 0.12203, + "grad_norm": 0.9190332889556885, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 12203 + }, + { + "epoch": 0.12204, + "grad_norm": 0.9638498425483704, + "learning_rate": 0.003, + "loss": 4.049, + "step": 12204 + }, + { + "epoch": 0.12205, + "grad_norm": 1.0316067934036255, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 12205 + }, + { + "epoch": 0.12206, + "grad_norm": 1.0103806257247925, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 12206 + }, + { + "epoch": 0.12207, + "grad_norm": 0.9632631540298462, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 12207 + }, + { + "epoch": 0.12208, + "grad_norm": 0.9950851798057556, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 12208 + }, + { + "epoch": 0.12209, + "grad_norm": 0.9666954874992371, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 12209 + }, + { + "epoch": 0.1221, + "grad_norm": 0.8264901041984558, + "learning_rate": 0.003, + "loss": 4.037, + "step": 12210 + }, + { + "epoch": 0.12211, + "grad_norm": 0.9638129472732544, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 12211 + }, + { + "epoch": 0.12212, + "grad_norm": 1.0826435089111328, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 12212 + }, + { + "epoch": 0.12213, + "grad_norm": 0.922379195690155, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 12213 + }, + { + "epoch": 0.12214, + "grad_norm": 1.0430341958999634, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 12214 + }, + { + "epoch": 0.12215, + "grad_norm": 1.1599080562591553, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 12215 + }, + { + "epoch": 0.12216, + "grad_norm": 0.8667312264442444, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 12216 + }, + { + "epoch": 0.12217, + "grad_norm": 0.7546870708465576, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 12217 + }, + { + "epoch": 0.12218, + "grad_norm": 0.7160872220993042, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 12218 + }, + { + "epoch": 0.12219, + "grad_norm": 0.6566637754440308, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 12219 + }, + { + "epoch": 0.1222, + "grad_norm": 0.621151864528656, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 12220 + }, + { + "epoch": 0.12221, + "grad_norm": 0.6620752811431885, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 12221 + }, + { + "epoch": 0.12222, + "grad_norm": 0.6338918805122375, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 12222 + }, + { + "epoch": 0.12223, + "grad_norm": 0.5584115982055664, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 12223 + }, + { + "epoch": 0.12224, + "grad_norm": 0.6339457035064697, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 12224 + }, + { + "epoch": 0.12225, + "grad_norm": 0.706930935382843, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 12225 + }, + { + "epoch": 0.12226, + "grad_norm": 0.7364891767501831, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 12226 + }, + { + "epoch": 0.12227, + "grad_norm": 0.8302417397499084, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 12227 + }, + { + "epoch": 0.12228, + "grad_norm": 0.9262897968292236, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 12228 + }, + { + "epoch": 0.12229, + "grad_norm": 0.8998088240623474, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 12229 + }, + { + "epoch": 0.1223, + "grad_norm": 0.9295552968978882, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 12230 + }, + { + "epoch": 0.12231, + "grad_norm": 1.1946221590042114, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 12231 + }, + { + "epoch": 0.12232, + "grad_norm": 1.2436331510543823, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 12232 + }, + { + "epoch": 0.12233, + "grad_norm": 0.6863802075386047, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 12233 + }, + { + "epoch": 0.12234, + "grad_norm": 0.6205759644508362, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 12234 + }, + { + "epoch": 0.12235, + "grad_norm": 0.7436956167221069, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 12235 + }, + { + "epoch": 0.12236, + "grad_norm": 0.698631227016449, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 12236 + }, + { + "epoch": 0.12237, + "grad_norm": 0.7856922745704651, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 12237 + }, + { + "epoch": 0.12238, + "grad_norm": 0.7397446632385254, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 12238 + }, + { + "epoch": 0.12239, + "grad_norm": 0.6234411597251892, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 12239 + }, + { + "epoch": 0.1224, + "grad_norm": 0.6826146245002747, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 12240 + }, + { + "epoch": 0.12241, + "grad_norm": 0.8209196329116821, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 12241 + }, + { + "epoch": 0.12242, + "grad_norm": 0.8299884796142578, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 12242 + }, + { + "epoch": 0.12243, + "grad_norm": 0.8798038959503174, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 12243 + }, + { + "epoch": 0.12244, + "grad_norm": 0.8742157220840454, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 12244 + }, + { + "epoch": 0.12245, + "grad_norm": 0.8342897891998291, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 12245 + }, + { + "epoch": 0.12246, + "grad_norm": 0.8192770481109619, + "learning_rate": 0.003, + "loss": 4.02, + "step": 12246 + }, + { + "epoch": 0.12247, + "grad_norm": 0.7069940567016602, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 12247 + }, + { + "epoch": 0.12248, + "grad_norm": 0.6147398948669434, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 12248 + }, + { + "epoch": 0.12249, + "grad_norm": 0.5641793012619019, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 12249 + }, + { + "epoch": 0.1225, + "grad_norm": 0.5511481165885925, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 12250 + }, + { + "epoch": 0.12251, + "grad_norm": 0.5728287100791931, + "learning_rate": 0.003, + "loss": 4.033, + "step": 12251 + }, + { + "epoch": 0.12252, + "grad_norm": 0.6058318614959717, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 12252 + }, + { + "epoch": 0.12253, + "grad_norm": 0.7094419002532959, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 12253 + }, + { + "epoch": 0.12254, + "grad_norm": 0.7600876092910767, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 12254 + }, + { + "epoch": 0.12255, + "grad_norm": 0.8422216773033142, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 12255 + }, + { + "epoch": 0.12256, + "grad_norm": 1.0142340660095215, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 12256 + }, + { + "epoch": 0.12257, + "grad_norm": 1.1331077814102173, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 12257 + }, + { + "epoch": 0.12258, + "grad_norm": 0.7469304800033569, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 12258 + }, + { + "epoch": 0.12259, + "grad_norm": 0.706351637840271, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 12259 + }, + { + "epoch": 0.1226, + "grad_norm": 0.6740030646324158, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 12260 + }, + { + "epoch": 0.12261, + "grad_norm": 0.6421547532081604, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 12261 + }, + { + "epoch": 0.12262, + "grad_norm": 0.6817286014556885, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 12262 + }, + { + "epoch": 0.12263, + "grad_norm": 0.7440471649169922, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 12263 + }, + { + "epoch": 0.12264, + "grad_norm": 0.6655385494232178, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 12264 + }, + { + "epoch": 0.12265, + "grad_norm": 0.6313167214393616, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 12265 + }, + { + "epoch": 0.12266, + "grad_norm": 0.806358814239502, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 12266 + }, + { + "epoch": 0.12267, + "grad_norm": 1.0218111276626587, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 12267 + }, + { + "epoch": 0.12268, + "grad_norm": 1.006636381149292, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 12268 + }, + { + "epoch": 0.12269, + "grad_norm": 0.9273249506950378, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 12269 + }, + { + "epoch": 0.1227, + "grad_norm": 0.9001882076263428, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 12270 + }, + { + "epoch": 0.12271, + "grad_norm": 0.8164517283439636, + "learning_rate": 0.003, + "loss": 4.011, + "step": 12271 + }, + { + "epoch": 0.12272, + "grad_norm": 0.9012036919593811, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 12272 + }, + { + "epoch": 0.12273, + "grad_norm": 1.025341272354126, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 12273 + }, + { + "epoch": 0.12274, + "grad_norm": 1.055772066116333, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 12274 + }, + { + "epoch": 0.12275, + "grad_norm": 1.0038049221038818, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 12275 + }, + { + "epoch": 0.12276, + "grad_norm": 0.9755121469497681, + "learning_rate": 0.003, + "loss": 4.027, + "step": 12276 + }, + { + "epoch": 0.12277, + "grad_norm": 0.9044497013092041, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 12277 + }, + { + "epoch": 0.12278, + "grad_norm": 0.8723435997962952, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 12278 + }, + { + "epoch": 0.12279, + "grad_norm": 0.7940369248390198, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 12279 + }, + { + "epoch": 0.1228, + "grad_norm": 0.8326817750930786, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 12280 + }, + { + "epoch": 0.12281, + "grad_norm": 0.8071943521499634, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 12281 + }, + { + "epoch": 0.12282, + "grad_norm": 0.7996351718902588, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 12282 + }, + { + "epoch": 0.12283, + "grad_norm": 0.7421732544898987, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 12283 + }, + { + "epoch": 0.12284, + "grad_norm": 0.7162188291549683, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 12284 + }, + { + "epoch": 0.12285, + "grad_norm": 0.6978950500488281, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 12285 + }, + { + "epoch": 0.12286, + "grad_norm": 0.6271545886993408, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 12286 + }, + { + "epoch": 0.12287, + "grad_norm": 0.6546214818954468, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 12287 + }, + { + "epoch": 0.12288, + "grad_norm": 0.6622276306152344, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 12288 + }, + { + "epoch": 0.12289, + "grad_norm": 0.7192988991737366, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 12289 + }, + { + "epoch": 0.1229, + "grad_norm": 0.8424289226531982, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 12290 + }, + { + "epoch": 0.12291, + "grad_norm": 1.0219829082489014, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 12291 + }, + { + "epoch": 0.12292, + "grad_norm": 1.1891382932662964, + "learning_rate": 0.003, + "loss": 4.019, + "step": 12292 + }, + { + "epoch": 0.12293, + "grad_norm": 0.7496522068977356, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 12293 + }, + { + "epoch": 0.12294, + "grad_norm": 0.6887698173522949, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 12294 + }, + { + "epoch": 0.12295, + "grad_norm": 0.6139264702796936, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 12295 + }, + { + "epoch": 0.12296, + "grad_norm": 0.6319611072540283, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 12296 + }, + { + "epoch": 0.12297, + "grad_norm": 0.8084327578544617, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 12297 + }, + { + "epoch": 0.12298, + "grad_norm": 0.946826159954071, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 12298 + }, + { + "epoch": 0.12299, + "grad_norm": 1.009784460067749, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 12299 + }, + { + "epoch": 0.123, + "grad_norm": 1.0070241689682007, + "learning_rate": 0.003, + "loss": 4.035, + "step": 12300 + }, + { + "epoch": 0.12301, + "grad_norm": 1.0530861616134644, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 12301 + }, + { + "epoch": 0.12302, + "grad_norm": 0.8976438045501709, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 12302 + }, + { + "epoch": 0.12303, + "grad_norm": 0.8210477828979492, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 12303 + }, + { + "epoch": 0.12304, + "grad_norm": 0.8757627606391907, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 12304 + }, + { + "epoch": 0.12305, + "grad_norm": 0.8765901327133179, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 12305 + }, + { + "epoch": 0.12306, + "grad_norm": 0.8396662473678589, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 12306 + }, + { + "epoch": 0.12307, + "grad_norm": 0.6701096892356873, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 12307 + }, + { + "epoch": 0.12308, + "grad_norm": 0.6225115060806274, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 12308 + }, + { + "epoch": 0.12309, + "grad_norm": 0.5880732536315918, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 12309 + }, + { + "epoch": 0.1231, + "grad_norm": 0.5994876027107239, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 12310 + }, + { + "epoch": 0.12311, + "grad_norm": 0.5422573685646057, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 12311 + }, + { + "epoch": 0.12312, + "grad_norm": 0.5807183384895325, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 12312 + }, + { + "epoch": 0.12313, + "grad_norm": 0.5875729322433472, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 12313 + }, + { + "epoch": 0.12314, + "grad_norm": 0.590406596660614, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 12314 + }, + { + "epoch": 0.12315, + "grad_norm": 0.663951575756073, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 12315 + }, + { + "epoch": 0.12316, + "grad_norm": 0.7856951951980591, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 12316 + }, + { + "epoch": 0.12317, + "grad_norm": 0.9424818754196167, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 12317 + }, + { + "epoch": 0.12318, + "grad_norm": 1.0489387512207031, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 12318 + }, + { + "epoch": 0.12319, + "grad_norm": 0.9710455536842346, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 12319 + }, + { + "epoch": 0.1232, + "grad_norm": 0.9531890153884888, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 12320 + }, + { + "epoch": 0.12321, + "grad_norm": 0.8863466382026672, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 12321 + }, + { + "epoch": 0.12322, + "grad_norm": 0.8031570911407471, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 12322 + }, + { + "epoch": 0.12323, + "grad_norm": 0.8672590255737305, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 12323 + }, + { + "epoch": 0.12324, + "grad_norm": 0.8071771264076233, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 12324 + }, + { + "epoch": 0.12325, + "grad_norm": 0.6649886965751648, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 12325 + }, + { + "epoch": 0.12326, + "grad_norm": 0.6209746599197388, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 12326 + }, + { + "epoch": 0.12327, + "grad_norm": 0.7036540508270264, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 12327 + }, + { + "epoch": 0.12328, + "grad_norm": 1.0158931016921997, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 12328 + }, + { + "epoch": 0.12329, + "grad_norm": 1.3318747282028198, + "learning_rate": 0.003, + "loss": 4.043, + "step": 12329 + }, + { + "epoch": 0.1233, + "grad_norm": 0.6423041224479675, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 12330 + }, + { + "epoch": 0.12331, + "grad_norm": 0.6106191873550415, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 12331 + }, + { + "epoch": 0.12332, + "grad_norm": 0.7630592584609985, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 12332 + }, + { + "epoch": 0.12333, + "grad_norm": 0.7819821834564209, + "learning_rate": 0.003, + "loss": 4.004, + "step": 12333 + }, + { + "epoch": 0.12334, + "grad_norm": 0.7031112313270569, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 12334 + }, + { + "epoch": 0.12335, + "grad_norm": 0.7386513948440552, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 12335 + }, + { + "epoch": 0.12336, + "grad_norm": 0.7526112794876099, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 12336 + }, + { + "epoch": 0.12337, + "grad_norm": 0.7977548837661743, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 12337 + }, + { + "epoch": 0.12338, + "grad_norm": 0.8067567944526672, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 12338 + }, + { + "epoch": 0.12339, + "grad_norm": 0.8627603650093079, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 12339 + }, + { + "epoch": 0.1234, + "grad_norm": 0.9207614660263062, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 12340 + }, + { + "epoch": 0.12341, + "grad_norm": 0.74089115858078, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 12341 + }, + { + "epoch": 0.12342, + "grad_norm": 0.6935558319091797, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 12342 + }, + { + "epoch": 0.12343, + "grad_norm": 0.7786235809326172, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 12343 + }, + { + "epoch": 0.12344, + "grad_norm": 0.9137173891067505, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 12344 + }, + { + "epoch": 0.12345, + "grad_norm": 0.8964434266090393, + "learning_rate": 0.003, + "loss": 3.992, + "step": 12345 + }, + { + "epoch": 0.12346, + "grad_norm": 0.9500121474266052, + "learning_rate": 0.003, + "loss": 4.011, + "step": 12346 + }, + { + "epoch": 0.12347, + "grad_norm": 0.8546897172927856, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 12347 + }, + { + "epoch": 0.12348, + "grad_norm": 0.7947580218315125, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 12348 + }, + { + "epoch": 0.12349, + "grad_norm": 0.8268088698387146, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 12349 + }, + { + "epoch": 0.1235, + "grad_norm": 0.7504370808601379, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 12350 + }, + { + "epoch": 0.12351, + "grad_norm": 0.7369322180747986, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 12351 + }, + { + "epoch": 0.12352, + "grad_norm": 0.7135762572288513, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 12352 + }, + { + "epoch": 0.12353, + "grad_norm": 0.7480922341346741, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 12353 + }, + { + "epoch": 0.12354, + "grad_norm": 0.7950556874275208, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 12354 + }, + { + "epoch": 0.12355, + "grad_norm": 0.8724504709243774, + "learning_rate": 0.003, + "loss": 4.016, + "step": 12355 + }, + { + "epoch": 0.12356, + "grad_norm": 1.0912940502166748, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 12356 + }, + { + "epoch": 0.12357, + "grad_norm": 0.9967378973960876, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 12357 + }, + { + "epoch": 0.12358, + "grad_norm": 0.8651845455169678, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 12358 + }, + { + "epoch": 0.12359, + "grad_norm": 0.7292317152023315, + "learning_rate": 0.003, + "loss": 3.996, + "step": 12359 + }, + { + "epoch": 0.1236, + "grad_norm": 0.7191388010978699, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 12360 + }, + { + "epoch": 0.12361, + "grad_norm": 0.645366370677948, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 12361 + }, + { + "epoch": 0.12362, + "grad_norm": 0.6226277947425842, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 12362 + }, + { + "epoch": 0.12363, + "grad_norm": 0.5864716172218323, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 12363 + }, + { + "epoch": 0.12364, + "grad_norm": 0.5753576159477234, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 12364 + }, + { + "epoch": 0.12365, + "grad_norm": 0.6819441914558411, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 12365 + }, + { + "epoch": 0.12366, + "grad_norm": 0.7048933506011963, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 12366 + }, + { + "epoch": 0.12367, + "grad_norm": 0.7060320377349854, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 12367 + }, + { + "epoch": 0.12368, + "grad_norm": 0.6548412442207336, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 12368 + }, + { + "epoch": 0.12369, + "grad_norm": 0.6585349440574646, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 12369 + }, + { + "epoch": 0.1237, + "grad_norm": 0.751133143901825, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 12370 + }, + { + "epoch": 0.12371, + "grad_norm": 0.8336089849472046, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 12371 + }, + { + "epoch": 0.12372, + "grad_norm": 1.0661908388137817, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 12372 + }, + { + "epoch": 0.12373, + "grad_norm": 1.1019768714904785, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 12373 + }, + { + "epoch": 0.12374, + "grad_norm": 0.7196155190467834, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 12374 + }, + { + "epoch": 0.12375, + "grad_norm": 0.737911581993103, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 12375 + }, + { + "epoch": 0.12376, + "grad_norm": 0.854030191898346, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 12376 + }, + { + "epoch": 0.12377, + "grad_norm": 0.8438823819160461, + "learning_rate": 0.003, + "loss": 3.986, + "step": 12377 + }, + { + "epoch": 0.12378, + "grad_norm": 0.9753403067588806, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 12378 + }, + { + "epoch": 0.12379, + "grad_norm": 1.1652812957763672, + "learning_rate": 0.003, + "loss": 4.022, + "step": 12379 + }, + { + "epoch": 0.1238, + "grad_norm": 0.9180656671524048, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 12380 + }, + { + "epoch": 0.12381, + "grad_norm": 0.8215830326080322, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 12381 + }, + { + "epoch": 0.12382, + "grad_norm": 0.8445789813995361, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 12382 + }, + { + "epoch": 0.12383, + "grad_norm": 0.8105951547622681, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 12383 + }, + { + "epoch": 0.12384, + "grad_norm": 0.8429214954376221, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 12384 + }, + { + "epoch": 0.12385, + "grad_norm": 1.0361195802688599, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 12385 + }, + { + "epoch": 0.12386, + "grad_norm": 1.1223865747451782, + "learning_rate": 0.003, + "loss": 4.034, + "step": 12386 + }, + { + "epoch": 0.12387, + "grad_norm": 0.7196579575538635, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 12387 + }, + { + "epoch": 0.12388, + "grad_norm": 0.6705121994018555, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 12388 + }, + { + "epoch": 0.12389, + "grad_norm": 0.9487857818603516, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 12389 + }, + { + "epoch": 0.1239, + "grad_norm": 1.1736794710159302, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 12390 + }, + { + "epoch": 0.12391, + "grad_norm": 0.8355658054351807, + "learning_rate": 0.003, + "loss": 3.986, + "step": 12391 + }, + { + "epoch": 0.12392, + "grad_norm": 0.7802602052688599, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 12392 + }, + { + "epoch": 0.12393, + "grad_norm": 0.8559476137161255, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 12393 + }, + { + "epoch": 0.12394, + "grad_norm": 0.8642109632492065, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 12394 + }, + { + "epoch": 0.12395, + "grad_norm": 0.8027004599571228, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 12395 + }, + { + "epoch": 0.12396, + "grad_norm": 0.6828938126564026, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 12396 + }, + { + "epoch": 0.12397, + "grad_norm": 0.7188079357147217, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 12397 + }, + { + "epoch": 0.12398, + "grad_norm": 0.745271623134613, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 12398 + }, + { + "epoch": 0.12399, + "grad_norm": 0.8426857590675354, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 12399 + }, + { + "epoch": 0.124, + "grad_norm": 0.8381713032722473, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 12400 + }, + { + "epoch": 0.12401, + "grad_norm": 0.712110698223114, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 12401 + }, + { + "epoch": 0.12402, + "grad_norm": 0.6172171831130981, + "learning_rate": 0.003, + "loss": 4.032, + "step": 12402 + }, + { + "epoch": 0.12403, + "grad_norm": 0.633280336856842, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 12403 + }, + { + "epoch": 0.12404, + "grad_norm": 0.5696246027946472, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 12404 + }, + { + "epoch": 0.12405, + "grad_norm": 0.6653860211372375, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 12405 + }, + { + "epoch": 0.12406, + "grad_norm": 0.6514864563941956, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 12406 + }, + { + "epoch": 0.12407, + "grad_norm": 0.7046161890029907, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 12407 + }, + { + "epoch": 0.12408, + "grad_norm": 0.8537508845329285, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 12408 + }, + { + "epoch": 0.12409, + "grad_norm": 0.8411065340042114, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 12409 + }, + { + "epoch": 0.1241, + "grad_norm": 1.0898795127868652, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 12410 + }, + { + "epoch": 0.12411, + "grad_norm": 1.1494314670562744, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 12411 + }, + { + "epoch": 0.12412, + "grad_norm": 0.8649919629096985, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 12412 + }, + { + "epoch": 0.12413, + "grad_norm": 0.6732476949691772, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 12413 + }, + { + "epoch": 0.12414, + "grad_norm": 0.6648367643356323, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 12414 + }, + { + "epoch": 0.12415, + "grad_norm": 0.690162181854248, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 12415 + }, + { + "epoch": 0.12416, + "grad_norm": 0.7361937165260315, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 12416 + }, + { + "epoch": 0.12417, + "grad_norm": 0.7241788506507874, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 12417 + }, + { + "epoch": 0.12418, + "grad_norm": 0.8302240371704102, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 12418 + }, + { + "epoch": 0.12419, + "grad_norm": 0.7926172018051147, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 12419 + }, + { + "epoch": 0.1242, + "grad_norm": 0.6780438423156738, + "learning_rate": 0.003, + "loss": 3.984, + "step": 12420 + }, + { + "epoch": 0.12421, + "grad_norm": 0.6866170167922974, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 12421 + }, + { + "epoch": 0.12422, + "grad_norm": 0.6301003098487854, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 12422 + }, + { + "epoch": 0.12423, + "grad_norm": 0.5220256447792053, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 12423 + }, + { + "epoch": 0.12424, + "grad_norm": 0.5333865284919739, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 12424 + }, + { + "epoch": 0.12425, + "grad_norm": 0.5312225818634033, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 12425 + }, + { + "epoch": 0.12426, + "grad_norm": 0.7749300599098206, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 12426 + }, + { + "epoch": 0.12427, + "grad_norm": 1.0293408632278442, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 12427 + }, + { + "epoch": 0.12428, + "grad_norm": 1.2746484279632568, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 12428 + }, + { + "epoch": 0.12429, + "grad_norm": 0.5662277936935425, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 12429 + }, + { + "epoch": 0.1243, + "grad_norm": 0.6462495923042297, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 12430 + }, + { + "epoch": 0.12431, + "grad_norm": 0.7909497618675232, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 12431 + }, + { + "epoch": 0.12432, + "grad_norm": 0.828307032585144, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 12432 + }, + { + "epoch": 0.12433, + "grad_norm": 0.7236911058425903, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 12433 + }, + { + "epoch": 0.12434, + "grad_norm": 0.6816155314445496, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 12434 + }, + { + "epoch": 0.12435, + "grad_norm": 0.6432563662528992, + "learning_rate": 0.003, + "loss": 4.011, + "step": 12435 + }, + { + "epoch": 0.12436, + "grad_norm": 0.6814159154891968, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 12436 + }, + { + "epoch": 0.12437, + "grad_norm": 0.6854045987129211, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 12437 + }, + { + "epoch": 0.12438, + "grad_norm": 0.7913328409194946, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 12438 + }, + { + "epoch": 0.12439, + "grad_norm": 1.0051794052124023, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 12439 + }, + { + "epoch": 0.1244, + "grad_norm": 1.0895280838012695, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 12440 + }, + { + "epoch": 0.12441, + "grad_norm": 1.0590616464614868, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 12441 + }, + { + "epoch": 0.12442, + "grad_norm": 1.073089361190796, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 12442 + }, + { + "epoch": 0.12443, + "grad_norm": 1.0370190143585205, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 12443 + }, + { + "epoch": 0.12444, + "grad_norm": 0.8426869511604309, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 12444 + }, + { + "epoch": 0.12445, + "grad_norm": 0.8574836850166321, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 12445 + }, + { + "epoch": 0.12446, + "grad_norm": 0.8478577136993408, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 12446 + }, + { + "epoch": 0.12447, + "grad_norm": 1.0601985454559326, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 12447 + }, + { + "epoch": 0.12448, + "grad_norm": 1.0860538482666016, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 12448 + }, + { + "epoch": 0.12449, + "grad_norm": 1.008545994758606, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 12449 + }, + { + "epoch": 0.1245, + "grad_norm": 1.0802119970321655, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 12450 + }, + { + "epoch": 0.12451, + "grad_norm": 0.8374972343444824, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 12451 + }, + { + "epoch": 0.12452, + "grad_norm": 0.9035298228263855, + "learning_rate": 0.003, + "loss": 4.048, + "step": 12452 + }, + { + "epoch": 0.12453, + "grad_norm": 0.7676633596420288, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 12453 + }, + { + "epoch": 0.12454, + "grad_norm": 0.6823439598083496, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 12454 + }, + { + "epoch": 0.12455, + "grad_norm": 0.6423712968826294, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 12455 + }, + { + "epoch": 0.12456, + "grad_norm": 0.6407604217529297, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 12456 + }, + { + "epoch": 0.12457, + "grad_norm": 0.7209190726280212, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 12457 + }, + { + "epoch": 0.12458, + "grad_norm": 0.7898573875427246, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 12458 + }, + { + "epoch": 0.12459, + "grad_norm": 0.7464539408683777, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 12459 + }, + { + "epoch": 0.1246, + "grad_norm": 0.7210266590118408, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 12460 + }, + { + "epoch": 0.12461, + "grad_norm": 0.6472708582878113, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 12461 + }, + { + "epoch": 0.12462, + "grad_norm": 0.6778544187545776, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 12462 + }, + { + "epoch": 0.12463, + "grad_norm": 0.7389536499977112, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 12463 + }, + { + "epoch": 0.12464, + "grad_norm": 0.9307025074958801, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 12464 + }, + { + "epoch": 0.12465, + "grad_norm": 1.2131528854370117, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 12465 + }, + { + "epoch": 0.12466, + "grad_norm": 0.8831672072410583, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 12466 + }, + { + "epoch": 0.12467, + "grad_norm": 0.7936322093009949, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 12467 + }, + { + "epoch": 0.12468, + "grad_norm": 0.7008469104766846, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 12468 + }, + { + "epoch": 0.12469, + "grad_norm": 0.6694667339324951, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 12469 + }, + { + "epoch": 0.1247, + "grad_norm": 0.581159770488739, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 12470 + }, + { + "epoch": 0.12471, + "grad_norm": 0.5600728392601013, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 12471 + }, + { + "epoch": 0.12472, + "grad_norm": 0.5877519845962524, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 12472 + }, + { + "epoch": 0.12473, + "grad_norm": 0.595278263092041, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 12473 + }, + { + "epoch": 0.12474, + "grad_norm": 0.5744599103927612, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 12474 + }, + { + "epoch": 0.12475, + "grad_norm": 0.6423681378364563, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 12475 + }, + { + "epoch": 0.12476, + "grad_norm": 0.8026760220527649, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 12476 + }, + { + "epoch": 0.12477, + "grad_norm": 1.138680338859558, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 12477 + }, + { + "epoch": 0.12478, + "grad_norm": 1.0252193212509155, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 12478 + }, + { + "epoch": 0.12479, + "grad_norm": 0.7640153765678406, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 12479 + }, + { + "epoch": 0.1248, + "grad_norm": 0.6829314231872559, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 12480 + }, + { + "epoch": 0.12481, + "grad_norm": 0.7560384273529053, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 12481 + }, + { + "epoch": 0.12482, + "grad_norm": 0.9146919846534729, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 12482 + }, + { + "epoch": 0.12483, + "grad_norm": 1.131492257118225, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 12483 + }, + { + "epoch": 0.12484, + "grad_norm": 0.9777736067771912, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 12484 + }, + { + "epoch": 0.12485, + "grad_norm": 0.8896247744560242, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 12485 + }, + { + "epoch": 0.12486, + "grad_norm": 0.7884039282798767, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 12486 + }, + { + "epoch": 0.12487, + "grad_norm": 0.74324631690979, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 12487 + }, + { + "epoch": 0.12488, + "grad_norm": 0.7056796550750732, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 12488 + }, + { + "epoch": 0.12489, + "grad_norm": 0.6776674389839172, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 12489 + }, + { + "epoch": 0.1249, + "grad_norm": 0.7072422504425049, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 12490 + }, + { + "epoch": 0.12491, + "grad_norm": 0.7262662649154663, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 12491 + }, + { + "epoch": 0.12492, + "grad_norm": 0.8424429297447205, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 12492 + }, + { + "epoch": 0.12493, + "grad_norm": 0.9678304195404053, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 12493 + }, + { + "epoch": 0.12494, + "grad_norm": 0.8881115317344666, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 12494 + }, + { + "epoch": 0.12495, + "grad_norm": 0.8806186318397522, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 12495 + }, + { + "epoch": 0.12496, + "grad_norm": 0.8792669773101807, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 12496 + }, + { + "epoch": 0.12497, + "grad_norm": 0.9512883424758911, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 12497 + }, + { + "epoch": 0.12498, + "grad_norm": 0.9773924946784973, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 12498 + }, + { + "epoch": 0.12499, + "grad_norm": 0.9709060192108154, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 12499 + }, + { + "epoch": 0.125, + "grad_norm": 1.2080605030059814, + "learning_rate": 0.003, + "loss": 4.079, + "step": 12500 + }, + { + "epoch": 0.12501, + "grad_norm": 0.8987671136856079, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 12501 + }, + { + "epoch": 0.12502, + "grad_norm": 0.8190386295318604, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 12502 + }, + { + "epoch": 0.12503, + "grad_norm": 0.9141505360603333, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 12503 + }, + { + "epoch": 0.12504, + "grad_norm": 0.8598799705505371, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 12504 + }, + { + "epoch": 0.12505, + "grad_norm": 0.9165339469909668, + "learning_rate": 0.003, + "loss": 4.054, + "step": 12505 + }, + { + "epoch": 0.12506, + "grad_norm": 1.0256816148757935, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 12506 + }, + { + "epoch": 0.12507, + "grad_norm": 1.0075628757476807, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 12507 + }, + { + "epoch": 0.12508, + "grad_norm": 0.8303059935569763, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 12508 + }, + { + "epoch": 0.12509, + "grad_norm": 0.9752023220062256, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 12509 + }, + { + "epoch": 0.1251, + "grad_norm": 0.9722163677215576, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 12510 + }, + { + "epoch": 0.12511, + "grad_norm": 1.1083474159240723, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 12511 + }, + { + "epoch": 0.12512, + "grad_norm": 1.0361601114273071, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 12512 + }, + { + "epoch": 0.12513, + "grad_norm": 1.0454269647598267, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 12513 + }, + { + "epoch": 0.12514, + "grad_norm": 0.9313620924949646, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 12514 + }, + { + "epoch": 0.12515, + "grad_norm": 0.9025012850761414, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 12515 + }, + { + "epoch": 0.12516, + "grad_norm": 0.8324545621871948, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 12516 + }, + { + "epoch": 0.12517, + "grad_norm": 0.729629397392273, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 12517 + }, + { + "epoch": 0.12518, + "grad_norm": 0.7196733355522156, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 12518 + }, + { + "epoch": 0.12519, + "grad_norm": 0.7338614463806152, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 12519 + }, + { + "epoch": 0.1252, + "grad_norm": 0.7173320055007935, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 12520 + }, + { + "epoch": 0.12521, + "grad_norm": 0.660810649394989, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 12521 + }, + { + "epoch": 0.12522, + "grad_norm": 0.6969593167304993, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 12522 + }, + { + "epoch": 0.12523, + "grad_norm": 0.603926956653595, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 12523 + }, + { + "epoch": 0.12524, + "grad_norm": 0.5833172798156738, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 12524 + }, + { + "epoch": 0.12525, + "grad_norm": 0.7020481824874878, + "learning_rate": 0.003, + "loss": 3.991, + "step": 12525 + }, + { + "epoch": 0.12526, + "grad_norm": 0.8667402863502502, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 12526 + }, + { + "epoch": 0.12527, + "grad_norm": 0.9506835341453552, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 12527 + }, + { + "epoch": 0.12528, + "grad_norm": 0.9923872351646423, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 12528 + }, + { + "epoch": 0.12529, + "grad_norm": 0.7969499230384827, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 12529 + }, + { + "epoch": 0.1253, + "grad_norm": 0.6928942799568176, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 12530 + }, + { + "epoch": 0.12531, + "grad_norm": 0.6717604994773865, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 12531 + }, + { + "epoch": 0.12532, + "grad_norm": 0.6549240350723267, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 12532 + }, + { + "epoch": 0.12533, + "grad_norm": 0.6569952964782715, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 12533 + }, + { + "epoch": 0.12534, + "grad_norm": 0.6018800735473633, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 12534 + }, + { + "epoch": 0.12535, + "grad_norm": 0.7109904289245605, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 12535 + }, + { + "epoch": 0.12536, + "grad_norm": 0.8113789558410645, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 12536 + }, + { + "epoch": 0.12537, + "grad_norm": 0.9771004319190979, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 12537 + }, + { + "epoch": 0.12538, + "grad_norm": 1.0944777727127075, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 12538 + }, + { + "epoch": 0.12539, + "grad_norm": 0.8347179293632507, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 12539 + }, + { + "epoch": 0.1254, + "grad_norm": 0.652307391166687, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 12540 + }, + { + "epoch": 0.12541, + "grad_norm": 0.5324810743331909, + "learning_rate": 0.003, + "loss": 3.99, + "step": 12541 + }, + { + "epoch": 0.12542, + "grad_norm": 0.5968160033226013, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 12542 + }, + { + "epoch": 0.12543, + "grad_norm": 0.6760523915290833, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 12543 + }, + { + "epoch": 0.12544, + "grad_norm": 0.7631462812423706, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 12544 + }, + { + "epoch": 0.12545, + "grad_norm": 0.829358696937561, + "learning_rate": 0.003, + "loss": 4.013, + "step": 12545 + }, + { + "epoch": 0.12546, + "grad_norm": 0.8343631029129028, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 12546 + }, + { + "epoch": 0.12547, + "grad_norm": 0.7265992760658264, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 12547 + }, + { + "epoch": 0.12548, + "grad_norm": 0.6316820383071899, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 12548 + }, + { + "epoch": 0.12549, + "grad_norm": 0.7243136167526245, + "learning_rate": 0.003, + "loss": 4.022, + "step": 12549 + }, + { + "epoch": 0.1255, + "grad_norm": 0.6993957757949829, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 12550 + }, + { + "epoch": 0.12551, + "grad_norm": 0.7077828645706177, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 12551 + }, + { + "epoch": 0.12552, + "grad_norm": 0.8004039525985718, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 12552 + }, + { + "epoch": 0.12553, + "grad_norm": 0.7996693849563599, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 12553 + }, + { + "epoch": 0.12554, + "grad_norm": 0.7634584903717041, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 12554 + }, + { + "epoch": 0.12555, + "grad_norm": 0.7199156880378723, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 12555 + }, + { + "epoch": 0.12556, + "grad_norm": 0.7700698971748352, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 12556 + }, + { + "epoch": 0.12557, + "grad_norm": 0.9102762341499329, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 12557 + }, + { + "epoch": 0.12558, + "grad_norm": 0.9445754885673523, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 12558 + }, + { + "epoch": 0.12559, + "grad_norm": 0.8067466020584106, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 12559 + }, + { + "epoch": 0.1256, + "grad_norm": 0.9157754778862, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 12560 + }, + { + "epoch": 0.12561, + "grad_norm": 1.014098882675171, + "learning_rate": 0.003, + "loss": 4.006, + "step": 12561 + }, + { + "epoch": 0.12562, + "grad_norm": 0.9550620913505554, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 12562 + }, + { + "epoch": 0.12563, + "grad_norm": 0.8556691408157349, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 12563 + }, + { + "epoch": 0.12564, + "grad_norm": 0.791690468788147, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 12564 + }, + { + "epoch": 0.12565, + "grad_norm": 0.7778317928314209, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 12565 + }, + { + "epoch": 0.12566, + "grad_norm": 0.7928282618522644, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 12566 + }, + { + "epoch": 0.12567, + "grad_norm": 0.8807454705238342, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 12567 + }, + { + "epoch": 0.12568, + "grad_norm": 1.1370134353637695, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 12568 + }, + { + "epoch": 0.12569, + "grad_norm": 1.082935094833374, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 12569 + }, + { + "epoch": 0.1257, + "grad_norm": 0.9118308424949646, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 12570 + }, + { + "epoch": 0.12571, + "grad_norm": 0.8950235843658447, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 12571 + }, + { + "epoch": 0.12572, + "grad_norm": 0.8408292531967163, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 12572 + }, + { + "epoch": 0.12573, + "grad_norm": 0.7380292415618896, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 12573 + }, + { + "epoch": 0.12574, + "grad_norm": 0.6289229989051819, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 12574 + }, + { + "epoch": 0.12575, + "grad_norm": 0.8124949932098389, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 12575 + }, + { + "epoch": 0.12576, + "grad_norm": 1.0226274728775024, + "learning_rate": 0.003, + "loss": 4.032, + "step": 12576 + }, + { + "epoch": 0.12577, + "grad_norm": 1.118428349494934, + "learning_rate": 0.003, + "loss": 4.017, + "step": 12577 + }, + { + "epoch": 0.12578, + "grad_norm": 0.7549068927764893, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 12578 + }, + { + "epoch": 0.12579, + "grad_norm": 0.6412855982780457, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 12579 + }, + { + "epoch": 0.1258, + "grad_norm": 0.6058320999145508, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 12580 + }, + { + "epoch": 0.12581, + "grad_norm": 0.7913022041320801, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 12581 + }, + { + "epoch": 0.12582, + "grad_norm": 0.9635642766952515, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 12582 + }, + { + "epoch": 0.12583, + "grad_norm": 0.9899559617042542, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 12583 + }, + { + "epoch": 0.12584, + "grad_norm": 0.816019594669342, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 12584 + }, + { + "epoch": 0.12585, + "grad_norm": 0.7499985098838806, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 12585 + }, + { + "epoch": 0.12586, + "grad_norm": 0.8531391024589539, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 12586 + }, + { + "epoch": 0.12587, + "grad_norm": 1.0371969938278198, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 12587 + }, + { + "epoch": 0.12588, + "grad_norm": 0.95607590675354, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 12588 + }, + { + "epoch": 0.12589, + "grad_norm": 0.9678847789764404, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 12589 + }, + { + "epoch": 0.1259, + "grad_norm": 0.974133312702179, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 12590 + }, + { + "epoch": 0.12591, + "grad_norm": 0.8166702389717102, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 12591 + }, + { + "epoch": 0.12592, + "grad_norm": 0.6931419372558594, + "learning_rate": 0.003, + "loss": 4.025, + "step": 12592 + }, + { + "epoch": 0.12593, + "grad_norm": 0.7652658224105835, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 12593 + }, + { + "epoch": 0.12594, + "grad_norm": 0.9171327352523804, + "learning_rate": 0.003, + "loss": 4.037, + "step": 12594 + }, + { + "epoch": 0.12595, + "grad_norm": 1.0293341875076294, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 12595 + }, + { + "epoch": 0.12596, + "grad_norm": 0.9116888046264648, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 12596 + }, + { + "epoch": 0.12597, + "grad_norm": 0.8362652659416199, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 12597 + }, + { + "epoch": 0.12598, + "grad_norm": 0.7781234383583069, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 12598 + }, + { + "epoch": 0.12599, + "grad_norm": 0.8064082860946655, + "learning_rate": 0.003, + "loss": 4.05, + "step": 12599 + }, + { + "epoch": 0.126, + "grad_norm": 0.8930453062057495, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 12600 + }, + { + "epoch": 0.12601, + "grad_norm": 0.8855689764022827, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 12601 + }, + { + "epoch": 0.12602, + "grad_norm": 0.7009050250053406, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 12602 + }, + { + "epoch": 0.12603, + "grad_norm": 0.6489673256874084, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 12603 + }, + { + "epoch": 0.12604, + "grad_norm": 0.5806393027305603, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 12604 + }, + { + "epoch": 0.12605, + "grad_norm": 0.6146126389503479, + "learning_rate": 0.003, + "loss": 4.015, + "step": 12605 + }, + { + "epoch": 0.12606, + "grad_norm": 0.6746604442596436, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 12606 + }, + { + "epoch": 0.12607, + "grad_norm": 0.8226182460784912, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 12607 + }, + { + "epoch": 0.12608, + "grad_norm": 0.9086608290672302, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 12608 + }, + { + "epoch": 0.12609, + "grad_norm": 1.0627309083938599, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 12609 + }, + { + "epoch": 0.1261, + "grad_norm": 1.0476856231689453, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 12610 + }, + { + "epoch": 0.12611, + "grad_norm": 0.8482815623283386, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 12611 + }, + { + "epoch": 0.12612, + "grad_norm": 0.7633053660392761, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 12612 + }, + { + "epoch": 0.12613, + "grad_norm": 0.6689841151237488, + "learning_rate": 0.003, + "loss": 3.987, + "step": 12613 + }, + { + "epoch": 0.12614, + "grad_norm": 0.6084244251251221, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 12614 + }, + { + "epoch": 0.12615, + "grad_norm": 0.6043604612350464, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 12615 + }, + { + "epoch": 0.12616, + "grad_norm": 0.6392335891723633, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 12616 + }, + { + "epoch": 0.12617, + "grad_norm": 0.707719087600708, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 12617 + }, + { + "epoch": 0.12618, + "grad_norm": 0.7622061371803284, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 12618 + }, + { + "epoch": 0.12619, + "grad_norm": 0.9297196865081787, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 12619 + }, + { + "epoch": 0.1262, + "grad_norm": 0.9986196160316467, + "learning_rate": 0.003, + "loss": 4.03, + "step": 12620 + }, + { + "epoch": 0.12621, + "grad_norm": 0.8918259143829346, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 12621 + }, + { + "epoch": 0.12622, + "grad_norm": 0.6821439862251282, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 12622 + }, + { + "epoch": 0.12623, + "grad_norm": 0.7973731160163879, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 12623 + }, + { + "epoch": 0.12624, + "grad_norm": 0.8734741806983948, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 12624 + }, + { + "epoch": 0.12625, + "grad_norm": 0.8574873805046082, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 12625 + }, + { + "epoch": 0.12626, + "grad_norm": 0.9886305928230286, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 12626 + }, + { + "epoch": 0.12627, + "grad_norm": 0.9929408431053162, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 12627 + }, + { + "epoch": 0.12628, + "grad_norm": 0.7818512916564941, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 12628 + }, + { + "epoch": 0.12629, + "grad_norm": 0.6975765824317932, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 12629 + }, + { + "epoch": 0.1263, + "grad_norm": 0.6380565166473389, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 12630 + }, + { + "epoch": 0.12631, + "grad_norm": 0.728340208530426, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 12631 + }, + { + "epoch": 0.12632, + "grad_norm": 0.8124565482139587, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 12632 + }, + { + "epoch": 0.12633, + "grad_norm": 0.8869121670722961, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 12633 + }, + { + "epoch": 0.12634, + "grad_norm": 0.8852589130401611, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 12634 + }, + { + "epoch": 0.12635, + "grad_norm": 0.6980815529823303, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 12635 + }, + { + "epoch": 0.12636, + "grad_norm": 0.5582261681556702, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 12636 + }, + { + "epoch": 0.12637, + "grad_norm": 0.6569713950157166, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 12637 + }, + { + "epoch": 0.12638, + "grad_norm": 0.6073877811431885, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 12638 + }, + { + "epoch": 0.12639, + "grad_norm": 0.6059941649436951, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 12639 + }, + { + "epoch": 0.1264, + "grad_norm": 0.6994345188140869, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 12640 + }, + { + "epoch": 0.12641, + "grad_norm": 0.6538774371147156, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 12641 + }, + { + "epoch": 0.12642, + "grad_norm": 0.5538105368614197, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 12642 + }, + { + "epoch": 0.12643, + "grad_norm": 0.5192658305168152, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 12643 + }, + { + "epoch": 0.12644, + "grad_norm": 0.44997328519821167, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 12644 + }, + { + "epoch": 0.12645, + "grad_norm": 0.5587505102157593, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 12645 + }, + { + "epoch": 0.12646, + "grad_norm": 0.7219934463500977, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 12646 + }, + { + "epoch": 0.12647, + "grad_norm": 0.9880595803260803, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 12647 + }, + { + "epoch": 0.12648, + "grad_norm": 1.2066372632980347, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 12648 + }, + { + "epoch": 0.12649, + "grad_norm": 0.6333531737327576, + "learning_rate": 0.003, + "loss": 3.984, + "step": 12649 + }, + { + "epoch": 0.1265, + "grad_norm": 0.6311098337173462, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 12650 + }, + { + "epoch": 0.12651, + "grad_norm": 0.6966622471809387, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 12651 + }, + { + "epoch": 0.12652, + "grad_norm": 0.7308709621429443, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 12652 + }, + { + "epoch": 0.12653, + "grad_norm": 0.7971374988555908, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 12653 + }, + { + "epoch": 0.12654, + "grad_norm": 0.8622496128082275, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 12654 + }, + { + "epoch": 0.12655, + "grad_norm": 0.8505637049674988, + "learning_rate": 0.003, + "loss": 4.025, + "step": 12655 + }, + { + "epoch": 0.12656, + "grad_norm": 0.8776390552520752, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 12656 + }, + { + "epoch": 0.12657, + "grad_norm": 1.0906471014022827, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 12657 + }, + { + "epoch": 0.12658, + "grad_norm": 1.083461880683899, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 12658 + }, + { + "epoch": 0.12659, + "grad_norm": 1.0041580200195312, + "learning_rate": 0.003, + "loss": 3.998, + "step": 12659 + }, + { + "epoch": 0.1266, + "grad_norm": 0.8595044612884521, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 12660 + }, + { + "epoch": 0.12661, + "grad_norm": 0.9145925045013428, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 12661 + }, + { + "epoch": 0.12662, + "grad_norm": 0.9530350565910339, + "learning_rate": 0.003, + "loss": 4.019, + "step": 12662 + }, + { + "epoch": 0.12663, + "grad_norm": 0.9560208320617676, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 12663 + }, + { + "epoch": 0.12664, + "grad_norm": 0.794122040271759, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 12664 + }, + { + "epoch": 0.12665, + "grad_norm": 0.7163031697273254, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 12665 + }, + { + "epoch": 0.12666, + "grad_norm": 0.7502034306526184, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 12666 + }, + { + "epoch": 0.12667, + "grad_norm": 0.8145994544029236, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 12667 + }, + { + "epoch": 0.12668, + "grad_norm": 0.9863829016685486, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 12668 + }, + { + "epoch": 0.12669, + "grad_norm": 1.2059465646743774, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 12669 + }, + { + "epoch": 0.1267, + "grad_norm": 0.9510130286216736, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 12670 + }, + { + "epoch": 0.12671, + "grad_norm": 1.0077412128448486, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 12671 + }, + { + "epoch": 0.12672, + "grad_norm": 1.0883513689041138, + "learning_rate": 0.003, + "loss": 4.024, + "step": 12672 + }, + { + "epoch": 0.12673, + "grad_norm": 1.0037307739257812, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 12673 + }, + { + "epoch": 0.12674, + "grad_norm": 0.8705130219459534, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 12674 + }, + { + "epoch": 0.12675, + "grad_norm": 0.7188978791236877, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 12675 + }, + { + "epoch": 0.12676, + "grad_norm": 0.8157229423522949, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 12676 + }, + { + "epoch": 0.12677, + "grad_norm": 1.175352692604065, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 12677 + }, + { + "epoch": 0.12678, + "grad_norm": 0.9241523742675781, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 12678 + }, + { + "epoch": 0.12679, + "grad_norm": 0.935789942741394, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 12679 + }, + { + "epoch": 0.1268, + "grad_norm": 1.0995280742645264, + "learning_rate": 0.003, + "loss": 4.043, + "step": 12680 + }, + { + "epoch": 0.12681, + "grad_norm": 1.0176584720611572, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 12681 + }, + { + "epoch": 0.12682, + "grad_norm": 0.9173998832702637, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 12682 + }, + { + "epoch": 0.12683, + "grad_norm": 0.8360102772712708, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 12683 + }, + { + "epoch": 0.12684, + "grad_norm": 0.8676466941833496, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 12684 + }, + { + "epoch": 0.12685, + "grad_norm": 0.8226211071014404, + "learning_rate": 0.003, + "loss": 4.037, + "step": 12685 + }, + { + "epoch": 0.12686, + "grad_norm": 0.7982207536697388, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 12686 + }, + { + "epoch": 0.12687, + "grad_norm": 0.7003099918365479, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 12687 + }, + { + "epoch": 0.12688, + "grad_norm": 0.6146358251571655, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 12688 + }, + { + "epoch": 0.12689, + "grad_norm": 0.7245614528656006, + "learning_rate": 0.003, + "loss": 4.011, + "step": 12689 + }, + { + "epoch": 0.1269, + "grad_norm": 0.6969562768936157, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 12690 + }, + { + "epoch": 0.12691, + "grad_norm": 0.646485447883606, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 12691 + }, + { + "epoch": 0.12692, + "grad_norm": 0.6295223236083984, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 12692 + }, + { + "epoch": 0.12693, + "grad_norm": 0.6015465259552002, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 12693 + }, + { + "epoch": 0.12694, + "grad_norm": 0.6565688848495483, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 12694 + }, + { + "epoch": 0.12695, + "grad_norm": 0.8570446968078613, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 12695 + }, + { + "epoch": 0.12696, + "grad_norm": 1.0074641704559326, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 12696 + }, + { + "epoch": 0.12697, + "grad_norm": 0.9953659772872925, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 12697 + }, + { + "epoch": 0.12698, + "grad_norm": 0.7940637469291687, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 12698 + }, + { + "epoch": 0.12699, + "grad_norm": 0.679948091506958, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 12699 + }, + { + "epoch": 0.127, + "grad_norm": 0.752021312713623, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 12700 + }, + { + "epoch": 0.12701, + "grad_norm": 0.9325186610221863, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 12701 + }, + { + "epoch": 0.12702, + "grad_norm": 1.2120436429977417, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 12702 + }, + { + "epoch": 0.12703, + "grad_norm": 0.6374262571334839, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 12703 + }, + { + "epoch": 0.12704, + "grad_norm": 0.6450442671775818, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 12704 + }, + { + "epoch": 0.12705, + "grad_norm": 0.6463040113449097, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 12705 + }, + { + "epoch": 0.12706, + "grad_norm": 0.5338553786277771, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 12706 + }, + { + "epoch": 0.12707, + "grad_norm": 0.511511504650116, + "learning_rate": 0.003, + "loss": 4.018, + "step": 12707 + }, + { + "epoch": 0.12708, + "grad_norm": 0.5351196527481079, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 12708 + }, + { + "epoch": 0.12709, + "grad_norm": 0.5771192908287048, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 12709 + }, + { + "epoch": 0.1271, + "grad_norm": 0.5235174298286438, + "learning_rate": 0.003, + "loss": 4.011, + "step": 12710 + }, + { + "epoch": 0.12711, + "grad_norm": 0.5067513585090637, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 12711 + }, + { + "epoch": 0.12712, + "grad_norm": 0.4862622320652008, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 12712 + }, + { + "epoch": 0.12713, + "grad_norm": 0.4507077932357788, + "learning_rate": 0.003, + "loss": 4.003, + "step": 12713 + }, + { + "epoch": 0.12714, + "grad_norm": 0.4960992932319641, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 12714 + }, + { + "epoch": 0.12715, + "grad_norm": 0.4874277710914612, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 12715 + }, + { + "epoch": 0.12716, + "grad_norm": 0.45987147092819214, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 12716 + }, + { + "epoch": 0.12717, + "grad_norm": 0.5444669127464294, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 12717 + }, + { + "epoch": 0.12718, + "grad_norm": 0.6519801616668701, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 12718 + }, + { + "epoch": 0.12719, + "grad_norm": 0.846007764339447, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 12719 + }, + { + "epoch": 0.1272, + "grad_norm": 1.1677876710891724, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 12720 + }, + { + "epoch": 0.12721, + "grad_norm": 0.9765973091125488, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 12721 + }, + { + "epoch": 0.12722, + "grad_norm": 0.9912946820259094, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 12722 + }, + { + "epoch": 0.12723, + "grad_norm": 1.0117685794830322, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 12723 + }, + { + "epoch": 0.12724, + "grad_norm": 1.0825741291046143, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 12724 + }, + { + "epoch": 0.12725, + "grad_norm": 1.1411371231079102, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 12725 + }, + { + "epoch": 0.12726, + "grad_norm": 0.8763951063156128, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 12726 + }, + { + "epoch": 0.12727, + "grad_norm": 1.0362228155136108, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 12727 + }, + { + "epoch": 0.12728, + "grad_norm": 1.0936086177825928, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 12728 + }, + { + "epoch": 0.12729, + "grad_norm": 0.8342863321304321, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 12729 + }, + { + "epoch": 0.1273, + "grad_norm": 0.8730246424674988, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 12730 + }, + { + "epoch": 0.12731, + "grad_norm": 0.8429976105690002, + "learning_rate": 0.003, + "loss": 4.006, + "step": 12731 + }, + { + "epoch": 0.12732, + "grad_norm": 0.9439195394515991, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 12732 + }, + { + "epoch": 0.12733, + "grad_norm": 1.016707420349121, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 12733 + }, + { + "epoch": 0.12734, + "grad_norm": 1.1399983167648315, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 12734 + }, + { + "epoch": 0.12735, + "grad_norm": 1.092917561531067, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 12735 + }, + { + "epoch": 0.12736, + "grad_norm": 1.0256966352462769, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 12736 + }, + { + "epoch": 0.12737, + "grad_norm": 0.9440638422966003, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 12737 + }, + { + "epoch": 0.12738, + "grad_norm": 0.8956122994422913, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 12738 + }, + { + "epoch": 0.12739, + "grad_norm": 0.7672011256217957, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 12739 + }, + { + "epoch": 0.1274, + "grad_norm": 0.8364803791046143, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 12740 + }, + { + "epoch": 0.12741, + "grad_norm": 0.9540492296218872, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 12741 + }, + { + "epoch": 0.12742, + "grad_norm": 1.300502061843872, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 12742 + }, + { + "epoch": 0.12743, + "grad_norm": 0.5849984884262085, + "learning_rate": 0.003, + "loss": 4.011, + "step": 12743 + }, + { + "epoch": 0.12744, + "grad_norm": 0.5924078822135925, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 12744 + }, + { + "epoch": 0.12745, + "grad_norm": 0.6661898493766785, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 12745 + }, + { + "epoch": 0.12746, + "grad_norm": 0.6874447464942932, + "learning_rate": 0.003, + "loss": 4.044, + "step": 12746 + }, + { + "epoch": 0.12747, + "grad_norm": 0.7282478213310242, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 12747 + }, + { + "epoch": 0.12748, + "grad_norm": 0.6811204552650452, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 12748 + }, + { + "epoch": 0.12749, + "grad_norm": 0.5884941816329956, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 12749 + }, + { + "epoch": 0.1275, + "grad_norm": 0.5128979682922363, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 12750 + }, + { + "epoch": 0.12751, + "grad_norm": 0.542918860912323, + "learning_rate": 0.003, + "loss": 4.033, + "step": 12751 + }, + { + "epoch": 0.12752, + "grad_norm": 0.6755454540252686, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 12752 + }, + { + "epoch": 0.12753, + "grad_norm": 0.7383790612220764, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 12753 + }, + { + "epoch": 0.12754, + "grad_norm": 0.752862811088562, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 12754 + }, + { + "epoch": 0.12755, + "grad_norm": 0.7278972864151001, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 12755 + }, + { + "epoch": 0.12756, + "grad_norm": 0.6267310976982117, + "learning_rate": 0.003, + "loss": 4.025, + "step": 12756 + }, + { + "epoch": 0.12757, + "grad_norm": 0.5327348113059998, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 12757 + }, + { + "epoch": 0.12758, + "grad_norm": 0.5505622029304504, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 12758 + }, + { + "epoch": 0.12759, + "grad_norm": 0.6469417214393616, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 12759 + }, + { + "epoch": 0.1276, + "grad_norm": 0.7604227066040039, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 12760 + }, + { + "epoch": 0.12761, + "grad_norm": 0.9358194470405579, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 12761 + }, + { + "epoch": 0.12762, + "grad_norm": 1.2231742143630981, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 12762 + }, + { + "epoch": 0.12763, + "grad_norm": 0.6989595890045166, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 12763 + }, + { + "epoch": 0.12764, + "grad_norm": 0.6533869504928589, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 12764 + }, + { + "epoch": 0.12765, + "grad_norm": 0.6442908644676208, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 12765 + }, + { + "epoch": 0.12766, + "grad_norm": 0.6153333783149719, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 12766 + }, + { + "epoch": 0.12767, + "grad_norm": 0.6329177618026733, + "learning_rate": 0.003, + "loss": 4.008, + "step": 12767 + }, + { + "epoch": 0.12768, + "grad_norm": 0.6864577531814575, + "learning_rate": 0.003, + "loss": 4.031, + "step": 12768 + }, + { + "epoch": 0.12769, + "grad_norm": 0.8032853007316589, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 12769 + }, + { + "epoch": 0.1277, + "grad_norm": 0.8559294939041138, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 12770 + }, + { + "epoch": 0.12771, + "grad_norm": 0.8253934383392334, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 12771 + }, + { + "epoch": 0.12772, + "grad_norm": 0.8039045333862305, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 12772 + }, + { + "epoch": 0.12773, + "grad_norm": 0.7011158466339111, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 12773 + }, + { + "epoch": 0.12774, + "grad_norm": 0.8235172629356384, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 12774 + }, + { + "epoch": 0.12775, + "grad_norm": 0.8840197324752808, + "learning_rate": 0.003, + "loss": 4.002, + "step": 12775 + }, + { + "epoch": 0.12776, + "grad_norm": 0.9510916471481323, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 12776 + }, + { + "epoch": 0.12777, + "grad_norm": 0.9348542094230652, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 12777 + }, + { + "epoch": 0.12778, + "grad_norm": 0.8219630122184753, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 12778 + }, + { + "epoch": 0.12779, + "grad_norm": 0.7910003662109375, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 12779 + }, + { + "epoch": 0.1278, + "grad_norm": 0.8482397794723511, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 12780 + }, + { + "epoch": 0.12781, + "grad_norm": 0.9158820509910583, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 12781 + }, + { + "epoch": 0.12782, + "grad_norm": 1.0353529453277588, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 12782 + }, + { + "epoch": 0.12783, + "grad_norm": 1.087900161743164, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 12783 + }, + { + "epoch": 0.12784, + "grad_norm": 1.1465449333190918, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 12784 + }, + { + "epoch": 0.12785, + "grad_norm": 0.8442012667655945, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 12785 + }, + { + "epoch": 0.12786, + "grad_norm": 0.6666450500488281, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 12786 + }, + { + "epoch": 0.12787, + "grad_norm": 0.6549657583236694, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 12787 + }, + { + "epoch": 0.12788, + "grad_norm": 0.694678008556366, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 12788 + }, + { + "epoch": 0.12789, + "grad_norm": 0.6600160002708435, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 12789 + }, + { + "epoch": 0.1279, + "grad_norm": 0.75733482837677, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 12790 + }, + { + "epoch": 0.12791, + "grad_norm": 0.7874218225479126, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 12791 + }, + { + "epoch": 0.12792, + "grad_norm": 0.7725749015808105, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 12792 + }, + { + "epoch": 0.12793, + "grad_norm": 0.7349872589111328, + "learning_rate": 0.003, + "loss": 4.021, + "step": 12793 + }, + { + "epoch": 0.12794, + "grad_norm": 0.7539064884185791, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 12794 + }, + { + "epoch": 0.12795, + "grad_norm": 0.8775387406349182, + "learning_rate": 0.003, + "loss": 4.018, + "step": 12795 + }, + { + "epoch": 0.12796, + "grad_norm": 0.9745922088623047, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 12796 + }, + { + "epoch": 0.12797, + "grad_norm": 0.9114720821380615, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 12797 + }, + { + "epoch": 0.12798, + "grad_norm": 1.044573187828064, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 12798 + }, + { + "epoch": 0.12799, + "grad_norm": 1.1686325073242188, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 12799 + }, + { + "epoch": 0.128, + "grad_norm": 0.8817925453186035, + "learning_rate": 0.003, + "loss": 4.006, + "step": 12800 + }, + { + "epoch": 0.12801, + "grad_norm": 0.800365149974823, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 12801 + }, + { + "epoch": 0.12802, + "grad_norm": 0.8279086351394653, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 12802 + }, + { + "epoch": 0.12803, + "grad_norm": 0.9438039660453796, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 12803 + }, + { + "epoch": 0.12804, + "grad_norm": 0.9062755703926086, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 12804 + }, + { + "epoch": 0.12805, + "grad_norm": 0.7312350273132324, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 12805 + }, + { + "epoch": 0.12806, + "grad_norm": 0.7411422729492188, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 12806 + }, + { + "epoch": 0.12807, + "grad_norm": 0.7521910667419434, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 12807 + }, + { + "epoch": 0.12808, + "grad_norm": 0.7977126836776733, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 12808 + }, + { + "epoch": 0.12809, + "grad_norm": 0.7871893644332886, + "learning_rate": 0.003, + "loss": 4.01, + "step": 12809 + }, + { + "epoch": 0.1281, + "grad_norm": 0.8420676589012146, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 12810 + }, + { + "epoch": 0.12811, + "grad_norm": 0.9533843994140625, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 12811 + }, + { + "epoch": 0.12812, + "grad_norm": 1.204431176185608, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 12812 + }, + { + "epoch": 0.12813, + "grad_norm": 0.9592463374137878, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 12813 + }, + { + "epoch": 0.12814, + "grad_norm": 0.9378253221511841, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 12814 + }, + { + "epoch": 0.12815, + "grad_norm": 1.095750331878662, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 12815 + }, + { + "epoch": 0.12816, + "grad_norm": 0.9643769860267639, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 12816 + }, + { + "epoch": 0.12817, + "grad_norm": 0.9230008721351624, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 12817 + }, + { + "epoch": 0.12818, + "grad_norm": 0.9425803422927856, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 12818 + }, + { + "epoch": 0.12819, + "grad_norm": 0.8780527114868164, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 12819 + }, + { + "epoch": 0.1282, + "grad_norm": 0.82173752784729, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 12820 + }, + { + "epoch": 0.12821, + "grad_norm": 0.7313708662986755, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 12821 + }, + { + "epoch": 0.12822, + "grad_norm": 0.5953492522239685, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 12822 + }, + { + "epoch": 0.12823, + "grad_norm": 0.5389003753662109, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 12823 + }, + { + "epoch": 0.12824, + "grad_norm": 0.4850378930568695, + "learning_rate": 0.003, + "loss": 4.016, + "step": 12824 + }, + { + "epoch": 0.12825, + "grad_norm": 0.479413777589798, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 12825 + }, + { + "epoch": 0.12826, + "grad_norm": 0.4634591341018677, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 12826 + }, + { + "epoch": 0.12827, + "grad_norm": 0.46316298842430115, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 12827 + }, + { + "epoch": 0.12828, + "grad_norm": 0.5380398631095886, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 12828 + }, + { + "epoch": 0.12829, + "grad_norm": 0.6179541945457458, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 12829 + }, + { + "epoch": 0.1283, + "grad_norm": 0.7865729928016663, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 12830 + }, + { + "epoch": 0.12831, + "grad_norm": 1.0153639316558838, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 12831 + }, + { + "epoch": 0.12832, + "grad_norm": 1.1865519285202026, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 12832 + }, + { + "epoch": 0.12833, + "grad_norm": 0.7247180938720703, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 12833 + }, + { + "epoch": 0.12834, + "grad_norm": 0.6804682612419128, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 12834 + }, + { + "epoch": 0.12835, + "grad_norm": 0.5373378396034241, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 12835 + }, + { + "epoch": 0.12836, + "grad_norm": 0.6478649973869324, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 12836 + }, + { + "epoch": 0.12837, + "grad_norm": 0.7635132670402527, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 12837 + }, + { + "epoch": 0.12838, + "grad_norm": 0.8292766809463501, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 12838 + }, + { + "epoch": 0.12839, + "grad_norm": 0.8596310019493103, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 12839 + }, + { + "epoch": 0.1284, + "grad_norm": 0.9096194505691528, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 12840 + }, + { + "epoch": 0.12841, + "grad_norm": 0.9546847939491272, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 12841 + }, + { + "epoch": 0.12842, + "grad_norm": 1.0239142179489136, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 12842 + }, + { + "epoch": 0.12843, + "grad_norm": 1.1351287364959717, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 12843 + }, + { + "epoch": 0.12844, + "grad_norm": 0.8102722764015198, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 12844 + }, + { + "epoch": 0.12845, + "grad_norm": 0.6956878900527954, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 12845 + }, + { + "epoch": 0.12846, + "grad_norm": 0.8143348693847656, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 12846 + }, + { + "epoch": 0.12847, + "grad_norm": 1.0078028440475464, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 12847 + }, + { + "epoch": 0.12848, + "grad_norm": 1.0167229175567627, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 12848 + }, + { + "epoch": 0.12849, + "grad_norm": 0.9717152118682861, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 12849 + }, + { + "epoch": 0.1285, + "grad_norm": 0.8741232752799988, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 12850 + }, + { + "epoch": 0.12851, + "grad_norm": 0.7856922149658203, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 12851 + }, + { + "epoch": 0.12852, + "grad_norm": 0.7052040696144104, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 12852 + }, + { + "epoch": 0.12853, + "grad_norm": 0.6401236653327942, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 12853 + }, + { + "epoch": 0.12854, + "grad_norm": 0.6218764185905457, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 12854 + }, + { + "epoch": 0.12855, + "grad_norm": 0.5677884817123413, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 12855 + }, + { + "epoch": 0.12856, + "grad_norm": 0.6288309097290039, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 12856 + }, + { + "epoch": 0.12857, + "grad_norm": 0.9016016721725464, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 12857 + }, + { + "epoch": 0.12858, + "grad_norm": 1.2577475309371948, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 12858 + }, + { + "epoch": 0.12859, + "grad_norm": 0.9303752779960632, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 12859 + }, + { + "epoch": 0.1286, + "grad_norm": 0.8439638614654541, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 12860 + }, + { + "epoch": 0.12861, + "grad_norm": 0.945719838142395, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 12861 + }, + { + "epoch": 0.12862, + "grad_norm": 1.131162405014038, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 12862 + }, + { + "epoch": 0.12863, + "grad_norm": 0.9842260479927063, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 12863 + }, + { + "epoch": 0.12864, + "grad_norm": 0.8646676540374756, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 12864 + }, + { + "epoch": 0.12865, + "grad_norm": 0.9796521067619324, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 12865 + }, + { + "epoch": 0.12866, + "grad_norm": 0.8813204169273376, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 12866 + }, + { + "epoch": 0.12867, + "grad_norm": 0.7494421601295471, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 12867 + }, + { + "epoch": 0.12868, + "grad_norm": 0.75898277759552, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 12868 + }, + { + "epoch": 0.12869, + "grad_norm": 0.7349717020988464, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 12869 + }, + { + "epoch": 0.1287, + "grad_norm": 0.6646801829338074, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 12870 + }, + { + "epoch": 0.12871, + "grad_norm": 0.7402473092079163, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 12871 + }, + { + "epoch": 0.12872, + "grad_norm": 0.7680990099906921, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 12872 + }, + { + "epoch": 0.12873, + "grad_norm": 0.7846862077713013, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 12873 + }, + { + "epoch": 0.12874, + "grad_norm": 0.8325425982475281, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 12874 + }, + { + "epoch": 0.12875, + "grad_norm": 1.0272611379623413, + "learning_rate": 0.003, + "loss": 4.043, + "step": 12875 + }, + { + "epoch": 0.12876, + "grad_norm": 1.1894290447235107, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 12876 + }, + { + "epoch": 0.12877, + "grad_norm": 0.8812578916549683, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 12877 + }, + { + "epoch": 0.12878, + "grad_norm": 0.8838363885879517, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 12878 + }, + { + "epoch": 0.12879, + "grad_norm": 0.8199244141578674, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 12879 + }, + { + "epoch": 0.1288, + "grad_norm": 0.6704943776130676, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 12880 + }, + { + "epoch": 0.12881, + "grad_norm": 0.603921115398407, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 12881 + }, + { + "epoch": 0.12882, + "grad_norm": 0.47761204838752747, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 12882 + }, + { + "epoch": 0.12883, + "grad_norm": 0.5151971578598022, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 12883 + }, + { + "epoch": 0.12884, + "grad_norm": 0.5642871260643005, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 12884 + }, + { + "epoch": 0.12885, + "grad_norm": 0.5510008931159973, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 12885 + }, + { + "epoch": 0.12886, + "grad_norm": 0.5852599740028381, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 12886 + }, + { + "epoch": 0.12887, + "grad_norm": 0.6851714253425598, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 12887 + }, + { + "epoch": 0.12888, + "grad_norm": 0.8603548407554626, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 12888 + }, + { + "epoch": 0.12889, + "grad_norm": 1.0281792879104614, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 12889 + }, + { + "epoch": 0.1289, + "grad_norm": 1.1843347549438477, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 12890 + }, + { + "epoch": 0.12891, + "grad_norm": 0.5978385806083679, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 12891 + }, + { + "epoch": 0.12892, + "grad_norm": 0.6088024973869324, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 12892 + }, + { + "epoch": 0.12893, + "grad_norm": 0.7430006265640259, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 12893 + }, + { + "epoch": 0.12894, + "grad_norm": 0.9343671202659607, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 12894 + }, + { + "epoch": 0.12895, + "grad_norm": 1.1110566854476929, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 12895 + }, + { + "epoch": 0.12896, + "grad_norm": 0.9754167199134827, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 12896 + }, + { + "epoch": 0.12897, + "grad_norm": 1.0744414329528809, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 12897 + }, + { + "epoch": 0.12898, + "grad_norm": 0.8748178482055664, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 12898 + }, + { + "epoch": 0.12899, + "grad_norm": 0.9084308743476868, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 12899 + }, + { + "epoch": 0.129, + "grad_norm": 0.8756034970283508, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 12900 + }, + { + "epoch": 0.12901, + "grad_norm": 0.8914014101028442, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 12901 + }, + { + "epoch": 0.12902, + "grad_norm": 0.943042516708374, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 12902 + }, + { + "epoch": 0.12903, + "grad_norm": 1.0388507843017578, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 12903 + }, + { + "epoch": 0.12904, + "grad_norm": 0.8804526925086975, + "learning_rate": 0.003, + "loss": 3.999, + "step": 12904 + }, + { + "epoch": 0.12905, + "grad_norm": 0.7760341763496399, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 12905 + }, + { + "epoch": 0.12906, + "grad_norm": 0.7428815960884094, + "learning_rate": 0.003, + "loss": 4.021, + "step": 12906 + }, + { + "epoch": 0.12907, + "grad_norm": 0.8633037805557251, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 12907 + }, + { + "epoch": 0.12908, + "grad_norm": 1.0847586393356323, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 12908 + }, + { + "epoch": 0.12909, + "grad_norm": 1.10981023311615, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 12909 + }, + { + "epoch": 0.1291, + "grad_norm": 0.8015402555465698, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 12910 + }, + { + "epoch": 0.12911, + "grad_norm": 0.6852465867996216, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 12911 + }, + { + "epoch": 0.12912, + "grad_norm": 0.6617090106010437, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 12912 + }, + { + "epoch": 0.12913, + "grad_norm": 0.6093372702598572, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 12913 + }, + { + "epoch": 0.12914, + "grad_norm": 0.5485365986824036, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 12914 + }, + { + "epoch": 0.12915, + "grad_norm": 0.5607293844223022, + "learning_rate": 0.003, + "loss": 3.991, + "step": 12915 + }, + { + "epoch": 0.12916, + "grad_norm": 0.7262622117996216, + "learning_rate": 0.003, + "loss": 4.033, + "step": 12916 + }, + { + "epoch": 0.12917, + "grad_norm": 1.0545352697372437, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 12917 + }, + { + "epoch": 0.12918, + "grad_norm": 1.1793196201324463, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 12918 + }, + { + "epoch": 0.12919, + "grad_norm": 0.710120439529419, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 12919 + }, + { + "epoch": 0.1292, + "grad_norm": 0.6058240532875061, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 12920 + }, + { + "epoch": 0.12921, + "grad_norm": 0.6739874482154846, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 12921 + }, + { + "epoch": 0.12922, + "grad_norm": 0.7524260878562927, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 12922 + }, + { + "epoch": 0.12923, + "grad_norm": 0.8127276301383972, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 12923 + }, + { + "epoch": 0.12924, + "grad_norm": 0.8501728177070618, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 12924 + }, + { + "epoch": 0.12925, + "grad_norm": 0.7651922702789307, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 12925 + }, + { + "epoch": 0.12926, + "grad_norm": 0.8245254755020142, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 12926 + }, + { + "epoch": 0.12927, + "grad_norm": 0.8267773389816284, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 12927 + }, + { + "epoch": 0.12928, + "grad_norm": 0.9031170606613159, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 12928 + }, + { + "epoch": 0.12929, + "grad_norm": 1.1104865074157715, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 12929 + }, + { + "epoch": 0.1293, + "grad_norm": 0.9228253364562988, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 12930 + }, + { + "epoch": 0.12931, + "grad_norm": 0.6955811381340027, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 12931 + }, + { + "epoch": 0.12932, + "grad_norm": 0.5862828493118286, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 12932 + }, + { + "epoch": 0.12933, + "grad_norm": 0.6926164031028748, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 12933 + }, + { + "epoch": 0.12934, + "grad_norm": 0.8853495717048645, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 12934 + }, + { + "epoch": 0.12935, + "grad_norm": 0.9204506874084473, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 12935 + }, + { + "epoch": 0.12936, + "grad_norm": 0.852302610874176, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 12936 + }, + { + "epoch": 0.12937, + "grad_norm": 0.6635742783546448, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 12937 + }, + { + "epoch": 0.12938, + "grad_norm": 0.6472227573394775, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 12938 + }, + { + "epoch": 0.12939, + "grad_norm": 0.7804316282272339, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 12939 + }, + { + "epoch": 0.1294, + "grad_norm": 0.8684574365615845, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 12940 + }, + { + "epoch": 0.12941, + "grad_norm": 0.8802387118339539, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 12941 + }, + { + "epoch": 0.12942, + "grad_norm": 0.796864926815033, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 12942 + }, + { + "epoch": 0.12943, + "grad_norm": 0.6880064606666565, + "learning_rate": 0.003, + "loss": 4.004, + "step": 12943 + }, + { + "epoch": 0.12944, + "grad_norm": 0.5624730587005615, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 12944 + }, + { + "epoch": 0.12945, + "grad_norm": 0.5585161447525024, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 12945 + }, + { + "epoch": 0.12946, + "grad_norm": 0.5746771097183228, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 12946 + }, + { + "epoch": 0.12947, + "grad_norm": 0.637833297252655, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 12947 + }, + { + "epoch": 0.12948, + "grad_norm": 0.6900153756141663, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 12948 + }, + { + "epoch": 0.12949, + "grad_norm": 0.7411880493164062, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 12949 + }, + { + "epoch": 0.1295, + "grad_norm": 0.7695903182029724, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 12950 + }, + { + "epoch": 0.12951, + "grad_norm": 0.8654025793075562, + "learning_rate": 0.003, + "loss": 4.01, + "step": 12951 + }, + { + "epoch": 0.12952, + "grad_norm": 0.9959903955459595, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 12952 + }, + { + "epoch": 0.12953, + "grad_norm": 1.0637086629867554, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 12953 + }, + { + "epoch": 0.12954, + "grad_norm": 0.8980008363723755, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 12954 + }, + { + "epoch": 0.12955, + "grad_norm": 0.9990094900131226, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 12955 + }, + { + "epoch": 0.12956, + "grad_norm": 1.0221704244613647, + "learning_rate": 0.003, + "loss": 4.022, + "step": 12956 + }, + { + "epoch": 0.12957, + "grad_norm": 1.0424240827560425, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 12957 + }, + { + "epoch": 0.12958, + "grad_norm": 0.9421285390853882, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 12958 + }, + { + "epoch": 0.12959, + "grad_norm": 0.8220925331115723, + "learning_rate": 0.003, + "loss": 4.051, + "step": 12959 + }, + { + "epoch": 0.1296, + "grad_norm": 0.9669578075408936, + "learning_rate": 0.003, + "loss": 4.026, + "step": 12960 + }, + { + "epoch": 0.12961, + "grad_norm": 1.1217392683029175, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 12961 + }, + { + "epoch": 0.12962, + "grad_norm": 1.0248963832855225, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 12962 + }, + { + "epoch": 0.12963, + "grad_norm": 1.0191049575805664, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 12963 + }, + { + "epoch": 0.12964, + "grad_norm": 1.205246090888977, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 12964 + }, + { + "epoch": 0.12965, + "grad_norm": 0.8745798468589783, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 12965 + }, + { + "epoch": 0.12966, + "grad_norm": 0.8760191798210144, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 12966 + }, + { + "epoch": 0.12967, + "grad_norm": 0.9490057826042175, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 12967 + }, + { + "epoch": 0.12968, + "grad_norm": 1.0541231632232666, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 12968 + }, + { + "epoch": 0.12969, + "grad_norm": 1.1173850297927856, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 12969 + }, + { + "epoch": 0.1297, + "grad_norm": 0.7659215331077576, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 12970 + }, + { + "epoch": 0.12971, + "grad_norm": 0.7066275477409363, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 12971 + }, + { + "epoch": 0.12972, + "grad_norm": 0.6963605880737305, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 12972 + }, + { + "epoch": 0.12973, + "grad_norm": 0.756811797618866, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 12973 + }, + { + "epoch": 0.12974, + "grad_norm": 0.801081120967865, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 12974 + }, + { + "epoch": 0.12975, + "grad_norm": 0.9290609955787659, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 12975 + }, + { + "epoch": 0.12976, + "grad_norm": 0.8734400272369385, + "learning_rate": 0.003, + "loss": 4.038, + "step": 12976 + }, + { + "epoch": 0.12977, + "grad_norm": 0.7204127907752991, + "learning_rate": 0.003, + "loss": 4.037, + "step": 12977 + }, + { + "epoch": 0.12978, + "grad_norm": 0.6100812554359436, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 12978 + }, + { + "epoch": 0.12979, + "grad_norm": 0.493714302778244, + "learning_rate": 0.003, + "loss": 3.989, + "step": 12979 + }, + { + "epoch": 0.1298, + "grad_norm": 0.6417368650436401, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 12980 + }, + { + "epoch": 0.12981, + "grad_norm": 0.8046885132789612, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 12981 + }, + { + "epoch": 0.12982, + "grad_norm": 0.8287491202354431, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 12982 + }, + { + "epoch": 0.12983, + "grad_norm": 0.8081839680671692, + "learning_rate": 0.003, + "loss": 3.993, + "step": 12983 + }, + { + "epoch": 0.12984, + "grad_norm": 0.6881992220878601, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 12984 + }, + { + "epoch": 0.12985, + "grad_norm": 0.5813509225845337, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 12985 + }, + { + "epoch": 0.12986, + "grad_norm": 0.6606971621513367, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 12986 + }, + { + "epoch": 0.12987, + "grad_norm": 0.6416282653808594, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 12987 + }, + { + "epoch": 0.12988, + "grad_norm": 0.6557727456092834, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 12988 + }, + { + "epoch": 0.12989, + "grad_norm": 0.7388274669647217, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 12989 + }, + { + "epoch": 0.1299, + "grad_norm": 0.7772365212440491, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 12990 + }, + { + "epoch": 0.12991, + "grad_norm": 0.824487030506134, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 12991 + }, + { + "epoch": 0.12992, + "grad_norm": 0.9061369895935059, + "learning_rate": 0.003, + "loss": 4.028, + "step": 12992 + }, + { + "epoch": 0.12993, + "grad_norm": 0.9541290402412415, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 12993 + }, + { + "epoch": 0.12994, + "grad_norm": 0.9159508347511292, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 12994 + }, + { + "epoch": 0.12995, + "grad_norm": 0.8194580078125, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 12995 + }, + { + "epoch": 0.12996, + "grad_norm": 0.799264669418335, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 12996 + }, + { + "epoch": 0.12997, + "grad_norm": 0.88383948802948, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 12997 + }, + { + "epoch": 0.12998, + "grad_norm": 0.8685938715934753, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 12998 + }, + { + "epoch": 0.12999, + "grad_norm": 0.8577209115028381, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 12999 + }, + { + "epoch": 0.13, + "grad_norm": 0.8677167296409607, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 13000 + }, + { + "epoch": 0.13001, + "grad_norm": 0.9226064085960388, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 13001 + }, + { + "epoch": 0.13002, + "grad_norm": 0.8852340579032898, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 13002 + }, + { + "epoch": 0.13003, + "grad_norm": 0.9692376852035522, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 13003 + }, + { + "epoch": 0.13004, + "grad_norm": 0.9139446020126343, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13004 + }, + { + "epoch": 0.13005, + "grad_norm": 0.7610192894935608, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 13005 + }, + { + "epoch": 0.13006, + "grad_norm": 0.7845240235328674, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 13006 + }, + { + "epoch": 0.13007, + "grad_norm": 0.826093852519989, + "learning_rate": 0.003, + "loss": 4.042, + "step": 13007 + }, + { + "epoch": 0.13008, + "grad_norm": 0.8527858853340149, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 13008 + }, + { + "epoch": 0.13009, + "grad_norm": 0.8826994895935059, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 13009 + }, + { + "epoch": 0.1301, + "grad_norm": 0.8605583906173706, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 13010 + }, + { + "epoch": 0.13011, + "grad_norm": 0.905889093875885, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 13011 + }, + { + "epoch": 0.13012, + "grad_norm": 0.9393042325973511, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 13012 + }, + { + "epoch": 0.13013, + "grad_norm": 0.8369143605232239, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 13013 + }, + { + "epoch": 0.13014, + "grad_norm": 0.8191674947738647, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 13014 + }, + { + "epoch": 0.13015, + "grad_norm": 0.8748950362205505, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 13015 + }, + { + "epoch": 0.13016, + "grad_norm": 0.9429699778556824, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 13016 + }, + { + "epoch": 0.13017, + "grad_norm": 0.8982075452804565, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 13017 + }, + { + "epoch": 0.13018, + "grad_norm": 0.9404157400131226, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 13018 + }, + { + "epoch": 0.13019, + "grad_norm": 1.006630539894104, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 13019 + }, + { + "epoch": 0.1302, + "grad_norm": 0.8865049481391907, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 13020 + }, + { + "epoch": 0.13021, + "grad_norm": 0.7027060985565186, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 13021 + }, + { + "epoch": 0.13022, + "grad_norm": 0.6569135785102844, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 13022 + }, + { + "epoch": 0.13023, + "grad_norm": 0.6144932508468628, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 13023 + }, + { + "epoch": 0.13024, + "grad_norm": 0.6117271780967712, + "learning_rate": 0.003, + "loss": 3.97, + "step": 13024 + }, + { + "epoch": 0.13025, + "grad_norm": 0.5788269639015198, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 13025 + }, + { + "epoch": 0.13026, + "grad_norm": 0.6472201347351074, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 13026 + }, + { + "epoch": 0.13027, + "grad_norm": 0.7238059043884277, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 13027 + }, + { + "epoch": 0.13028, + "grad_norm": 0.7223124504089355, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 13028 + }, + { + "epoch": 0.13029, + "grad_norm": 0.7082488536834717, + "learning_rate": 0.003, + "loss": 3.993, + "step": 13029 + }, + { + "epoch": 0.1303, + "grad_norm": 0.8084279298782349, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 13030 + }, + { + "epoch": 0.13031, + "grad_norm": 0.9165552854537964, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 13031 + }, + { + "epoch": 0.13032, + "grad_norm": 0.9037119150161743, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 13032 + }, + { + "epoch": 0.13033, + "grad_norm": 0.9617390036582947, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 13033 + }, + { + "epoch": 0.13034, + "grad_norm": 0.9257889986038208, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 13034 + }, + { + "epoch": 0.13035, + "grad_norm": 0.8421040773391724, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 13035 + }, + { + "epoch": 0.13036, + "grad_norm": 0.8228592276573181, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 13036 + }, + { + "epoch": 0.13037, + "grad_norm": 0.8626160025596619, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 13037 + }, + { + "epoch": 0.13038, + "grad_norm": 0.7961760759353638, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 13038 + }, + { + "epoch": 0.13039, + "grad_norm": 0.8328917026519775, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 13039 + }, + { + "epoch": 0.1304, + "grad_norm": 0.858963668346405, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 13040 + }, + { + "epoch": 0.13041, + "grad_norm": 1.034338355064392, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 13041 + }, + { + "epoch": 0.13042, + "grad_norm": 1.2494183778762817, + "learning_rate": 0.003, + "loss": 4.034, + "step": 13042 + }, + { + "epoch": 0.13043, + "grad_norm": 0.5884882807731628, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 13043 + }, + { + "epoch": 0.13044, + "grad_norm": 0.7131565809249878, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 13044 + }, + { + "epoch": 0.13045, + "grad_norm": 0.7175974249839783, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 13045 + }, + { + "epoch": 0.13046, + "grad_norm": 0.746105432510376, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 13046 + }, + { + "epoch": 0.13047, + "grad_norm": 0.8767604231834412, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 13047 + }, + { + "epoch": 0.13048, + "grad_norm": 0.8941057920455933, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 13048 + }, + { + "epoch": 0.13049, + "grad_norm": 0.8328637480735779, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 13049 + }, + { + "epoch": 0.1305, + "grad_norm": 0.8061503171920776, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 13050 + }, + { + "epoch": 0.13051, + "grad_norm": 0.8265740871429443, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 13051 + }, + { + "epoch": 0.13052, + "grad_norm": 0.845345139503479, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 13052 + }, + { + "epoch": 0.13053, + "grad_norm": 0.7674865126609802, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 13053 + }, + { + "epoch": 0.13054, + "grad_norm": 0.9645304679870605, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 13054 + }, + { + "epoch": 0.13055, + "grad_norm": 1.078338623046875, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 13055 + }, + { + "epoch": 0.13056, + "grad_norm": 0.8772696852684021, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 13056 + }, + { + "epoch": 0.13057, + "grad_norm": 0.8021833300590515, + "learning_rate": 0.003, + "loss": 4.025, + "step": 13057 + }, + { + "epoch": 0.13058, + "grad_norm": 0.9574021697044373, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 13058 + }, + { + "epoch": 0.13059, + "grad_norm": 1.120952844619751, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 13059 + }, + { + "epoch": 0.1306, + "grad_norm": 0.8463264107704163, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 13060 + }, + { + "epoch": 0.13061, + "grad_norm": 0.8074550032615662, + "learning_rate": 0.003, + "loss": 4.0, + "step": 13061 + }, + { + "epoch": 0.13062, + "grad_norm": 0.7267398238182068, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 13062 + }, + { + "epoch": 0.13063, + "grad_norm": 0.7274813652038574, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 13063 + }, + { + "epoch": 0.13064, + "grad_norm": 0.6505253314971924, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 13064 + }, + { + "epoch": 0.13065, + "grad_norm": 0.6174700260162354, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 13065 + }, + { + "epoch": 0.13066, + "grad_norm": 0.6143016219139099, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 13066 + }, + { + "epoch": 0.13067, + "grad_norm": 0.6251732707023621, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 13067 + }, + { + "epoch": 0.13068, + "grad_norm": 0.6309815645217896, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 13068 + }, + { + "epoch": 0.13069, + "grad_norm": 0.6709125638008118, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 13069 + }, + { + "epoch": 0.1307, + "grad_norm": 0.6948953866958618, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 13070 + }, + { + "epoch": 0.13071, + "grad_norm": 0.7670696377754211, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 13071 + }, + { + "epoch": 0.13072, + "grad_norm": 0.8004587292671204, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 13072 + }, + { + "epoch": 0.13073, + "grad_norm": 0.8366334438323975, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 13073 + }, + { + "epoch": 0.13074, + "grad_norm": 0.8868482112884521, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 13074 + }, + { + "epoch": 0.13075, + "grad_norm": 0.8429307341575623, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 13075 + }, + { + "epoch": 0.13076, + "grad_norm": 0.7455659508705139, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 13076 + }, + { + "epoch": 0.13077, + "grad_norm": 0.8017807602882385, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 13077 + }, + { + "epoch": 0.13078, + "grad_norm": 0.9632897973060608, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 13078 + }, + { + "epoch": 0.13079, + "grad_norm": 1.227366328239441, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 13079 + }, + { + "epoch": 0.1308, + "grad_norm": 0.9098268747329712, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 13080 + }, + { + "epoch": 0.13081, + "grad_norm": 1.0370805263519287, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 13081 + }, + { + "epoch": 0.13082, + "grad_norm": 1.001566767692566, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 13082 + }, + { + "epoch": 0.13083, + "grad_norm": 0.8821815848350525, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 13083 + }, + { + "epoch": 0.13084, + "grad_norm": 0.7804445028305054, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 13084 + }, + { + "epoch": 0.13085, + "grad_norm": 0.8916219472885132, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 13085 + }, + { + "epoch": 0.13086, + "grad_norm": 0.7816018462181091, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 13086 + }, + { + "epoch": 0.13087, + "grad_norm": 0.7173853516578674, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 13087 + }, + { + "epoch": 0.13088, + "grad_norm": 0.6855860948562622, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 13088 + }, + { + "epoch": 0.13089, + "grad_norm": 0.6942515969276428, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 13089 + }, + { + "epoch": 0.1309, + "grad_norm": 0.7197878956794739, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 13090 + }, + { + "epoch": 0.13091, + "grad_norm": 0.8413582444190979, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 13091 + }, + { + "epoch": 0.13092, + "grad_norm": 0.9425943493843079, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 13092 + }, + { + "epoch": 0.13093, + "grad_norm": 1.069011926651001, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 13093 + }, + { + "epoch": 0.13094, + "grad_norm": 0.8811655044555664, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 13094 + }, + { + "epoch": 0.13095, + "grad_norm": 0.6359062194824219, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 13095 + }, + { + "epoch": 0.13096, + "grad_norm": 0.619958221912384, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 13096 + }, + { + "epoch": 0.13097, + "grad_norm": 0.5777302384376526, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 13097 + }, + { + "epoch": 0.13098, + "grad_norm": 0.6459702849388123, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 13098 + }, + { + "epoch": 0.13099, + "grad_norm": 0.6265745162963867, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 13099 + }, + { + "epoch": 0.131, + "grad_norm": 0.6241823434829712, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 13100 + }, + { + "epoch": 0.13101, + "grad_norm": 0.6626433730125427, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 13101 + }, + { + "epoch": 0.13102, + "grad_norm": 0.7558318376541138, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 13102 + }, + { + "epoch": 0.13103, + "grad_norm": 0.7066266536712646, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 13103 + }, + { + "epoch": 0.13104, + "grad_norm": 0.7672896981239319, + "learning_rate": 0.003, + "loss": 4.02, + "step": 13104 + }, + { + "epoch": 0.13105, + "grad_norm": 0.9246180653572083, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 13105 + }, + { + "epoch": 0.13106, + "grad_norm": 0.9987875819206238, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 13106 + }, + { + "epoch": 0.13107, + "grad_norm": 1.169500470161438, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13107 + }, + { + "epoch": 0.13108, + "grad_norm": 0.9271288514137268, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 13108 + }, + { + "epoch": 0.13109, + "grad_norm": 0.7768493890762329, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 13109 + }, + { + "epoch": 0.1311, + "grad_norm": 0.705818772315979, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 13110 + }, + { + "epoch": 0.13111, + "grad_norm": 0.6956490278244019, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 13111 + }, + { + "epoch": 0.13112, + "grad_norm": 0.7605606913566589, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 13112 + }, + { + "epoch": 0.13113, + "grad_norm": 0.7887161374092102, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 13113 + }, + { + "epoch": 0.13114, + "grad_norm": 0.82375168800354, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 13114 + }, + { + "epoch": 0.13115, + "grad_norm": 0.8668541312217712, + "learning_rate": 0.003, + "loss": 4.024, + "step": 13115 + }, + { + "epoch": 0.13116, + "grad_norm": 1.0358824729919434, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 13116 + }, + { + "epoch": 0.13117, + "grad_norm": 1.1184041500091553, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 13117 + }, + { + "epoch": 0.13118, + "grad_norm": 0.7959602475166321, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 13118 + }, + { + "epoch": 0.13119, + "grad_norm": 0.6026773452758789, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 13119 + }, + { + "epoch": 0.1312, + "grad_norm": 0.7496844530105591, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 13120 + }, + { + "epoch": 0.13121, + "grad_norm": 0.7308691143989563, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 13121 + }, + { + "epoch": 0.13122, + "grad_norm": 0.6883700489997864, + "learning_rate": 0.003, + "loss": 4.008, + "step": 13122 + }, + { + "epoch": 0.13123, + "grad_norm": 0.6974976062774658, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 13123 + }, + { + "epoch": 0.13124, + "grad_norm": 0.8474065065383911, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 13124 + }, + { + "epoch": 0.13125, + "grad_norm": 1.0197933912277222, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 13125 + }, + { + "epoch": 0.13126, + "grad_norm": 1.2125210762023926, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 13126 + }, + { + "epoch": 0.13127, + "grad_norm": 0.7846187949180603, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 13127 + }, + { + "epoch": 0.13128, + "grad_norm": 0.8484559655189514, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 13128 + }, + { + "epoch": 0.13129, + "grad_norm": 1.1045933961868286, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 13129 + }, + { + "epoch": 0.1313, + "grad_norm": 0.9178270101547241, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 13130 + }, + { + "epoch": 0.13131, + "grad_norm": 0.8832055330276489, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 13131 + }, + { + "epoch": 0.13132, + "grad_norm": 0.8153164386749268, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 13132 + }, + { + "epoch": 0.13133, + "grad_norm": 0.9219346046447754, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 13133 + }, + { + "epoch": 0.13134, + "grad_norm": 0.8553265929222107, + "learning_rate": 0.003, + "loss": 4.009, + "step": 13134 + }, + { + "epoch": 0.13135, + "grad_norm": 0.8976668119430542, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 13135 + }, + { + "epoch": 0.13136, + "grad_norm": 0.882498562335968, + "learning_rate": 0.003, + "loss": 4.026, + "step": 13136 + }, + { + "epoch": 0.13137, + "grad_norm": 0.9435257911682129, + "learning_rate": 0.003, + "loss": 4.006, + "step": 13137 + }, + { + "epoch": 0.13138, + "grad_norm": 0.9073676466941833, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 13138 + }, + { + "epoch": 0.13139, + "grad_norm": 0.9786491394042969, + "learning_rate": 0.003, + "loss": 4.038, + "step": 13139 + }, + { + "epoch": 0.1314, + "grad_norm": 0.9401513934135437, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 13140 + }, + { + "epoch": 0.13141, + "grad_norm": 0.8031672835350037, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 13141 + }, + { + "epoch": 0.13142, + "grad_norm": 0.9604483842849731, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 13142 + }, + { + "epoch": 0.13143, + "grad_norm": 0.9621740579605103, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 13143 + }, + { + "epoch": 0.13144, + "grad_norm": 0.7710050344467163, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 13144 + }, + { + "epoch": 0.13145, + "grad_norm": 0.8008609414100647, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 13145 + }, + { + "epoch": 0.13146, + "grad_norm": 0.6710706353187561, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 13146 + }, + { + "epoch": 0.13147, + "grad_norm": 0.7560577392578125, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 13147 + }, + { + "epoch": 0.13148, + "grad_norm": 0.781263530254364, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 13148 + }, + { + "epoch": 0.13149, + "grad_norm": 0.7743945717811584, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 13149 + }, + { + "epoch": 0.1315, + "grad_norm": 0.7937132716178894, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 13150 + }, + { + "epoch": 0.13151, + "grad_norm": 0.8247299194335938, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 13151 + }, + { + "epoch": 0.13152, + "grad_norm": 0.8840051889419556, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 13152 + }, + { + "epoch": 0.13153, + "grad_norm": 0.8960859775543213, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 13153 + }, + { + "epoch": 0.13154, + "grad_norm": 0.9697312712669373, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 13154 + }, + { + "epoch": 0.13155, + "grad_norm": 1.0689184665679932, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 13155 + }, + { + "epoch": 0.13156, + "grad_norm": 1.0230175256729126, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 13156 + }, + { + "epoch": 0.13157, + "grad_norm": 0.8293696045875549, + "learning_rate": 0.003, + "loss": 4.061, + "step": 13157 + }, + { + "epoch": 0.13158, + "grad_norm": 0.6961458325386047, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 13158 + }, + { + "epoch": 0.13159, + "grad_norm": 0.6905359625816345, + "learning_rate": 0.003, + "loss": 4.007, + "step": 13159 + }, + { + "epoch": 0.1316, + "grad_norm": 0.6143713593482971, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 13160 + }, + { + "epoch": 0.13161, + "grad_norm": 0.5593964457511902, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 13161 + }, + { + "epoch": 0.13162, + "grad_norm": 0.5873188376426697, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 13162 + }, + { + "epoch": 0.13163, + "grad_norm": 0.6039851307868958, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 13163 + }, + { + "epoch": 0.13164, + "grad_norm": 0.6437452435493469, + "learning_rate": 0.003, + "loss": 3.995, + "step": 13164 + }, + { + "epoch": 0.13165, + "grad_norm": 0.696509599685669, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 13165 + }, + { + "epoch": 0.13166, + "grad_norm": 0.6366514563560486, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 13166 + }, + { + "epoch": 0.13167, + "grad_norm": 0.5992265939712524, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 13167 + }, + { + "epoch": 0.13168, + "grad_norm": 0.7489793300628662, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 13168 + }, + { + "epoch": 0.13169, + "grad_norm": 0.8105339407920837, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 13169 + }, + { + "epoch": 0.1317, + "grad_norm": 0.8892659544944763, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 13170 + }, + { + "epoch": 0.13171, + "grad_norm": 0.9990162253379822, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 13171 + }, + { + "epoch": 0.13172, + "grad_norm": 1.0135401487350464, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 13172 + }, + { + "epoch": 0.13173, + "grad_norm": 0.829917311668396, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 13173 + }, + { + "epoch": 0.13174, + "grad_norm": 0.6570572853088379, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 13174 + }, + { + "epoch": 0.13175, + "grad_norm": 0.654499888420105, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 13175 + }, + { + "epoch": 0.13176, + "grad_norm": 0.6734559535980225, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 13176 + }, + { + "epoch": 0.13177, + "grad_norm": 0.708061695098877, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 13177 + }, + { + "epoch": 0.13178, + "grad_norm": 0.6476017832756042, + "learning_rate": 0.003, + "loss": 4.007, + "step": 13178 + }, + { + "epoch": 0.13179, + "grad_norm": 0.7670444846153259, + "learning_rate": 0.003, + "loss": 4.011, + "step": 13179 + }, + { + "epoch": 0.1318, + "grad_norm": 0.9231152534484863, + "learning_rate": 0.003, + "loss": 4.001, + "step": 13180 + }, + { + "epoch": 0.13181, + "grad_norm": 1.0558562278747559, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 13181 + }, + { + "epoch": 0.13182, + "grad_norm": 1.0957084894180298, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 13182 + }, + { + "epoch": 0.13183, + "grad_norm": 0.754009485244751, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 13183 + }, + { + "epoch": 0.13184, + "grad_norm": 0.5575379133224487, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 13184 + }, + { + "epoch": 0.13185, + "grad_norm": 0.7497402429580688, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 13185 + }, + { + "epoch": 0.13186, + "grad_norm": 0.9239612221717834, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 13186 + }, + { + "epoch": 0.13187, + "grad_norm": 1.0049750804901123, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 13187 + }, + { + "epoch": 0.13188, + "grad_norm": 0.9554564952850342, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 13188 + }, + { + "epoch": 0.13189, + "grad_norm": 0.8094347715377808, + "learning_rate": 0.003, + "loss": 3.982, + "step": 13189 + }, + { + "epoch": 0.1319, + "grad_norm": 0.7651543617248535, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 13190 + }, + { + "epoch": 0.13191, + "grad_norm": 0.7921789288520813, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 13191 + }, + { + "epoch": 0.13192, + "grad_norm": 0.7744830846786499, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 13192 + }, + { + "epoch": 0.13193, + "grad_norm": 0.8157005310058594, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 13193 + }, + { + "epoch": 0.13194, + "grad_norm": 0.7987657785415649, + "learning_rate": 0.003, + "loss": 4.016, + "step": 13194 + }, + { + "epoch": 0.13195, + "grad_norm": 0.7527452111244202, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 13195 + }, + { + "epoch": 0.13196, + "grad_norm": 0.6971560120582581, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 13196 + }, + { + "epoch": 0.13197, + "grad_norm": 0.6103990077972412, + "learning_rate": 0.003, + "loss": 4.024, + "step": 13197 + }, + { + "epoch": 0.13198, + "grad_norm": 0.5640221834182739, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 13198 + }, + { + "epoch": 0.13199, + "grad_norm": 0.512528657913208, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 13199 + }, + { + "epoch": 0.132, + "grad_norm": 0.535915195941925, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 13200 + }, + { + "epoch": 0.13201, + "grad_norm": 0.5550373196601868, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 13201 + }, + { + "epoch": 0.13202, + "grad_norm": 0.821470320224762, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 13202 + }, + { + "epoch": 0.13203, + "grad_norm": 1.1169755458831787, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 13203 + }, + { + "epoch": 0.13204, + "grad_norm": 0.9754448533058167, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 13204 + }, + { + "epoch": 0.13205, + "grad_norm": 1.0438214540481567, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 13205 + }, + { + "epoch": 0.13206, + "grad_norm": 0.8361873626708984, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 13206 + }, + { + "epoch": 0.13207, + "grad_norm": 0.7784366011619568, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 13207 + }, + { + "epoch": 0.13208, + "grad_norm": 0.7746293544769287, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 13208 + }, + { + "epoch": 0.13209, + "grad_norm": 0.7999920845031738, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 13209 + }, + { + "epoch": 0.1321, + "grad_norm": 0.8070949912071228, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 13210 + }, + { + "epoch": 0.13211, + "grad_norm": 0.9598238468170166, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 13211 + }, + { + "epoch": 0.13212, + "grad_norm": 1.0701309442520142, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 13212 + }, + { + "epoch": 0.13213, + "grad_norm": 1.023281216621399, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 13213 + }, + { + "epoch": 0.13214, + "grad_norm": 0.9652174115180969, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 13214 + }, + { + "epoch": 0.13215, + "grad_norm": 0.9821644425392151, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 13215 + }, + { + "epoch": 0.13216, + "grad_norm": 1.14763605594635, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 13216 + }, + { + "epoch": 0.13217, + "grad_norm": 0.7339428067207336, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 13217 + }, + { + "epoch": 0.13218, + "grad_norm": 0.7249696254730225, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 13218 + }, + { + "epoch": 0.13219, + "grad_norm": 0.7884764075279236, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 13219 + }, + { + "epoch": 0.1322, + "grad_norm": 0.7975813150405884, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 13220 + }, + { + "epoch": 0.13221, + "grad_norm": 0.8157402276992798, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 13221 + }, + { + "epoch": 0.13222, + "grad_norm": 0.6767211556434631, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 13222 + }, + { + "epoch": 0.13223, + "grad_norm": 0.6843516826629639, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 13223 + }, + { + "epoch": 0.13224, + "grad_norm": 0.5921446084976196, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 13224 + }, + { + "epoch": 0.13225, + "grad_norm": 0.5578984022140503, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 13225 + }, + { + "epoch": 0.13226, + "grad_norm": 0.5266581177711487, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 13226 + }, + { + "epoch": 0.13227, + "grad_norm": 0.5994659662246704, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 13227 + }, + { + "epoch": 0.13228, + "grad_norm": 0.8346339464187622, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 13228 + }, + { + "epoch": 0.13229, + "grad_norm": 1.1519432067871094, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 13229 + }, + { + "epoch": 0.1323, + "grad_norm": 0.9563078284263611, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 13230 + }, + { + "epoch": 0.13231, + "grad_norm": 0.9070895314216614, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 13231 + }, + { + "epoch": 0.13232, + "grad_norm": 0.8742591738700867, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 13232 + }, + { + "epoch": 0.13233, + "grad_norm": 0.8203812837600708, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 13233 + }, + { + "epoch": 0.13234, + "grad_norm": 0.8483673930168152, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 13234 + }, + { + "epoch": 0.13235, + "grad_norm": 0.8297470808029175, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 13235 + }, + { + "epoch": 0.13236, + "grad_norm": 0.8401219844818115, + "learning_rate": 0.003, + "loss": 3.986, + "step": 13236 + }, + { + "epoch": 0.13237, + "grad_norm": 0.899707019329071, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 13237 + }, + { + "epoch": 0.13238, + "grad_norm": 0.8262010216712952, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 13238 + }, + { + "epoch": 0.13239, + "grad_norm": 0.691097617149353, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 13239 + }, + { + "epoch": 0.1324, + "grad_norm": 0.6713021993637085, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 13240 + }, + { + "epoch": 0.13241, + "grad_norm": 0.7988129258155823, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 13241 + }, + { + "epoch": 0.13242, + "grad_norm": 0.9341169595718384, + "learning_rate": 0.003, + "loss": 4.039, + "step": 13242 + }, + { + "epoch": 0.13243, + "grad_norm": 1.187171459197998, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 13243 + }, + { + "epoch": 0.13244, + "grad_norm": 0.8251987099647522, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 13244 + }, + { + "epoch": 0.13245, + "grad_norm": 0.6872908473014832, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 13245 + }, + { + "epoch": 0.13246, + "grad_norm": 0.6644535064697266, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 13246 + }, + { + "epoch": 0.13247, + "grad_norm": 0.7135847210884094, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 13247 + }, + { + "epoch": 0.13248, + "grad_norm": 0.7650418281555176, + "learning_rate": 0.003, + "loss": 4.037, + "step": 13248 + }, + { + "epoch": 0.13249, + "grad_norm": 0.7449178695678711, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 13249 + }, + { + "epoch": 0.1325, + "grad_norm": 0.664030134677887, + "learning_rate": 0.003, + "loss": 4.008, + "step": 13250 + }, + { + "epoch": 0.13251, + "grad_norm": 0.6319637298583984, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 13251 + }, + { + "epoch": 0.13252, + "grad_norm": 0.7340186834335327, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 13252 + }, + { + "epoch": 0.13253, + "grad_norm": 0.7993084788322449, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 13253 + }, + { + "epoch": 0.13254, + "grad_norm": 0.986703097820282, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 13254 + }, + { + "epoch": 0.13255, + "grad_norm": 1.1799412965774536, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 13255 + }, + { + "epoch": 0.13256, + "grad_norm": 0.7233451008796692, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 13256 + }, + { + "epoch": 0.13257, + "grad_norm": 0.5764203071594238, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 13257 + }, + { + "epoch": 0.13258, + "grad_norm": 0.6633540391921997, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 13258 + }, + { + "epoch": 0.13259, + "grad_norm": 0.6555531024932861, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 13259 + }, + { + "epoch": 0.1326, + "grad_norm": 0.6837939023971558, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 13260 + }, + { + "epoch": 0.13261, + "grad_norm": 0.7089996337890625, + "learning_rate": 0.003, + "loss": 4.015, + "step": 13261 + }, + { + "epoch": 0.13262, + "grad_norm": 0.8657823204994202, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 13262 + }, + { + "epoch": 0.13263, + "grad_norm": 1.006532073020935, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 13263 + }, + { + "epoch": 0.13264, + "grad_norm": 0.9170596599578857, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 13264 + }, + { + "epoch": 0.13265, + "grad_norm": 0.8759198784828186, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 13265 + }, + { + "epoch": 0.13266, + "grad_norm": 0.9278095960617065, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 13266 + }, + { + "epoch": 0.13267, + "grad_norm": 0.9916574358940125, + "learning_rate": 0.003, + "loss": 4.021, + "step": 13267 + }, + { + "epoch": 0.13268, + "grad_norm": 0.9431864023208618, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 13268 + }, + { + "epoch": 0.13269, + "grad_norm": 0.8153677582740784, + "learning_rate": 0.003, + "loss": 3.993, + "step": 13269 + }, + { + "epoch": 0.1327, + "grad_norm": 0.8011820316314697, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 13270 + }, + { + "epoch": 0.13271, + "grad_norm": 0.7391328811645508, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 13271 + }, + { + "epoch": 0.13272, + "grad_norm": 0.7164938449859619, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 13272 + }, + { + "epoch": 0.13273, + "grad_norm": 0.6879814267158508, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 13273 + }, + { + "epoch": 0.13274, + "grad_norm": 0.6241834759712219, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 13274 + }, + { + "epoch": 0.13275, + "grad_norm": 0.6053652167320251, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 13275 + }, + { + "epoch": 0.13276, + "grad_norm": 0.6461558938026428, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 13276 + }, + { + "epoch": 0.13277, + "grad_norm": 0.6877732276916504, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 13277 + }, + { + "epoch": 0.13278, + "grad_norm": 0.9395847320556641, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 13278 + }, + { + "epoch": 0.13279, + "grad_norm": 1.2979737520217896, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 13279 + }, + { + "epoch": 0.1328, + "grad_norm": 0.7131104469299316, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 13280 + }, + { + "epoch": 0.13281, + "grad_norm": 0.7071772813796997, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 13281 + }, + { + "epoch": 0.13282, + "grad_norm": 0.7539332509040833, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 13282 + }, + { + "epoch": 0.13283, + "grad_norm": 0.7766580581665039, + "learning_rate": 0.003, + "loss": 4.024, + "step": 13283 + }, + { + "epoch": 0.13284, + "grad_norm": 0.8711881041526794, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 13284 + }, + { + "epoch": 0.13285, + "grad_norm": 0.9808143377304077, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 13285 + }, + { + "epoch": 0.13286, + "grad_norm": 1.0515544414520264, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 13286 + }, + { + "epoch": 0.13287, + "grad_norm": 0.9690415859222412, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 13287 + }, + { + "epoch": 0.13288, + "grad_norm": 0.9867044687271118, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 13288 + }, + { + "epoch": 0.13289, + "grad_norm": 1.0373833179473877, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 13289 + }, + { + "epoch": 0.1329, + "grad_norm": 0.9455235600471497, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 13290 + }, + { + "epoch": 0.13291, + "grad_norm": 0.9202231168746948, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 13291 + }, + { + "epoch": 0.13292, + "grad_norm": 1.135408878326416, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 13292 + }, + { + "epoch": 0.13293, + "grad_norm": 1.058363914489746, + "learning_rate": 0.003, + "loss": 4.024, + "step": 13293 + }, + { + "epoch": 0.13294, + "grad_norm": 0.9352374076843262, + "learning_rate": 0.003, + "loss": 4.033, + "step": 13294 + }, + { + "epoch": 0.13295, + "grad_norm": 0.8602332472801208, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 13295 + }, + { + "epoch": 0.13296, + "grad_norm": 0.9056118130683899, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 13296 + }, + { + "epoch": 0.13297, + "grad_norm": 0.8816995024681091, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 13297 + }, + { + "epoch": 0.13298, + "grad_norm": 0.8176944851875305, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 13298 + }, + { + "epoch": 0.13299, + "grad_norm": 0.7336651086807251, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 13299 + }, + { + "epoch": 0.133, + "grad_norm": 0.7953958511352539, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 13300 + }, + { + "epoch": 0.13301, + "grad_norm": 0.6614307761192322, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 13301 + }, + { + "epoch": 0.13302, + "grad_norm": 0.6963691115379333, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 13302 + }, + { + "epoch": 0.13303, + "grad_norm": 0.7363566756248474, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 13303 + }, + { + "epoch": 0.13304, + "grad_norm": 0.8909085988998413, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 13304 + }, + { + "epoch": 0.13305, + "grad_norm": 0.9495358467102051, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 13305 + }, + { + "epoch": 0.13306, + "grad_norm": 0.7758327126502991, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 13306 + }, + { + "epoch": 0.13307, + "grad_norm": 0.6987218260765076, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 13307 + }, + { + "epoch": 0.13308, + "grad_norm": 0.7181601524353027, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 13308 + }, + { + "epoch": 0.13309, + "grad_norm": 0.621520459651947, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 13309 + }, + { + "epoch": 0.1331, + "grad_norm": 0.5995641350746155, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 13310 + }, + { + "epoch": 0.13311, + "grad_norm": 0.6506985425949097, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 13311 + }, + { + "epoch": 0.13312, + "grad_norm": 0.7738531827926636, + "learning_rate": 0.003, + "loss": 4.027, + "step": 13312 + }, + { + "epoch": 0.13313, + "grad_norm": 0.8469628691673279, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 13313 + }, + { + "epoch": 0.13314, + "grad_norm": 0.8940180540084839, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 13314 + }, + { + "epoch": 0.13315, + "grad_norm": 0.8369051218032837, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 13315 + }, + { + "epoch": 0.13316, + "grad_norm": 0.8768254518508911, + "learning_rate": 0.003, + "loss": 4.006, + "step": 13316 + }, + { + "epoch": 0.13317, + "grad_norm": 0.8258373141288757, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 13317 + }, + { + "epoch": 0.13318, + "grad_norm": 0.9554781913757324, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 13318 + }, + { + "epoch": 0.13319, + "grad_norm": 1.2251650094985962, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 13319 + }, + { + "epoch": 0.1332, + "grad_norm": 1.1571624279022217, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 13320 + }, + { + "epoch": 0.13321, + "grad_norm": 0.7736831307411194, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 13321 + }, + { + "epoch": 0.13322, + "grad_norm": 0.7312729954719543, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 13322 + }, + { + "epoch": 0.13323, + "grad_norm": 0.7697762250900269, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 13323 + }, + { + "epoch": 0.13324, + "grad_norm": 0.8291617035865784, + "learning_rate": 0.003, + "loss": 4.032, + "step": 13324 + }, + { + "epoch": 0.13325, + "grad_norm": 0.773373544216156, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 13325 + }, + { + "epoch": 0.13326, + "grad_norm": 0.8538793325424194, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 13326 + }, + { + "epoch": 0.13327, + "grad_norm": 0.8671174049377441, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 13327 + }, + { + "epoch": 0.13328, + "grad_norm": 0.8660413026809692, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 13328 + }, + { + "epoch": 0.13329, + "grad_norm": 0.937990128993988, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 13329 + }, + { + "epoch": 0.1333, + "grad_norm": 0.9119920134544373, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 13330 + }, + { + "epoch": 0.13331, + "grad_norm": 0.8983891606330872, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 13331 + }, + { + "epoch": 0.13332, + "grad_norm": 0.7864333987236023, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 13332 + }, + { + "epoch": 0.13333, + "grad_norm": 0.6092193126678467, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 13333 + }, + { + "epoch": 0.13334, + "grad_norm": 0.6237526535987854, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 13334 + }, + { + "epoch": 0.13335, + "grad_norm": 0.6614526510238647, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 13335 + }, + { + "epoch": 0.13336, + "grad_norm": 0.6877928376197815, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 13336 + }, + { + "epoch": 0.13337, + "grad_norm": 0.6516103744506836, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 13337 + }, + { + "epoch": 0.13338, + "grad_norm": 0.6226836442947388, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 13338 + }, + { + "epoch": 0.13339, + "grad_norm": 0.6081013679504395, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 13339 + }, + { + "epoch": 0.1334, + "grad_norm": 0.6231227517127991, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 13340 + }, + { + "epoch": 0.13341, + "grad_norm": 0.6606029868125916, + "learning_rate": 0.003, + "loss": 4.051, + "step": 13341 + }, + { + "epoch": 0.13342, + "grad_norm": 0.6467459797859192, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 13342 + }, + { + "epoch": 0.13343, + "grad_norm": 0.5785483121871948, + "learning_rate": 0.003, + "loss": 4.002, + "step": 13343 + }, + { + "epoch": 0.13344, + "grad_norm": 0.6745023727416992, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 13344 + }, + { + "epoch": 0.13345, + "grad_norm": 0.6457123756408691, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 13345 + }, + { + "epoch": 0.13346, + "grad_norm": 0.6489463448524475, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 13346 + }, + { + "epoch": 0.13347, + "grad_norm": 0.82561856508255, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 13347 + }, + { + "epoch": 0.13348, + "grad_norm": 1.219636082649231, + "learning_rate": 0.003, + "loss": 4.036, + "step": 13348 + }, + { + "epoch": 0.13349, + "grad_norm": 0.9405218362808228, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 13349 + }, + { + "epoch": 0.1335, + "grad_norm": 0.7687895894050598, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 13350 + }, + { + "epoch": 0.13351, + "grad_norm": 0.7437965869903564, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 13351 + }, + { + "epoch": 0.13352, + "grad_norm": 0.7874672412872314, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 13352 + }, + { + "epoch": 0.13353, + "grad_norm": 0.7944535613059998, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 13353 + }, + { + "epoch": 0.13354, + "grad_norm": 0.8526052236557007, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 13354 + }, + { + "epoch": 0.13355, + "grad_norm": 0.9036006331443787, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 13355 + }, + { + "epoch": 0.13356, + "grad_norm": 1.0086641311645508, + "learning_rate": 0.003, + "loss": 4.034, + "step": 13356 + }, + { + "epoch": 0.13357, + "grad_norm": 1.0628443956375122, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 13357 + }, + { + "epoch": 0.13358, + "grad_norm": 0.9245560169219971, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 13358 + }, + { + "epoch": 0.13359, + "grad_norm": 0.9113062024116516, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 13359 + }, + { + "epoch": 0.1336, + "grad_norm": 0.8313310742378235, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 13360 + }, + { + "epoch": 0.13361, + "grad_norm": 0.7733218669891357, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13361 + }, + { + "epoch": 0.13362, + "grad_norm": 0.8257316946983337, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 13362 + }, + { + "epoch": 0.13363, + "grad_norm": 0.9762942790985107, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 13363 + }, + { + "epoch": 0.13364, + "grad_norm": 1.0518579483032227, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 13364 + }, + { + "epoch": 0.13365, + "grad_norm": 1.0676583051681519, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 13365 + }, + { + "epoch": 0.13366, + "grad_norm": 1.1223315000534058, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 13366 + }, + { + "epoch": 0.13367, + "grad_norm": 0.8500106930732727, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 13367 + }, + { + "epoch": 0.13368, + "grad_norm": 0.793196976184845, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 13368 + }, + { + "epoch": 0.13369, + "grad_norm": 0.8577861785888672, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 13369 + }, + { + "epoch": 0.1337, + "grad_norm": 0.792363166809082, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 13370 + }, + { + "epoch": 0.13371, + "grad_norm": 0.7480350136756897, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 13371 + }, + { + "epoch": 0.13372, + "grad_norm": 0.7009800672531128, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 13372 + }, + { + "epoch": 0.13373, + "grad_norm": 0.6068524122238159, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 13373 + }, + { + "epoch": 0.13374, + "grad_norm": 0.5666672587394714, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 13374 + }, + { + "epoch": 0.13375, + "grad_norm": 0.5108823776245117, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 13375 + }, + { + "epoch": 0.13376, + "grad_norm": 0.5774194598197937, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 13376 + }, + { + "epoch": 0.13377, + "grad_norm": 0.6002748608589172, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 13377 + }, + { + "epoch": 0.13378, + "grad_norm": 0.7773520946502686, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 13378 + }, + { + "epoch": 0.13379, + "grad_norm": 1.0793273448944092, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 13379 + }, + { + "epoch": 0.1338, + "grad_norm": 1.0280603170394897, + "learning_rate": 0.003, + "loss": 4.009, + "step": 13380 + }, + { + "epoch": 0.13381, + "grad_norm": 1.089820384979248, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 13381 + }, + { + "epoch": 0.13382, + "grad_norm": 0.8872218132019043, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 13382 + }, + { + "epoch": 0.13383, + "grad_norm": 0.7349159121513367, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 13383 + }, + { + "epoch": 0.13384, + "grad_norm": 0.6866400837898254, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 13384 + }, + { + "epoch": 0.13385, + "grad_norm": 0.7582387328147888, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 13385 + }, + { + "epoch": 0.13386, + "grad_norm": 0.8372118473052979, + "learning_rate": 0.003, + "loss": 4.027, + "step": 13386 + }, + { + "epoch": 0.13387, + "grad_norm": 0.9449887871742249, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 13387 + }, + { + "epoch": 0.13388, + "grad_norm": 1.0124119520187378, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 13388 + }, + { + "epoch": 0.13389, + "grad_norm": 1.1869573593139648, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 13389 + }, + { + "epoch": 0.1339, + "grad_norm": 0.8177801370620728, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 13390 + }, + { + "epoch": 0.13391, + "grad_norm": 0.8380702137947083, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 13391 + }, + { + "epoch": 0.13392, + "grad_norm": 0.9253681898117065, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 13392 + }, + { + "epoch": 0.13393, + "grad_norm": 0.8196813464164734, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 13393 + }, + { + "epoch": 0.13394, + "grad_norm": 0.7437666058540344, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 13394 + }, + { + "epoch": 0.13395, + "grad_norm": 0.7029215097427368, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 13395 + }, + { + "epoch": 0.13396, + "grad_norm": 0.6601882576942444, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 13396 + }, + { + "epoch": 0.13397, + "grad_norm": 0.5576168894767761, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 13397 + }, + { + "epoch": 0.13398, + "grad_norm": 0.5518843531608582, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 13398 + }, + { + "epoch": 0.13399, + "grad_norm": 0.6258888840675354, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 13399 + }, + { + "epoch": 0.134, + "grad_norm": 0.6261287331581116, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 13400 + }, + { + "epoch": 0.13401, + "grad_norm": 0.6984511017799377, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 13401 + }, + { + "epoch": 0.13402, + "grad_norm": 0.9650695323944092, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 13402 + }, + { + "epoch": 0.13403, + "grad_norm": 1.528144121170044, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 13403 + }, + { + "epoch": 0.13404, + "grad_norm": 0.5469996333122253, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 13404 + }, + { + "epoch": 0.13405, + "grad_norm": 0.8591877818107605, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 13405 + }, + { + "epoch": 0.13406, + "grad_norm": 1.1252491474151611, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 13406 + }, + { + "epoch": 0.13407, + "grad_norm": 0.895357608795166, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 13407 + }, + { + "epoch": 0.13408, + "grad_norm": 0.8792118430137634, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 13408 + }, + { + "epoch": 0.13409, + "grad_norm": 0.7852404117584229, + "learning_rate": 0.003, + "loss": 4.018, + "step": 13409 + }, + { + "epoch": 0.1341, + "grad_norm": 0.6954460740089417, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 13410 + }, + { + "epoch": 0.13411, + "grad_norm": 0.6863563060760498, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 13411 + }, + { + "epoch": 0.13412, + "grad_norm": 0.7757440209388733, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 13412 + }, + { + "epoch": 0.13413, + "grad_norm": 0.7826729416847229, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 13413 + }, + { + "epoch": 0.13414, + "grad_norm": 0.833393394947052, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 13414 + }, + { + "epoch": 0.13415, + "grad_norm": 0.8553380966186523, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 13415 + }, + { + "epoch": 0.13416, + "grad_norm": 0.7370104789733887, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 13416 + }, + { + "epoch": 0.13417, + "grad_norm": 0.594001054763794, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 13417 + }, + { + "epoch": 0.13418, + "grad_norm": 0.6473716497421265, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 13418 + }, + { + "epoch": 0.13419, + "grad_norm": 0.6357209086418152, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 13419 + }, + { + "epoch": 0.1342, + "grad_norm": 0.761445939540863, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 13420 + }, + { + "epoch": 0.13421, + "grad_norm": 1.2044970989227295, + "learning_rate": 0.003, + "loss": 4.033, + "step": 13421 + }, + { + "epoch": 0.13422, + "grad_norm": 1.1725409030914307, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 13422 + }, + { + "epoch": 0.13423, + "grad_norm": 0.7587118744850159, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 13423 + }, + { + "epoch": 0.13424, + "grad_norm": 0.7179146409034729, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 13424 + }, + { + "epoch": 0.13425, + "grad_norm": 0.8006987571716309, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 13425 + }, + { + "epoch": 0.13426, + "grad_norm": 0.7377433776855469, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 13426 + }, + { + "epoch": 0.13427, + "grad_norm": 0.68075031042099, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 13427 + }, + { + "epoch": 0.13428, + "grad_norm": 0.7437992691993713, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 13428 + }, + { + "epoch": 0.13429, + "grad_norm": 0.7679684162139893, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 13429 + }, + { + "epoch": 0.1343, + "grad_norm": 0.8779183626174927, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 13430 + }, + { + "epoch": 0.13431, + "grad_norm": 0.8668930530548096, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 13431 + }, + { + "epoch": 0.13432, + "grad_norm": 0.75432950258255, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 13432 + }, + { + "epoch": 0.13433, + "grad_norm": 0.9304035902023315, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 13433 + }, + { + "epoch": 0.13434, + "grad_norm": 1.0698155164718628, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 13434 + }, + { + "epoch": 0.13435, + "grad_norm": 1.0459843873977661, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 13435 + }, + { + "epoch": 0.13436, + "grad_norm": 0.9980152249336243, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 13436 + }, + { + "epoch": 0.13437, + "grad_norm": 1.0447239875793457, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 13437 + }, + { + "epoch": 0.13438, + "grad_norm": 0.9681057929992676, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 13438 + }, + { + "epoch": 0.13439, + "grad_norm": 0.8853546380996704, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 13439 + }, + { + "epoch": 0.1344, + "grad_norm": 0.910854697227478, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 13440 + }, + { + "epoch": 0.13441, + "grad_norm": 0.8217945098876953, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 13441 + }, + { + "epoch": 0.13442, + "grad_norm": 0.7678465247154236, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 13442 + }, + { + "epoch": 0.13443, + "grad_norm": 0.6415266394615173, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 13443 + }, + { + "epoch": 0.13444, + "grad_norm": 0.6081373691558838, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 13444 + }, + { + "epoch": 0.13445, + "grad_norm": 0.6811102628707886, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 13445 + }, + { + "epoch": 0.13446, + "grad_norm": 0.9398423433303833, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 13446 + }, + { + "epoch": 0.13447, + "grad_norm": 1.2884865999221802, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 13447 + }, + { + "epoch": 0.13448, + "grad_norm": 0.7476019859313965, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 13448 + }, + { + "epoch": 0.13449, + "grad_norm": 0.6717467904090881, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 13449 + }, + { + "epoch": 0.1345, + "grad_norm": 0.6408231854438782, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 13450 + }, + { + "epoch": 0.13451, + "grad_norm": 0.7041977643966675, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13451 + }, + { + "epoch": 0.13452, + "grad_norm": 0.6291084885597229, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 13452 + }, + { + "epoch": 0.13453, + "grad_norm": 0.6171227097511292, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 13453 + }, + { + "epoch": 0.13454, + "grad_norm": 0.6809126734733582, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 13454 + }, + { + "epoch": 0.13455, + "grad_norm": 0.7555040121078491, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 13455 + }, + { + "epoch": 0.13456, + "grad_norm": 0.7336331009864807, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 13456 + }, + { + "epoch": 0.13457, + "grad_norm": 0.8007299900054932, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 13457 + }, + { + "epoch": 0.13458, + "grad_norm": 1.040239691734314, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 13458 + }, + { + "epoch": 0.13459, + "grad_norm": 1.1279948949813843, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 13459 + }, + { + "epoch": 0.1346, + "grad_norm": 0.5545344352722168, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 13460 + }, + { + "epoch": 0.13461, + "grad_norm": 0.584706723690033, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 13461 + }, + { + "epoch": 0.13462, + "grad_norm": 0.731682300567627, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 13462 + }, + { + "epoch": 0.13463, + "grad_norm": 0.8271353840827942, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 13463 + }, + { + "epoch": 0.13464, + "grad_norm": 0.979057252407074, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 13464 + }, + { + "epoch": 0.13465, + "grad_norm": 0.8582794666290283, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 13465 + }, + { + "epoch": 0.13466, + "grad_norm": 0.6173996925354004, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 13466 + }, + { + "epoch": 0.13467, + "grad_norm": 0.593360424041748, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 13467 + }, + { + "epoch": 0.13468, + "grad_norm": 0.5905532836914062, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 13468 + }, + { + "epoch": 0.13469, + "grad_norm": 0.6375170946121216, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 13469 + }, + { + "epoch": 0.1347, + "grad_norm": 0.6386033296585083, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 13470 + }, + { + "epoch": 0.13471, + "grad_norm": 0.630262553691864, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 13471 + }, + { + "epoch": 0.13472, + "grad_norm": 0.6448857188224792, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 13472 + }, + { + "epoch": 0.13473, + "grad_norm": 0.7303140759468079, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 13473 + }, + { + "epoch": 0.13474, + "grad_norm": 0.8613793253898621, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 13474 + }, + { + "epoch": 0.13475, + "grad_norm": 0.9755415916442871, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 13475 + }, + { + "epoch": 0.13476, + "grad_norm": 0.942423403263092, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 13476 + }, + { + "epoch": 0.13477, + "grad_norm": 0.9543547034263611, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 13477 + }, + { + "epoch": 0.13478, + "grad_norm": 0.9376398324966431, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 13478 + }, + { + "epoch": 0.13479, + "grad_norm": 0.8613263964653015, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 13479 + }, + { + "epoch": 0.1348, + "grad_norm": 0.8967022895812988, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 13480 + }, + { + "epoch": 0.13481, + "grad_norm": 0.9113709926605225, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 13481 + }, + { + "epoch": 0.13482, + "grad_norm": 1.0850682258605957, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 13482 + }, + { + "epoch": 0.13483, + "grad_norm": 1.093895673751831, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 13483 + }, + { + "epoch": 0.13484, + "grad_norm": 1.1075501441955566, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 13484 + }, + { + "epoch": 0.13485, + "grad_norm": 1.0485846996307373, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 13485 + }, + { + "epoch": 0.13486, + "grad_norm": 0.9366323351860046, + "learning_rate": 0.003, + "loss": 4.028, + "step": 13486 + }, + { + "epoch": 0.13487, + "grad_norm": 0.8190991878509521, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 13487 + }, + { + "epoch": 0.13488, + "grad_norm": 0.8177504539489746, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 13488 + }, + { + "epoch": 0.13489, + "grad_norm": 0.8176032900810242, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 13489 + }, + { + "epoch": 0.1349, + "grad_norm": 0.8497923612594604, + "learning_rate": 0.003, + "loss": 4.025, + "step": 13490 + }, + { + "epoch": 0.13491, + "grad_norm": 1.0562101602554321, + "learning_rate": 0.003, + "loss": 4.041, + "step": 13491 + }, + { + "epoch": 0.13492, + "grad_norm": 1.167596459388733, + "learning_rate": 0.003, + "loss": 4.043, + "step": 13492 + }, + { + "epoch": 0.13493, + "grad_norm": 1.1683759689331055, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 13493 + }, + { + "epoch": 0.13494, + "grad_norm": 1.0237256288528442, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 13494 + }, + { + "epoch": 0.13495, + "grad_norm": 0.8731281757354736, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 13495 + }, + { + "epoch": 0.13496, + "grad_norm": 0.8121966123580933, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 13496 + }, + { + "epoch": 0.13497, + "grad_norm": 0.8778128623962402, + "learning_rate": 0.003, + "loss": 4.016, + "step": 13497 + }, + { + "epoch": 0.13498, + "grad_norm": 0.7970010638237, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 13498 + }, + { + "epoch": 0.13499, + "grad_norm": 0.7046995759010315, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 13499 + }, + { + "epoch": 0.135, + "grad_norm": 0.6215248107910156, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 13500 + }, + { + "epoch": 0.13501, + "grad_norm": 0.6347829699516296, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 13501 + }, + { + "epoch": 0.13502, + "grad_norm": 0.6442795991897583, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 13502 + }, + { + "epoch": 0.13503, + "grad_norm": 0.6396309733390808, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 13503 + }, + { + "epoch": 0.13504, + "grad_norm": 0.6402974128723145, + "learning_rate": 0.003, + "loss": 4.009, + "step": 13504 + }, + { + "epoch": 0.13505, + "grad_norm": 0.7731344103813171, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 13505 + }, + { + "epoch": 0.13506, + "grad_norm": 0.9753412008285522, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 13506 + }, + { + "epoch": 0.13507, + "grad_norm": 1.1997851133346558, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 13507 + }, + { + "epoch": 0.13508, + "grad_norm": 0.8190879821777344, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 13508 + }, + { + "epoch": 0.13509, + "grad_norm": 0.5890142321586609, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 13509 + }, + { + "epoch": 0.1351, + "grad_norm": 0.5968877077102661, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 13510 + }, + { + "epoch": 0.13511, + "grad_norm": 0.6396861672401428, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 13511 + }, + { + "epoch": 0.13512, + "grad_norm": 0.6868215799331665, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 13512 + }, + { + "epoch": 0.13513, + "grad_norm": 0.7117477059364319, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 13513 + }, + { + "epoch": 0.13514, + "grad_norm": 0.782675564289093, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 13514 + }, + { + "epoch": 0.13515, + "grad_norm": 0.8306724429130554, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 13515 + }, + { + "epoch": 0.13516, + "grad_norm": 0.7327220439910889, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 13516 + }, + { + "epoch": 0.13517, + "grad_norm": 0.5924178957939148, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 13517 + }, + { + "epoch": 0.13518, + "grad_norm": 0.5788958668708801, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 13518 + }, + { + "epoch": 0.13519, + "grad_norm": 0.602343738079071, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 13519 + }, + { + "epoch": 0.1352, + "grad_norm": 0.671697199344635, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 13520 + }, + { + "epoch": 0.13521, + "grad_norm": 0.6452050805091858, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 13521 + }, + { + "epoch": 0.13522, + "grad_norm": 0.7343513369560242, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 13522 + }, + { + "epoch": 0.13523, + "grad_norm": 0.8468841910362244, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 13523 + }, + { + "epoch": 0.13524, + "grad_norm": 0.941436767578125, + "learning_rate": 0.003, + "loss": 4.026, + "step": 13524 + }, + { + "epoch": 0.13525, + "grad_norm": 1.0155689716339111, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 13525 + }, + { + "epoch": 0.13526, + "grad_norm": 1.1248732805252075, + "learning_rate": 0.003, + "loss": 4.019, + "step": 13526 + }, + { + "epoch": 0.13527, + "grad_norm": 0.9863359332084656, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 13527 + }, + { + "epoch": 0.13528, + "grad_norm": 0.922791063785553, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 13528 + }, + { + "epoch": 0.13529, + "grad_norm": 1.0490010976791382, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 13529 + }, + { + "epoch": 0.1353, + "grad_norm": 1.0780727863311768, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 13530 + }, + { + "epoch": 0.13531, + "grad_norm": 0.9660634398460388, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 13531 + }, + { + "epoch": 0.13532, + "grad_norm": 0.9875202775001526, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 13532 + }, + { + "epoch": 0.13533, + "grad_norm": 1.06731379032135, + "learning_rate": 0.003, + "loss": 4.017, + "step": 13533 + }, + { + "epoch": 0.13534, + "grad_norm": 0.8824008107185364, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 13534 + }, + { + "epoch": 0.13535, + "grad_norm": 0.8004950881004333, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 13535 + }, + { + "epoch": 0.13536, + "grad_norm": 0.7252044081687927, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 13536 + }, + { + "epoch": 0.13537, + "grad_norm": 0.7182325720787048, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 13537 + }, + { + "epoch": 0.13538, + "grad_norm": 0.6931709051132202, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 13538 + }, + { + "epoch": 0.13539, + "grad_norm": 0.6495668888092041, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 13539 + }, + { + "epoch": 0.1354, + "grad_norm": 0.6988504528999329, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 13540 + }, + { + "epoch": 0.13541, + "grad_norm": 0.8158469200134277, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 13541 + }, + { + "epoch": 0.13542, + "grad_norm": 0.989193320274353, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 13542 + }, + { + "epoch": 0.13543, + "grad_norm": 1.1236380338668823, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 13543 + }, + { + "epoch": 0.13544, + "grad_norm": 0.9561525583267212, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 13544 + }, + { + "epoch": 0.13545, + "grad_norm": 0.9818124175071716, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 13545 + }, + { + "epoch": 0.13546, + "grad_norm": 1.0886470079421997, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 13546 + }, + { + "epoch": 0.13547, + "grad_norm": 0.9130642414093018, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 13547 + }, + { + "epoch": 0.13548, + "grad_norm": 0.7905890345573425, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 13548 + }, + { + "epoch": 0.13549, + "grad_norm": 0.7697553634643555, + "learning_rate": 0.003, + "loss": 4.007, + "step": 13549 + }, + { + "epoch": 0.1355, + "grad_norm": 0.6934955716133118, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 13550 + }, + { + "epoch": 0.13551, + "grad_norm": 0.6529156565666199, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 13551 + }, + { + "epoch": 0.13552, + "grad_norm": 0.5677207112312317, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 13552 + }, + { + "epoch": 0.13553, + "grad_norm": 0.6107994318008423, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 13553 + }, + { + "epoch": 0.13554, + "grad_norm": 0.7828263640403748, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 13554 + }, + { + "epoch": 0.13555, + "grad_norm": 1.0647815465927124, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 13555 + }, + { + "epoch": 0.13556, + "grad_norm": 1.0612136125564575, + "learning_rate": 0.003, + "loss": 4.01, + "step": 13556 + }, + { + "epoch": 0.13557, + "grad_norm": 0.8446271419525146, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 13557 + }, + { + "epoch": 0.13558, + "grad_norm": 0.6697544455528259, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 13558 + }, + { + "epoch": 0.13559, + "grad_norm": 0.6345346570014954, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 13559 + }, + { + "epoch": 0.1356, + "grad_norm": 0.9283767938613892, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 13560 + }, + { + "epoch": 0.13561, + "grad_norm": 0.9697877168655396, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 13561 + }, + { + "epoch": 0.13562, + "grad_norm": 0.7506445050239563, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 13562 + }, + { + "epoch": 0.13563, + "grad_norm": 0.6054522395133972, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 13563 + }, + { + "epoch": 0.13564, + "grad_norm": 0.6194526553153992, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 13564 + }, + { + "epoch": 0.13565, + "grad_norm": 0.6902199983596802, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 13565 + }, + { + "epoch": 0.13566, + "grad_norm": 0.7207865118980408, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 13566 + }, + { + "epoch": 0.13567, + "grad_norm": 0.7407782077789307, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 13567 + }, + { + "epoch": 0.13568, + "grad_norm": 0.7367984652519226, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 13568 + }, + { + "epoch": 0.13569, + "grad_norm": 0.703450620174408, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13569 + }, + { + "epoch": 0.1357, + "grad_norm": 0.6739119291305542, + "learning_rate": 0.003, + "loss": 4.016, + "step": 13570 + }, + { + "epoch": 0.13571, + "grad_norm": 0.5856640934944153, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 13571 + }, + { + "epoch": 0.13572, + "grad_norm": 0.6790838241577148, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 13572 + }, + { + "epoch": 0.13573, + "grad_norm": 0.6438860297203064, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 13573 + }, + { + "epoch": 0.13574, + "grad_norm": 0.6685054898262024, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 13574 + }, + { + "epoch": 0.13575, + "grad_norm": 0.8931761980056763, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 13575 + }, + { + "epoch": 0.13576, + "grad_norm": 1.385057806968689, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 13576 + }, + { + "epoch": 0.13577, + "grad_norm": 0.7037860751152039, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 13577 + }, + { + "epoch": 0.13578, + "grad_norm": 0.6238915324211121, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 13578 + }, + { + "epoch": 0.13579, + "grad_norm": 0.8868461847305298, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 13579 + }, + { + "epoch": 0.1358, + "grad_norm": 1.170206904411316, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 13580 + }, + { + "epoch": 0.13581, + "grad_norm": 0.9208158254623413, + "learning_rate": 0.003, + "loss": 3.997, + "step": 13581 + }, + { + "epoch": 0.13582, + "grad_norm": 0.778101921081543, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 13582 + }, + { + "epoch": 0.13583, + "grad_norm": 0.8657846450805664, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 13583 + }, + { + "epoch": 0.13584, + "grad_norm": 0.9851948022842407, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 13584 + }, + { + "epoch": 0.13585, + "grad_norm": 1.0144891738891602, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 13585 + }, + { + "epoch": 0.13586, + "grad_norm": 0.9492523670196533, + "learning_rate": 0.003, + "loss": 4.064, + "step": 13586 + }, + { + "epoch": 0.13587, + "grad_norm": 0.8685741424560547, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 13587 + }, + { + "epoch": 0.13588, + "grad_norm": 0.8987164497375488, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 13588 + }, + { + "epoch": 0.13589, + "grad_norm": 0.8334383368492126, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 13589 + }, + { + "epoch": 0.1359, + "grad_norm": 0.7396984696388245, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 13590 + }, + { + "epoch": 0.13591, + "grad_norm": 0.8303385972976685, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 13591 + }, + { + "epoch": 0.13592, + "grad_norm": 0.9425806403160095, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 13592 + }, + { + "epoch": 0.13593, + "grad_norm": 0.928742527961731, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 13593 + }, + { + "epoch": 0.13594, + "grad_norm": 0.7929375767707825, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 13594 + }, + { + "epoch": 0.13595, + "grad_norm": 0.7942190766334534, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 13595 + }, + { + "epoch": 0.13596, + "grad_norm": 0.9037315845489502, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 13596 + }, + { + "epoch": 0.13597, + "grad_norm": 0.718514621257782, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 13597 + }, + { + "epoch": 0.13598, + "grad_norm": 0.8128697276115417, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 13598 + }, + { + "epoch": 0.13599, + "grad_norm": 0.9463288187980652, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 13599 + }, + { + "epoch": 0.136, + "grad_norm": 1.0681328773498535, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 13600 + }, + { + "epoch": 0.13601, + "grad_norm": 0.9706233143806458, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 13601 + }, + { + "epoch": 0.13602, + "grad_norm": 0.9826554656028748, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 13602 + }, + { + "epoch": 0.13603, + "grad_norm": 0.9344239830970764, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 13603 + }, + { + "epoch": 0.13604, + "grad_norm": 0.8888978958129883, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 13604 + }, + { + "epoch": 0.13605, + "grad_norm": 0.8250786066055298, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 13605 + }, + { + "epoch": 0.13606, + "grad_norm": 0.734150230884552, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 13606 + }, + { + "epoch": 0.13607, + "grad_norm": 0.7302347421646118, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 13607 + }, + { + "epoch": 0.13608, + "grad_norm": 0.8481457829475403, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 13608 + }, + { + "epoch": 0.13609, + "grad_norm": 1.174586296081543, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 13609 + }, + { + "epoch": 0.1361, + "grad_norm": 1.094225525856018, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 13610 + }, + { + "epoch": 0.13611, + "grad_norm": 0.7953914403915405, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 13611 + }, + { + "epoch": 0.13612, + "grad_norm": 0.6411198973655701, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 13612 + }, + { + "epoch": 0.13613, + "grad_norm": 0.5927622318267822, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 13613 + }, + { + "epoch": 0.13614, + "grad_norm": 0.5709002614021301, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 13614 + }, + { + "epoch": 0.13615, + "grad_norm": 0.6412385702133179, + "learning_rate": 0.003, + "loss": 4.023, + "step": 13615 + }, + { + "epoch": 0.13616, + "grad_norm": 0.6297585964202881, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 13616 + }, + { + "epoch": 0.13617, + "grad_norm": 0.5881394147872925, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 13617 + }, + { + "epoch": 0.13618, + "grad_norm": 0.5443843007087708, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 13618 + }, + { + "epoch": 0.13619, + "grad_norm": 0.541292130947113, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 13619 + }, + { + "epoch": 0.1362, + "grad_norm": 0.5641786456108093, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 13620 + }, + { + "epoch": 0.13621, + "grad_norm": 0.6552765369415283, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 13621 + }, + { + "epoch": 0.13622, + "grad_norm": 0.7206015586853027, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 13622 + }, + { + "epoch": 0.13623, + "grad_norm": 0.7735071778297424, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 13623 + }, + { + "epoch": 0.13624, + "grad_norm": 0.9413561820983887, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 13624 + }, + { + "epoch": 0.13625, + "grad_norm": 1.1914006471633911, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 13625 + }, + { + "epoch": 0.13626, + "grad_norm": 0.9235355257987976, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 13626 + }, + { + "epoch": 0.13627, + "grad_norm": 0.8000466227531433, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 13627 + }, + { + "epoch": 0.13628, + "grad_norm": 0.7122133374214172, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 13628 + }, + { + "epoch": 0.13629, + "grad_norm": 0.8387061953544617, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 13629 + }, + { + "epoch": 0.1363, + "grad_norm": 0.9526289701461792, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 13630 + }, + { + "epoch": 0.13631, + "grad_norm": 0.9225990772247314, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 13631 + }, + { + "epoch": 0.13632, + "grad_norm": 0.8726639151573181, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 13632 + }, + { + "epoch": 0.13633, + "grad_norm": 0.7738573551177979, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 13633 + }, + { + "epoch": 0.13634, + "grad_norm": 0.7793677449226379, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 13634 + }, + { + "epoch": 0.13635, + "grad_norm": 0.9074026346206665, + "learning_rate": 0.003, + "loss": 4.039, + "step": 13635 + }, + { + "epoch": 0.13636, + "grad_norm": 1.1774895191192627, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 13636 + }, + { + "epoch": 0.13637, + "grad_norm": 0.9255838394165039, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 13637 + }, + { + "epoch": 0.13638, + "grad_norm": 0.7061364054679871, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 13638 + }, + { + "epoch": 0.13639, + "grad_norm": 0.6312673091888428, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 13639 + }, + { + "epoch": 0.1364, + "grad_norm": 0.7434985637664795, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 13640 + }, + { + "epoch": 0.13641, + "grad_norm": 0.6780279874801636, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 13641 + }, + { + "epoch": 0.13642, + "grad_norm": 0.6205322742462158, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 13642 + }, + { + "epoch": 0.13643, + "grad_norm": 0.7035835385322571, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 13643 + }, + { + "epoch": 0.13644, + "grad_norm": 0.8032388091087341, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 13644 + }, + { + "epoch": 0.13645, + "grad_norm": 0.8594980239868164, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 13645 + }, + { + "epoch": 0.13646, + "grad_norm": 0.9656379222869873, + "learning_rate": 0.003, + "loss": 3.981, + "step": 13646 + }, + { + "epoch": 0.13647, + "grad_norm": 1.038041114807129, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 13647 + }, + { + "epoch": 0.13648, + "grad_norm": 0.907473087310791, + "learning_rate": 0.003, + "loss": 3.979, + "step": 13648 + }, + { + "epoch": 0.13649, + "grad_norm": 0.8047930002212524, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 13649 + }, + { + "epoch": 0.1365, + "grad_norm": 0.7641589641571045, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 13650 + }, + { + "epoch": 0.13651, + "grad_norm": 0.732158362865448, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 13651 + }, + { + "epoch": 0.13652, + "grad_norm": 0.6533657908439636, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 13652 + }, + { + "epoch": 0.13653, + "grad_norm": 0.7759801149368286, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 13653 + }, + { + "epoch": 0.13654, + "grad_norm": 0.8410548567771912, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 13654 + }, + { + "epoch": 0.13655, + "grad_norm": 0.907723605632782, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 13655 + }, + { + "epoch": 0.13656, + "grad_norm": 0.9699224829673767, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 13656 + }, + { + "epoch": 0.13657, + "grad_norm": 1.0853382349014282, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 13657 + }, + { + "epoch": 0.13658, + "grad_norm": 1.0450563430786133, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 13658 + }, + { + "epoch": 0.13659, + "grad_norm": 1.0690622329711914, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 13659 + }, + { + "epoch": 0.1366, + "grad_norm": 0.9868205785751343, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 13660 + }, + { + "epoch": 0.13661, + "grad_norm": 1.0094572305679321, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 13661 + }, + { + "epoch": 0.13662, + "grad_norm": 0.900059700012207, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 13662 + }, + { + "epoch": 0.13663, + "grad_norm": 0.9126373529434204, + "learning_rate": 0.003, + "loss": 4.044, + "step": 13663 + }, + { + "epoch": 0.13664, + "grad_norm": 1.091404914855957, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 13664 + }, + { + "epoch": 0.13665, + "grad_norm": 0.9754030108451843, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 13665 + }, + { + "epoch": 0.13666, + "grad_norm": 0.9318715929985046, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 13666 + }, + { + "epoch": 0.13667, + "grad_norm": 0.9125456213951111, + "learning_rate": 0.003, + "loss": 4.047, + "step": 13667 + }, + { + "epoch": 0.13668, + "grad_norm": 1.003765344619751, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 13668 + }, + { + "epoch": 0.13669, + "grad_norm": 0.789003312587738, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 13669 + }, + { + "epoch": 0.1367, + "grad_norm": 0.7328260540962219, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 13670 + }, + { + "epoch": 0.13671, + "grad_norm": 0.8168613314628601, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 13671 + }, + { + "epoch": 0.13672, + "grad_norm": 0.6888318657875061, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 13672 + }, + { + "epoch": 0.13673, + "grad_norm": 0.7158035039901733, + "learning_rate": 0.003, + "loss": 4.024, + "step": 13673 + }, + { + "epoch": 0.13674, + "grad_norm": 0.6708112955093384, + "learning_rate": 0.003, + "loss": 4.03, + "step": 13674 + }, + { + "epoch": 0.13675, + "grad_norm": 0.9656121134757996, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 13675 + }, + { + "epoch": 0.13676, + "grad_norm": 1.0979527235031128, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 13676 + }, + { + "epoch": 0.13677, + "grad_norm": 1.1144235134124756, + "learning_rate": 0.003, + "loss": 4.033, + "step": 13677 + }, + { + "epoch": 0.13678, + "grad_norm": 0.8657492399215698, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 13678 + }, + { + "epoch": 0.13679, + "grad_norm": 0.707968533039093, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 13679 + }, + { + "epoch": 0.1368, + "grad_norm": 0.6362190246582031, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 13680 + }, + { + "epoch": 0.13681, + "grad_norm": 0.7352656722068787, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 13681 + }, + { + "epoch": 0.13682, + "grad_norm": 0.7660700082778931, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 13682 + }, + { + "epoch": 0.13683, + "grad_norm": 0.8414221405982971, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 13683 + }, + { + "epoch": 0.13684, + "grad_norm": 0.9656125903129578, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 13684 + }, + { + "epoch": 0.13685, + "grad_norm": 1.0350044965744019, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 13685 + }, + { + "epoch": 0.13686, + "grad_norm": 0.9132952094078064, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 13686 + }, + { + "epoch": 0.13687, + "grad_norm": 0.6878951191902161, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 13687 + }, + { + "epoch": 0.13688, + "grad_norm": 0.5668317079544067, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 13688 + }, + { + "epoch": 0.13689, + "grad_norm": 0.6665008068084717, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 13689 + }, + { + "epoch": 0.1369, + "grad_norm": 0.672523558139801, + "learning_rate": 0.003, + "loss": 4.029, + "step": 13690 + }, + { + "epoch": 0.13691, + "grad_norm": 0.6515674591064453, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 13691 + }, + { + "epoch": 0.13692, + "grad_norm": 0.4998486042022705, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 13692 + }, + { + "epoch": 0.13693, + "grad_norm": 0.4743835926055908, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 13693 + }, + { + "epoch": 0.13694, + "grad_norm": 0.5076819658279419, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 13694 + }, + { + "epoch": 0.13695, + "grad_norm": 0.4468526542186737, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 13695 + }, + { + "epoch": 0.13696, + "grad_norm": 0.5420965552330017, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 13696 + }, + { + "epoch": 0.13697, + "grad_norm": 0.6205175518989563, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 13697 + }, + { + "epoch": 0.13698, + "grad_norm": 0.6296148896217346, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 13698 + }, + { + "epoch": 0.13699, + "grad_norm": 0.6779153943061829, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 13699 + }, + { + "epoch": 0.137, + "grad_norm": 0.8325897455215454, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 13700 + }, + { + "epoch": 0.13701, + "grad_norm": 0.9757177829742432, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 13701 + }, + { + "epoch": 0.13702, + "grad_norm": 1.009141206741333, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 13702 + }, + { + "epoch": 0.13703, + "grad_norm": 1.0573129653930664, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 13703 + }, + { + "epoch": 0.13704, + "grad_norm": 0.9576799273490906, + "learning_rate": 0.003, + "loss": 4.017, + "step": 13704 + }, + { + "epoch": 0.13705, + "grad_norm": 0.8672870397567749, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 13705 + }, + { + "epoch": 0.13706, + "grad_norm": 0.799553394317627, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 13706 + }, + { + "epoch": 0.13707, + "grad_norm": 0.7652454376220703, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 13707 + }, + { + "epoch": 0.13708, + "grad_norm": 0.8123028874397278, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 13708 + }, + { + "epoch": 0.13709, + "grad_norm": 0.7795895338058472, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 13709 + }, + { + "epoch": 0.1371, + "grad_norm": 0.8851892948150635, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 13710 + }, + { + "epoch": 0.13711, + "grad_norm": 1.0329574346542358, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 13711 + }, + { + "epoch": 0.13712, + "grad_norm": 0.9941471815109253, + "learning_rate": 0.003, + "loss": 3.996, + "step": 13712 + }, + { + "epoch": 0.13713, + "grad_norm": 0.9089716672897339, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 13713 + }, + { + "epoch": 0.13714, + "grad_norm": 0.8332324028015137, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 13714 + }, + { + "epoch": 0.13715, + "grad_norm": 0.9458366632461548, + "learning_rate": 0.003, + "loss": 4.033, + "step": 13715 + }, + { + "epoch": 0.13716, + "grad_norm": 1.0176984071731567, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 13716 + }, + { + "epoch": 0.13717, + "grad_norm": 1.2446688413619995, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 13717 + }, + { + "epoch": 0.13718, + "grad_norm": 0.681408166885376, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 13718 + }, + { + "epoch": 0.13719, + "grad_norm": 0.7224704623222351, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 13719 + }, + { + "epoch": 0.1372, + "grad_norm": 0.6912409663200378, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 13720 + }, + { + "epoch": 0.13721, + "grad_norm": 0.7295917868614197, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 13721 + }, + { + "epoch": 0.13722, + "grad_norm": 0.8005015254020691, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 13722 + }, + { + "epoch": 0.13723, + "grad_norm": 0.8219122886657715, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 13723 + }, + { + "epoch": 0.13724, + "grad_norm": 0.7524207830429077, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 13724 + }, + { + "epoch": 0.13725, + "grad_norm": 0.6833975911140442, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 13725 + }, + { + "epoch": 0.13726, + "grad_norm": 0.6006239652633667, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 13726 + }, + { + "epoch": 0.13727, + "grad_norm": 0.6788160800933838, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 13727 + }, + { + "epoch": 0.13728, + "grad_norm": 0.8558680415153503, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 13728 + }, + { + "epoch": 0.13729, + "grad_norm": 1.0257190465927124, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 13729 + }, + { + "epoch": 0.1373, + "grad_norm": 0.9806662797927856, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 13730 + }, + { + "epoch": 0.13731, + "grad_norm": 0.8801664113998413, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13731 + }, + { + "epoch": 0.13732, + "grad_norm": 0.7832850813865662, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 13732 + }, + { + "epoch": 0.13733, + "grad_norm": 0.9347167611122131, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 13733 + }, + { + "epoch": 0.13734, + "grad_norm": 0.9872972369194031, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 13734 + }, + { + "epoch": 0.13735, + "grad_norm": 0.8403304219245911, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 13735 + }, + { + "epoch": 0.13736, + "grad_norm": 0.8516517877578735, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 13736 + }, + { + "epoch": 0.13737, + "grad_norm": 0.9545196294784546, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 13737 + }, + { + "epoch": 0.13738, + "grad_norm": 1.3185926675796509, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 13738 + }, + { + "epoch": 0.13739, + "grad_norm": 0.832253634929657, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 13739 + }, + { + "epoch": 0.1374, + "grad_norm": 0.6808013916015625, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 13740 + }, + { + "epoch": 0.13741, + "grad_norm": 0.6344847679138184, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 13741 + }, + { + "epoch": 0.13742, + "grad_norm": 0.6468703746795654, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 13742 + }, + { + "epoch": 0.13743, + "grad_norm": 0.5937367677688599, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 13743 + }, + { + "epoch": 0.13744, + "grad_norm": 0.6512554883956909, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 13744 + }, + { + "epoch": 0.13745, + "grad_norm": 0.7392717003822327, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 13745 + }, + { + "epoch": 0.13746, + "grad_norm": 0.7809626460075378, + "learning_rate": 0.003, + "loss": 4.01, + "step": 13746 + }, + { + "epoch": 0.13747, + "grad_norm": 0.8470854163169861, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 13747 + }, + { + "epoch": 0.13748, + "grad_norm": 0.8827667832374573, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 13748 + }, + { + "epoch": 0.13749, + "grad_norm": 0.7677375078201294, + "learning_rate": 0.003, + "loss": 4.002, + "step": 13749 + }, + { + "epoch": 0.1375, + "grad_norm": 0.6403930187225342, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 13750 + }, + { + "epoch": 0.13751, + "grad_norm": 0.5701520442962646, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 13751 + }, + { + "epoch": 0.13752, + "grad_norm": 0.6364880204200745, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 13752 + }, + { + "epoch": 0.13753, + "grad_norm": 0.7677833437919617, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 13753 + }, + { + "epoch": 0.13754, + "grad_norm": 0.8433175683021545, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 13754 + }, + { + "epoch": 0.13755, + "grad_norm": 0.8671547770500183, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 13755 + }, + { + "epoch": 0.13756, + "grad_norm": 0.7540720701217651, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 13756 + }, + { + "epoch": 0.13757, + "grad_norm": 0.7149525284767151, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 13757 + }, + { + "epoch": 0.13758, + "grad_norm": 0.658466100692749, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 13758 + }, + { + "epoch": 0.13759, + "grad_norm": 0.6632937788963318, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 13759 + }, + { + "epoch": 0.1376, + "grad_norm": 0.7641159892082214, + "learning_rate": 0.003, + "loss": 4.03, + "step": 13760 + }, + { + "epoch": 0.13761, + "grad_norm": 0.8422225117683411, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 13761 + }, + { + "epoch": 0.13762, + "grad_norm": 0.9423866868019104, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 13762 + }, + { + "epoch": 0.13763, + "grad_norm": 1.0230610370635986, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 13763 + }, + { + "epoch": 0.13764, + "grad_norm": 1.2755424976348877, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 13764 + }, + { + "epoch": 0.13765, + "grad_norm": 0.8471209406852722, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 13765 + }, + { + "epoch": 0.13766, + "grad_norm": 0.8734785318374634, + "learning_rate": 0.003, + "loss": 3.994, + "step": 13766 + }, + { + "epoch": 0.13767, + "grad_norm": 0.7789347171783447, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 13767 + }, + { + "epoch": 0.13768, + "grad_norm": 0.9051489233970642, + "learning_rate": 0.003, + "loss": 4.027, + "step": 13768 + }, + { + "epoch": 0.13769, + "grad_norm": 0.7554457187652588, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 13769 + }, + { + "epoch": 0.1377, + "grad_norm": 0.7369177937507629, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 13770 + }, + { + "epoch": 0.13771, + "grad_norm": 1.0584901571273804, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 13771 + }, + { + "epoch": 0.13772, + "grad_norm": 1.3210872411727905, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 13772 + }, + { + "epoch": 0.13773, + "grad_norm": 0.5406081676483154, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 13773 + }, + { + "epoch": 0.13774, + "grad_norm": 0.6872454881668091, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 13774 + }, + { + "epoch": 0.13775, + "grad_norm": 0.7570769190788269, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 13775 + }, + { + "epoch": 0.13776, + "grad_norm": 0.9164384603500366, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 13776 + }, + { + "epoch": 0.13777, + "grad_norm": 1.1320087909698486, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 13777 + }, + { + "epoch": 0.13778, + "grad_norm": 0.8633060455322266, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 13778 + }, + { + "epoch": 0.13779, + "grad_norm": 0.7784515023231506, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 13779 + }, + { + "epoch": 0.1378, + "grad_norm": 0.7161611318588257, + "learning_rate": 0.003, + "loss": 4.02, + "step": 13780 + }, + { + "epoch": 0.13781, + "grad_norm": 0.7441836595535278, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 13781 + }, + { + "epoch": 0.13782, + "grad_norm": 0.711620032787323, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 13782 + }, + { + "epoch": 0.13783, + "grad_norm": 0.862492024898529, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 13783 + }, + { + "epoch": 0.13784, + "grad_norm": 0.9315939545631409, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 13784 + }, + { + "epoch": 0.13785, + "grad_norm": 1.0295634269714355, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 13785 + }, + { + "epoch": 0.13786, + "grad_norm": 0.981230616569519, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 13786 + }, + { + "epoch": 0.13787, + "grad_norm": 0.9886899590492249, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 13787 + }, + { + "epoch": 0.13788, + "grad_norm": 0.9177408218383789, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 13788 + }, + { + "epoch": 0.13789, + "grad_norm": 0.8614755868911743, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 13789 + }, + { + "epoch": 0.1379, + "grad_norm": 0.7738600373268127, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 13790 + }, + { + "epoch": 0.13791, + "grad_norm": 0.7703483700752258, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 13791 + }, + { + "epoch": 0.13792, + "grad_norm": 0.7503597736358643, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 13792 + }, + { + "epoch": 0.13793, + "grad_norm": 0.8256163001060486, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 13793 + }, + { + "epoch": 0.13794, + "grad_norm": 0.7746387720108032, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 13794 + }, + { + "epoch": 0.13795, + "grad_norm": 0.7241289615631104, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 13795 + }, + { + "epoch": 0.13796, + "grad_norm": 0.6929647326469421, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 13796 + }, + { + "epoch": 0.13797, + "grad_norm": 0.7155910730361938, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 13797 + }, + { + "epoch": 0.13798, + "grad_norm": 0.8154659867286682, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 13798 + }, + { + "epoch": 0.13799, + "grad_norm": 0.9948515892028809, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 13799 + }, + { + "epoch": 0.138, + "grad_norm": 1.161720633506775, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 13800 + }, + { + "epoch": 0.13801, + "grad_norm": 0.8052655458450317, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 13801 + }, + { + "epoch": 0.13802, + "grad_norm": 0.7105312347412109, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 13802 + }, + { + "epoch": 0.13803, + "grad_norm": 0.7463851571083069, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 13803 + }, + { + "epoch": 0.13804, + "grad_norm": 0.8518344163894653, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 13804 + }, + { + "epoch": 0.13805, + "grad_norm": 0.926234781742096, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 13805 + }, + { + "epoch": 0.13806, + "grad_norm": 1.0436285734176636, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 13806 + }, + { + "epoch": 0.13807, + "grad_norm": 1.0926405191421509, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 13807 + }, + { + "epoch": 0.13808, + "grad_norm": 0.9309552311897278, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 13808 + }, + { + "epoch": 0.13809, + "grad_norm": 0.7687803506851196, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 13809 + }, + { + "epoch": 0.1381, + "grad_norm": 0.6450755000114441, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 13810 + }, + { + "epoch": 0.13811, + "grad_norm": 0.6558342576026917, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 13811 + }, + { + "epoch": 0.13812, + "grad_norm": 0.7857155203819275, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 13812 + }, + { + "epoch": 0.13813, + "grad_norm": 0.9008277058601379, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 13813 + }, + { + "epoch": 0.13814, + "grad_norm": 1.1033518314361572, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 13814 + }, + { + "epoch": 0.13815, + "grad_norm": 0.8633435964584351, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 13815 + }, + { + "epoch": 0.13816, + "grad_norm": 0.5964910984039307, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 13816 + }, + { + "epoch": 0.13817, + "grad_norm": 0.5598768591880798, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 13817 + }, + { + "epoch": 0.13818, + "grad_norm": 0.5990583896636963, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 13818 + }, + { + "epoch": 0.13819, + "grad_norm": 0.5614003539085388, + "learning_rate": 0.003, + "loss": 3.992, + "step": 13819 + }, + { + "epoch": 0.1382, + "grad_norm": 0.5601891279220581, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 13820 + }, + { + "epoch": 0.13821, + "grad_norm": 0.6764342784881592, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 13821 + }, + { + "epoch": 0.13822, + "grad_norm": 0.7970566153526306, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 13822 + }, + { + "epoch": 0.13823, + "grad_norm": 1.0505768060684204, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 13823 + }, + { + "epoch": 0.13824, + "grad_norm": 1.0456247329711914, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 13824 + }, + { + "epoch": 0.13825, + "grad_norm": 0.8650031685829163, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 13825 + }, + { + "epoch": 0.13826, + "grad_norm": 0.7379238605499268, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 13826 + }, + { + "epoch": 0.13827, + "grad_norm": 0.7937077283859253, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 13827 + }, + { + "epoch": 0.13828, + "grad_norm": 0.8907710909843445, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 13828 + }, + { + "epoch": 0.13829, + "grad_norm": 0.9939072728157043, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 13829 + }, + { + "epoch": 0.1383, + "grad_norm": 1.0405255556106567, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 13830 + }, + { + "epoch": 0.13831, + "grad_norm": 0.8047432899475098, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 13831 + }, + { + "epoch": 0.13832, + "grad_norm": 0.7326861619949341, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 13832 + }, + { + "epoch": 0.13833, + "grad_norm": 0.8576661944389343, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 13833 + }, + { + "epoch": 0.13834, + "grad_norm": 0.8865624666213989, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 13834 + }, + { + "epoch": 0.13835, + "grad_norm": 1.0433152914047241, + "learning_rate": 0.003, + "loss": 4.007, + "step": 13835 + }, + { + "epoch": 0.13836, + "grad_norm": 1.1211081743240356, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 13836 + }, + { + "epoch": 0.13837, + "grad_norm": 0.8115255236625671, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 13837 + }, + { + "epoch": 0.13838, + "grad_norm": 0.7494043707847595, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 13838 + }, + { + "epoch": 0.13839, + "grad_norm": 0.7975366115570068, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 13839 + }, + { + "epoch": 0.1384, + "grad_norm": 0.9419084191322327, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 13840 + }, + { + "epoch": 0.13841, + "grad_norm": 1.0205785036087036, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 13841 + }, + { + "epoch": 0.13842, + "grad_norm": 0.9009459018707275, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 13842 + }, + { + "epoch": 0.13843, + "grad_norm": 0.8871148228645325, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 13843 + }, + { + "epoch": 0.13844, + "grad_norm": 0.8316652774810791, + "learning_rate": 0.003, + "loss": 4.006, + "step": 13844 + }, + { + "epoch": 0.13845, + "grad_norm": 0.8803573250770569, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 13845 + }, + { + "epoch": 0.13846, + "grad_norm": 0.9513351917266846, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 13846 + }, + { + "epoch": 0.13847, + "grad_norm": 0.9584240317344666, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 13847 + }, + { + "epoch": 0.13848, + "grad_norm": 0.9925170540809631, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 13848 + }, + { + "epoch": 0.13849, + "grad_norm": 1.1117236614227295, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 13849 + }, + { + "epoch": 0.1385, + "grad_norm": 0.8875993490219116, + "learning_rate": 0.003, + "loss": 4.026, + "step": 13850 + }, + { + "epoch": 0.13851, + "grad_norm": 1.1059731245040894, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 13851 + }, + { + "epoch": 0.13852, + "grad_norm": 1.0340826511383057, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 13852 + }, + { + "epoch": 0.13853, + "grad_norm": 0.8967835307121277, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 13853 + }, + { + "epoch": 0.13854, + "grad_norm": 0.7612320184707642, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 13854 + }, + { + "epoch": 0.13855, + "grad_norm": 0.7583867907524109, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13855 + }, + { + "epoch": 0.13856, + "grad_norm": 0.8329528570175171, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 13856 + }, + { + "epoch": 0.13857, + "grad_norm": 0.8736461997032166, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 13857 + }, + { + "epoch": 0.13858, + "grad_norm": 1.0216503143310547, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 13858 + }, + { + "epoch": 0.13859, + "grad_norm": 0.9922423958778381, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 13859 + }, + { + "epoch": 0.1386, + "grad_norm": 0.7935214042663574, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 13860 + }, + { + "epoch": 0.13861, + "grad_norm": 0.7321152687072754, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 13861 + }, + { + "epoch": 0.13862, + "grad_norm": 0.6173765659332275, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 13862 + }, + { + "epoch": 0.13863, + "grad_norm": 0.5426373481750488, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 13863 + }, + { + "epoch": 0.13864, + "grad_norm": 0.5262988209724426, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 13864 + }, + { + "epoch": 0.13865, + "grad_norm": 0.49531736969947815, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 13865 + }, + { + "epoch": 0.13866, + "grad_norm": 0.6078068017959595, + "learning_rate": 0.003, + "loss": 3.992, + "step": 13866 + }, + { + "epoch": 0.13867, + "grad_norm": 0.7791035175323486, + "learning_rate": 0.003, + "loss": 4.004, + "step": 13867 + }, + { + "epoch": 0.13868, + "grad_norm": 1.068627953529358, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 13868 + }, + { + "epoch": 0.13869, + "grad_norm": 1.1645622253417969, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 13869 + }, + { + "epoch": 0.1387, + "grad_norm": 0.6107794642448425, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 13870 + }, + { + "epoch": 0.13871, + "grad_norm": 0.6094905734062195, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 13871 + }, + { + "epoch": 0.13872, + "grad_norm": 0.7466140985488892, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 13872 + }, + { + "epoch": 0.13873, + "grad_norm": 0.7020432949066162, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 13873 + }, + { + "epoch": 0.13874, + "grad_norm": 0.7007889747619629, + "learning_rate": 0.003, + "loss": 4.004, + "step": 13874 + }, + { + "epoch": 0.13875, + "grad_norm": 0.6750438809394836, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 13875 + }, + { + "epoch": 0.13876, + "grad_norm": 0.5738099217414856, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 13876 + }, + { + "epoch": 0.13877, + "grad_norm": 0.6238150596618652, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 13877 + }, + { + "epoch": 0.13878, + "grad_norm": 0.6997177600860596, + "learning_rate": 0.003, + "loss": 3.982, + "step": 13878 + }, + { + "epoch": 0.13879, + "grad_norm": 0.7597888708114624, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 13879 + }, + { + "epoch": 0.1388, + "grad_norm": 0.7653550505638123, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 13880 + }, + { + "epoch": 0.13881, + "grad_norm": 0.7531750798225403, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 13881 + }, + { + "epoch": 0.13882, + "grad_norm": 0.7508596181869507, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 13882 + }, + { + "epoch": 0.13883, + "grad_norm": 0.762292742729187, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 13883 + }, + { + "epoch": 0.13884, + "grad_norm": 0.8650839328765869, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 13884 + }, + { + "epoch": 0.13885, + "grad_norm": 0.9082847237586975, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 13885 + }, + { + "epoch": 0.13886, + "grad_norm": 1.0475541353225708, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 13886 + }, + { + "epoch": 0.13887, + "grad_norm": 0.8007717728614807, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 13887 + }, + { + "epoch": 0.13888, + "grad_norm": 0.5885525941848755, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 13888 + }, + { + "epoch": 0.13889, + "grad_norm": 0.5717440247535706, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 13889 + }, + { + "epoch": 0.1389, + "grad_norm": 0.5407517552375793, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 13890 + }, + { + "epoch": 0.13891, + "grad_norm": 0.5301905274391174, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 13891 + }, + { + "epoch": 0.13892, + "grad_norm": 0.6643052697181702, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 13892 + }, + { + "epoch": 0.13893, + "grad_norm": 0.7942160367965698, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 13893 + }, + { + "epoch": 0.13894, + "grad_norm": 1.1539918184280396, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 13894 + }, + { + "epoch": 0.13895, + "grad_norm": 1.0380985736846924, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 13895 + }, + { + "epoch": 0.13896, + "grad_norm": 1.0376297235488892, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 13896 + }, + { + "epoch": 0.13897, + "grad_norm": 0.9031848311424255, + "learning_rate": 0.003, + "loss": 4.011, + "step": 13897 + }, + { + "epoch": 0.13898, + "grad_norm": 0.782201886177063, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 13898 + }, + { + "epoch": 0.13899, + "grad_norm": 0.8607664108276367, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 13899 + }, + { + "epoch": 0.139, + "grad_norm": 0.9175698161125183, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 13900 + }, + { + "epoch": 0.13901, + "grad_norm": 0.8617496490478516, + "learning_rate": 0.003, + "loss": 4.034, + "step": 13901 + }, + { + "epoch": 0.13902, + "grad_norm": 1.0018264055252075, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 13902 + }, + { + "epoch": 0.13903, + "grad_norm": 1.1453460454940796, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 13903 + }, + { + "epoch": 0.13904, + "grad_norm": 0.8639815449714661, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 13904 + }, + { + "epoch": 0.13905, + "grad_norm": 0.874984860420227, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 13905 + }, + { + "epoch": 0.13906, + "grad_norm": 0.9996791481971741, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 13906 + }, + { + "epoch": 0.13907, + "grad_norm": 1.052925944328308, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 13907 + }, + { + "epoch": 0.13908, + "grad_norm": 0.8876622915267944, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 13908 + }, + { + "epoch": 0.13909, + "grad_norm": 0.859562873840332, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 13909 + }, + { + "epoch": 0.1391, + "grad_norm": 0.8640323877334595, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 13910 + }, + { + "epoch": 0.13911, + "grad_norm": 0.8964762091636658, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 13911 + }, + { + "epoch": 0.13912, + "grad_norm": 0.9361962676048279, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 13912 + }, + { + "epoch": 0.13913, + "grad_norm": 0.9901366829872131, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 13913 + }, + { + "epoch": 0.13914, + "grad_norm": 0.8960790634155273, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 13914 + }, + { + "epoch": 0.13915, + "grad_norm": 0.8571736812591553, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 13915 + }, + { + "epoch": 0.13916, + "grad_norm": 0.7628933191299438, + "learning_rate": 0.003, + "loss": 4.027, + "step": 13916 + }, + { + "epoch": 0.13917, + "grad_norm": 0.6896006464958191, + "learning_rate": 0.003, + "loss": 4.035, + "step": 13917 + }, + { + "epoch": 0.13918, + "grad_norm": 0.7465510368347168, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 13918 + }, + { + "epoch": 0.13919, + "grad_norm": 0.6486213803291321, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 13919 + }, + { + "epoch": 0.1392, + "grad_norm": 0.6836336851119995, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 13920 + }, + { + "epoch": 0.13921, + "grad_norm": 0.7179591655731201, + "learning_rate": 0.003, + "loss": 4.013, + "step": 13921 + }, + { + "epoch": 0.13922, + "grad_norm": 0.9003704190254211, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 13922 + }, + { + "epoch": 0.13923, + "grad_norm": 1.1060301065444946, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 13923 + }, + { + "epoch": 0.13924, + "grad_norm": 0.8880802989006042, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 13924 + }, + { + "epoch": 0.13925, + "grad_norm": 0.6416323184967041, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 13925 + }, + { + "epoch": 0.13926, + "grad_norm": 0.5431221127510071, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 13926 + }, + { + "epoch": 0.13927, + "grad_norm": 0.6355161666870117, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 13927 + }, + { + "epoch": 0.13928, + "grad_norm": 0.7978395819664001, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 13928 + }, + { + "epoch": 0.13929, + "grad_norm": 1.029797077178955, + "learning_rate": 0.003, + "loss": 3.993, + "step": 13929 + }, + { + "epoch": 0.1393, + "grad_norm": 1.049689531326294, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 13930 + }, + { + "epoch": 0.13931, + "grad_norm": 0.8194695115089417, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 13931 + }, + { + "epoch": 0.13932, + "grad_norm": 0.7146439552307129, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 13932 + }, + { + "epoch": 0.13933, + "grad_norm": 0.7814133167266846, + "learning_rate": 0.003, + "loss": 3.986, + "step": 13933 + }, + { + "epoch": 0.13934, + "grad_norm": 0.9935269951820374, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 13934 + }, + { + "epoch": 0.13935, + "grad_norm": 0.9926442503929138, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 13935 + }, + { + "epoch": 0.13936, + "grad_norm": 0.7810591459274292, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 13936 + }, + { + "epoch": 0.13937, + "grad_norm": 0.6529386043548584, + "learning_rate": 0.003, + "loss": 3.991, + "step": 13937 + }, + { + "epoch": 0.13938, + "grad_norm": 0.6165884137153625, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 13938 + }, + { + "epoch": 0.13939, + "grad_norm": 0.6040536165237427, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 13939 + }, + { + "epoch": 0.1394, + "grad_norm": 0.6759775280952454, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 13940 + }, + { + "epoch": 0.13941, + "grad_norm": 0.7123100161552429, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 13941 + }, + { + "epoch": 0.13942, + "grad_norm": 0.7399691939353943, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 13942 + }, + { + "epoch": 0.13943, + "grad_norm": 0.8146520853042603, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 13943 + }, + { + "epoch": 0.13944, + "grad_norm": 0.7907037138938904, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 13944 + }, + { + "epoch": 0.13945, + "grad_norm": 0.8921141624450684, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 13945 + }, + { + "epoch": 0.13946, + "grad_norm": 0.9739939570426941, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 13946 + }, + { + "epoch": 0.13947, + "grad_norm": 0.9629883170127869, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 13947 + }, + { + "epoch": 0.13948, + "grad_norm": 0.9390729665756226, + "learning_rate": 0.003, + "loss": 4.036, + "step": 13948 + }, + { + "epoch": 0.13949, + "grad_norm": 0.9858075380325317, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 13949 + }, + { + "epoch": 0.1395, + "grad_norm": 0.9481609463691711, + "learning_rate": 0.003, + "loss": 4.047, + "step": 13950 + }, + { + "epoch": 0.13951, + "grad_norm": 1.255966305732727, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 13951 + }, + { + "epoch": 0.13952, + "grad_norm": 1.0052669048309326, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 13952 + }, + { + "epoch": 0.13953, + "grad_norm": 1.0262274742126465, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 13953 + }, + { + "epoch": 0.13954, + "grad_norm": 1.0653098821640015, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 13954 + }, + { + "epoch": 0.13955, + "grad_norm": 0.9458314180374146, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 13955 + }, + { + "epoch": 0.13956, + "grad_norm": 0.7262864112854004, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 13956 + }, + { + "epoch": 0.13957, + "grad_norm": 0.6911898851394653, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 13957 + }, + { + "epoch": 0.13958, + "grad_norm": 0.7391590476036072, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 13958 + }, + { + "epoch": 0.13959, + "grad_norm": 0.7500927448272705, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 13959 + }, + { + "epoch": 0.1396, + "grad_norm": 0.7304272055625916, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 13960 + }, + { + "epoch": 0.13961, + "grad_norm": 0.786148726940155, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 13961 + }, + { + "epoch": 0.13962, + "grad_norm": 0.7240628600120544, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 13962 + }, + { + "epoch": 0.13963, + "grad_norm": 0.7036923170089722, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 13963 + }, + { + "epoch": 0.13964, + "grad_norm": 0.7554377913475037, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 13964 + }, + { + "epoch": 0.13965, + "grad_norm": 0.8028139472007751, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 13965 + }, + { + "epoch": 0.13966, + "grad_norm": 0.7784460783004761, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 13966 + }, + { + "epoch": 0.13967, + "grad_norm": 0.6501883268356323, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 13967 + }, + { + "epoch": 0.13968, + "grad_norm": 0.6861026883125305, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 13968 + }, + { + "epoch": 0.13969, + "grad_norm": 0.7006701231002808, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 13969 + }, + { + "epoch": 0.1397, + "grad_norm": 0.5880072712898254, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 13970 + }, + { + "epoch": 0.13971, + "grad_norm": 0.6021777391433716, + "learning_rate": 0.003, + "loss": 4.025, + "step": 13971 + }, + { + "epoch": 0.13972, + "grad_norm": 0.6727344393730164, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 13972 + }, + { + "epoch": 0.13973, + "grad_norm": 0.8657001256942749, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 13973 + }, + { + "epoch": 0.13974, + "grad_norm": 1.2038829326629639, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 13974 + }, + { + "epoch": 0.13975, + "grad_norm": 1.0898510217666626, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 13975 + }, + { + "epoch": 0.13976, + "grad_norm": 0.7038004994392395, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 13976 + }, + { + "epoch": 0.13977, + "grad_norm": 0.5771560072898865, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 13977 + }, + { + "epoch": 0.13978, + "grad_norm": 0.7867242693901062, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 13978 + }, + { + "epoch": 0.13979, + "grad_norm": 0.9191820621490479, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 13979 + }, + { + "epoch": 0.1398, + "grad_norm": 1.0388456583023071, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 13980 + }, + { + "epoch": 0.13981, + "grad_norm": 0.8392547369003296, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 13981 + }, + { + "epoch": 0.13982, + "grad_norm": 0.7875909805297852, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 13982 + }, + { + "epoch": 0.13983, + "grad_norm": 0.9248232245445251, + "learning_rate": 0.003, + "loss": 4.028, + "step": 13983 + }, + { + "epoch": 0.13984, + "grad_norm": 0.9973124861717224, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 13984 + }, + { + "epoch": 0.13985, + "grad_norm": 0.9626933932304382, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 13985 + }, + { + "epoch": 0.13986, + "grad_norm": 0.8206491470336914, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 13986 + }, + { + "epoch": 0.13987, + "grad_norm": 0.8485277891159058, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 13987 + }, + { + "epoch": 0.13988, + "grad_norm": 1.0558282136917114, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 13988 + }, + { + "epoch": 0.13989, + "grad_norm": 1.1129555702209473, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13989 + }, + { + "epoch": 0.1399, + "grad_norm": 0.8321244120597839, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 13990 + }, + { + "epoch": 0.13991, + "grad_norm": 0.7786542773246765, + "learning_rate": 0.003, + "loss": 4.026, + "step": 13991 + }, + { + "epoch": 0.13992, + "grad_norm": 0.6897526383399963, + "learning_rate": 0.003, + "loss": 4.016, + "step": 13992 + }, + { + "epoch": 0.13993, + "grad_norm": 0.7743588089942932, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 13993 + }, + { + "epoch": 0.13994, + "grad_norm": 0.8591880798339844, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 13994 + }, + { + "epoch": 0.13995, + "grad_norm": 0.8250160813331604, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 13995 + }, + { + "epoch": 0.13996, + "grad_norm": 0.7697608470916748, + "learning_rate": 0.003, + "loss": 4.012, + "step": 13996 + }, + { + "epoch": 0.13997, + "grad_norm": 0.7532920241355896, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13997 + }, + { + "epoch": 0.13998, + "grad_norm": 0.7935050129890442, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 13998 + }, + { + "epoch": 0.13999, + "grad_norm": 0.8093975186347961, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 13999 + }, + { + "epoch": 0.14, + "grad_norm": 0.6865020990371704, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 14000 + }, + { + "epoch": 0.14001, + "grad_norm": 0.6700130105018616, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 14001 + }, + { + "epoch": 0.14002, + "grad_norm": 0.6797581911087036, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 14002 + }, + { + "epoch": 0.14003, + "grad_norm": 0.6515060663223267, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 14003 + }, + { + "epoch": 0.14004, + "grad_norm": 0.8096653819084167, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 14004 + }, + { + "epoch": 0.14005, + "grad_norm": 0.997570276260376, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 14005 + }, + { + "epoch": 0.14006, + "grad_norm": 1.2135899066925049, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 14006 + }, + { + "epoch": 0.14007, + "grad_norm": 0.7099087834358215, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 14007 + }, + { + "epoch": 0.14008, + "grad_norm": 0.7432337403297424, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 14008 + }, + { + "epoch": 0.14009, + "grad_norm": 0.8485451936721802, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 14009 + }, + { + "epoch": 0.1401, + "grad_norm": 0.7696965932846069, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 14010 + }, + { + "epoch": 0.14011, + "grad_norm": 0.7383472323417664, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 14011 + }, + { + "epoch": 0.14012, + "grad_norm": 0.8596447110176086, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 14012 + }, + { + "epoch": 0.14013, + "grad_norm": 1.019231915473938, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 14013 + }, + { + "epoch": 0.14014, + "grad_norm": 1.083799123764038, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 14014 + }, + { + "epoch": 0.14015, + "grad_norm": 0.7324127554893494, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 14015 + }, + { + "epoch": 0.14016, + "grad_norm": 0.6100013852119446, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 14016 + }, + { + "epoch": 0.14017, + "grad_norm": 0.5659640431404114, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 14017 + }, + { + "epoch": 0.14018, + "grad_norm": 0.5738541483879089, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 14018 + }, + { + "epoch": 0.14019, + "grad_norm": 0.5823083519935608, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 14019 + }, + { + "epoch": 0.1402, + "grad_norm": 0.575903594493866, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 14020 + }, + { + "epoch": 0.14021, + "grad_norm": 0.585667610168457, + "learning_rate": 0.003, + "loss": 4.022, + "step": 14021 + }, + { + "epoch": 0.14022, + "grad_norm": 0.6648114323616028, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 14022 + }, + { + "epoch": 0.14023, + "grad_norm": 0.6566023826599121, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 14023 + }, + { + "epoch": 0.14024, + "grad_norm": 0.7724025249481201, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 14024 + }, + { + "epoch": 0.14025, + "grad_norm": 0.9104265570640564, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 14025 + }, + { + "epoch": 0.14026, + "grad_norm": 0.9596461057662964, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 14026 + }, + { + "epoch": 0.14027, + "grad_norm": 1.0541963577270508, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 14027 + }, + { + "epoch": 0.14028, + "grad_norm": 0.9626950025558472, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 14028 + }, + { + "epoch": 0.14029, + "grad_norm": 0.9601374268531799, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 14029 + }, + { + "epoch": 0.1403, + "grad_norm": 0.92144376039505, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 14030 + }, + { + "epoch": 0.14031, + "grad_norm": 0.8757461905479431, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 14031 + }, + { + "epoch": 0.14032, + "grad_norm": 0.8189901113510132, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 14032 + }, + { + "epoch": 0.14033, + "grad_norm": 0.8232151865959167, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 14033 + }, + { + "epoch": 0.14034, + "grad_norm": 0.8041811585426331, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 14034 + }, + { + "epoch": 0.14035, + "grad_norm": 0.895202100276947, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 14035 + }, + { + "epoch": 0.14036, + "grad_norm": 0.7802772521972656, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 14036 + }, + { + "epoch": 0.14037, + "grad_norm": 0.8878588676452637, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 14037 + }, + { + "epoch": 0.14038, + "grad_norm": 0.8776265382766724, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 14038 + }, + { + "epoch": 0.14039, + "grad_norm": 0.7140911221504211, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 14039 + }, + { + "epoch": 0.1404, + "grad_norm": 0.7636114358901978, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 14040 + }, + { + "epoch": 0.14041, + "grad_norm": 0.8802103400230408, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 14041 + }, + { + "epoch": 0.14042, + "grad_norm": 0.8500376343727112, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 14042 + }, + { + "epoch": 0.14043, + "grad_norm": 0.9757203459739685, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 14043 + }, + { + "epoch": 0.14044, + "grad_norm": 1.1028854846954346, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 14044 + }, + { + "epoch": 0.14045, + "grad_norm": 1.1399049758911133, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 14045 + }, + { + "epoch": 0.14046, + "grad_norm": 0.7455863356590271, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 14046 + }, + { + "epoch": 0.14047, + "grad_norm": 0.6432068347930908, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 14047 + }, + { + "epoch": 0.14048, + "grad_norm": 0.6292878985404968, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 14048 + }, + { + "epoch": 0.14049, + "grad_norm": 0.6780155897140503, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 14049 + }, + { + "epoch": 0.1405, + "grad_norm": 0.8050853610038757, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 14050 + }, + { + "epoch": 0.14051, + "grad_norm": 0.8014886975288391, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 14051 + }, + { + "epoch": 0.14052, + "grad_norm": 0.8241047859191895, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 14052 + }, + { + "epoch": 0.14053, + "grad_norm": 0.8780802488327026, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 14053 + }, + { + "epoch": 0.14054, + "grad_norm": 0.9953709840774536, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 14054 + }, + { + "epoch": 0.14055, + "grad_norm": 1.020336389541626, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 14055 + }, + { + "epoch": 0.14056, + "grad_norm": 0.9320582151412964, + "learning_rate": 0.003, + "loss": 4.034, + "step": 14056 + }, + { + "epoch": 0.14057, + "grad_norm": 0.7158468961715698, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 14057 + }, + { + "epoch": 0.14058, + "grad_norm": 0.6542931795120239, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 14058 + }, + { + "epoch": 0.14059, + "grad_norm": 0.5777246952056885, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 14059 + }, + { + "epoch": 0.1406, + "grad_norm": 0.5632390379905701, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 14060 + }, + { + "epoch": 0.14061, + "grad_norm": 0.6434266567230225, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 14061 + }, + { + "epoch": 0.14062, + "grad_norm": 0.6584779024124146, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 14062 + }, + { + "epoch": 0.14063, + "grad_norm": 0.8279368281364441, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 14063 + }, + { + "epoch": 0.14064, + "grad_norm": 1.1137088537216187, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 14064 + }, + { + "epoch": 0.14065, + "grad_norm": 0.9908633828163147, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 14065 + }, + { + "epoch": 0.14066, + "grad_norm": 0.814708411693573, + "learning_rate": 0.003, + "loss": 4.011, + "step": 14066 + }, + { + "epoch": 0.14067, + "grad_norm": 0.6335097551345825, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 14067 + }, + { + "epoch": 0.14068, + "grad_norm": 0.6922863721847534, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 14068 + }, + { + "epoch": 0.14069, + "grad_norm": 0.8013014197349548, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 14069 + }, + { + "epoch": 0.1407, + "grad_norm": 0.8348655700683594, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 14070 + }, + { + "epoch": 0.14071, + "grad_norm": 0.8111964464187622, + "learning_rate": 0.003, + "loss": 4.019, + "step": 14071 + }, + { + "epoch": 0.14072, + "grad_norm": 0.8467102646827698, + "learning_rate": 0.003, + "loss": 4.032, + "step": 14072 + }, + { + "epoch": 0.14073, + "grad_norm": 1.0120962858200073, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 14073 + }, + { + "epoch": 0.14074, + "grad_norm": 1.0943655967712402, + "learning_rate": 0.003, + "loss": 4.042, + "step": 14074 + }, + { + "epoch": 0.14075, + "grad_norm": 1.0920785665512085, + "learning_rate": 0.003, + "loss": 4.035, + "step": 14075 + }, + { + "epoch": 0.14076, + "grad_norm": 0.9106650948524475, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 14076 + }, + { + "epoch": 0.14077, + "grad_norm": 0.8137316107749939, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 14077 + }, + { + "epoch": 0.14078, + "grad_norm": 0.8589572310447693, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 14078 + }, + { + "epoch": 0.14079, + "grad_norm": 0.8999586701393127, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 14079 + }, + { + "epoch": 0.1408, + "grad_norm": 0.6625589728355408, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 14080 + }, + { + "epoch": 0.14081, + "grad_norm": 0.6258305311203003, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 14081 + }, + { + "epoch": 0.14082, + "grad_norm": 0.653458297252655, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 14082 + }, + { + "epoch": 0.14083, + "grad_norm": 0.6369495987892151, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 14083 + }, + { + "epoch": 0.14084, + "grad_norm": 0.6380630731582642, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 14084 + }, + { + "epoch": 0.14085, + "grad_norm": 0.6367437839508057, + "learning_rate": 0.003, + "loss": 3.998, + "step": 14085 + }, + { + "epoch": 0.14086, + "grad_norm": 0.6577797532081604, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 14086 + }, + { + "epoch": 0.14087, + "grad_norm": 0.6576122641563416, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 14087 + }, + { + "epoch": 0.14088, + "grad_norm": 0.6562979817390442, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 14088 + }, + { + "epoch": 0.14089, + "grad_norm": 0.7040120959281921, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 14089 + }, + { + "epoch": 0.1409, + "grad_norm": 0.7400263547897339, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 14090 + }, + { + "epoch": 0.14091, + "grad_norm": 1.0668511390686035, + "learning_rate": 0.003, + "loss": 3.999, + "step": 14091 + }, + { + "epoch": 0.14092, + "grad_norm": 1.3164247274398804, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 14092 + }, + { + "epoch": 0.14093, + "grad_norm": 0.7127968072891235, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 14093 + }, + { + "epoch": 0.14094, + "grad_norm": 0.5614868998527527, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 14094 + }, + { + "epoch": 0.14095, + "grad_norm": 0.603585958480835, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 14095 + }, + { + "epoch": 0.14096, + "grad_norm": 0.6809159517288208, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 14096 + }, + { + "epoch": 0.14097, + "grad_norm": 0.8328123688697815, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 14097 + }, + { + "epoch": 0.14098, + "grad_norm": 0.9252586364746094, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 14098 + }, + { + "epoch": 0.14099, + "grad_norm": 0.9418660402297974, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 14099 + }, + { + "epoch": 0.141, + "grad_norm": 0.8653552532196045, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 14100 + }, + { + "epoch": 0.14101, + "grad_norm": 0.9184288382530212, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 14101 + }, + { + "epoch": 0.14102, + "grad_norm": 1.0402677059173584, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 14102 + }, + { + "epoch": 0.14103, + "grad_norm": 1.1628098487854004, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 14103 + }, + { + "epoch": 0.14104, + "grad_norm": 0.7976023554801941, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 14104 + }, + { + "epoch": 0.14105, + "grad_norm": 0.6668494343757629, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 14105 + }, + { + "epoch": 0.14106, + "grad_norm": 0.6816210746765137, + "learning_rate": 0.003, + "loss": 4.012, + "step": 14106 + }, + { + "epoch": 0.14107, + "grad_norm": 0.7886625528335571, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 14107 + }, + { + "epoch": 0.14108, + "grad_norm": 0.7702372670173645, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 14108 + }, + { + "epoch": 0.14109, + "grad_norm": 0.8172313570976257, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 14109 + }, + { + "epoch": 0.1411, + "grad_norm": 0.8660810589790344, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 14110 + }, + { + "epoch": 0.14111, + "grad_norm": 0.9497108459472656, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 14111 + }, + { + "epoch": 0.14112, + "grad_norm": 1.007104516029358, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 14112 + }, + { + "epoch": 0.14113, + "grad_norm": 1.1830443143844604, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 14113 + }, + { + "epoch": 0.14114, + "grad_norm": 0.8512828350067139, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 14114 + }, + { + "epoch": 0.14115, + "grad_norm": 0.6526487469673157, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14115 + }, + { + "epoch": 0.14116, + "grad_norm": 0.6798277497291565, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 14116 + }, + { + "epoch": 0.14117, + "grad_norm": 0.7448130249977112, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 14117 + }, + { + "epoch": 0.14118, + "grad_norm": 0.7307347655296326, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 14118 + }, + { + "epoch": 0.14119, + "grad_norm": 0.7131176590919495, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 14119 + }, + { + "epoch": 0.1412, + "grad_norm": 0.7631412744522095, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 14120 + }, + { + "epoch": 0.14121, + "grad_norm": 0.9671663641929626, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 14121 + }, + { + "epoch": 0.14122, + "grad_norm": 1.2928788661956787, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 14122 + }, + { + "epoch": 0.14123, + "grad_norm": 0.661189615726471, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 14123 + }, + { + "epoch": 0.14124, + "grad_norm": 0.6652116179466248, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 14124 + }, + { + "epoch": 0.14125, + "grad_norm": 0.8989524245262146, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 14125 + }, + { + "epoch": 0.14126, + "grad_norm": 1.0922940969467163, + "learning_rate": 0.003, + "loss": 3.97, + "step": 14126 + }, + { + "epoch": 0.14127, + "grad_norm": 1.0371108055114746, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 14127 + }, + { + "epoch": 0.14128, + "grad_norm": 1.0985851287841797, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 14128 + }, + { + "epoch": 0.14129, + "grad_norm": 0.8594481945037842, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 14129 + }, + { + "epoch": 0.1413, + "grad_norm": 0.8762556910514832, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 14130 + }, + { + "epoch": 0.14131, + "grad_norm": 1.2780417203903198, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 14131 + }, + { + "epoch": 0.14132, + "grad_norm": 0.8962858319282532, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 14132 + }, + { + "epoch": 0.14133, + "grad_norm": 0.6861137747764587, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 14133 + }, + { + "epoch": 0.14134, + "grad_norm": 0.637040913105011, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 14134 + }, + { + "epoch": 0.14135, + "grad_norm": 0.6423055529594421, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 14135 + }, + { + "epoch": 0.14136, + "grad_norm": 0.619636058807373, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 14136 + }, + { + "epoch": 0.14137, + "grad_norm": 0.5906887650489807, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 14137 + }, + { + "epoch": 0.14138, + "grad_norm": 0.5638257265090942, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 14138 + }, + { + "epoch": 0.14139, + "grad_norm": 0.5778006315231323, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 14139 + }, + { + "epoch": 0.1414, + "grad_norm": 0.5948111414909363, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 14140 + }, + { + "epoch": 0.14141, + "grad_norm": 0.6248180866241455, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 14141 + }, + { + "epoch": 0.14142, + "grad_norm": 0.7527813911437988, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 14142 + }, + { + "epoch": 0.14143, + "grad_norm": 0.7687675952911377, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 14143 + }, + { + "epoch": 0.14144, + "grad_norm": 0.7792863249778748, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 14144 + }, + { + "epoch": 0.14145, + "grad_norm": 0.9133181571960449, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 14145 + }, + { + "epoch": 0.14146, + "grad_norm": 1.0635042190551758, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 14146 + }, + { + "epoch": 0.14147, + "grad_norm": 0.9550064206123352, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 14147 + }, + { + "epoch": 0.14148, + "grad_norm": 0.9956814646720886, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 14148 + }, + { + "epoch": 0.14149, + "grad_norm": 0.9905039072036743, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 14149 + }, + { + "epoch": 0.1415, + "grad_norm": 0.981390118598938, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 14150 + }, + { + "epoch": 0.14151, + "grad_norm": 0.8436192274093628, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 14151 + }, + { + "epoch": 0.14152, + "grad_norm": 0.8707951307296753, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 14152 + }, + { + "epoch": 0.14153, + "grad_norm": 0.9241832494735718, + "learning_rate": 0.003, + "loss": 3.992, + "step": 14153 + }, + { + "epoch": 0.14154, + "grad_norm": 0.9518395066261292, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 14154 + }, + { + "epoch": 0.14155, + "grad_norm": 1.0943546295166016, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 14155 + }, + { + "epoch": 0.14156, + "grad_norm": 0.8065584897994995, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 14156 + }, + { + "epoch": 0.14157, + "grad_norm": 0.7145382165908813, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 14157 + }, + { + "epoch": 0.14158, + "grad_norm": 0.8641098141670227, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 14158 + }, + { + "epoch": 0.14159, + "grad_norm": 0.7735960483551025, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 14159 + }, + { + "epoch": 0.1416, + "grad_norm": 0.7756565809249878, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 14160 + }, + { + "epoch": 0.14161, + "grad_norm": 0.7883841395378113, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 14161 + }, + { + "epoch": 0.14162, + "grad_norm": 0.7832257747650146, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 14162 + }, + { + "epoch": 0.14163, + "grad_norm": 0.8378828167915344, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 14163 + }, + { + "epoch": 0.14164, + "grad_norm": 0.8510462641716003, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 14164 + }, + { + "epoch": 0.14165, + "grad_norm": 0.9078649878501892, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 14165 + }, + { + "epoch": 0.14166, + "grad_norm": 0.8352487683296204, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 14166 + }, + { + "epoch": 0.14167, + "grad_norm": 0.7029122710227966, + "learning_rate": 0.003, + "loss": 3.991, + "step": 14167 + }, + { + "epoch": 0.14168, + "grad_norm": 0.7005764842033386, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 14168 + }, + { + "epoch": 0.14169, + "grad_norm": 0.699022650718689, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 14169 + }, + { + "epoch": 0.1417, + "grad_norm": 0.6358689069747925, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 14170 + }, + { + "epoch": 0.14171, + "grad_norm": 0.8033818006515503, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 14171 + }, + { + "epoch": 0.14172, + "grad_norm": 1.1550252437591553, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 14172 + }, + { + "epoch": 0.14173, + "grad_norm": 1.2971549034118652, + "learning_rate": 0.003, + "loss": 4.03, + "step": 14173 + }, + { + "epoch": 0.14174, + "grad_norm": 0.7195215821266174, + "learning_rate": 0.003, + "loss": 4.018, + "step": 14174 + }, + { + "epoch": 0.14175, + "grad_norm": 0.5937075614929199, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 14175 + }, + { + "epoch": 0.14176, + "grad_norm": 0.7166873216629028, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 14176 + }, + { + "epoch": 0.14177, + "grad_norm": 0.7563587427139282, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 14177 + }, + { + "epoch": 0.14178, + "grad_norm": 0.9330248832702637, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 14178 + }, + { + "epoch": 0.14179, + "grad_norm": 0.9753488302230835, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 14179 + }, + { + "epoch": 0.1418, + "grad_norm": 0.8488276600837708, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 14180 + }, + { + "epoch": 0.14181, + "grad_norm": 0.6426655650138855, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 14181 + }, + { + "epoch": 0.14182, + "grad_norm": 0.6621089577674866, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 14182 + }, + { + "epoch": 0.14183, + "grad_norm": 0.8314296007156372, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 14183 + }, + { + "epoch": 0.14184, + "grad_norm": 1.0920621156692505, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 14184 + }, + { + "epoch": 0.14185, + "grad_norm": 1.1957831382751465, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 14185 + }, + { + "epoch": 0.14186, + "grad_norm": 0.726463794708252, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 14186 + }, + { + "epoch": 0.14187, + "grad_norm": 0.7537030577659607, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 14187 + }, + { + "epoch": 0.14188, + "grad_norm": 0.7797044515609741, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 14188 + }, + { + "epoch": 0.14189, + "grad_norm": 0.7619060277938843, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 14189 + }, + { + "epoch": 0.1419, + "grad_norm": 0.7681478261947632, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 14190 + }, + { + "epoch": 0.14191, + "grad_norm": 0.737553060054779, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 14191 + }, + { + "epoch": 0.14192, + "grad_norm": 0.5535001754760742, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 14192 + }, + { + "epoch": 0.14193, + "grad_norm": 0.6274293661117554, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 14193 + }, + { + "epoch": 0.14194, + "grad_norm": 0.640678882598877, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 14194 + }, + { + "epoch": 0.14195, + "grad_norm": 0.6661068201065063, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 14195 + }, + { + "epoch": 0.14196, + "grad_norm": 0.7455634474754333, + "learning_rate": 0.003, + "loss": 4.011, + "step": 14196 + }, + { + "epoch": 0.14197, + "grad_norm": 0.8102667331695557, + "learning_rate": 0.003, + "loss": 4.019, + "step": 14197 + }, + { + "epoch": 0.14198, + "grad_norm": 0.8579203486442566, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 14198 + }, + { + "epoch": 0.14199, + "grad_norm": 1.149722695350647, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 14199 + }, + { + "epoch": 0.142, + "grad_norm": 1.1549619436264038, + "learning_rate": 0.003, + "loss": 4.028, + "step": 14200 + }, + { + "epoch": 0.14201, + "grad_norm": 0.8290915489196777, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 14201 + }, + { + "epoch": 0.14202, + "grad_norm": 0.706985354423523, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 14202 + }, + { + "epoch": 0.14203, + "grad_norm": 0.7158949971199036, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 14203 + }, + { + "epoch": 0.14204, + "grad_norm": 0.7190188765525818, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 14204 + }, + { + "epoch": 0.14205, + "grad_norm": 0.7959105968475342, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 14205 + }, + { + "epoch": 0.14206, + "grad_norm": 1.0919005870819092, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 14206 + }, + { + "epoch": 0.14207, + "grad_norm": 1.1565816402435303, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 14207 + }, + { + "epoch": 0.14208, + "grad_norm": 0.8859974145889282, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 14208 + }, + { + "epoch": 0.14209, + "grad_norm": 0.8769922852516174, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 14209 + }, + { + "epoch": 0.1421, + "grad_norm": 0.7491868138313293, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 14210 + }, + { + "epoch": 0.14211, + "grad_norm": 0.7215003371238708, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 14211 + }, + { + "epoch": 0.14212, + "grad_norm": 0.7080946564674377, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 14212 + }, + { + "epoch": 0.14213, + "grad_norm": 0.7395340800285339, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 14213 + }, + { + "epoch": 0.14214, + "grad_norm": 0.7478938698768616, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 14214 + }, + { + "epoch": 0.14215, + "grad_norm": 0.6703423261642456, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 14215 + }, + { + "epoch": 0.14216, + "grad_norm": 0.7728683352470398, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 14216 + }, + { + "epoch": 0.14217, + "grad_norm": 0.8863434791564941, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 14217 + }, + { + "epoch": 0.14218, + "grad_norm": 0.9171063899993896, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 14218 + }, + { + "epoch": 0.14219, + "grad_norm": 0.8939700126647949, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 14219 + }, + { + "epoch": 0.1422, + "grad_norm": 0.9751589298248291, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 14220 + }, + { + "epoch": 0.14221, + "grad_norm": 1.1756740808486938, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 14221 + }, + { + "epoch": 0.14222, + "grad_norm": 0.7842356562614441, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 14222 + }, + { + "epoch": 0.14223, + "grad_norm": 0.8838635087013245, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 14223 + }, + { + "epoch": 0.14224, + "grad_norm": 1.165834665298462, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 14224 + }, + { + "epoch": 0.14225, + "grad_norm": 1.0083396434783936, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 14225 + }, + { + "epoch": 0.14226, + "grad_norm": 1.2527788877487183, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 14226 + }, + { + "epoch": 0.14227, + "grad_norm": 0.6788728833198547, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 14227 + }, + { + "epoch": 0.14228, + "grad_norm": 0.6345340013504028, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 14228 + }, + { + "epoch": 0.14229, + "grad_norm": 0.7083902955055237, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 14229 + }, + { + "epoch": 0.1423, + "grad_norm": 0.6958598494529724, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 14230 + }, + { + "epoch": 0.14231, + "grad_norm": 0.6795510649681091, + "learning_rate": 0.003, + "loss": 4.031, + "step": 14231 + }, + { + "epoch": 0.14232, + "grad_norm": 0.5360814332962036, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 14232 + }, + { + "epoch": 0.14233, + "grad_norm": 0.5661422610282898, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 14233 + }, + { + "epoch": 0.14234, + "grad_norm": 0.5672509670257568, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 14234 + }, + { + "epoch": 0.14235, + "grad_norm": 0.6105342507362366, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 14235 + }, + { + "epoch": 0.14236, + "grad_norm": 0.6176836490631104, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 14236 + }, + { + "epoch": 0.14237, + "grad_norm": 0.6107972860336304, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 14237 + }, + { + "epoch": 0.14238, + "grad_norm": 0.7120127081871033, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 14238 + }, + { + "epoch": 0.14239, + "grad_norm": 0.9503802061080933, + "learning_rate": 0.003, + "loss": 4.043, + "step": 14239 + }, + { + "epoch": 0.1424, + "grad_norm": 1.435011863708496, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 14240 + }, + { + "epoch": 0.14241, + "grad_norm": 0.6943506598472595, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 14241 + }, + { + "epoch": 0.14242, + "grad_norm": 0.7207216620445251, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 14242 + }, + { + "epoch": 0.14243, + "grad_norm": 0.8685964941978455, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 14243 + }, + { + "epoch": 0.14244, + "grad_norm": 0.8263289332389832, + "learning_rate": 0.003, + "loss": 4.01, + "step": 14244 + }, + { + "epoch": 0.14245, + "grad_norm": 0.8320058584213257, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 14245 + }, + { + "epoch": 0.14246, + "grad_norm": 0.8826883435249329, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 14246 + }, + { + "epoch": 0.14247, + "grad_norm": 0.9846150279045105, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 14247 + }, + { + "epoch": 0.14248, + "grad_norm": 0.9439970850944519, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 14248 + }, + { + "epoch": 0.14249, + "grad_norm": 0.7696685194969177, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 14249 + }, + { + "epoch": 0.1425, + "grad_norm": 0.631126344203949, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 14250 + }, + { + "epoch": 0.14251, + "grad_norm": 0.7179323434829712, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 14251 + }, + { + "epoch": 0.14252, + "grad_norm": 0.8308407068252563, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 14252 + }, + { + "epoch": 0.14253, + "grad_norm": 0.8950212597846985, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 14253 + }, + { + "epoch": 0.14254, + "grad_norm": 1.186442494392395, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 14254 + }, + { + "epoch": 0.14255, + "grad_norm": 1.0353965759277344, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 14255 + }, + { + "epoch": 0.14256, + "grad_norm": 1.0124056339263916, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 14256 + }, + { + "epoch": 0.14257, + "grad_norm": 0.9364923238754272, + "learning_rate": 0.003, + "loss": 4.018, + "step": 14257 + }, + { + "epoch": 0.14258, + "grad_norm": 0.7490666508674622, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 14258 + }, + { + "epoch": 0.14259, + "grad_norm": 0.6310158967971802, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 14259 + }, + { + "epoch": 0.1426, + "grad_norm": 0.6345019936561584, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 14260 + }, + { + "epoch": 0.14261, + "grad_norm": 0.6502352952957153, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 14261 + }, + { + "epoch": 0.14262, + "grad_norm": 0.7433469295501709, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 14262 + }, + { + "epoch": 0.14263, + "grad_norm": 0.9879776835441589, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 14263 + }, + { + "epoch": 0.14264, + "grad_norm": 1.222913384437561, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 14264 + }, + { + "epoch": 0.14265, + "grad_norm": 0.7845995426177979, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 14265 + }, + { + "epoch": 0.14266, + "grad_norm": 0.7524864077568054, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 14266 + }, + { + "epoch": 0.14267, + "grad_norm": 0.757966160774231, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 14267 + }, + { + "epoch": 0.14268, + "grad_norm": 0.7970731258392334, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 14268 + }, + { + "epoch": 0.14269, + "grad_norm": 0.7954204082489014, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 14269 + }, + { + "epoch": 0.1427, + "grad_norm": 0.864578366279602, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 14270 + }, + { + "epoch": 0.14271, + "grad_norm": 0.9321130514144897, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 14271 + }, + { + "epoch": 0.14272, + "grad_norm": 0.9318530559539795, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 14272 + }, + { + "epoch": 0.14273, + "grad_norm": 0.9119703769683838, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 14273 + }, + { + "epoch": 0.14274, + "grad_norm": 0.9916678071022034, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 14274 + }, + { + "epoch": 0.14275, + "grad_norm": 0.8041375875473022, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 14275 + }, + { + "epoch": 0.14276, + "grad_norm": 0.6699801683425903, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 14276 + }, + { + "epoch": 0.14277, + "grad_norm": 0.7691702842712402, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 14277 + }, + { + "epoch": 0.14278, + "grad_norm": 0.9321637749671936, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 14278 + }, + { + "epoch": 0.14279, + "grad_norm": 1.0950120687484741, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 14279 + }, + { + "epoch": 0.1428, + "grad_norm": 0.8870145678520203, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 14280 + }, + { + "epoch": 0.14281, + "grad_norm": 0.7608538866043091, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 14281 + }, + { + "epoch": 0.14282, + "grad_norm": 0.7831116914749146, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 14282 + }, + { + "epoch": 0.14283, + "grad_norm": 0.7721794247627258, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 14283 + }, + { + "epoch": 0.14284, + "grad_norm": 0.7339931726455688, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 14284 + }, + { + "epoch": 0.14285, + "grad_norm": 0.7375816702842712, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 14285 + }, + { + "epoch": 0.14286, + "grad_norm": 0.7514878511428833, + "learning_rate": 0.003, + "loss": 4.003, + "step": 14286 + }, + { + "epoch": 0.14287, + "grad_norm": 0.7763694524765015, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 14287 + }, + { + "epoch": 0.14288, + "grad_norm": 0.7558686137199402, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 14288 + }, + { + "epoch": 0.14289, + "grad_norm": 0.6009389162063599, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 14289 + }, + { + "epoch": 0.1429, + "grad_norm": 0.6581655144691467, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 14290 + }, + { + "epoch": 0.14291, + "grad_norm": 0.9600313901901245, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 14291 + }, + { + "epoch": 0.14292, + "grad_norm": 1.3686339855194092, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 14292 + }, + { + "epoch": 0.14293, + "grad_norm": 0.6834939122200012, + "learning_rate": 0.003, + "loss": 3.986, + "step": 14293 + }, + { + "epoch": 0.14294, + "grad_norm": 0.7326520681381226, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 14294 + }, + { + "epoch": 0.14295, + "grad_norm": 0.863531231880188, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 14295 + }, + { + "epoch": 0.14296, + "grad_norm": 0.99330735206604, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 14296 + }, + { + "epoch": 0.14297, + "grad_norm": 1.054336667060852, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 14297 + }, + { + "epoch": 0.14298, + "grad_norm": 1.0437896251678467, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 14298 + }, + { + "epoch": 0.14299, + "grad_norm": 0.9987184405326843, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 14299 + }, + { + "epoch": 0.143, + "grad_norm": 1.0736325979232788, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 14300 + }, + { + "epoch": 0.14301, + "grad_norm": 1.1124069690704346, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 14301 + }, + { + "epoch": 0.14302, + "grad_norm": 1.0295443534851074, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 14302 + }, + { + "epoch": 0.14303, + "grad_norm": 1.1448150873184204, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 14303 + }, + { + "epoch": 0.14304, + "grad_norm": 0.9082779884338379, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 14304 + }, + { + "epoch": 0.14305, + "grad_norm": 1.0699198246002197, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 14305 + }, + { + "epoch": 0.14306, + "grad_norm": 0.8107674717903137, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 14306 + }, + { + "epoch": 0.14307, + "grad_norm": 0.7893140316009521, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 14307 + }, + { + "epoch": 0.14308, + "grad_norm": 0.9483770132064819, + "learning_rate": 0.003, + "loss": 4.048, + "step": 14308 + }, + { + "epoch": 0.14309, + "grad_norm": 1.1574307680130005, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 14309 + }, + { + "epoch": 0.1431, + "grad_norm": 0.8233171105384827, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 14310 + }, + { + "epoch": 0.14311, + "grad_norm": 0.7954649925231934, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 14311 + }, + { + "epoch": 0.14312, + "grad_norm": 0.7902868390083313, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 14312 + }, + { + "epoch": 0.14313, + "grad_norm": 0.8275642395019531, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 14313 + }, + { + "epoch": 0.14314, + "grad_norm": 0.9443146586418152, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 14314 + }, + { + "epoch": 0.14315, + "grad_norm": 0.9062085747718811, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 14315 + }, + { + "epoch": 0.14316, + "grad_norm": 0.8800166249275208, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 14316 + }, + { + "epoch": 0.14317, + "grad_norm": 0.7463732361793518, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 14317 + }, + { + "epoch": 0.14318, + "grad_norm": 0.6335029602050781, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 14318 + }, + { + "epoch": 0.14319, + "grad_norm": 0.6070913672447205, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 14319 + }, + { + "epoch": 0.1432, + "grad_norm": 0.6634745597839355, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 14320 + }, + { + "epoch": 0.14321, + "grad_norm": 0.5926946997642517, + "learning_rate": 0.003, + "loss": 3.987, + "step": 14321 + }, + { + "epoch": 0.14322, + "grad_norm": 0.5566388964653015, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 14322 + }, + { + "epoch": 0.14323, + "grad_norm": 0.5370780825614929, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 14323 + }, + { + "epoch": 0.14324, + "grad_norm": 0.5878376364707947, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 14324 + }, + { + "epoch": 0.14325, + "grad_norm": 0.6978987455368042, + "learning_rate": 0.003, + "loss": 3.98, + "step": 14325 + }, + { + "epoch": 0.14326, + "grad_norm": 0.8885076642036438, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 14326 + }, + { + "epoch": 0.14327, + "grad_norm": 1.2482155561447144, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 14327 + }, + { + "epoch": 0.14328, + "grad_norm": 0.8148013949394226, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 14328 + }, + { + "epoch": 0.14329, + "grad_norm": 0.5906567573547363, + "learning_rate": 0.003, + "loss": 3.995, + "step": 14329 + }, + { + "epoch": 0.1433, + "grad_norm": 0.5495211482048035, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 14330 + }, + { + "epoch": 0.14331, + "grad_norm": 0.5542047023773193, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 14331 + }, + { + "epoch": 0.14332, + "grad_norm": 0.6755203008651733, + "learning_rate": 0.003, + "loss": 4.001, + "step": 14332 + }, + { + "epoch": 0.14333, + "grad_norm": 0.8951558470726013, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 14333 + }, + { + "epoch": 0.14334, + "grad_norm": 0.9805142283439636, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 14334 + }, + { + "epoch": 0.14335, + "grad_norm": 0.8546414375305176, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 14335 + }, + { + "epoch": 0.14336, + "grad_norm": 0.6646180748939514, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 14336 + }, + { + "epoch": 0.14337, + "grad_norm": 0.6930346488952637, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 14337 + }, + { + "epoch": 0.14338, + "grad_norm": 0.8854844570159912, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 14338 + }, + { + "epoch": 0.14339, + "grad_norm": 1.0480695962905884, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 14339 + }, + { + "epoch": 0.1434, + "grad_norm": 0.8314961791038513, + "learning_rate": 0.003, + "loss": 3.993, + "step": 14340 + }, + { + "epoch": 0.14341, + "grad_norm": 0.6699071526527405, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 14341 + }, + { + "epoch": 0.14342, + "grad_norm": 0.6851991415023804, + "learning_rate": 0.003, + "loss": 4.005, + "step": 14342 + }, + { + "epoch": 0.14343, + "grad_norm": 0.8025005459785461, + "learning_rate": 0.003, + "loss": 4.048, + "step": 14343 + }, + { + "epoch": 0.14344, + "grad_norm": 0.7941917181015015, + "learning_rate": 0.003, + "loss": 4.052, + "step": 14344 + }, + { + "epoch": 0.14345, + "grad_norm": 0.8650415539741516, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 14345 + }, + { + "epoch": 0.14346, + "grad_norm": 1.114258050918579, + "learning_rate": 0.003, + "loss": 4.025, + "step": 14346 + }, + { + "epoch": 0.14347, + "grad_norm": 0.9780334234237671, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 14347 + }, + { + "epoch": 0.14348, + "grad_norm": 0.8963751196861267, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 14348 + }, + { + "epoch": 0.14349, + "grad_norm": 1.0181679725646973, + "learning_rate": 0.003, + "loss": 4.012, + "step": 14349 + }, + { + "epoch": 0.1435, + "grad_norm": 0.9988883137702942, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 14350 + }, + { + "epoch": 0.14351, + "grad_norm": 0.9089468717575073, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 14351 + }, + { + "epoch": 0.14352, + "grad_norm": 0.77008455991745, + "learning_rate": 0.003, + "loss": 4.018, + "step": 14352 + }, + { + "epoch": 0.14353, + "grad_norm": 0.8111112117767334, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 14353 + }, + { + "epoch": 0.14354, + "grad_norm": 0.9026668667793274, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 14354 + }, + { + "epoch": 0.14355, + "grad_norm": 1.0534589290618896, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 14355 + }, + { + "epoch": 0.14356, + "grad_norm": 1.1436132192611694, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 14356 + }, + { + "epoch": 0.14357, + "grad_norm": 0.9831972718238831, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 14357 + }, + { + "epoch": 0.14358, + "grad_norm": 1.1651567220687866, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 14358 + }, + { + "epoch": 0.14359, + "grad_norm": 0.7918037176132202, + "learning_rate": 0.003, + "loss": 4.027, + "step": 14359 + }, + { + "epoch": 0.1436, + "grad_norm": 0.6700632572174072, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 14360 + }, + { + "epoch": 0.14361, + "grad_norm": 0.5784865021705627, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 14361 + }, + { + "epoch": 0.14362, + "grad_norm": 0.5571222901344299, + "learning_rate": 0.003, + "loss": 4.051, + "step": 14362 + }, + { + "epoch": 0.14363, + "grad_norm": 0.5615730285644531, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 14363 + }, + { + "epoch": 0.14364, + "grad_norm": 0.5923025012016296, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 14364 + }, + { + "epoch": 0.14365, + "grad_norm": 0.6931462287902832, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 14365 + }, + { + "epoch": 0.14366, + "grad_norm": 0.718263566493988, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 14366 + }, + { + "epoch": 0.14367, + "grad_norm": 0.8923155665397644, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 14367 + }, + { + "epoch": 0.14368, + "grad_norm": 1.072655439376831, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 14368 + }, + { + "epoch": 0.14369, + "grad_norm": 0.9154521822929382, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 14369 + }, + { + "epoch": 0.1437, + "grad_norm": 0.7606717348098755, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 14370 + }, + { + "epoch": 0.14371, + "grad_norm": 0.6659698486328125, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 14371 + }, + { + "epoch": 0.14372, + "grad_norm": 0.770995557308197, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 14372 + }, + { + "epoch": 0.14373, + "grad_norm": 0.865330696105957, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 14373 + }, + { + "epoch": 0.14374, + "grad_norm": 0.9684982895851135, + "learning_rate": 0.003, + "loss": 4.017, + "step": 14374 + }, + { + "epoch": 0.14375, + "grad_norm": 0.9319028258323669, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 14375 + }, + { + "epoch": 0.14376, + "grad_norm": 0.7594645619392395, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 14376 + }, + { + "epoch": 0.14377, + "grad_norm": 0.7884442210197449, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 14377 + }, + { + "epoch": 0.14378, + "grad_norm": 0.8902500867843628, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 14378 + }, + { + "epoch": 0.14379, + "grad_norm": 1.033056378364563, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 14379 + }, + { + "epoch": 0.1438, + "grad_norm": 0.8748496174812317, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 14380 + }, + { + "epoch": 0.14381, + "grad_norm": 0.9056290984153748, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 14381 + }, + { + "epoch": 0.14382, + "grad_norm": 1.0608599185943604, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 14382 + }, + { + "epoch": 0.14383, + "grad_norm": 1.0192004442214966, + "learning_rate": 0.003, + "loss": 4.024, + "step": 14383 + }, + { + "epoch": 0.14384, + "grad_norm": 0.8027071952819824, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 14384 + }, + { + "epoch": 0.14385, + "grad_norm": 0.7244923710823059, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 14385 + }, + { + "epoch": 0.14386, + "grad_norm": 0.7831423878669739, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 14386 + }, + { + "epoch": 0.14387, + "grad_norm": 0.7867714762687683, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 14387 + }, + { + "epoch": 0.14388, + "grad_norm": 0.917858362197876, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14388 + }, + { + "epoch": 0.14389, + "grad_norm": 1.0930110216140747, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 14389 + }, + { + "epoch": 0.1439, + "grad_norm": 0.9966136813163757, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 14390 + }, + { + "epoch": 0.14391, + "grad_norm": 0.8886557817459106, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 14391 + }, + { + "epoch": 0.14392, + "grad_norm": 0.7261216044425964, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 14392 + }, + { + "epoch": 0.14393, + "grad_norm": 0.6868066787719727, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 14393 + }, + { + "epoch": 0.14394, + "grad_norm": 0.6649624705314636, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 14394 + }, + { + "epoch": 0.14395, + "grad_norm": 0.7228597402572632, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 14395 + }, + { + "epoch": 0.14396, + "grad_norm": 0.7414789795875549, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 14396 + }, + { + "epoch": 0.14397, + "grad_norm": 0.7033632397651672, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 14397 + }, + { + "epoch": 0.14398, + "grad_norm": 0.6831393241882324, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 14398 + }, + { + "epoch": 0.14399, + "grad_norm": 0.6953455209732056, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 14399 + }, + { + "epoch": 0.144, + "grad_norm": 0.6803028583526611, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 14400 + }, + { + "epoch": 0.14401, + "grad_norm": 0.7951777577400208, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 14401 + }, + { + "epoch": 0.14402, + "grad_norm": 0.8658850789070129, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 14402 + }, + { + "epoch": 0.14403, + "grad_norm": 0.9823598861694336, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 14403 + }, + { + "epoch": 0.14404, + "grad_norm": 1.1010268926620483, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 14404 + }, + { + "epoch": 0.14405, + "grad_norm": 0.9394732117652893, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 14405 + }, + { + "epoch": 0.14406, + "grad_norm": 0.840592622756958, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 14406 + }, + { + "epoch": 0.14407, + "grad_norm": 0.7738531231880188, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 14407 + }, + { + "epoch": 0.14408, + "grad_norm": 0.7063372731208801, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 14408 + }, + { + "epoch": 0.14409, + "grad_norm": 0.6962958574295044, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 14409 + }, + { + "epoch": 0.1441, + "grad_norm": 0.6501365303993225, + "learning_rate": 0.003, + "loss": 3.98, + "step": 14410 + }, + { + "epoch": 0.14411, + "grad_norm": 0.5795419216156006, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 14411 + }, + { + "epoch": 0.14412, + "grad_norm": 0.6202126145362854, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 14412 + }, + { + "epoch": 0.14413, + "grad_norm": 0.8158420920372009, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 14413 + }, + { + "epoch": 0.14414, + "grad_norm": 0.8209929466247559, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 14414 + }, + { + "epoch": 0.14415, + "grad_norm": 0.8057379126548767, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 14415 + }, + { + "epoch": 0.14416, + "grad_norm": 0.8939820528030396, + "learning_rate": 0.003, + "loss": 3.986, + "step": 14416 + }, + { + "epoch": 0.14417, + "grad_norm": 1.0734862089157104, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 14417 + }, + { + "epoch": 0.14418, + "grad_norm": 0.9874979257583618, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 14418 + }, + { + "epoch": 0.14419, + "grad_norm": 1.05538809299469, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 14419 + }, + { + "epoch": 0.1442, + "grad_norm": 0.8185986280441284, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 14420 + }, + { + "epoch": 0.14421, + "grad_norm": 0.6607561111450195, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 14421 + }, + { + "epoch": 0.14422, + "grad_norm": 0.5887096524238586, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 14422 + }, + { + "epoch": 0.14423, + "grad_norm": 0.5062503218650818, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 14423 + }, + { + "epoch": 0.14424, + "grad_norm": 0.5057479739189148, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 14424 + }, + { + "epoch": 0.14425, + "grad_norm": 0.5376363396644592, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 14425 + }, + { + "epoch": 0.14426, + "grad_norm": 0.647079586982727, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 14426 + }, + { + "epoch": 0.14427, + "grad_norm": 0.7828883528709412, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 14427 + }, + { + "epoch": 0.14428, + "grad_norm": 0.7896987795829773, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 14428 + }, + { + "epoch": 0.14429, + "grad_norm": 0.7318724989891052, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 14429 + }, + { + "epoch": 0.1443, + "grad_norm": 0.7879105806350708, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 14430 + }, + { + "epoch": 0.14431, + "grad_norm": 0.8891569972038269, + "learning_rate": 0.003, + "loss": 4.041, + "step": 14431 + }, + { + "epoch": 0.14432, + "grad_norm": 1.0650385618209839, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 14432 + }, + { + "epoch": 0.14433, + "grad_norm": 0.9432113170623779, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 14433 + }, + { + "epoch": 0.14434, + "grad_norm": 0.8969363570213318, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 14434 + }, + { + "epoch": 0.14435, + "grad_norm": 0.855061948299408, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 14435 + }, + { + "epoch": 0.14436, + "grad_norm": 0.7581377029418945, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 14436 + }, + { + "epoch": 0.14437, + "grad_norm": 0.7667735815048218, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 14437 + }, + { + "epoch": 0.14438, + "grad_norm": 0.8322454690933228, + "learning_rate": 0.003, + "loss": 4.032, + "step": 14438 + }, + { + "epoch": 0.14439, + "grad_norm": 0.914036214351654, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 14439 + }, + { + "epoch": 0.1444, + "grad_norm": 1.0237014293670654, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 14440 + }, + { + "epoch": 0.14441, + "grad_norm": 1.2672991752624512, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 14441 + }, + { + "epoch": 0.14442, + "grad_norm": 0.8671845197677612, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 14442 + }, + { + "epoch": 0.14443, + "grad_norm": 0.6996704339981079, + "learning_rate": 0.003, + "loss": 3.994, + "step": 14443 + }, + { + "epoch": 0.14444, + "grad_norm": 0.6526155471801758, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 14444 + }, + { + "epoch": 0.14445, + "grad_norm": 0.6436783075332642, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 14445 + }, + { + "epoch": 0.14446, + "grad_norm": 0.600752592086792, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 14446 + }, + { + "epoch": 0.14447, + "grad_norm": 0.6957166194915771, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 14447 + }, + { + "epoch": 0.14448, + "grad_norm": 0.9329416155815125, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 14448 + }, + { + "epoch": 0.14449, + "grad_norm": 1.3451831340789795, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 14449 + }, + { + "epoch": 0.1445, + "grad_norm": 0.9525917172431946, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 14450 + }, + { + "epoch": 0.14451, + "grad_norm": 0.9183449745178223, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 14451 + }, + { + "epoch": 0.14452, + "grad_norm": 0.8432508707046509, + "learning_rate": 0.003, + "loss": 4.009, + "step": 14452 + }, + { + "epoch": 0.14453, + "grad_norm": 0.7553740739822388, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 14453 + }, + { + "epoch": 0.14454, + "grad_norm": 0.7473167777061462, + "learning_rate": 0.003, + "loss": 4.003, + "step": 14454 + }, + { + "epoch": 0.14455, + "grad_norm": 0.7682895660400391, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 14455 + }, + { + "epoch": 0.14456, + "grad_norm": 0.7811801433563232, + "learning_rate": 0.003, + "loss": 3.981, + "step": 14456 + }, + { + "epoch": 0.14457, + "grad_norm": 0.6911787986755371, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 14457 + }, + { + "epoch": 0.14458, + "grad_norm": 0.6579571962356567, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 14458 + }, + { + "epoch": 0.14459, + "grad_norm": 0.7433202266693115, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 14459 + }, + { + "epoch": 0.1446, + "grad_norm": 0.8381054401397705, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 14460 + }, + { + "epoch": 0.14461, + "grad_norm": 1.0219905376434326, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 14461 + }, + { + "epoch": 0.14462, + "grad_norm": 1.1736960411071777, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 14462 + }, + { + "epoch": 0.14463, + "grad_norm": 0.8547998666763306, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 14463 + }, + { + "epoch": 0.14464, + "grad_norm": 0.7964674234390259, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 14464 + }, + { + "epoch": 0.14465, + "grad_norm": 0.7108013033866882, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 14465 + }, + { + "epoch": 0.14466, + "grad_norm": 0.7695088982582092, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 14466 + }, + { + "epoch": 0.14467, + "grad_norm": 0.7569155693054199, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 14467 + }, + { + "epoch": 0.14468, + "grad_norm": 0.7204357981681824, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 14468 + }, + { + "epoch": 0.14469, + "grad_norm": 0.7813732624053955, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 14469 + }, + { + "epoch": 0.1447, + "grad_norm": 0.8003991842269897, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 14470 + }, + { + "epoch": 0.14471, + "grad_norm": 0.8164728283882141, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 14471 + }, + { + "epoch": 0.14472, + "grad_norm": 0.696689248085022, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 14472 + }, + { + "epoch": 0.14473, + "grad_norm": 0.7590498328208923, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 14473 + }, + { + "epoch": 0.14474, + "grad_norm": 0.7519974708557129, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 14474 + }, + { + "epoch": 0.14475, + "grad_norm": 0.7030192613601685, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 14475 + }, + { + "epoch": 0.14476, + "grad_norm": 0.7600857019424438, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 14476 + }, + { + "epoch": 0.14477, + "grad_norm": 0.7709656953811646, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 14477 + }, + { + "epoch": 0.14478, + "grad_norm": 0.8543557524681091, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 14478 + }, + { + "epoch": 0.14479, + "grad_norm": 1.0360087156295776, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 14479 + }, + { + "epoch": 0.1448, + "grad_norm": 1.1634830236434937, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 14480 + }, + { + "epoch": 0.14481, + "grad_norm": 0.8620505332946777, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 14481 + }, + { + "epoch": 0.14482, + "grad_norm": 0.9569733738899231, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 14482 + }, + { + "epoch": 0.14483, + "grad_norm": 1.3167060613632202, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 14483 + }, + { + "epoch": 0.14484, + "grad_norm": 0.745167076587677, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 14484 + }, + { + "epoch": 0.14485, + "grad_norm": 0.7294960021972656, + "learning_rate": 0.003, + "loss": 4.002, + "step": 14485 + }, + { + "epoch": 0.14486, + "grad_norm": 0.7164823412895203, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 14486 + }, + { + "epoch": 0.14487, + "grad_norm": 0.752485990524292, + "learning_rate": 0.003, + "loss": 4.004, + "step": 14487 + }, + { + "epoch": 0.14488, + "grad_norm": 0.6875482201576233, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 14488 + }, + { + "epoch": 0.14489, + "grad_norm": 0.5566693544387817, + "learning_rate": 0.003, + "loss": 3.999, + "step": 14489 + }, + { + "epoch": 0.1449, + "grad_norm": 0.579632043838501, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 14490 + }, + { + "epoch": 0.14491, + "grad_norm": 0.6094452738761902, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 14491 + }, + { + "epoch": 0.14492, + "grad_norm": 0.6271584630012512, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 14492 + }, + { + "epoch": 0.14493, + "grad_norm": 0.5886869430541992, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 14493 + }, + { + "epoch": 0.14494, + "grad_norm": 0.7035474181175232, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 14494 + }, + { + "epoch": 0.14495, + "grad_norm": 0.880418062210083, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 14495 + }, + { + "epoch": 0.14496, + "grad_norm": 0.9390040040016174, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 14496 + }, + { + "epoch": 0.14497, + "grad_norm": 1.0518993139266968, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 14497 + }, + { + "epoch": 0.14498, + "grad_norm": 1.1795552968978882, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 14498 + }, + { + "epoch": 0.14499, + "grad_norm": 0.9480694532394409, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 14499 + }, + { + "epoch": 0.145, + "grad_norm": 0.9626632928848267, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 14500 + }, + { + "epoch": 0.14501, + "grad_norm": 0.9182538390159607, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 14501 + }, + { + "epoch": 0.14502, + "grad_norm": 0.7662861943244934, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 14502 + }, + { + "epoch": 0.14503, + "grad_norm": 0.6444974541664124, + "learning_rate": 0.003, + "loss": 4.033, + "step": 14503 + }, + { + "epoch": 0.14504, + "grad_norm": 0.6466870903968811, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 14504 + }, + { + "epoch": 0.14505, + "grad_norm": 0.6860358715057373, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 14505 + }, + { + "epoch": 0.14506, + "grad_norm": 0.777827262878418, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 14506 + }, + { + "epoch": 0.14507, + "grad_norm": 0.8074339032173157, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 14507 + }, + { + "epoch": 0.14508, + "grad_norm": 0.7027100920677185, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 14508 + }, + { + "epoch": 0.14509, + "grad_norm": 0.690327525138855, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 14509 + }, + { + "epoch": 0.1451, + "grad_norm": 0.8751059770584106, + "learning_rate": 0.003, + "loss": 4.027, + "step": 14510 + }, + { + "epoch": 0.14511, + "grad_norm": 1.0037294626235962, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 14511 + }, + { + "epoch": 0.14512, + "grad_norm": 1.1385389566421509, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 14512 + }, + { + "epoch": 0.14513, + "grad_norm": 0.8328064680099487, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 14513 + }, + { + "epoch": 0.14514, + "grad_norm": 0.750587522983551, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 14514 + }, + { + "epoch": 0.14515, + "grad_norm": 0.780562162399292, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 14515 + }, + { + "epoch": 0.14516, + "grad_norm": 0.8911523222923279, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 14516 + }, + { + "epoch": 0.14517, + "grad_norm": 0.8570557832717896, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 14517 + }, + { + "epoch": 0.14518, + "grad_norm": 0.8679510354995728, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 14518 + }, + { + "epoch": 0.14519, + "grad_norm": 0.9858636856079102, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 14519 + }, + { + "epoch": 0.1452, + "grad_norm": 1.074411392211914, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 14520 + }, + { + "epoch": 0.14521, + "grad_norm": 0.8335927128791809, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 14521 + }, + { + "epoch": 0.14522, + "grad_norm": 0.780289888381958, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 14522 + }, + { + "epoch": 0.14523, + "grad_norm": 0.843913197517395, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 14523 + }, + { + "epoch": 0.14524, + "grad_norm": 0.8714582920074463, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 14524 + }, + { + "epoch": 0.14525, + "grad_norm": 0.8275744318962097, + "learning_rate": 0.003, + "loss": 3.994, + "step": 14525 + }, + { + "epoch": 0.14526, + "grad_norm": 0.9582111239433289, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 14526 + }, + { + "epoch": 0.14527, + "grad_norm": 0.9758346676826477, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 14527 + }, + { + "epoch": 0.14528, + "grad_norm": 1.0356850624084473, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 14528 + }, + { + "epoch": 0.14529, + "grad_norm": 0.9981045722961426, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 14529 + }, + { + "epoch": 0.1453, + "grad_norm": 0.9965315461158752, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 14530 + }, + { + "epoch": 0.14531, + "grad_norm": 0.9737066626548767, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 14531 + }, + { + "epoch": 0.14532, + "grad_norm": 0.8926593065261841, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 14532 + }, + { + "epoch": 0.14533, + "grad_norm": 0.7353343367576599, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 14533 + }, + { + "epoch": 0.14534, + "grad_norm": 0.817808985710144, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 14534 + }, + { + "epoch": 0.14535, + "grad_norm": 0.8736345767974854, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 14535 + }, + { + "epoch": 0.14536, + "grad_norm": 0.9836347103118896, + "learning_rate": 0.003, + "loss": 4.034, + "step": 14536 + }, + { + "epoch": 0.14537, + "grad_norm": 1.0643627643585205, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 14537 + }, + { + "epoch": 0.14538, + "grad_norm": 1.0713415145874023, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 14538 + }, + { + "epoch": 0.14539, + "grad_norm": 0.8724493980407715, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 14539 + }, + { + "epoch": 0.1454, + "grad_norm": 0.7486632466316223, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 14540 + }, + { + "epoch": 0.14541, + "grad_norm": 0.7692328095436096, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 14541 + }, + { + "epoch": 0.14542, + "grad_norm": 0.7887241840362549, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 14542 + }, + { + "epoch": 0.14543, + "grad_norm": 0.7101812362670898, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 14543 + }, + { + "epoch": 0.14544, + "grad_norm": 0.6259065270423889, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 14544 + }, + { + "epoch": 0.14545, + "grad_norm": 0.6179724931716919, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 14545 + }, + { + "epoch": 0.14546, + "grad_norm": 0.6964648962020874, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 14546 + }, + { + "epoch": 0.14547, + "grad_norm": 0.736311674118042, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 14547 + }, + { + "epoch": 0.14548, + "grad_norm": 0.7651875615119934, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 14548 + }, + { + "epoch": 0.14549, + "grad_norm": 0.6706742644309998, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 14549 + }, + { + "epoch": 0.1455, + "grad_norm": 0.6436061859130859, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 14550 + }, + { + "epoch": 0.14551, + "grad_norm": 0.8034910559654236, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 14551 + }, + { + "epoch": 0.14552, + "grad_norm": 1.0899924039840698, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 14552 + }, + { + "epoch": 0.14553, + "grad_norm": 1.0492063760757446, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 14553 + }, + { + "epoch": 0.14554, + "grad_norm": 0.9474644064903259, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 14554 + }, + { + "epoch": 0.14555, + "grad_norm": 0.8426516652107239, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 14555 + }, + { + "epoch": 0.14556, + "grad_norm": 0.7306226491928101, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 14556 + }, + { + "epoch": 0.14557, + "grad_norm": 0.8061223030090332, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 14557 + }, + { + "epoch": 0.14558, + "grad_norm": 1.023594617843628, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 14558 + }, + { + "epoch": 0.14559, + "grad_norm": 1.2255849838256836, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 14559 + }, + { + "epoch": 0.1456, + "grad_norm": 0.7095961570739746, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 14560 + }, + { + "epoch": 0.14561, + "grad_norm": 0.6970982551574707, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 14561 + }, + { + "epoch": 0.14562, + "grad_norm": 0.830156683921814, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 14562 + }, + { + "epoch": 0.14563, + "grad_norm": 0.7684354186058044, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 14563 + }, + { + "epoch": 0.14564, + "grad_norm": 0.8628973364830017, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 14564 + }, + { + "epoch": 0.14565, + "grad_norm": 0.9758632183074951, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 14565 + }, + { + "epoch": 0.14566, + "grad_norm": 0.9080876708030701, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 14566 + }, + { + "epoch": 0.14567, + "grad_norm": 0.8077095746994019, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 14567 + }, + { + "epoch": 0.14568, + "grad_norm": 0.7247617244720459, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 14568 + }, + { + "epoch": 0.14569, + "grad_norm": 0.6292919516563416, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 14569 + }, + { + "epoch": 0.1457, + "grad_norm": 0.7047914266586304, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 14570 + }, + { + "epoch": 0.14571, + "grad_norm": 0.8658979535102844, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 14571 + }, + { + "epoch": 0.14572, + "grad_norm": 0.9502711296081543, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 14572 + }, + { + "epoch": 0.14573, + "grad_norm": 1.1412993669509888, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 14573 + }, + { + "epoch": 0.14574, + "grad_norm": 0.8256980776786804, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 14574 + }, + { + "epoch": 0.14575, + "grad_norm": 0.745071530342102, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 14575 + }, + { + "epoch": 0.14576, + "grad_norm": 0.6846836805343628, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 14576 + }, + { + "epoch": 0.14577, + "grad_norm": 0.6639923453330994, + "learning_rate": 0.003, + "loss": 3.997, + "step": 14577 + }, + { + "epoch": 0.14578, + "grad_norm": 0.6641462445259094, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 14578 + }, + { + "epoch": 0.14579, + "grad_norm": 0.6584463715553284, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 14579 + }, + { + "epoch": 0.1458, + "grad_norm": 0.7167730927467346, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 14580 + }, + { + "epoch": 0.14581, + "grad_norm": 0.7999163866043091, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 14581 + }, + { + "epoch": 0.14582, + "grad_norm": 0.7912095785140991, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 14582 + }, + { + "epoch": 0.14583, + "grad_norm": 0.8575613498687744, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 14583 + }, + { + "epoch": 0.14584, + "grad_norm": 1.1373587846755981, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 14584 + }, + { + "epoch": 0.14585, + "grad_norm": 1.114978313446045, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 14585 + }, + { + "epoch": 0.14586, + "grad_norm": 0.8075107336044312, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 14586 + }, + { + "epoch": 0.14587, + "grad_norm": 0.7308522462844849, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 14587 + }, + { + "epoch": 0.14588, + "grad_norm": 0.8998515009880066, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 14588 + }, + { + "epoch": 0.14589, + "grad_norm": 1.0372451543807983, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 14589 + }, + { + "epoch": 0.1459, + "grad_norm": 1.1466875076293945, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 14590 + }, + { + "epoch": 0.14591, + "grad_norm": 0.7109951972961426, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 14591 + }, + { + "epoch": 0.14592, + "grad_norm": 0.612815797328949, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 14592 + }, + { + "epoch": 0.14593, + "grad_norm": 0.6005706787109375, + "learning_rate": 0.003, + "loss": 4.033, + "step": 14593 + }, + { + "epoch": 0.14594, + "grad_norm": 0.6634933948516846, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 14594 + }, + { + "epoch": 0.14595, + "grad_norm": 0.6079758405685425, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 14595 + }, + { + "epoch": 0.14596, + "grad_norm": 0.6029581427574158, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 14596 + }, + { + "epoch": 0.14597, + "grad_norm": 0.6305979490280151, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 14597 + }, + { + "epoch": 0.14598, + "grad_norm": 0.6574081182479858, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 14598 + }, + { + "epoch": 0.14599, + "grad_norm": 0.6735368371009827, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 14599 + }, + { + "epoch": 0.146, + "grad_norm": 0.651794970035553, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 14600 + }, + { + "epoch": 0.14601, + "grad_norm": 0.6393558382987976, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 14601 + }, + { + "epoch": 0.14602, + "grad_norm": 0.7819718718528748, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 14602 + }, + { + "epoch": 0.14603, + "grad_norm": 0.9885872006416321, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 14603 + }, + { + "epoch": 0.14604, + "grad_norm": 1.2221794128417969, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 14604 + }, + { + "epoch": 0.14605, + "grad_norm": 0.6630940437316895, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 14605 + }, + { + "epoch": 0.14606, + "grad_norm": 0.5590291023254395, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 14606 + }, + { + "epoch": 0.14607, + "grad_norm": 0.6604478359222412, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 14607 + }, + { + "epoch": 0.14608, + "grad_norm": 0.7741831541061401, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 14608 + }, + { + "epoch": 0.14609, + "grad_norm": 0.9608388543128967, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 14609 + }, + { + "epoch": 0.1461, + "grad_norm": 1.0947613716125488, + "learning_rate": 0.003, + "loss": 4.012, + "step": 14610 + }, + { + "epoch": 0.14611, + "grad_norm": 0.9668039679527283, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 14611 + }, + { + "epoch": 0.14612, + "grad_norm": 1.0010789632797241, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 14612 + }, + { + "epoch": 0.14613, + "grad_norm": 1.0474330186843872, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 14613 + }, + { + "epoch": 0.14614, + "grad_norm": 0.8920873999595642, + "learning_rate": 0.003, + "loss": 4.024, + "step": 14614 + }, + { + "epoch": 0.14615, + "grad_norm": 0.8349394798278809, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 14615 + }, + { + "epoch": 0.14616, + "grad_norm": 0.883679211139679, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 14616 + }, + { + "epoch": 0.14617, + "grad_norm": 0.8766427040100098, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 14617 + }, + { + "epoch": 0.14618, + "grad_norm": 1.103816032409668, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 14618 + }, + { + "epoch": 0.14619, + "grad_norm": 1.0742919445037842, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 14619 + }, + { + "epoch": 0.1462, + "grad_norm": 1.0519556999206543, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 14620 + }, + { + "epoch": 0.14621, + "grad_norm": 1.0646867752075195, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 14621 + }, + { + "epoch": 0.14622, + "grad_norm": 1.0145447254180908, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 14622 + }, + { + "epoch": 0.14623, + "grad_norm": 0.9972551465034485, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 14623 + }, + { + "epoch": 0.14624, + "grad_norm": 1.0853716135025024, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 14624 + }, + { + "epoch": 0.14625, + "grad_norm": 0.8484166264533997, + "learning_rate": 0.003, + "loss": 4.043, + "step": 14625 + }, + { + "epoch": 0.14626, + "grad_norm": 0.7008185386657715, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 14626 + }, + { + "epoch": 0.14627, + "grad_norm": 0.7344235181808472, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 14627 + }, + { + "epoch": 0.14628, + "grad_norm": 0.7429527640342712, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 14628 + }, + { + "epoch": 0.14629, + "grad_norm": 0.6254894137382507, + "learning_rate": 0.003, + "loss": 4.001, + "step": 14629 + }, + { + "epoch": 0.1463, + "grad_norm": 0.486186683177948, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 14630 + }, + { + "epoch": 0.14631, + "grad_norm": 0.5246158838272095, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 14631 + }, + { + "epoch": 0.14632, + "grad_norm": 0.5575757026672363, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 14632 + }, + { + "epoch": 0.14633, + "grad_norm": 0.6978687644004822, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 14633 + }, + { + "epoch": 0.14634, + "grad_norm": 0.8760482668876648, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 14634 + }, + { + "epoch": 0.14635, + "grad_norm": 1.2433359622955322, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 14635 + }, + { + "epoch": 0.14636, + "grad_norm": 0.7502328753471375, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 14636 + }, + { + "epoch": 0.14637, + "grad_norm": 0.6649960279464722, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 14637 + }, + { + "epoch": 0.14638, + "grad_norm": 0.7799578309059143, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 14638 + }, + { + "epoch": 0.14639, + "grad_norm": 0.857184886932373, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 14639 + }, + { + "epoch": 0.1464, + "grad_norm": 1.1021485328674316, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 14640 + }, + { + "epoch": 0.14641, + "grad_norm": 0.9773330688476562, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 14641 + }, + { + "epoch": 0.14642, + "grad_norm": 0.8717167973518372, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 14642 + }, + { + "epoch": 0.14643, + "grad_norm": 0.8049965500831604, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 14643 + }, + { + "epoch": 0.14644, + "grad_norm": 0.8108642101287842, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 14644 + }, + { + "epoch": 0.14645, + "grad_norm": 0.8317939639091492, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 14645 + }, + { + "epoch": 0.14646, + "grad_norm": 0.8724066019058228, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 14646 + }, + { + "epoch": 0.14647, + "grad_norm": 0.9462814331054688, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 14647 + }, + { + "epoch": 0.14648, + "grad_norm": 0.9973888993263245, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 14648 + }, + { + "epoch": 0.14649, + "grad_norm": 0.8211222887039185, + "learning_rate": 0.003, + "loss": 4.02, + "step": 14649 + }, + { + "epoch": 0.1465, + "grad_norm": 0.7676868438720703, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 14650 + }, + { + "epoch": 0.14651, + "grad_norm": 0.7277308106422424, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 14651 + }, + { + "epoch": 0.14652, + "grad_norm": 0.7895736694335938, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 14652 + }, + { + "epoch": 0.14653, + "grad_norm": 0.7760275602340698, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 14653 + }, + { + "epoch": 0.14654, + "grad_norm": 0.9125950336456299, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 14654 + }, + { + "epoch": 0.14655, + "grad_norm": 1.0902222394943237, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 14655 + }, + { + "epoch": 0.14656, + "grad_norm": 0.9675227999687195, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 14656 + }, + { + "epoch": 0.14657, + "grad_norm": 0.8446677923202515, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 14657 + }, + { + "epoch": 0.14658, + "grad_norm": 0.7147864699363708, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 14658 + }, + { + "epoch": 0.14659, + "grad_norm": 0.72620689868927, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 14659 + }, + { + "epoch": 0.1466, + "grad_norm": 0.6665274500846863, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 14660 + }, + { + "epoch": 0.14661, + "grad_norm": 0.6929038763046265, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 14661 + }, + { + "epoch": 0.14662, + "grad_norm": 0.6481709480285645, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 14662 + }, + { + "epoch": 0.14663, + "grad_norm": 0.5627266764640808, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 14663 + }, + { + "epoch": 0.14664, + "grad_norm": 0.5991783738136292, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 14664 + }, + { + "epoch": 0.14665, + "grad_norm": 0.5894344449043274, + "learning_rate": 0.003, + "loss": 3.995, + "step": 14665 + }, + { + "epoch": 0.14666, + "grad_norm": 0.5698339939117432, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 14666 + }, + { + "epoch": 0.14667, + "grad_norm": 0.6349946856498718, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 14667 + }, + { + "epoch": 0.14668, + "grad_norm": 0.7414714097976685, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 14668 + }, + { + "epoch": 0.14669, + "grad_norm": 0.7325112223625183, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 14669 + }, + { + "epoch": 0.1467, + "grad_norm": 0.7556993365287781, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 14670 + }, + { + "epoch": 0.14671, + "grad_norm": 0.8011714816093445, + "learning_rate": 0.003, + "loss": 4.027, + "step": 14671 + }, + { + "epoch": 0.14672, + "grad_norm": 0.8728906512260437, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 14672 + }, + { + "epoch": 0.14673, + "grad_norm": 0.9702844023704529, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 14673 + }, + { + "epoch": 0.14674, + "grad_norm": 1.0082310438156128, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 14674 + }, + { + "epoch": 0.14675, + "grad_norm": 1.2419564723968506, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 14675 + }, + { + "epoch": 0.14676, + "grad_norm": 0.9880674481391907, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 14676 + }, + { + "epoch": 0.14677, + "grad_norm": 0.8718649744987488, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 14677 + }, + { + "epoch": 0.14678, + "grad_norm": 0.8813468217849731, + "learning_rate": 0.003, + "loss": 4.023, + "step": 14678 + }, + { + "epoch": 0.14679, + "grad_norm": 0.7749958634376526, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 14679 + }, + { + "epoch": 0.1468, + "grad_norm": 0.7995460033416748, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 14680 + }, + { + "epoch": 0.14681, + "grad_norm": 0.6909242272377014, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 14681 + }, + { + "epoch": 0.14682, + "grad_norm": 0.736655592918396, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 14682 + }, + { + "epoch": 0.14683, + "grad_norm": 0.7802330851554871, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 14683 + }, + { + "epoch": 0.14684, + "grad_norm": 0.8775361180305481, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 14684 + }, + { + "epoch": 0.14685, + "grad_norm": 1.0075017213821411, + "learning_rate": 0.003, + "loss": 4.032, + "step": 14685 + }, + { + "epoch": 0.14686, + "grad_norm": 1.010310173034668, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 14686 + }, + { + "epoch": 0.14687, + "grad_norm": 0.7436519265174866, + "learning_rate": 0.003, + "loss": 4.016, + "step": 14687 + }, + { + "epoch": 0.14688, + "grad_norm": 0.788584291934967, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 14688 + }, + { + "epoch": 0.14689, + "grad_norm": 0.776448130607605, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 14689 + }, + { + "epoch": 0.1469, + "grad_norm": 0.7722798585891724, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 14690 + }, + { + "epoch": 0.14691, + "grad_norm": 0.710889995098114, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 14691 + }, + { + "epoch": 0.14692, + "grad_norm": 0.8781268000602722, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 14692 + }, + { + "epoch": 0.14693, + "grad_norm": 1.0924203395843506, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 14693 + }, + { + "epoch": 0.14694, + "grad_norm": 1.026205062866211, + "learning_rate": 0.003, + "loss": 4.006, + "step": 14694 + }, + { + "epoch": 0.14695, + "grad_norm": 0.9521628022193909, + "learning_rate": 0.003, + "loss": 4.039, + "step": 14695 + }, + { + "epoch": 0.14696, + "grad_norm": 0.9316068887710571, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 14696 + }, + { + "epoch": 0.14697, + "grad_norm": 0.9264803528785706, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 14697 + }, + { + "epoch": 0.14698, + "grad_norm": 1.1187646389007568, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 14698 + }, + { + "epoch": 0.14699, + "grad_norm": 0.8024663329124451, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 14699 + }, + { + "epoch": 0.147, + "grad_norm": 0.7368111610412598, + "learning_rate": 0.003, + "loss": 4.03, + "step": 14700 + }, + { + "epoch": 0.14701, + "grad_norm": 0.7123501300811768, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 14701 + }, + { + "epoch": 0.14702, + "grad_norm": 0.7903427481651306, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 14702 + }, + { + "epoch": 0.14703, + "grad_norm": 0.8391876220703125, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 14703 + }, + { + "epoch": 0.14704, + "grad_norm": 0.9227834939956665, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 14704 + }, + { + "epoch": 0.14705, + "grad_norm": 1.0087929964065552, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 14705 + }, + { + "epoch": 0.14706, + "grad_norm": 0.8531105518341064, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 14706 + }, + { + "epoch": 0.14707, + "grad_norm": 0.7724258899688721, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 14707 + }, + { + "epoch": 0.14708, + "grad_norm": 0.7460688352584839, + "learning_rate": 0.003, + "loss": 4.008, + "step": 14708 + }, + { + "epoch": 0.14709, + "grad_norm": 0.7925257086753845, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 14709 + }, + { + "epoch": 0.1471, + "grad_norm": 0.8145542740821838, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 14710 + }, + { + "epoch": 0.14711, + "grad_norm": 0.8683342933654785, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 14711 + }, + { + "epoch": 0.14712, + "grad_norm": 1.0504648685455322, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 14712 + }, + { + "epoch": 0.14713, + "grad_norm": 1.0538825988769531, + "learning_rate": 0.003, + "loss": 3.995, + "step": 14713 + }, + { + "epoch": 0.14714, + "grad_norm": 0.9523059129714966, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 14714 + }, + { + "epoch": 0.14715, + "grad_norm": 0.9349806308746338, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 14715 + }, + { + "epoch": 0.14716, + "grad_norm": 0.9929547309875488, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 14716 + }, + { + "epoch": 0.14717, + "grad_norm": 0.9924283027648926, + "learning_rate": 0.003, + "loss": 4.039, + "step": 14717 + }, + { + "epoch": 0.14718, + "grad_norm": 0.9050933718681335, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 14718 + }, + { + "epoch": 0.14719, + "grad_norm": 0.8614309430122375, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 14719 + }, + { + "epoch": 0.1472, + "grad_norm": 0.8637762665748596, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 14720 + }, + { + "epoch": 0.14721, + "grad_norm": 0.7075213193893433, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 14721 + }, + { + "epoch": 0.14722, + "grad_norm": 0.6924384832382202, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 14722 + }, + { + "epoch": 0.14723, + "grad_norm": 0.7364385724067688, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 14723 + }, + { + "epoch": 0.14724, + "grad_norm": 0.8216361403465271, + "learning_rate": 0.003, + "loss": 4.018, + "step": 14724 + }, + { + "epoch": 0.14725, + "grad_norm": 0.8571340441703796, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 14725 + }, + { + "epoch": 0.14726, + "grad_norm": 0.947347104549408, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 14726 + }, + { + "epoch": 0.14727, + "grad_norm": 1.083392858505249, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 14727 + }, + { + "epoch": 0.14728, + "grad_norm": 1.0357502698898315, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 14728 + }, + { + "epoch": 0.14729, + "grad_norm": 0.7941818237304688, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 14729 + }, + { + "epoch": 0.1473, + "grad_norm": 0.692725658416748, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 14730 + }, + { + "epoch": 0.14731, + "grad_norm": 0.6918818354606628, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 14731 + }, + { + "epoch": 0.14732, + "grad_norm": 0.7324870228767395, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 14732 + }, + { + "epoch": 0.14733, + "grad_norm": 0.8361362218856812, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 14733 + }, + { + "epoch": 0.14734, + "grad_norm": 0.8729264736175537, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14734 + }, + { + "epoch": 0.14735, + "grad_norm": 0.653428316116333, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 14735 + }, + { + "epoch": 0.14736, + "grad_norm": 0.6323630809783936, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 14736 + }, + { + "epoch": 0.14737, + "grad_norm": 0.6049655079841614, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 14737 + }, + { + "epoch": 0.14738, + "grad_norm": 0.5639153718948364, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 14738 + }, + { + "epoch": 0.14739, + "grad_norm": 0.6139823794364929, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 14739 + }, + { + "epoch": 0.1474, + "grad_norm": 0.7757695913314819, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 14740 + }, + { + "epoch": 0.14741, + "grad_norm": 0.9068511128425598, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 14741 + }, + { + "epoch": 0.14742, + "grad_norm": 1.1613571643829346, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 14742 + }, + { + "epoch": 0.14743, + "grad_norm": 0.7322368025779724, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 14743 + }, + { + "epoch": 0.14744, + "grad_norm": 0.5892424583435059, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 14744 + }, + { + "epoch": 0.14745, + "grad_norm": 0.7783081531524658, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 14745 + }, + { + "epoch": 0.14746, + "grad_norm": 0.8132655620574951, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 14746 + }, + { + "epoch": 0.14747, + "grad_norm": 0.8056081533432007, + "learning_rate": 0.003, + "loss": 4.014, + "step": 14747 + }, + { + "epoch": 0.14748, + "grad_norm": 0.8169584274291992, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 14748 + }, + { + "epoch": 0.14749, + "grad_norm": 0.6788761019706726, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 14749 + }, + { + "epoch": 0.1475, + "grad_norm": 0.6912890076637268, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 14750 + }, + { + "epoch": 0.14751, + "grad_norm": 0.791756272315979, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 14751 + }, + { + "epoch": 0.14752, + "grad_norm": 0.6415024995803833, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 14752 + }, + { + "epoch": 0.14753, + "grad_norm": 0.5531638264656067, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 14753 + }, + { + "epoch": 0.14754, + "grad_norm": 0.6161421537399292, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 14754 + }, + { + "epoch": 0.14755, + "grad_norm": 0.7031818628311157, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 14755 + }, + { + "epoch": 0.14756, + "grad_norm": 0.7122676968574524, + "learning_rate": 0.003, + "loss": 3.998, + "step": 14756 + }, + { + "epoch": 0.14757, + "grad_norm": 0.7026879191398621, + "learning_rate": 0.003, + "loss": 4.013, + "step": 14757 + }, + { + "epoch": 0.14758, + "grad_norm": 0.847398579120636, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 14758 + }, + { + "epoch": 0.14759, + "grad_norm": 1.0359034538269043, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 14759 + }, + { + "epoch": 0.1476, + "grad_norm": 1.1722480058670044, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 14760 + }, + { + "epoch": 0.14761, + "grad_norm": 1.0708212852478027, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 14761 + }, + { + "epoch": 0.14762, + "grad_norm": 1.000401496887207, + "learning_rate": 0.003, + "loss": 4.009, + "step": 14762 + }, + { + "epoch": 0.14763, + "grad_norm": 0.9598147869110107, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 14763 + }, + { + "epoch": 0.14764, + "grad_norm": 0.9209848046302795, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 14764 + }, + { + "epoch": 0.14765, + "grad_norm": 0.834210216999054, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 14765 + }, + { + "epoch": 0.14766, + "grad_norm": 0.902065098285675, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 14766 + }, + { + "epoch": 0.14767, + "grad_norm": 0.9380612969398499, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 14767 + }, + { + "epoch": 0.14768, + "grad_norm": 1.0841375589370728, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 14768 + }, + { + "epoch": 0.14769, + "grad_norm": 0.9712322950363159, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 14769 + }, + { + "epoch": 0.1477, + "grad_norm": 0.988662600517273, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 14770 + }, + { + "epoch": 0.14771, + "grad_norm": 1.0296447277069092, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 14771 + }, + { + "epoch": 0.14772, + "grad_norm": 0.9799277186393738, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 14772 + }, + { + "epoch": 0.14773, + "grad_norm": 0.9216949343681335, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 14773 + }, + { + "epoch": 0.14774, + "grad_norm": 0.9277490377426147, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 14774 + }, + { + "epoch": 0.14775, + "grad_norm": 1.1422898769378662, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 14775 + }, + { + "epoch": 0.14776, + "grad_norm": 0.8414909839630127, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 14776 + }, + { + "epoch": 0.14777, + "grad_norm": 0.7519655823707581, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 14777 + }, + { + "epoch": 0.14778, + "grad_norm": 0.6935589909553528, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 14778 + }, + { + "epoch": 0.14779, + "grad_norm": 0.632209300994873, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 14779 + }, + { + "epoch": 0.1478, + "grad_norm": 0.6803045272827148, + "learning_rate": 0.003, + "loss": 4.011, + "step": 14780 + }, + { + "epoch": 0.14781, + "grad_norm": 0.7468591928482056, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 14781 + }, + { + "epoch": 0.14782, + "grad_norm": 0.7592949271202087, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 14782 + }, + { + "epoch": 0.14783, + "grad_norm": 0.8823490142822266, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 14783 + }, + { + "epoch": 0.14784, + "grad_norm": 0.8962957262992859, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 14784 + }, + { + "epoch": 0.14785, + "grad_norm": 0.9850898385047913, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 14785 + }, + { + "epoch": 0.14786, + "grad_norm": 1.0215193033218384, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 14786 + }, + { + "epoch": 0.14787, + "grad_norm": 0.8413836359977722, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 14787 + }, + { + "epoch": 0.14788, + "grad_norm": 0.740068256855011, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 14788 + }, + { + "epoch": 0.14789, + "grad_norm": 0.7006083130836487, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 14789 + }, + { + "epoch": 0.1479, + "grad_norm": 0.7278753519058228, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 14790 + }, + { + "epoch": 0.14791, + "grad_norm": 0.8917559385299683, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 14791 + }, + { + "epoch": 0.14792, + "grad_norm": 0.9600374102592468, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 14792 + }, + { + "epoch": 0.14793, + "grad_norm": 0.9399484992027283, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 14793 + }, + { + "epoch": 0.14794, + "grad_norm": 0.9212305545806885, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 14794 + }, + { + "epoch": 0.14795, + "grad_norm": 0.8454551696777344, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 14795 + }, + { + "epoch": 0.14796, + "grad_norm": 0.8320006728172302, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 14796 + }, + { + "epoch": 0.14797, + "grad_norm": 0.8420716524124146, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 14797 + }, + { + "epoch": 0.14798, + "grad_norm": 0.8744964003562927, + "learning_rate": 0.003, + "loss": 4.032, + "step": 14798 + }, + { + "epoch": 0.14799, + "grad_norm": 0.9001518487930298, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 14799 + }, + { + "epoch": 0.148, + "grad_norm": 0.9593254327774048, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 14800 + }, + { + "epoch": 0.14801, + "grad_norm": 1.0270856618881226, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 14801 + }, + { + "epoch": 0.14802, + "grad_norm": 0.909136950969696, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 14802 + }, + { + "epoch": 0.14803, + "grad_norm": 0.8870131969451904, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 14803 + }, + { + "epoch": 0.14804, + "grad_norm": 0.972500741481781, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 14804 + }, + { + "epoch": 0.14805, + "grad_norm": 1.041571855545044, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 14805 + }, + { + "epoch": 0.14806, + "grad_norm": 0.8508731722831726, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 14806 + }, + { + "epoch": 0.14807, + "grad_norm": 0.7221386432647705, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 14807 + }, + { + "epoch": 0.14808, + "grad_norm": 0.6646813750267029, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 14808 + }, + { + "epoch": 0.14809, + "grad_norm": 0.5410613417625427, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 14809 + }, + { + "epoch": 0.1481, + "grad_norm": 0.5607827305793762, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 14810 + }, + { + "epoch": 0.14811, + "grad_norm": 0.5264371037483215, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 14811 + }, + { + "epoch": 0.14812, + "grad_norm": 0.507453203201294, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 14812 + }, + { + "epoch": 0.14813, + "grad_norm": 0.5576320290565491, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 14813 + }, + { + "epoch": 0.14814, + "grad_norm": 0.6028345227241516, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 14814 + }, + { + "epoch": 0.14815, + "grad_norm": 0.5697914361953735, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 14815 + }, + { + "epoch": 0.14816, + "grad_norm": 0.6025079488754272, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 14816 + }, + { + "epoch": 0.14817, + "grad_norm": 0.6084979772567749, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 14817 + }, + { + "epoch": 0.14818, + "grad_norm": 0.7108429074287415, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 14818 + }, + { + "epoch": 0.14819, + "grad_norm": 0.8598563075065613, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 14819 + }, + { + "epoch": 0.1482, + "grad_norm": 1.1866936683654785, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 14820 + }, + { + "epoch": 0.14821, + "grad_norm": 1.1063987016677856, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 14821 + }, + { + "epoch": 0.14822, + "grad_norm": 0.8805261850357056, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 14822 + }, + { + "epoch": 0.14823, + "grad_norm": 0.7659177780151367, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 14823 + }, + { + "epoch": 0.14824, + "grad_norm": 0.7803784012794495, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 14824 + }, + { + "epoch": 0.14825, + "grad_norm": 0.8470391035079956, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 14825 + }, + { + "epoch": 0.14826, + "grad_norm": 0.8442140817642212, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 14826 + }, + { + "epoch": 0.14827, + "grad_norm": 0.9312122464179993, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 14827 + }, + { + "epoch": 0.14828, + "grad_norm": 1.031015157699585, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 14828 + }, + { + "epoch": 0.14829, + "grad_norm": 1.0510855913162231, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 14829 + }, + { + "epoch": 0.1483, + "grad_norm": 1.0041213035583496, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 14830 + }, + { + "epoch": 0.14831, + "grad_norm": 1.0243691205978394, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 14831 + }, + { + "epoch": 0.14832, + "grad_norm": 0.8999918103218079, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 14832 + }, + { + "epoch": 0.14833, + "grad_norm": 0.7126042246818542, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 14833 + }, + { + "epoch": 0.14834, + "grad_norm": 0.6217748522758484, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 14834 + }, + { + "epoch": 0.14835, + "grad_norm": 0.5969050526618958, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 14835 + }, + { + "epoch": 0.14836, + "grad_norm": 0.6038386821746826, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 14836 + }, + { + "epoch": 0.14837, + "grad_norm": 0.5933994054794312, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 14837 + }, + { + "epoch": 0.14838, + "grad_norm": 0.4869672954082489, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 14838 + }, + { + "epoch": 0.14839, + "grad_norm": 0.5288945436477661, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 14839 + }, + { + "epoch": 0.1484, + "grad_norm": 0.5412532687187195, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 14840 + }, + { + "epoch": 0.14841, + "grad_norm": 0.6282064318656921, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 14841 + }, + { + "epoch": 0.14842, + "grad_norm": 0.7172026634216309, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 14842 + }, + { + "epoch": 0.14843, + "grad_norm": 0.7796453833580017, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 14843 + }, + { + "epoch": 0.14844, + "grad_norm": 0.8600711226463318, + "learning_rate": 0.003, + "loss": 4.007, + "step": 14844 + }, + { + "epoch": 0.14845, + "grad_norm": 0.8411175608634949, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 14845 + }, + { + "epoch": 0.14846, + "grad_norm": 1.10649836063385, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 14846 + }, + { + "epoch": 0.14847, + "grad_norm": 1.4016904830932617, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 14847 + }, + { + "epoch": 0.14848, + "grad_norm": 0.7915355563163757, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 14848 + }, + { + "epoch": 0.14849, + "grad_norm": 0.7058914303779602, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 14849 + }, + { + "epoch": 0.1485, + "grad_norm": 0.7597223520278931, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 14850 + }, + { + "epoch": 0.14851, + "grad_norm": 0.8670496940612793, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 14851 + }, + { + "epoch": 0.14852, + "grad_norm": 0.7875985503196716, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 14852 + }, + { + "epoch": 0.14853, + "grad_norm": 0.8555645942687988, + "learning_rate": 0.003, + "loss": 4.022, + "step": 14853 + }, + { + "epoch": 0.14854, + "grad_norm": 1.1159355640411377, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 14854 + }, + { + "epoch": 0.14855, + "grad_norm": 1.0782500505447388, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 14855 + }, + { + "epoch": 0.14856, + "grad_norm": 0.8411812782287598, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 14856 + }, + { + "epoch": 0.14857, + "grad_norm": 0.6487085223197937, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 14857 + }, + { + "epoch": 0.14858, + "grad_norm": 0.6623062491416931, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 14858 + }, + { + "epoch": 0.14859, + "grad_norm": 0.7163119316101074, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 14859 + }, + { + "epoch": 0.1486, + "grad_norm": 0.6611106991767883, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 14860 + }, + { + "epoch": 0.14861, + "grad_norm": 0.6478426456451416, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 14861 + }, + { + "epoch": 0.14862, + "grad_norm": 0.7446329593658447, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 14862 + }, + { + "epoch": 0.14863, + "grad_norm": 0.9106122851371765, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 14863 + }, + { + "epoch": 0.14864, + "grad_norm": 1.1429930925369263, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 14864 + }, + { + "epoch": 0.14865, + "grad_norm": 0.8995509743690491, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 14865 + }, + { + "epoch": 0.14866, + "grad_norm": 0.761750340461731, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 14866 + }, + { + "epoch": 0.14867, + "grad_norm": 0.732969343662262, + "learning_rate": 0.003, + "loss": 3.979, + "step": 14867 + }, + { + "epoch": 0.14868, + "grad_norm": 0.7871980667114258, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 14868 + }, + { + "epoch": 0.14869, + "grad_norm": 0.7820833921432495, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 14869 + }, + { + "epoch": 0.1487, + "grad_norm": 0.8342604637145996, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 14870 + }, + { + "epoch": 0.14871, + "grad_norm": 0.9378725290298462, + "learning_rate": 0.003, + "loss": 4.022, + "step": 14871 + }, + { + "epoch": 0.14872, + "grad_norm": 1.1117609739303589, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 14872 + }, + { + "epoch": 0.14873, + "grad_norm": 0.8830320835113525, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 14873 + }, + { + "epoch": 0.14874, + "grad_norm": 0.9355973601341248, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 14874 + }, + { + "epoch": 0.14875, + "grad_norm": 0.9096114635467529, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 14875 + }, + { + "epoch": 0.14876, + "grad_norm": 0.902562141418457, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 14876 + }, + { + "epoch": 0.14877, + "grad_norm": 0.868760883808136, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 14877 + }, + { + "epoch": 0.14878, + "grad_norm": 0.8453792929649353, + "learning_rate": 0.003, + "loss": 4.023, + "step": 14878 + }, + { + "epoch": 0.14879, + "grad_norm": 0.9315659403800964, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 14879 + }, + { + "epoch": 0.1488, + "grad_norm": 0.824686586856842, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 14880 + }, + { + "epoch": 0.14881, + "grad_norm": 0.7330214381217957, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 14881 + }, + { + "epoch": 0.14882, + "grad_norm": 0.7923099994659424, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 14882 + }, + { + "epoch": 0.14883, + "grad_norm": 0.8784119486808777, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 14883 + }, + { + "epoch": 0.14884, + "grad_norm": 1.2640788555145264, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 14884 + }, + { + "epoch": 0.14885, + "grad_norm": 0.874704897403717, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 14885 + }, + { + "epoch": 0.14886, + "grad_norm": 0.8032217025756836, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 14886 + }, + { + "epoch": 0.14887, + "grad_norm": 0.8753021955490112, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 14887 + }, + { + "epoch": 0.14888, + "grad_norm": 0.8627346158027649, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 14888 + }, + { + "epoch": 0.14889, + "grad_norm": 0.821229875087738, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 14889 + }, + { + "epoch": 0.1489, + "grad_norm": 0.6824096441268921, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 14890 + }, + { + "epoch": 0.14891, + "grad_norm": 0.6843130588531494, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 14891 + }, + { + "epoch": 0.14892, + "grad_norm": 0.8635883927345276, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 14892 + }, + { + "epoch": 0.14893, + "grad_norm": 0.8831451535224915, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 14893 + }, + { + "epoch": 0.14894, + "grad_norm": 0.9102339744567871, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 14894 + }, + { + "epoch": 0.14895, + "grad_norm": 0.9062256217002869, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 14895 + }, + { + "epoch": 0.14896, + "grad_norm": 0.7547155618667603, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 14896 + }, + { + "epoch": 0.14897, + "grad_norm": 0.8735215067863464, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 14897 + }, + { + "epoch": 0.14898, + "grad_norm": 1.1400668621063232, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 14898 + }, + { + "epoch": 0.14899, + "grad_norm": 0.9181110858917236, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 14899 + }, + { + "epoch": 0.149, + "grad_norm": 0.7998493313789368, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 14900 + }, + { + "epoch": 0.14901, + "grad_norm": 0.8677207231521606, + "learning_rate": 0.003, + "loss": 4.024, + "step": 14901 + }, + { + "epoch": 0.14902, + "grad_norm": 0.833134651184082, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 14902 + }, + { + "epoch": 0.14903, + "grad_norm": 0.7749120593070984, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 14903 + }, + { + "epoch": 0.14904, + "grad_norm": 0.9325809478759766, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 14904 + }, + { + "epoch": 0.14905, + "grad_norm": 1.0706454515457153, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 14905 + }, + { + "epoch": 0.14906, + "grad_norm": 0.9616243839263916, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 14906 + }, + { + "epoch": 0.14907, + "grad_norm": 0.8771450519561768, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 14907 + }, + { + "epoch": 0.14908, + "grad_norm": 0.824774444103241, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 14908 + }, + { + "epoch": 0.14909, + "grad_norm": 0.8146010041236877, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 14909 + }, + { + "epoch": 0.1491, + "grad_norm": 0.9311041235923767, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 14910 + }, + { + "epoch": 0.14911, + "grad_norm": 0.9459924697875977, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 14911 + }, + { + "epoch": 0.14912, + "grad_norm": 0.9101096391677856, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 14912 + }, + { + "epoch": 0.14913, + "grad_norm": 0.9328768849372864, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 14913 + }, + { + "epoch": 0.14914, + "grad_norm": 0.8582000732421875, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 14914 + }, + { + "epoch": 0.14915, + "grad_norm": 0.8020802736282349, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 14915 + }, + { + "epoch": 0.14916, + "grad_norm": 0.6678415536880493, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 14916 + }, + { + "epoch": 0.14917, + "grad_norm": 0.58196622133255, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 14917 + }, + { + "epoch": 0.14918, + "grad_norm": 0.5718095302581787, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 14918 + }, + { + "epoch": 0.14919, + "grad_norm": 0.5286774039268494, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 14919 + }, + { + "epoch": 0.1492, + "grad_norm": 0.5560582280158997, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 14920 + }, + { + "epoch": 0.14921, + "grad_norm": 0.5656267404556274, + "learning_rate": 0.003, + "loss": 4.018, + "step": 14921 + }, + { + "epoch": 0.14922, + "grad_norm": 0.583981454372406, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 14922 + }, + { + "epoch": 0.14923, + "grad_norm": 0.6026471257209778, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 14923 + }, + { + "epoch": 0.14924, + "grad_norm": 0.6006498336791992, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 14924 + }, + { + "epoch": 0.14925, + "grad_norm": 0.7609928846359253, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 14925 + }, + { + "epoch": 0.14926, + "grad_norm": 1.1219854354858398, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 14926 + }, + { + "epoch": 0.14927, + "grad_norm": 1.2104982137680054, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 14927 + }, + { + "epoch": 0.14928, + "grad_norm": 0.5168392062187195, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 14928 + }, + { + "epoch": 0.14929, + "grad_norm": 0.7282378077507019, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 14929 + }, + { + "epoch": 0.1493, + "grad_norm": 1.0722233057022095, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 14930 + }, + { + "epoch": 0.14931, + "grad_norm": 1.0388180017471313, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 14931 + }, + { + "epoch": 0.14932, + "grad_norm": 0.8689343929290771, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 14932 + }, + { + "epoch": 0.14933, + "grad_norm": 0.8487166166305542, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 14933 + }, + { + "epoch": 0.14934, + "grad_norm": 0.8658933043479919, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 14934 + }, + { + "epoch": 0.14935, + "grad_norm": 0.9383997321128845, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 14935 + }, + { + "epoch": 0.14936, + "grad_norm": 0.9382808208465576, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 14936 + }, + { + "epoch": 0.14937, + "grad_norm": 0.9391462206840515, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 14937 + }, + { + "epoch": 0.14938, + "grad_norm": 0.8973220586776733, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 14938 + }, + { + "epoch": 0.14939, + "grad_norm": 0.8659802675247192, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 14939 + }, + { + "epoch": 0.1494, + "grad_norm": 0.8968918919563293, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 14940 + }, + { + "epoch": 0.14941, + "grad_norm": 0.8753356337547302, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 14941 + }, + { + "epoch": 0.14942, + "grad_norm": 0.8973208069801331, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 14942 + }, + { + "epoch": 0.14943, + "grad_norm": 0.7997957468032837, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 14943 + }, + { + "epoch": 0.14944, + "grad_norm": 0.6905195116996765, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 14944 + }, + { + "epoch": 0.14945, + "grad_norm": 0.6598191261291504, + "learning_rate": 0.003, + "loss": 3.989, + "step": 14945 + }, + { + "epoch": 0.14946, + "grad_norm": 0.7957265973091125, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 14946 + }, + { + "epoch": 0.14947, + "grad_norm": 0.7551441788673401, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 14947 + }, + { + "epoch": 0.14948, + "grad_norm": 0.6483982801437378, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 14948 + }, + { + "epoch": 0.14949, + "grad_norm": 0.7106829285621643, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 14949 + }, + { + "epoch": 0.1495, + "grad_norm": 0.8401058316230774, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 14950 + }, + { + "epoch": 0.14951, + "grad_norm": 0.8834985494613647, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 14951 + }, + { + "epoch": 0.14952, + "grad_norm": 1.0601893663406372, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 14952 + }, + { + "epoch": 0.14953, + "grad_norm": 1.2167599201202393, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 14953 + }, + { + "epoch": 0.14954, + "grad_norm": 0.8047480583190918, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 14954 + }, + { + "epoch": 0.14955, + "grad_norm": 0.6802546381950378, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 14955 + }, + { + "epoch": 0.14956, + "grad_norm": 0.6390823721885681, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 14956 + }, + { + "epoch": 0.14957, + "grad_norm": 0.7274983525276184, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 14957 + }, + { + "epoch": 0.14958, + "grad_norm": 0.7369818687438965, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 14958 + }, + { + "epoch": 0.14959, + "grad_norm": 0.7644141912460327, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 14959 + }, + { + "epoch": 0.1496, + "grad_norm": 0.674062192440033, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 14960 + }, + { + "epoch": 0.14961, + "grad_norm": 0.8109387755393982, + "learning_rate": 0.003, + "loss": 4.005, + "step": 14961 + }, + { + "epoch": 0.14962, + "grad_norm": 0.8895452618598938, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 14962 + }, + { + "epoch": 0.14963, + "grad_norm": 1.0448538064956665, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 14963 + }, + { + "epoch": 0.14964, + "grad_norm": 1.0570226907730103, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 14964 + }, + { + "epoch": 0.14965, + "grad_norm": 1.0361334085464478, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 14965 + }, + { + "epoch": 0.14966, + "grad_norm": 1.0691248178482056, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 14966 + }, + { + "epoch": 0.14967, + "grad_norm": 0.9369614124298096, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 14967 + }, + { + "epoch": 0.14968, + "grad_norm": 0.911221981048584, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 14968 + }, + { + "epoch": 0.14969, + "grad_norm": 0.8358284831047058, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 14969 + }, + { + "epoch": 0.1497, + "grad_norm": 0.7841851115226746, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 14970 + }, + { + "epoch": 0.14971, + "grad_norm": 0.8406400084495544, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 14971 + }, + { + "epoch": 0.14972, + "grad_norm": 0.9173609614372253, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 14972 + }, + { + "epoch": 0.14973, + "grad_norm": 0.8612219095230103, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 14973 + }, + { + "epoch": 0.14974, + "grad_norm": 1.0527944564819336, + "learning_rate": 0.003, + "loss": 4.002, + "step": 14974 + }, + { + "epoch": 0.14975, + "grad_norm": 1.1495803594589233, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 14975 + }, + { + "epoch": 0.14976, + "grad_norm": 0.8519269227981567, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 14976 + }, + { + "epoch": 0.14977, + "grad_norm": 0.6940068006515503, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 14977 + }, + { + "epoch": 0.14978, + "grad_norm": 0.6137828826904297, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 14978 + }, + { + "epoch": 0.14979, + "grad_norm": 0.7167901396751404, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 14979 + }, + { + "epoch": 0.1498, + "grad_norm": 0.736061692237854, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 14980 + }, + { + "epoch": 0.14981, + "grad_norm": 0.7785318493843079, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 14981 + }, + { + "epoch": 0.14982, + "grad_norm": 0.7325731515884399, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 14982 + }, + { + "epoch": 0.14983, + "grad_norm": 0.7698650360107422, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 14983 + }, + { + "epoch": 0.14984, + "grad_norm": 0.9054916501045227, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 14984 + }, + { + "epoch": 0.14985, + "grad_norm": 0.8448771834373474, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 14985 + }, + { + "epoch": 0.14986, + "grad_norm": 0.7488740682601929, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 14986 + }, + { + "epoch": 0.14987, + "grad_norm": 0.7946826219558716, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 14987 + }, + { + "epoch": 0.14988, + "grad_norm": 0.8177905082702637, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 14988 + }, + { + "epoch": 0.14989, + "grad_norm": 0.9731899499893188, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 14989 + }, + { + "epoch": 0.1499, + "grad_norm": 1.1797256469726562, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 14990 + }, + { + "epoch": 0.14991, + "grad_norm": 0.6459851264953613, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 14991 + }, + { + "epoch": 0.14992, + "grad_norm": 0.4921308755874634, + "learning_rate": 0.003, + "loss": 3.988, + "step": 14992 + }, + { + "epoch": 0.14993, + "grad_norm": 0.5904116630554199, + "learning_rate": 0.003, + "loss": 3.994, + "step": 14993 + }, + { + "epoch": 0.14994, + "grad_norm": 0.6307472586631775, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 14994 + }, + { + "epoch": 0.14995, + "grad_norm": 0.6794448494911194, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 14995 + }, + { + "epoch": 0.14996, + "grad_norm": 0.8052518963813782, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 14996 + }, + { + "epoch": 0.14997, + "grad_norm": 0.9949678778648376, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 14997 + }, + { + "epoch": 0.14998, + "grad_norm": 1.0166358947753906, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 14998 + }, + { + "epoch": 0.14999, + "grad_norm": 0.9890971779823303, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 14999 + }, + { + "epoch": 0.15, + "grad_norm": 1.0871403217315674, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 15000 + }, + { + "epoch": 0.15001, + "grad_norm": 0.8645810484886169, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 15001 + }, + { + "epoch": 0.15002, + "grad_norm": 0.702522873878479, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 15002 + }, + { + "epoch": 0.15003, + "grad_norm": 0.7483938932418823, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 15003 + }, + { + "epoch": 0.15004, + "grad_norm": 1.0146819353103638, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 15004 + }, + { + "epoch": 0.15005, + "grad_norm": 1.0388811826705933, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 15005 + }, + { + "epoch": 0.15006, + "grad_norm": 0.8578092455863953, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 15006 + }, + { + "epoch": 0.15007, + "grad_norm": 0.8312777876853943, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 15007 + }, + { + "epoch": 0.15008, + "grad_norm": 0.7816850543022156, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 15008 + }, + { + "epoch": 0.15009, + "grad_norm": 0.6976773142814636, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 15009 + }, + { + "epoch": 0.1501, + "grad_norm": 0.6751989126205444, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 15010 + }, + { + "epoch": 0.15011, + "grad_norm": 0.7083759307861328, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 15011 + }, + { + "epoch": 0.15012, + "grad_norm": 0.6603441834449768, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 15012 + }, + { + "epoch": 0.15013, + "grad_norm": 0.6130689978599548, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 15013 + }, + { + "epoch": 0.15014, + "grad_norm": 0.6485565900802612, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 15014 + }, + { + "epoch": 0.15015, + "grad_norm": 0.7764594554901123, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 15015 + }, + { + "epoch": 0.15016, + "grad_norm": 0.9251732230186462, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 15016 + }, + { + "epoch": 0.15017, + "grad_norm": 0.9680980443954468, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 15017 + }, + { + "epoch": 0.15018, + "grad_norm": 0.89886474609375, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 15018 + }, + { + "epoch": 0.15019, + "grad_norm": 0.8506682515144348, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 15019 + }, + { + "epoch": 0.1502, + "grad_norm": 0.9167236089706421, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 15020 + }, + { + "epoch": 0.15021, + "grad_norm": 0.9710371494293213, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 15021 + }, + { + "epoch": 0.15022, + "grad_norm": 0.9589751958847046, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 15022 + }, + { + "epoch": 0.15023, + "grad_norm": 0.9774499535560608, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 15023 + }, + { + "epoch": 0.15024, + "grad_norm": 0.9559521079063416, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 15024 + }, + { + "epoch": 0.15025, + "grad_norm": 1.0851106643676758, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 15025 + }, + { + "epoch": 0.15026, + "grad_norm": 0.9278112649917603, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 15026 + }, + { + "epoch": 0.15027, + "grad_norm": 0.900368332862854, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 15027 + }, + { + "epoch": 0.15028, + "grad_norm": 0.7963383793830872, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 15028 + }, + { + "epoch": 0.15029, + "grad_norm": 0.7205212116241455, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 15029 + }, + { + "epoch": 0.1503, + "grad_norm": 0.7486795783042908, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 15030 + }, + { + "epoch": 0.15031, + "grad_norm": 0.6926259398460388, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 15031 + }, + { + "epoch": 0.15032, + "grad_norm": 0.6792514324188232, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 15032 + }, + { + "epoch": 0.15033, + "grad_norm": 0.5921057462692261, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 15033 + }, + { + "epoch": 0.15034, + "grad_norm": 0.7386751770973206, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 15034 + }, + { + "epoch": 0.15035, + "grad_norm": 0.725128710269928, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 15035 + }, + { + "epoch": 0.15036, + "grad_norm": 0.8471537828445435, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 15036 + }, + { + "epoch": 0.15037, + "grad_norm": 0.8906320333480835, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 15037 + }, + { + "epoch": 0.15038, + "grad_norm": 0.9164702296257019, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 15038 + }, + { + "epoch": 0.15039, + "grad_norm": 0.9383320212364197, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 15039 + }, + { + "epoch": 0.1504, + "grad_norm": 1.0033183097839355, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 15040 + }, + { + "epoch": 0.15041, + "grad_norm": 1.1808382272720337, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 15041 + }, + { + "epoch": 0.15042, + "grad_norm": 0.9112935066223145, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 15042 + }, + { + "epoch": 0.15043, + "grad_norm": 0.8786137104034424, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 15043 + }, + { + "epoch": 0.15044, + "grad_norm": 0.9137707948684692, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 15044 + }, + { + "epoch": 0.15045, + "grad_norm": 0.8492369055747986, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 15045 + }, + { + "epoch": 0.15046, + "grad_norm": 0.8525924682617188, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 15046 + }, + { + "epoch": 0.15047, + "grad_norm": 0.8138546347618103, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 15047 + }, + { + "epoch": 0.15048, + "grad_norm": 0.8754577040672302, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 15048 + }, + { + "epoch": 0.15049, + "grad_norm": 1.0250927209854126, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 15049 + }, + { + "epoch": 0.1505, + "grad_norm": 1.0228712558746338, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 15050 + }, + { + "epoch": 0.15051, + "grad_norm": 0.8266756534576416, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 15051 + }, + { + "epoch": 0.15052, + "grad_norm": 0.6580228209495544, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 15052 + }, + { + "epoch": 0.15053, + "grad_norm": 0.5499899983406067, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 15053 + }, + { + "epoch": 0.15054, + "grad_norm": 0.7129097580909729, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 15054 + }, + { + "epoch": 0.15055, + "grad_norm": 0.7538021802902222, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 15055 + }, + { + "epoch": 0.15056, + "grad_norm": 0.7968039512634277, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 15056 + }, + { + "epoch": 0.15057, + "grad_norm": 0.8896482586860657, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 15057 + }, + { + "epoch": 0.15058, + "grad_norm": 0.9867523908615112, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 15058 + }, + { + "epoch": 0.15059, + "grad_norm": 1.057076334953308, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 15059 + }, + { + "epoch": 0.1506, + "grad_norm": 0.9541208148002625, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 15060 + }, + { + "epoch": 0.15061, + "grad_norm": 0.9236926436424255, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 15061 + }, + { + "epoch": 0.15062, + "grad_norm": 0.8116821050643921, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 15062 + }, + { + "epoch": 0.15063, + "grad_norm": 0.7856393456459045, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 15063 + }, + { + "epoch": 0.15064, + "grad_norm": 0.6679021120071411, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 15064 + }, + { + "epoch": 0.15065, + "grad_norm": 0.6571440100669861, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 15065 + }, + { + "epoch": 0.15066, + "grad_norm": 0.7009978890419006, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 15066 + }, + { + "epoch": 0.15067, + "grad_norm": 0.8476483821868896, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 15067 + }, + { + "epoch": 0.15068, + "grad_norm": 0.9449343681335449, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 15068 + }, + { + "epoch": 0.15069, + "grad_norm": 1.0391207933425903, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 15069 + }, + { + "epoch": 0.1507, + "grad_norm": 0.9836397767066956, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 15070 + }, + { + "epoch": 0.15071, + "grad_norm": 0.9063265323638916, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 15071 + }, + { + "epoch": 0.15072, + "grad_norm": 0.8169854879379272, + "learning_rate": 0.003, + "loss": 4.02, + "step": 15072 + }, + { + "epoch": 0.15073, + "grad_norm": 0.6318926811218262, + "learning_rate": 0.003, + "loss": 3.996, + "step": 15073 + }, + { + "epoch": 0.15074, + "grad_norm": 0.6150188446044922, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 15074 + }, + { + "epoch": 0.15075, + "grad_norm": 0.730800449848175, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 15075 + }, + { + "epoch": 0.15076, + "grad_norm": 0.7448334693908691, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 15076 + }, + { + "epoch": 0.15077, + "grad_norm": 0.6553254723548889, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 15077 + }, + { + "epoch": 0.15078, + "grad_norm": 0.7216061353683472, + "learning_rate": 0.003, + "loss": 3.988, + "step": 15078 + }, + { + "epoch": 0.15079, + "grad_norm": 0.9148663282394409, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 15079 + }, + { + "epoch": 0.1508, + "grad_norm": 1.18192458152771, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 15080 + }, + { + "epoch": 0.15081, + "grad_norm": 0.7886486053466797, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 15081 + }, + { + "epoch": 0.15082, + "grad_norm": 0.6921918988227844, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 15082 + }, + { + "epoch": 0.15083, + "grad_norm": 0.8153104186058044, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 15083 + }, + { + "epoch": 0.15084, + "grad_norm": 0.8307874798774719, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 15084 + }, + { + "epoch": 0.15085, + "grad_norm": 0.7925188541412354, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 15085 + }, + { + "epoch": 0.15086, + "grad_norm": 0.8499267101287842, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 15086 + }, + { + "epoch": 0.15087, + "grad_norm": 1.0056763887405396, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 15087 + }, + { + "epoch": 0.15088, + "grad_norm": 0.960319459438324, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 15088 + }, + { + "epoch": 0.15089, + "grad_norm": 0.9148436784744263, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 15089 + }, + { + "epoch": 0.1509, + "grad_norm": 0.9970728158950806, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 15090 + }, + { + "epoch": 0.15091, + "grad_norm": 0.9315593838691711, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 15091 + }, + { + "epoch": 0.15092, + "grad_norm": 0.8441429138183594, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 15092 + }, + { + "epoch": 0.15093, + "grad_norm": 0.646895706653595, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 15093 + }, + { + "epoch": 0.15094, + "grad_norm": 0.5070883631706238, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 15094 + }, + { + "epoch": 0.15095, + "grad_norm": 0.4848787486553192, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 15095 + }, + { + "epoch": 0.15096, + "grad_norm": 0.5808219909667969, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 15096 + }, + { + "epoch": 0.15097, + "grad_norm": 0.6804661750793457, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 15097 + }, + { + "epoch": 0.15098, + "grad_norm": 0.8220118880271912, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 15098 + }, + { + "epoch": 0.15099, + "grad_norm": 0.9969527721405029, + "learning_rate": 0.003, + "loss": 4.004, + "step": 15099 + }, + { + "epoch": 0.151, + "grad_norm": 1.0243529081344604, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 15100 + }, + { + "epoch": 0.15101, + "grad_norm": 0.8672932982444763, + "learning_rate": 0.003, + "loss": 4.006, + "step": 15101 + }, + { + "epoch": 0.15102, + "grad_norm": 1.051192283630371, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 15102 + }, + { + "epoch": 0.15103, + "grad_norm": 0.9607369899749756, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 15103 + }, + { + "epoch": 0.15104, + "grad_norm": 1.0513368844985962, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 15104 + }, + { + "epoch": 0.15105, + "grad_norm": 1.097847819328308, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 15105 + }, + { + "epoch": 0.15106, + "grad_norm": 0.9191159605979919, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 15106 + }, + { + "epoch": 0.15107, + "grad_norm": 0.8986817598342896, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 15107 + }, + { + "epoch": 0.15108, + "grad_norm": 0.8361642956733704, + "learning_rate": 0.003, + "loss": 4.026, + "step": 15108 + }, + { + "epoch": 0.15109, + "grad_norm": 0.8894073367118835, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 15109 + }, + { + "epoch": 0.1511, + "grad_norm": 0.9681217670440674, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 15110 + }, + { + "epoch": 0.15111, + "grad_norm": 1.0168449878692627, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 15111 + }, + { + "epoch": 0.15112, + "grad_norm": 0.839482843875885, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 15112 + }, + { + "epoch": 0.15113, + "grad_norm": 0.7267372012138367, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 15113 + }, + { + "epoch": 0.15114, + "grad_norm": 0.6676825881004333, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 15114 + }, + { + "epoch": 0.15115, + "grad_norm": 0.7883790135383606, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 15115 + }, + { + "epoch": 0.15116, + "grad_norm": 0.9439095258712769, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 15116 + }, + { + "epoch": 0.15117, + "grad_norm": 1.1422020196914673, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 15117 + }, + { + "epoch": 0.15118, + "grad_norm": 0.6147944331169128, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 15118 + }, + { + "epoch": 0.15119, + "grad_norm": 0.4783836603164673, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 15119 + }, + { + "epoch": 0.1512, + "grad_norm": 0.6159612536430359, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 15120 + }, + { + "epoch": 0.15121, + "grad_norm": 0.6156262159347534, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 15121 + }, + { + "epoch": 0.15122, + "grad_norm": 0.596961259841919, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 15122 + }, + { + "epoch": 0.15123, + "grad_norm": 0.5993465185165405, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 15123 + }, + { + "epoch": 0.15124, + "grad_norm": 0.4942002296447754, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 15124 + }, + { + "epoch": 0.15125, + "grad_norm": 0.5079962611198425, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 15125 + }, + { + "epoch": 0.15126, + "grad_norm": 0.5626096725463867, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 15126 + }, + { + "epoch": 0.15127, + "grad_norm": 0.6039233207702637, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 15127 + }, + { + "epoch": 0.15128, + "grad_norm": 0.5964253544807434, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 15128 + }, + { + "epoch": 0.15129, + "grad_norm": 0.7165008187294006, + "learning_rate": 0.003, + "loss": 4.026, + "step": 15129 + }, + { + "epoch": 0.1513, + "grad_norm": 0.9771832823753357, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 15130 + }, + { + "epoch": 0.15131, + "grad_norm": 1.2352522611618042, + "learning_rate": 0.003, + "loss": 4.003, + "step": 15131 + }, + { + "epoch": 0.15132, + "grad_norm": 0.7619431614875793, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 15132 + }, + { + "epoch": 0.15133, + "grad_norm": 0.6206051707267761, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 15133 + }, + { + "epoch": 0.15134, + "grad_norm": 0.6780843734741211, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 15134 + }, + { + "epoch": 0.15135, + "grad_norm": 0.8164312243461609, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 15135 + }, + { + "epoch": 0.15136, + "grad_norm": 0.9336850643157959, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 15136 + }, + { + "epoch": 0.15137, + "grad_norm": 1.0265147686004639, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 15137 + }, + { + "epoch": 0.15138, + "grad_norm": 1.0117080211639404, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 15138 + }, + { + "epoch": 0.15139, + "grad_norm": 0.806343138217926, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 15139 + }, + { + "epoch": 0.1514, + "grad_norm": 0.8026455044746399, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 15140 + }, + { + "epoch": 0.15141, + "grad_norm": 0.8498087525367737, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 15141 + }, + { + "epoch": 0.15142, + "grad_norm": 0.9997002482414246, + "learning_rate": 0.003, + "loss": 4.021, + "step": 15142 + }, + { + "epoch": 0.15143, + "grad_norm": 1.2353754043579102, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 15143 + }, + { + "epoch": 0.15144, + "grad_norm": 0.7741185426712036, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 15144 + }, + { + "epoch": 0.15145, + "grad_norm": 0.6971248388290405, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 15145 + }, + { + "epoch": 0.15146, + "grad_norm": 0.7391877770423889, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 15146 + }, + { + "epoch": 0.15147, + "grad_norm": 0.8225349187850952, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 15147 + }, + { + "epoch": 0.15148, + "grad_norm": 0.9799991250038147, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 15148 + }, + { + "epoch": 0.15149, + "grad_norm": 1.0774352550506592, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 15149 + }, + { + "epoch": 0.1515, + "grad_norm": 0.8947973251342773, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 15150 + }, + { + "epoch": 0.15151, + "grad_norm": 0.7971606850624084, + "learning_rate": 0.003, + "loss": 4.007, + "step": 15151 + }, + { + "epoch": 0.15152, + "grad_norm": 0.7421729564666748, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 15152 + }, + { + "epoch": 0.15153, + "grad_norm": 0.6884970664978027, + "learning_rate": 0.003, + "loss": 4.025, + "step": 15153 + }, + { + "epoch": 0.15154, + "grad_norm": 0.6970769166946411, + "learning_rate": 0.003, + "loss": 3.997, + "step": 15154 + }, + { + "epoch": 0.15155, + "grad_norm": 0.7541738748550415, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 15155 + }, + { + "epoch": 0.15156, + "grad_norm": 0.9252972602844238, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 15156 + }, + { + "epoch": 0.15157, + "grad_norm": 1.024196982383728, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 15157 + }, + { + "epoch": 0.15158, + "grad_norm": 0.998375415802002, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 15158 + }, + { + "epoch": 0.15159, + "grad_norm": 0.9763141870498657, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 15159 + }, + { + "epoch": 0.1516, + "grad_norm": 1.0023225545883179, + "learning_rate": 0.003, + "loss": 4.003, + "step": 15160 + }, + { + "epoch": 0.15161, + "grad_norm": 1.0663807392120361, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 15161 + }, + { + "epoch": 0.15162, + "grad_norm": 0.9437249302864075, + "learning_rate": 0.003, + "loss": 4.052, + "step": 15162 + }, + { + "epoch": 0.15163, + "grad_norm": 1.0163962841033936, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 15163 + }, + { + "epoch": 0.15164, + "grad_norm": 1.1673986911773682, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 15164 + }, + { + "epoch": 0.15165, + "grad_norm": 0.882068395614624, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 15165 + }, + { + "epoch": 0.15166, + "grad_norm": 0.7180541753768921, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 15166 + }, + { + "epoch": 0.15167, + "grad_norm": 0.6750436425209045, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 15167 + }, + { + "epoch": 0.15168, + "grad_norm": 0.6546518802642822, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 15168 + }, + { + "epoch": 0.15169, + "grad_norm": 0.6339498162269592, + "learning_rate": 0.003, + "loss": 4.002, + "step": 15169 + }, + { + "epoch": 0.1517, + "grad_norm": 0.5468326210975647, + "learning_rate": 0.003, + "loss": 3.99, + "step": 15170 + }, + { + "epoch": 0.15171, + "grad_norm": 0.5912162065505981, + "learning_rate": 0.003, + "loss": 4.012, + "step": 15171 + }, + { + "epoch": 0.15172, + "grad_norm": 0.6342115998268127, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 15172 + }, + { + "epoch": 0.15173, + "grad_norm": 0.6179934740066528, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 15173 + }, + { + "epoch": 0.15174, + "grad_norm": 0.630556583404541, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 15174 + }, + { + "epoch": 0.15175, + "grad_norm": 0.6733129024505615, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 15175 + }, + { + "epoch": 0.15176, + "grad_norm": 0.837215006351471, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 15176 + }, + { + "epoch": 0.15177, + "grad_norm": 1.103715181350708, + "learning_rate": 0.003, + "loss": 4.006, + "step": 15177 + }, + { + "epoch": 0.15178, + "grad_norm": 1.0142706632614136, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 15178 + }, + { + "epoch": 0.15179, + "grad_norm": 1.1049751043319702, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 15179 + }, + { + "epoch": 0.1518, + "grad_norm": 0.9215617775917053, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 15180 + }, + { + "epoch": 0.15181, + "grad_norm": 0.9071481823921204, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 15181 + }, + { + "epoch": 0.15182, + "grad_norm": 0.9090131521224976, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 15182 + }, + { + "epoch": 0.15183, + "grad_norm": 0.823883593082428, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 15183 + }, + { + "epoch": 0.15184, + "grad_norm": 0.8454837203025818, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 15184 + }, + { + "epoch": 0.15185, + "grad_norm": 0.7196727395057678, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 15185 + }, + { + "epoch": 0.15186, + "grad_norm": 0.8515127897262573, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 15186 + }, + { + "epoch": 0.15187, + "grad_norm": 0.847205638885498, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 15187 + }, + { + "epoch": 0.15188, + "grad_norm": 0.8074892163276672, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 15188 + }, + { + "epoch": 0.15189, + "grad_norm": 0.797053873538971, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 15189 + }, + { + "epoch": 0.1519, + "grad_norm": 0.9285748600959778, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 15190 + }, + { + "epoch": 0.15191, + "grad_norm": 0.9438667297363281, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 15191 + }, + { + "epoch": 0.15192, + "grad_norm": 0.7593942880630493, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 15192 + }, + { + "epoch": 0.15193, + "grad_norm": 0.7198943495750427, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 15193 + }, + { + "epoch": 0.15194, + "grad_norm": 0.6867886781692505, + "learning_rate": 0.003, + "loss": 4.01, + "step": 15194 + }, + { + "epoch": 0.15195, + "grad_norm": 0.6413894295692444, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 15195 + }, + { + "epoch": 0.15196, + "grad_norm": 0.8065673112869263, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 15196 + }, + { + "epoch": 0.15197, + "grad_norm": 0.8199344277381897, + "learning_rate": 0.003, + "loss": 4.0, + "step": 15197 + }, + { + "epoch": 0.15198, + "grad_norm": 0.7948529720306396, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 15198 + }, + { + "epoch": 0.15199, + "grad_norm": 0.7588329911231995, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15199 + }, + { + "epoch": 0.152, + "grad_norm": 0.759814441204071, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 15200 + }, + { + "epoch": 0.15201, + "grad_norm": 0.8657602071762085, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 15201 + }, + { + "epoch": 0.15202, + "grad_norm": 0.809902548789978, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 15202 + }, + { + "epoch": 0.15203, + "grad_norm": 0.7706008553504944, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 15203 + }, + { + "epoch": 0.15204, + "grad_norm": 0.8675158023834229, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 15204 + }, + { + "epoch": 0.15205, + "grad_norm": 0.9507919549942017, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 15205 + }, + { + "epoch": 0.15206, + "grad_norm": 0.9662113189697266, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 15206 + }, + { + "epoch": 0.15207, + "grad_norm": 0.886137843132019, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 15207 + }, + { + "epoch": 0.15208, + "grad_norm": 0.8959958553314209, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 15208 + }, + { + "epoch": 0.15209, + "grad_norm": 0.9112538695335388, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 15209 + }, + { + "epoch": 0.1521, + "grad_norm": 1.0017379522323608, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 15210 + }, + { + "epoch": 0.15211, + "grad_norm": 1.1133248805999756, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 15211 + }, + { + "epoch": 0.15212, + "grad_norm": 0.7938923239707947, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 15212 + }, + { + "epoch": 0.15213, + "grad_norm": 0.7378742694854736, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 15213 + }, + { + "epoch": 0.15214, + "grad_norm": 0.8953700661659241, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 15214 + }, + { + "epoch": 0.15215, + "grad_norm": 0.9183306694030762, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 15215 + }, + { + "epoch": 0.15216, + "grad_norm": 0.9317597150802612, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 15216 + }, + { + "epoch": 0.15217, + "grad_norm": 0.9850735664367676, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 15217 + }, + { + "epoch": 0.15218, + "grad_norm": 1.0636837482452393, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 15218 + }, + { + "epoch": 0.15219, + "grad_norm": 1.0994192361831665, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 15219 + }, + { + "epoch": 0.1522, + "grad_norm": 1.2374557256698608, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 15220 + }, + { + "epoch": 0.15221, + "grad_norm": 0.9231405258178711, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 15221 + }, + { + "epoch": 0.15222, + "grad_norm": 0.9628676772117615, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 15222 + }, + { + "epoch": 0.15223, + "grad_norm": 0.9543718695640564, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 15223 + }, + { + "epoch": 0.15224, + "grad_norm": 1.016852617263794, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 15224 + }, + { + "epoch": 0.15225, + "grad_norm": 0.9168770909309387, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 15225 + }, + { + "epoch": 0.15226, + "grad_norm": 0.6900375485420227, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 15226 + }, + { + "epoch": 0.15227, + "grad_norm": 0.5902383327484131, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 15227 + }, + { + "epoch": 0.15228, + "grad_norm": 0.5826788544654846, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 15228 + }, + { + "epoch": 0.15229, + "grad_norm": 0.5453304648399353, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 15229 + }, + { + "epoch": 0.1523, + "grad_norm": 0.4619797170162201, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 15230 + }, + { + "epoch": 0.15231, + "grad_norm": 0.45512986183166504, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 15231 + }, + { + "epoch": 0.15232, + "grad_norm": 0.4618683457374573, + "learning_rate": 0.003, + "loss": 3.991, + "step": 15232 + }, + { + "epoch": 0.15233, + "grad_norm": 0.5123746991157532, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 15233 + }, + { + "epoch": 0.15234, + "grad_norm": 0.6983823180198669, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 15234 + }, + { + "epoch": 0.15235, + "grad_norm": 1.0241881608963013, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 15235 + }, + { + "epoch": 0.15236, + "grad_norm": 1.2229256629943848, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 15236 + }, + { + "epoch": 0.15237, + "grad_norm": 0.7741057276725769, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 15237 + }, + { + "epoch": 0.15238, + "grad_norm": 0.6281371116638184, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 15238 + }, + { + "epoch": 0.15239, + "grad_norm": 0.6690221428871155, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 15239 + }, + { + "epoch": 0.1524, + "grad_norm": 0.7069966197013855, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 15240 + }, + { + "epoch": 0.15241, + "grad_norm": 0.6972704529762268, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 15241 + }, + { + "epoch": 0.15242, + "grad_norm": 0.6901689767837524, + "learning_rate": 0.003, + "loss": 4.028, + "step": 15242 + }, + { + "epoch": 0.15243, + "grad_norm": 0.7443487048149109, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 15243 + }, + { + "epoch": 0.15244, + "grad_norm": 0.7118276357650757, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 15244 + }, + { + "epoch": 0.15245, + "grad_norm": 0.7309918403625488, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 15245 + }, + { + "epoch": 0.15246, + "grad_norm": 0.7601751089096069, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 15246 + }, + { + "epoch": 0.15247, + "grad_norm": 0.8328369855880737, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 15247 + }, + { + "epoch": 0.15248, + "grad_norm": 0.9573512673377991, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 15248 + }, + { + "epoch": 0.15249, + "grad_norm": 0.9210628867149353, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 15249 + }, + { + "epoch": 0.1525, + "grad_norm": 0.8686948418617249, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 15250 + }, + { + "epoch": 0.15251, + "grad_norm": 0.8697372078895569, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 15251 + }, + { + "epoch": 0.15252, + "grad_norm": 0.8911150693893433, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 15252 + }, + { + "epoch": 0.15253, + "grad_norm": 0.8357546329498291, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 15253 + }, + { + "epoch": 0.15254, + "grad_norm": 0.7480366230010986, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 15254 + }, + { + "epoch": 0.15255, + "grad_norm": 0.6970673203468323, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 15255 + }, + { + "epoch": 0.15256, + "grad_norm": 0.6920578479766846, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 15256 + }, + { + "epoch": 0.15257, + "grad_norm": 0.7250205278396606, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 15257 + }, + { + "epoch": 0.15258, + "grad_norm": 0.7733274102210999, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 15258 + }, + { + "epoch": 0.15259, + "grad_norm": 0.8828164935112, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 15259 + }, + { + "epoch": 0.1526, + "grad_norm": 0.8882545828819275, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 15260 + }, + { + "epoch": 0.15261, + "grad_norm": 0.9990447163581848, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 15261 + }, + { + "epoch": 0.15262, + "grad_norm": 1.3572981357574463, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 15262 + }, + { + "epoch": 0.15263, + "grad_norm": 0.7818669080734253, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 15263 + }, + { + "epoch": 0.15264, + "grad_norm": 0.8175413608551025, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 15264 + }, + { + "epoch": 0.15265, + "grad_norm": 0.9246871471405029, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 15265 + }, + { + "epoch": 0.15266, + "grad_norm": 1.0359551906585693, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 15266 + }, + { + "epoch": 0.15267, + "grad_norm": 1.083351731300354, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 15267 + }, + { + "epoch": 0.15268, + "grad_norm": 0.9957295060157776, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 15268 + }, + { + "epoch": 0.15269, + "grad_norm": 0.999646008014679, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 15269 + }, + { + "epoch": 0.1527, + "grad_norm": 0.8644801378250122, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 15270 + }, + { + "epoch": 0.15271, + "grad_norm": 0.6981543302536011, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 15271 + }, + { + "epoch": 0.15272, + "grad_norm": 0.6990094184875488, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 15272 + }, + { + "epoch": 0.15273, + "grad_norm": 0.9216434955596924, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 15273 + }, + { + "epoch": 0.15274, + "grad_norm": 0.9980815052986145, + "learning_rate": 0.003, + "loss": 4.005, + "step": 15274 + }, + { + "epoch": 0.15275, + "grad_norm": 1.024815320968628, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 15275 + }, + { + "epoch": 0.15276, + "grad_norm": 0.9341852068901062, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 15276 + }, + { + "epoch": 0.15277, + "grad_norm": 0.952576756477356, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 15277 + }, + { + "epoch": 0.15278, + "grad_norm": 0.8232364654541016, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15278 + }, + { + "epoch": 0.15279, + "grad_norm": 0.6267227530479431, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 15279 + }, + { + "epoch": 0.1528, + "grad_norm": 0.706098198890686, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 15280 + }, + { + "epoch": 0.15281, + "grad_norm": 0.7639279365539551, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 15281 + }, + { + "epoch": 0.15282, + "grad_norm": 0.8171020746231079, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 15282 + }, + { + "epoch": 0.15283, + "grad_norm": 0.8467424511909485, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 15283 + }, + { + "epoch": 0.15284, + "grad_norm": 0.8294116258621216, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 15284 + }, + { + "epoch": 0.15285, + "grad_norm": 0.8115954399108887, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 15285 + }, + { + "epoch": 0.15286, + "grad_norm": 0.789169192314148, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 15286 + }, + { + "epoch": 0.15287, + "grad_norm": 0.6780423521995544, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 15287 + }, + { + "epoch": 0.15288, + "grad_norm": 0.5915562510490417, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 15288 + }, + { + "epoch": 0.15289, + "grad_norm": 0.5556778907775879, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 15289 + }, + { + "epoch": 0.1529, + "grad_norm": 0.7075419425964355, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 15290 + }, + { + "epoch": 0.15291, + "grad_norm": 0.82306969165802, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 15291 + }, + { + "epoch": 0.15292, + "grad_norm": 0.8850873112678528, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 15292 + }, + { + "epoch": 0.15293, + "grad_norm": 0.8600744009017944, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 15293 + }, + { + "epoch": 0.15294, + "grad_norm": 0.8272354602813721, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 15294 + }, + { + "epoch": 0.15295, + "grad_norm": 0.9683411717414856, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 15295 + }, + { + "epoch": 0.15296, + "grad_norm": 1.0481081008911133, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 15296 + }, + { + "epoch": 0.15297, + "grad_norm": 0.9039106369018555, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 15297 + }, + { + "epoch": 0.15298, + "grad_norm": 1.220619797706604, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 15298 + }, + { + "epoch": 0.15299, + "grad_norm": 0.8804208636283875, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 15299 + }, + { + "epoch": 0.153, + "grad_norm": 0.6494994759559631, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 15300 + }, + { + "epoch": 0.15301, + "grad_norm": 0.6240890026092529, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 15301 + }, + { + "epoch": 0.15302, + "grad_norm": 0.5636361837387085, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 15302 + }, + { + "epoch": 0.15303, + "grad_norm": 0.6072368025779724, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 15303 + }, + { + "epoch": 0.15304, + "grad_norm": 0.6338242292404175, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 15304 + }, + { + "epoch": 0.15305, + "grad_norm": 0.6941981315612793, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 15305 + }, + { + "epoch": 0.15306, + "grad_norm": 0.6970587968826294, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 15306 + }, + { + "epoch": 0.15307, + "grad_norm": 0.7661889791488647, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 15307 + }, + { + "epoch": 0.15308, + "grad_norm": 0.9541347622871399, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 15308 + }, + { + "epoch": 0.15309, + "grad_norm": 1.085623860359192, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 15309 + }, + { + "epoch": 0.1531, + "grad_norm": 0.8310011029243469, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 15310 + }, + { + "epoch": 0.15311, + "grad_norm": 0.7221589088439941, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 15311 + }, + { + "epoch": 0.15312, + "grad_norm": 0.747313380241394, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 15312 + }, + { + "epoch": 0.15313, + "grad_norm": 0.8484153151512146, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 15313 + }, + { + "epoch": 0.15314, + "grad_norm": 0.8482285141944885, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 15314 + }, + { + "epoch": 0.15315, + "grad_norm": 0.7901672720909119, + "learning_rate": 0.003, + "loss": 4.024, + "step": 15315 + }, + { + "epoch": 0.15316, + "grad_norm": 0.8714357018470764, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 15316 + }, + { + "epoch": 0.15317, + "grad_norm": 0.8399065136909485, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 15317 + }, + { + "epoch": 0.15318, + "grad_norm": 0.9220929145812988, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 15318 + }, + { + "epoch": 0.15319, + "grad_norm": 0.9695712924003601, + "learning_rate": 0.003, + "loss": 3.992, + "step": 15319 + }, + { + "epoch": 0.1532, + "grad_norm": 1.101618766784668, + "learning_rate": 0.003, + "loss": 4.017, + "step": 15320 + }, + { + "epoch": 0.15321, + "grad_norm": 0.9492210149765015, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 15321 + }, + { + "epoch": 0.15322, + "grad_norm": 0.9337678551673889, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 15322 + }, + { + "epoch": 0.15323, + "grad_norm": 1.0097078084945679, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 15323 + }, + { + "epoch": 0.15324, + "grad_norm": 0.9319543242454529, + "learning_rate": 0.003, + "loss": 4.015, + "step": 15324 + }, + { + "epoch": 0.15325, + "grad_norm": 0.8490461707115173, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 15325 + }, + { + "epoch": 0.15326, + "grad_norm": 0.806574285030365, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 15326 + }, + { + "epoch": 0.15327, + "grad_norm": 0.7265695333480835, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 15327 + }, + { + "epoch": 0.15328, + "grad_norm": 0.7705208659172058, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 15328 + }, + { + "epoch": 0.15329, + "grad_norm": 1.0969406366348267, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 15329 + }, + { + "epoch": 0.1533, + "grad_norm": 1.1445393562316895, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 15330 + }, + { + "epoch": 0.15331, + "grad_norm": 0.8153518438339233, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 15331 + }, + { + "epoch": 0.15332, + "grad_norm": 0.6289441585540771, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 15332 + }, + { + "epoch": 0.15333, + "grad_norm": 0.48444995284080505, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 15333 + }, + { + "epoch": 0.15334, + "grad_norm": 0.5958626866340637, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 15334 + }, + { + "epoch": 0.15335, + "grad_norm": 0.7594508528709412, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 15335 + }, + { + "epoch": 0.15336, + "grad_norm": 0.9328643083572388, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15336 + }, + { + "epoch": 0.15337, + "grad_norm": 1.0968477725982666, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 15337 + }, + { + "epoch": 0.15338, + "grad_norm": 0.8232316970825195, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 15338 + }, + { + "epoch": 0.15339, + "grad_norm": 0.6940597891807556, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 15339 + }, + { + "epoch": 0.1534, + "grad_norm": 0.6054121255874634, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 15340 + }, + { + "epoch": 0.15341, + "grad_norm": 0.6718443632125854, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 15341 + }, + { + "epoch": 0.15342, + "grad_norm": 0.8152159452438354, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 15342 + }, + { + "epoch": 0.15343, + "grad_norm": 1.1432894468307495, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 15343 + }, + { + "epoch": 0.15344, + "grad_norm": 0.9310418367385864, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 15344 + }, + { + "epoch": 0.15345, + "grad_norm": 0.7084609866142273, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 15345 + }, + { + "epoch": 0.15346, + "grad_norm": 0.6835572123527527, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 15346 + }, + { + "epoch": 0.15347, + "grad_norm": 0.6644852161407471, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 15347 + }, + { + "epoch": 0.15348, + "grad_norm": 0.6577886939048767, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 15348 + }, + { + "epoch": 0.15349, + "grad_norm": 0.7408700585365295, + "learning_rate": 0.003, + "loss": 4.008, + "step": 15349 + }, + { + "epoch": 0.1535, + "grad_norm": 0.8827539682388306, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 15350 + }, + { + "epoch": 0.15351, + "grad_norm": 0.8726819157600403, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 15351 + }, + { + "epoch": 0.15352, + "grad_norm": 0.8914332389831543, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 15352 + }, + { + "epoch": 0.15353, + "grad_norm": 0.9129746556282043, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 15353 + }, + { + "epoch": 0.15354, + "grad_norm": 0.9270063638687134, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 15354 + }, + { + "epoch": 0.15355, + "grad_norm": 0.9356761574745178, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 15355 + }, + { + "epoch": 0.15356, + "grad_norm": 0.947665810585022, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 15356 + }, + { + "epoch": 0.15357, + "grad_norm": 0.8571285009384155, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 15357 + }, + { + "epoch": 0.15358, + "grad_norm": 0.8978291749954224, + "learning_rate": 0.003, + "loss": 4.004, + "step": 15358 + }, + { + "epoch": 0.15359, + "grad_norm": 0.966492772102356, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 15359 + }, + { + "epoch": 0.1536, + "grad_norm": 0.8793907165527344, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 15360 + }, + { + "epoch": 0.15361, + "grad_norm": 0.8608678579330444, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 15361 + }, + { + "epoch": 0.15362, + "grad_norm": 0.7903480529785156, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 15362 + }, + { + "epoch": 0.15363, + "grad_norm": 0.8036578297615051, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 15363 + }, + { + "epoch": 0.15364, + "grad_norm": 0.8505967855453491, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 15364 + }, + { + "epoch": 0.15365, + "grad_norm": 0.9503588080406189, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 15365 + }, + { + "epoch": 0.15366, + "grad_norm": 0.759649395942688, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 15366 + }, + { + "epoch": 0.15367, + "grad_norm": 0.6503275036811829, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 15367 + }, + { + "epoch": 0.15368, + "grad_norm": 0.6055623888969421, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 15368 + }, + { + "epoch": 0.15369, + "grad_norm": 0.5819411277770996, + "learning_rate": 0.003, + "loss": 3.994, + "step": 15369 + }, + { + "epoch": 0.1537, + "grad_norm": 0.6257475018501282, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 15370 + }, + { + "epoch": 0.15371, + "grad_norm": 0.6066151261329651, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 15371 + }, + { + "epoch": 0.15372, + "grad_norm": 0.623354434967041, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 15372 + }, + { + "epoch": 0.15373, + "grad_norm": 0.6460627913475037, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 15373 + }, + { + "epoch": 0.15374, + "grad_norm": 0.741217851638794, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15374 + }, + { + "epoch": 0.15375, + "grad_norm": 0.9240460991859436, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 15375 + }, + { + "epoch": 0.15376, + "grad_norm": 1.118343472480774, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 15376 + }, + { + "epoch": 0.15377, + "grad_norm": 0.9606201648712158, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 15377 + }, + { + "epoch": 0.15378, + "grad_norm": 0.9432666301727295, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 15378 + }, + { + "epoch": 0.15379, + "grad_norm": 0.9836307764053345, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 15379 + }, + { + "epoch": 0.1538, + "grad_norm": 0.9370342493057251, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 15380 + }, + { + "epoch": 0.15381, + "grad_norm": 1.0143237113952637, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 15381 + }, + { + "epoch": 0.15382, + "grad_norm": 1.0634740591049194, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 15382 + }, + { + "epoch": 0.15383, + "grad_norm": 1.0176247358322144, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 15383 + }, + { + "epoch": 0.15384, + "grad_norm": 0.8548615574836731, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 15384 + }, + { + "epoch": 0.15385, + "grad_norm": 0.8597134947776794, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 15385 + }, + { + "epoch": 0.15386, + "grad_norm": 0.9391376376152039, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 15386 + }, + { + "epoch": 0.15387, + "grad_norm": 0.9654890894889832, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 15387 + }, + { + "epoch": 0.15388, + "grad_norm": 0.9590594172477722, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 15388 + }, + { + "epoch": 0.15389, + "grad_norm": 0.8145045638084412, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 15389 + }, + { + "epoch": 0.1539, + "grad_norm": 0.8208436369895935, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 15390 + }, + { + "epoch": 0.15391, + "grad_norm": 0.8954910635948181, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 15391 + }, + { + "epoch": 0.15392, + "grad_norm": 0.9844794869422913, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 15392 + }, + { + "epoch": 0.15393, + "grad_norm": 1.1418054103851318, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 15393 + }, + { + "epoch": 0.15394, + "grad_norm": 0.8355457186698914, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 15394 + }, + { + "epoch": 0.15395, + "grad_norm": 0.7603216171264648, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 15395 + }, + { + "epoch": 0.15396, + "grad_norm": 0.6772087216377258, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 15396 + }, + { + "epoch": 0.15397, + "grad_norm": 0.8091949820518494, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 15397 + }, + { + "epoch": 0.15398, + "grad_norm": 0.8611405491828918, + "learning_rate": 0.003, + "loss": 3.996, + "step": 15398 + }, + { + "epoch": 0.15399, + "grad_norm": 0.8038859963417053, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 15399 + }, + { + "epoch": 0.154, + "grad_norm": 0.7616530060768127, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 15400 + }, + { + "epoch": 0.15401, + "grad_norm": 0.846915602684021, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 15401 + }, + { + "epoch": 0.15402, + "grad_norm": 0.899276852607727, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 15402 + }, + { + "epoch": 0.15403, + "grad_norm": 1.0511224269866943, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 15403 + }, + { + "epoch": 0.15404, + "grad_norm": 1.0797245502471924, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 15404 + }, + { + "epoch": 0.15405, + "grad_norm": 0.9076430201530457, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 15405 + }, + { + "epoch": 0.15406, + "grad_norm": 0.7749274969100952, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 15406 + }, + { + "epoch": 0.15407, + "grad_norm": 0.6954168677330017, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 15407 + }, + { + "epoch": 0.15408, + "grad_norm": 0.6638758182525635, + "learning_rate": 0.003, + "loss": 4.023, + "step": 15408 + }, + { + "epoch": 0.15409, + "grad_norm": 0.6848050355911255, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 15409 + }, + { + "epoch": 0.1541, + "grad_norm": 0.7466670870780945, + "learning_rate": 0.003, + "loss": 4.022, + "step": 15410 + }, + { + "epoch": 0.15411, + "grad_norm": 0.7404317855834961, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 15411 + }, + { + "epoch": 0.15412, + "grad_norm": 0.7153498530387878, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 15412 + }, + { + "epoch": 0.15413, + "grad_norm": 0.7007918357849121, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 15413 + }, + { + "epoch": 0.15414, + "grad_norm": 0.5969216227531433, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 15414 + }, + { + "epoch": 0.15415, + "grad_norm": 0.5652239918708801, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 15415 + }, + { + "epoch": 0.15416, + "grad_norm": 0.5697609782218933, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 15416 + }, + { + "epoch": 0.15417, + "grad_norm": 0.6852931380271912, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 15417 + }, + { + "epoch": 0.15418, + "grad_norm": 0.8129041790962219, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 15418 + }, + { + "epoch": 0.15419, + "grad_norm": 0.8532951474189758, + "learning_rate": 0.003, + "loss": 3.996, + "step": 15419 + }, + { + "epoch": 0.1542, + "grad_norm": 0.8256349563598633, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 15420 + }, + { + "epoch": 0.15421, + "grad_norm": 0.7314229011535645, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 15421 + }, + { + "epoch": 0.15422, + "grad_norm": 0.7831211090087891, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 15422 + }, + { + "epoch": 0.15423, + "grad_norm": 0.7848354578018188, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 15423 + }, + { + "epoch": 0.15424, + "grad_norm": 0.8499462008476257, + "learning_rate": 0.003, + "loss": 4.008, + "step": 15424 + }, + { + "epoch": 0.15425, + "grad_norm": 0.8613630533218384, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 15425 + }, + { + "epoch": 0.15426, + "grad_norm": 0.9698076248168945, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 15426 + }, + { + "epoch": 0.15427, + "grad_norm": 1.0350285768508911, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 15427 + }, + { + "epoch": 0.15428, + "grad_norm": 1.0963178873062134, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 15428 + }, + { + "epoch": 0.15429, + "grad_norm": 0.8511303067207336, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 15429 + }, + { + "epoch": 0.1543, + "grad_norm": 0.6375939249992371, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 15430 + }, + { + "epoch": 0.15431, + "grad_norm": 0.6718809604644775, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 15431 + }, + { + "epoch": 0.15432, + "grad_norm": 0.7174335718154907, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 15432 + }, + { + "epoch": 0.15433, + "grad_norm": 0.7369046211242676, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 15433 + }, + { + "epoch": 0.15434, + "grad_norm": 0.7658149600028992, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 15434 + }, + { + "epoch": 0.15435, + "grad_norm": 0.9825096130371094, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 15435 + }, + { + "epoch": 0.15436, + "grad_norm": 1.0927093029022217, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 15436 + }, + { + "epoch": 0.15437, + "grad_norm": 0.9758725762367249, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 15437 + }, + { + "epoch": 0.15438, + "grad_norm": 1.1830312013626099, + "learning_rate": 0.003, + "loss": 4.047, + "step": 15438 + }, + { + "epoch": 0.15439, + "grad_norm": 1.0404443740844727, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 15439 + }, + { + "epoch": 0.1544, + "grad_norm": 0.884275496006012, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 15440 + }, + { + "epoch": 0.15441, + "grad_norm": 0.7949202060699463, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 15441 + }, + { + "epoch": 0.15442, + "grad_norm": 0.6556805372238159, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 15442 + }, + { + "epoch": 0.15443, + "grad_norm": 0.6673168540000916, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 15443 + }, + { + "epoch": 0.15444, + "grad_norm": 0.6797085404396057, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 15444 + }, + { + "epoch": 0.15445, + "grad_norm": 0.5978509187698364, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 15445 + }, + { + "epoch": 0.15446, + "grad_norm": 0.6056724190711975, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 15446 + }, + { + "epoch": 0.15447, + "grad_norm": 0.6114022731781006, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 15447 + }, + { + "epoch": 0.15448, + "grad_norm": 0.5594829320907593, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 15448 + }, + { + "epoch": 0.15449, + "grad_norm": 0.7436548471450806, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 15449 + }, + { + "epoch": 0.1545, + "grad_norm": 0.8702318072319031, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 15450 + }, + { + "epoch": 0.15451, + "grad_norm": 1.0084431171417236, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 15451 + }, + { + "epoch": 0.15452, + "grad_norm": 1.1599823236465454, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 15452 + }, + { + "epoch": 0.15453, + "grad_norm": 0.8092561364173889, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 15453 + }, + { + "epoch": 0.15454, + "grad_norm": 0.7046549320220947, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 15454 + }, + { + "epoch": 0.15455, + "grad_norm": 0.5919784903526306, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 15455 + }, + { + "epoch": 0.15456, + "grad_norm": 0.617529034614563, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 15456 + }, + { + "epoch": 0.15457, + "grad_norm": 0.6455297470092773, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 15457 + }, + { + "epoch": 0.15458, + "grad_norm": 0.631838858127594, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 15458 + }, + { + "epoch": 0.15459, + "grad_norm": 0.6798779368400574, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 15459 + }, + { + "epoch": 0.1546, + "grad_norm": 0.7333508133888245, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 15460 + }, + { + "epoch": 0.15461, + "grad_norm": 0.8497591614723206, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 15461 + }, + { + "epoch": 0.15462, + "grad_norm": 1.0627999305725098, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 15462 + }, + { + "epoch": 0.15463, + "grad_norm": 1.0860134363174438, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 15463 + }, + { + "epoch": 0.15464, + "grad_norm": 0.8445022106170654, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 15464 + }, + { + "epoch": 0.15465, + "grad_norm": 0.7499777674674988, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 15465 + }, + { + "epoch": 0.15466, + "grad_norm": 0.6469881534576416, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 15466 + }, + { + "epoch": 0.15467, + "grad_norm": 0.6957772374153137, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 15467 + }, + { + "epoch": 0.15468, + "grad_norm": 0.7566899657249451, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 15468 + }, + { + "epoch": 0.15469, + "grad_norm": 0.7406789064407349, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 15469 + }, + { + "epoch": 0.1547, + "grad_norm": 0.6400657296180725, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 15470 + }, + { + "epoch": 0.15471, + "grad_norm": 0.7208961844444275, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 15471 + }, + { + "epoch": 0.15472, + "grad_norm": 0.7523152828216553, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 15472 + }, + { + "epoch": 0.15473, + "grad_norm": 0.7218494415283203, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 15473 + }, + { + "epoch": 0.15474, + "grad_norm": 0.7980036735534668, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 15474 + }, + { + "epoch": 0.15475, + "grad_norm": 0.939830482006073, + "learning_rate": 0.003, + "loss": 4.015, + "step": 15475 + }, + { + "epoch": 0.15476, + "grad_norm": 1.0078206062316895, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 15476 + }, + { + "epoch": 0.15477, + "grad_norm": 0.9280356168746948, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 15477 + }, + { + "epoch": 0.15478, + "grad_norm": 0.9465794563293457, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 15478 + }, + { + "epoch": 0.15479, + "grad_norm": 0.8881992697715759, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 15479 + }, + { + "epoch": 0.1548, + "grad_norm": 1.0140552520751953, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 15480 + }, + { + "epoch": 0.15481, + "grad_norm": 1.0695966482162476, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 15481 + }, + { + "epoch": 0.15482, + "grad_norm": 0.9526455402374268, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 15482 + }, + { + "epoch": 0.15483, + "grad_norm": 1.1447579860687256, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 15483 + }, + { + "epoch": 0.15484, + "grad_norm": 0.8123161792755127, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 15484 + }, + { + "epoch": 0.15485, + "grad_norm": 0.8708415031433105, + "learning_rate": 0.003, + "loss": 4.039, + "step": 15485 + }, + { + "epoch": 0.15486, + "grad_norm": 0.9243865013122559, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 15486 + }, + { + "epoch": 0.15487, + "grad_norm": 1.0781525373458862, + "learning_rate": 0.003, + "loss": 4.022, + "step": 15487 + }, + { + "epoch": 0.15488, + "grad_norm": 0.9679509997367859, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 15488 + }, + { + "epoch": 0.15489, + "grad_norm": 0.8731271028518677, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 15489 + }, + { + "epoch": 0.1549, + "grad_norm": 0.7739214897155762, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 15490 + }, + { + "epoch": 0.15491, + "grad_norm": 0.9619302153587341, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 15491 + }, + { + "epoch": 0.15492, + "grad_norm": 0.9622506499290466, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 15492 + }, + { + "epoch": 0.15493, + "grad_norm": 0.9416058659553528, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 15493 + }, + { + "epoch": 0.15494, + "grad_norm": 0.911348283290863, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 15494 + }, + { + "epoch": 0.15495, + "grad_norm": 0.9651471972465515, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 15495 + }, + { + "epoch": 0.15496, + "grad_norm": 0.8875364661216736, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 15496 + }, + { + "epoch": 0.15497, + "grad_norm": 1.0809496641159058, + "learning_rate": 0.003, + "loss": 4.023, + "step": 15497 + }, + { + "epoch": 0.15498, + "grad_norm": 0.9345546960830688, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 15498 + }, + { + "epoch": 0.15499, + "grad_norm": 0.8077123165130615, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 15499 + }, + { + "epoch": 0.155, + "grad_norm": 0.8043227195739746, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 15500 + }, + { + "epoch": 0.15501, + "grad_norm": 0.916532576084137, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 15501 + }, + { + "epoch": 0.15502, + "grad_norm": 1.014280915260315, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 15502 + }, + { + "epoch": 0.15503, + "grad_norm": 1.0342527627944946, + "learning_rate": 0.003, + "loss": 4.014, + "step": 15503 + }, + { + "epoch": 0.15504, + "grad_norm": 0.8955778479576111, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 15504 + }, + { + "epoch": 0.15505, + "grad_norm": 0.8519617915153503, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 15505 + }, + { + "epoch": 0.15506, + "grad_norm": 0.7919250130653381, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 15506 + }, + { + "epoch": 0.15507, + "grad_norm": 0.7271054983139038, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 15507 + }, + { + "epoch": 0.15508, + "grad_norm": 0.7064192295074463, + "learning_rate": 0.003, + "loss": 4.007, + "step": 15508 + }, + { + "epoch": 0.15509, + "grad_norm": 0.7795794606208801, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 15509 + }, + { + "epoch": 0.1551, + "grad_norm": 0.8523615002632141, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 15510 + }, + { + "epoch": 0.15511, + "grad_norm": 0.8650283813476562, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 15511 + }, + { + "epoch": 0.15512, + "grad_norm": 0.9879773855209351, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 15512 + }, + { + "epoch": 0.15513, + "grad_norm": 1.2060025930404663, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 15513 + }, + { + "epoch": 0.15514, + "grad_norm": 0.6860678195953369, + "learning_rate": 0.003, + "loss": 4.04, + "step": 15514 + }, + { + "epoch": 0.15515, + "grad_norm": 0.6115677952766418, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 15515 + }, + { + "epoch": 0.15516, + "grad_norm": 0.5774584412574768, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 15516 + }, + { + "epoch": 0.15517, + "grad_norm": 0.6202374696731567, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 15517 + }, + { + "epoch": 0.15518, + "grad_norm": 0.5926021933555603, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 15518 + }, + { + "epoch": 0.15519, + "grad_norm": 0.6063708662986755, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 15519 + }, + { + "epoch": 0.1552, + "grad_norm": 0.5697833299636841, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 15520 + }, + { + "epoch": 0.15521, + "grad_norm": 0.6202701330184937, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 15521 + }, + { + "epoch": 0.15522, + "grad_norm": 0.7257983684539795, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 15522 + }, + { + "epoch": 0.15523, + "grad_norm": 0.8819434642791748, + "learning_rate": 0.003, + "loss": 4.026, + "step": 15523 + }, + { + "epoch": 0.15524, + "grad_norm": 0.9893710017204285, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 15524 + }, + { + "epoch": 0.15525, + "grad_norm": 0.8953197002410889, + "learning_rate": 0.003, + "loss": 4.012, + "step": 15525 + }, + { + "epoch": 0.15526, + "grad_norm": 0.7468194961547852, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 15526 + }, + { + "epoch": 0.15527, + "grad_norm": 0.7914648652076721, + "learning_rate": 0.003, + "loss": 3.998, + "step": 15527 + }, + { + "epoch": 0.15528, + "grad_norm": 0.6238877177238464, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 15528 + }, + { + "epoch": 0.15529, + "grad_norm": 0.6196391582489014, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 15529 + }, + { + "epoch": 0.1553, + "grad_norm": 0.726754903793335, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 15530 + }, + { + "epoch": 0.15531, + "grad_norm": 0.7787684202194214, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 15531 + }, + { + "epoch": 0.15532, + "grad_norm": 0.8006848096847534, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 15532 + }, + { + "epoch": 0.15533, + "grad_norm": 0.7540202736854553, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 15533 + }, + { + "epoch": 0.15534, + "grad_norm": 0.836521565914154, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 15534 + }, + { + "epoch": 0.15535, + "grad_norm": 0.9272993803024292, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 15535 + }, + { + "epoch": 0.15536, + "grad_norm": 1.074102759361267, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 15536 + }, + { + "epoch": 0.15537, + "grad_norm": 1.1279255151748657, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 15537 + }, + { + "epoch": 0.15538, + "grad_norm": 1.0611523389816284, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 15538 + }, + { + "epoch": 0.15539, + "grad_norm": 0.7072646617889404, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 15539 + }, + { + "epoch": 0.1554, + "grad_norm": 0.7035132050514221, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 15540 + }, + { + "epoch": 0.15541, + "grad_norm": 0.8227235078811646, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 15541 + }, + { + "epoch": 0.15542, + "grad_norm": 0.9182822108268738, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 15542 + }, + { + "epoch": 0.15543, + "grad_norm": 1.0358269214630127, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 15543 + }, + { + "epoch": 0.15544, + "grad_norm": 0.8956734538078308, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 15544 + }, + { + "epoch": 0.15545, + "grad_norm": 0.7539672255516052, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 15545 + }, + { + "epoch": 0.15546, + "grad_norm": 0.7827208638191223, + "learning_rate": 0.003, + "loss": 4.032, + "step": 15546 + }, + { + "epoch": 0.15547, + "grad_norm": 0.7323380708694458, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 15547 + }, + { + "epoch": 0.15548, + "grad_norm": 0.6001154184341431, + "learning_rate": 0.003, + "loss": 4.018, + "step": 15548 + }, + { + "epoch": 0.15549, + "grad_norm": 0.6410365700721741, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 15549 + }, + { + "epoch": 0.1555, + "grad_norm": 0.633104145526886, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 15550 + }, + { + "epoch": 0.15551, + "grad_norm": 0.6054565906524658, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 15551 + }, + { + "epoch": 0.15552, + "grad_norm": 0.6131342649459839, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 15552 + }, + { + "epoch": 0.15553, + "grad_norm": 0.6478719711303711, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 15553 + }, + { + "epoch": 0.15554, + "grad_norm": 0.6694060564041138, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 15554 + }, + { + "epoch": 0.15555, + "grad_norm": 0.6657654047012329, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 15555 + }, + { + "epoch": 0.15556, + "grad_norm": 0.7979192137718201, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 15556 + }, + { + "epoch": 0.15557, + "grad_norm": 1.0289673805236816, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 15557 + }, + { + "epoch": 0.15558, + "grad_norm": 1.0689125061035156, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 15558 + }, + { + "epoch": 0.15559, + "grad_norm": 0.9275684356689453, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 15559 + }, + { + "epoch": 0.1556, + "grad_norm": 0.934004008769989, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 15560 + }, + { + "epoch": 0.15561, + "grad_norm": 0.8707156777381897, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 15561 + }, + { + "epoch": 0.15562, + "grad_norm": 0.8118012547492981, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 15562 + }, + { + "epoch": 0.15563, + "grad_norm": 1.2029263973236084, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 15563 + }, + { + "epoch": 0.15564, + "grad_norm": 1.1070818901062012, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 15564 + }, + { + "epoch": 0.15565, + "grad_norm": 0.6871122717857361, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 15565 + }, + { + "epoch": 0.15566, + "grad_norm": 0.6614555716514587, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 15566 + }, + { + "epoch": 0.15567, + "grad_norm": 0.7378991842269897, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 15567 + }, + { + "epoch": 0.15568, + "grad_norm": 0.7864341139793396, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 15568 + }, + { + "epoch": 0.15569, + "grad_norm": 0.8094674348831177, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 15569 + }, + { + "epoch": 0.1557, + "grad_norm": 0.7100827097892761, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 15570 + }, + { + "epoch": 0.15571, + "grad_norm": 0.7554362416267395, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 15571 + }, + { + "epoch": 0.15572, + "grad_norm": 0.9113379120826721, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 15572 + }, + { + "epoch": 0.15573, + "grad_norm": 1.0028765201568604, + "learning_rate": 0.003, + "loss": 4.036, + "step": 15573 + }, + { + "epoch": 0.15574, + "grad_norm": 1.0085824728012085, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 15574 + }, + { + "epoch": 0.15575, + "grad_norm": 1.139907956123352, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 15575 + }, + { + "epoch": 0.15576, + "grad_norm": 0.9667785167694092, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 15576 + }, + { + "epoch": 0.15577, + "grad_norm": 1.0565452575683594, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 15577 + }, + { + "epoch": 0.15578, + "grad_norm": 0.8792425394058228, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 15578 + }, + { + "epoch": 0.15579, + "grad_norm": 0.895402729511261, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 15579 + }, + { + "epoch": 0.1558, + "grad_norm": 0.8785274028778076, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 15580 + }, + { + "epoch": 0.15581, + "grad_norm": 0.9280442595481873, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 15581 + }, + { + "epoch": 0.15582, + "grad_norm": 1.178812026977539, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 15582 + }, + { + "epoch": 0.15583, + "grad_norm": 0.9174395799636841, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 15583 + }, + { + "epoch": 0.15584, + "grad_norm": 0.9447359442710876, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 15584 + }, + { + "epoch": 0.15585, + "grad_norm": 1.1034443378448486, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 15585 + }, + { + "epoch": 0.15586, + "grad_norm": 1.0161769390106201, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 15586 + }, + { + "epoch": 0.15587, + "grad_norm": 0.8678581714630127, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 15587 + }, + { + "epoch": 0.15588, + "grad_norm": 0.6879012584686279, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 15588 + }, + { + "epoch": 0.15589, + "grad_norm": 0.6353170871734619, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 15589 + }, + { + "epoch": 0.1559, + "grad_norm": 0.5770264863967896, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 15590 + }, + { + "epoch": 0.15591, + "grad_norm": 0.5902904272079468, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 15591 + }, + { + "epoch": 0.15592, + "grad_norm": 0.5827825665473938, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 15592 + }, + { + "epoch": 0.15593, + "grad_norm": 0.5940733551979065, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 15593 + }, + { + "epoch": 0.15594, + "grad_norm": 0.6823462247848511, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 15594 + }, + { + "epoch": 0.15595, + "grad_norm": 0.6451441645622253, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 15595 + }, + { + "epoch": 0.15596, + "grad_norm": 0.5992090106010437, + "learning_rate": 0.003, + "loss": 3.98, + "step": 15596 + }, + { + "epoch": 0.15597, + "grad_norm": 0.6235384345054626, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 15597 + }, + { + "epoch": 0.15598, + "grad_norm": 0.6631515622138977, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 15598 + }, + { + "epoch": 0.15599, + "grad_norm": 0.9059483408927917, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 15599 + }, + { + "epoch": 0.156, + "grad_norm": 1.39214026927948, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 15600 + }, + { + "epoch": 0.15601, + "grad_norm": 0.6501911282539368, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 15601 + }, + { + "epoch": 0.15602, + "grad_norm": 0.6093357801437378, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 15602 + }, + { + "epoch": 0.15603, + "grad_norm": 0.6285752058029175, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 15603 + }, + { + "epoch": 0.15604, + "grad_norm": 0.7170791029930115, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 15604 + }, + { + "epoch": 0.15605, + "grad_norm": 0.9150353670120239, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 15605 + }, + { + "epoch": 0.15606, + "grad_norm": 1.0283564329147339, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 15606 + }, + { + "epoch": 0.15607, + "grad_norm": 0.8459843993186951, + "learning_rate": 0.003, + "loss": 3.993, + "step": 15607 + }, + { + "epoch": 0.15608, + "grad_norm": 0.6638538241386414, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 15608 + }, + { + "epoch": 0.15609, + "grad_norm": 0.6532926559448242, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 15609 + }, + { + "epoch": 0.1561, + "grad_norm": 0.6578401327133179, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 15610 + }, + { + "epoch": 0.15611, + "grad_norm": 0.6223822236061096, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 15611 + }, + { + "epoch": 0.15612, + "grad_norm": 0.6221069693565369, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 15612 + }, + { + "epoch": 0.15613, + "grad_norm": 0.7909890413284302, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 15613 + }, + { + "epoch": 0.15614, + "grad_norm": 0.8776581287384033, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 15614 + }, + { + "epoch": 0.15615, + "grad_norm": 1.013645052909851, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 15615 + }, + { + "epoch": 0.15616, + "grad_norm": 1.1218299865722656, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 15616 + }, + { + "epoch": 0.15617, + "grad_norm": 0.8555946350097656, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 15617 + }, + { + "epoch": 0.15618, + "grad_norm": 0.6954250931739807, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 15618 + }, + { + "epoch": 0.15619, + "grad_norm": 0.6700589060783386, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 15619 + }, + { + "epoch": 0.1562, + "grad_norm": 0.6501047015190125, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 15620 + }, + { + "epoch": 0.15621, + "grad_norm": 0.6431427001953125, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 15621 + }, + { + "epoch": 0.15622, + "grad_norm": 0.7406570315361023, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 15622 + }, + { + "epoch": 0.15623, + "grad_norm": 0.8946971893310547, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 15623 + }, + { + "epoch": 0.15624, + "grad_norm": 0.8742425441741943, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 15624 + }, + { + "epoch": 0.15625, + "grad_norm": 0.815301239490509, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 15625 + }, + { + "epoch": 0.15626, + "grad_norm": 0.7755097150802612, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 15626 + }, + { + "epoch": 0.15627, + "grad_norm": 0.711911141872406, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 15627 + }, + { + "epoch": 0.15628, + "grad_norm": 0.7888513207435608, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 15628 + }, + { + "epoch": 0.15629, + "grad_norm": 0.8034841418266296, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 15629 + }, + { + "epoch": 0.1563, + "grad_norm": 0.8269197344779968, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 15630 + }, + { + "epoch": 0.15631, + "grad_norm": 0.8813633918762207, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 15631 + }, + { + "epoch": 0.15632, + "grad_norm": 0.9256646633148193, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 15632 + }, + { + "epoch": 0.15633, + "grad_norm": 1.0113441944122314, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 15633 + }, + { + "epoch": 0.15634, + "grad_norm": 1.004540205001831, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 15634 + }, + { + "epoch": 0.15635, + "grad_norm": 1.0412640571594238, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 15635 + }, + { + "epoch": 0.15636, + "grad_norm": 0.9312543272972107, + "learning_rate": 0.003, + "loss": 4.016, + "step": 15636 + }, + { + "epoch": 0.15637, + "grad_norm": 0.9052963256835938, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 15637 + }, + { + "epoch": 0.15638, + "grad_norm": 0.8330254554748535, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 15638 + }, + { + "epoch": 0.15639, + "grad_norm": 1.111505389213562, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 15639 + }, + { + "epoch": 0.1564, + "grad_norm": 0.963144063949585, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 15640 + }, + { + "epoch": 0.15641, + "grad_norm": 1.0859062671661377, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 15641 + }, + { + "epoch": 0.15642, + "grad_norm": 1.0575289726257324, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 15642 + }, + { + "epoch": 0.15643, + "grad_norm": 0.8799792528152466, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 15643 + }, + { + "epoch": 0.15644, + "grad_norm": 0.7085080146789551, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 15644 + }, + { + "epoch": 0.15645, + "grad_norm": 0.6737964749336243, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 15645 + }, + { + "epoch": 0.15646, + "grad_norm": 0.707210898399353, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 15646 + }, + { + "epoch": 0.15647, + "grad_norm": 0.8392312526702881, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 15647 + }, + { + "epoch": 0.15648, + "grad_norm": 0.8704290986061096, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 15648 + }, + { + "epoch": 0.15649, + "grad_norm": 0.8335064053535461, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 15649 + }, + { + "epoch": 0.1565, + "grad_norm": 0.8319531083106995, + "learning_rate": 0.003, + "loss": 3.966, + "step": 15650 + }, + { + "epoch": 0.15651, + "grad_norm": 0.8499333262443542, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 15651 + }, + { + "epoch": 0.15652, + "grad_norm": 0.9454139471054077, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 15652 + }, + { + "epoch": 0.15653, + "grad_norm": 1.1151617765426636, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 15653 + }, + { + "epoch": 0.15654, + "grad_norm": 0.956888735294342, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 15654 + }, + { + "epoch": 0.15655, + "grad_norm": 0.9633231163024902, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 15655 + }, + { + "epoch": 0.15656, + "grad_norm": 0.875674843788147, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 15656 + }, + { + "epoch": 0.15657, + "grad_norm": 0.83741694688797, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 15657 + }, + { + "epoch": 0.15658, + "grad_norm": 0.7873344421386719, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 15658 + }, + { + "epoch": 0.15659, + "grad_norm": 0.8622631430625916, + "learning_rate": 0.003, + "loss": 4.033, + "step": 15659 + }, + { + "epoch": 0.1566, + "grad_norm": 0.9279429316520691, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 15660 + }, + { + "epoch": 0.15661, + "grad_norm": 1.0509636402130127, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 15661 + }, + { + "epoch": 0.15662, + "grad_norm": 1.0412518978118896, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 15662 + }, + { + "epoch": 0.15663, + "grad_norm": 0.8322525024414062, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 15663 + }, + { + "epoch": 0.15664, + "grad_norm": 0.6247098445892334, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 15664 + }, + { + "epoch": 0.15665, + "grad_norm": 0.6534708738327026, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 15665 + }, + { + "epoch": 0.15666, + "grad_norm": 0.75336754322052, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 15666 + }, + { + "epoch": 0.15667, + "grad_norm": 0.8604514598846436, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 15667 + }, + { + "epoch": 0.15668, + "grad_norm": 0.8654664158821106, + "learning_rate": 0.003, + "loss": 3.989, + "step": 15668 + }, + { + "epoch": 0.15669, + "grad_norm": 0.9390060305595398, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 15669 + }, + { + "epoch": 0.1567, + "grad_norm": 1.0578045845031738, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 15670 + }, + { + "epoch": 0.15671, + "grad_norm": 0.9121909737586975, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 15671 + }, + { + "epoch": 0.15672, + "grad_norm": 0.7117065191268921, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 15672 + }, + { + "epoch": 0.15673, + "grad_norm": 0.70128333568573, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 15673 + }, + { + "epoch": 0.15674, + "grad_norm": 0.6423056721687317, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 15674 + }, + { + "epoch": 0.15675, + "grad_norm": 0.6011069416999817, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 15675 + }, + { + "epoch": 0.15676, + "grad_norm": 0.5848559141159058, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 15676 + }, + { + "epoch": 0.15677, + "grad_norm": 0.5934710502624512, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 15677 + }, + { + "epoch": 0.15678, + "grad_norm": 0.5638435482978821, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 15678 + }, + { + "epoch": 0.15679, + "grad_norm": 0.6302936673164368, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 15679 + }, + { + "epoch": 0.1568, + "grad_norm": 0.6576014161109924, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 15680 + }, + { + "epoch": 0.15681, + "grad_norm": 0.732036292552948, + "learning_rate": 0.003, + "loss": 4.038, + "step": 15681 + }, + { + "epoch": 0.15682, + "grad_norm": 0.6981785297393799, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 15682 + }, + { + "epoch": 0.15683, + "grad_norm": 0.7637856006622314, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 15683 + }, + { + "epoch": 0.15684, + "grad_norm": 0.8975669145584106, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 15684 + }, + { + "epoch": 0.15685, + "grad_norm": 0.9504064917564392, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 15685 + }, + { + "epoch": 0.15686, + "grad_norm": 0.9674578309059143, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 15686 + }, + { + "epoch": 0.15687, + "grad_norm": 1.233467936515808, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 15687 + }, + { + "epoch": 0.15688, + "grad_norm": 1.1004282236099243, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 15688 + }, + { + "epoch": 0.15689, + "grad_norm": 0.9505053162574768, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 15689 + }, + { + "epoch": 0.1569, + "grad_norm": 1.0994209051132202, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 15690 + }, + { + "epoch": 0.15691, + "grad_norm": 1.0676054954528809, + "learning_rate": 0.003, + "loss": 4.02, + "step": 15691 + }, + { + "epoch": 0.15692, + "grad_norm": 0.883074939250946, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 15692 + }, + { + "epoch": 0.15693, + "grad_norm": 1.0565145015716553, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 15693 + }, + { + "epoch": 0.15694, + "grad_norm": 1.1039477586746216, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 15694 + }, + { + "epoch": 0.15695, + "grad_norm": 0.8456154465675354, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 15695 + }, + { + "epoch": 0.15696, + "grad_norm": 0.7581636905670166, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 15696 + }, + { + "epoch": 0.15697, + "grad_norm": 0.6965075135231018, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 15697 + }, + { + "epoch": 0.15698, + "grad_norm": 0.7575919032096863, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 15698 + }, + { + "epoch": 0.15699, + "grad_norm": 0.8254173398017883, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 15699 + }, + { + "epoch": 0.157, + "grad_norm": 0.8927112221717834, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 15700 + }, + { + "epoch": 0.15701, + "grad_norm": 0.8910102248191833, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 15701 + }, + { + "epoch": 0.15702, + "grad_norm": 0.8605186343193054, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 15702 + }, + { + "epoch": 0.15703, + "grad_norm": 0.8565143346786499, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 15703 + }, + { + "epoch": 0.15704, + "grad_norm": 0.8403797149658203, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 15704 + }, + { + "epoch": 0.15705, + "grad_norm": 0.8270075917243958, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 15705 + }, + { + "epoch": 0.15706, + "grad_norm": 0.7761716246604919, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 15706 + }, + { + "epoch": 0.15707, + "grad_norm": 0.7189915776252747, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 15707 + }, + { + "epoch": 0.15708, + "grad_norm": 0.7878168225288391, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15708 + }, + { + "epoch": 0.15709, + "grad_norm": 0.8748422861099243, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 15709 + }, + { + "epoch": 0.1571, + "grad_norm": 1.0179502964019775, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 15710 + }, + { + "epoch": 0.15711, + "grad_norm": 1.0124653577804565, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 15711 + }, + { + "epoch": 0.15712, + "grad_norm": 0.7594412565231323, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 15712 + }, + { + "epoch": 0.15713, + "grad_norm": 0.5174108743667603, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 15713 + }, + { + "epoch": 0.15714, + "grad_norm": 0.6995696425437927, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 15714 + }, + { + "epoch": 0.15715, + "grad_norm": 0.7402389645576477, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 15715 + }, + { + "epoch": 0.15716, + "grad_norm": 0.6621226072311401, + "learning_rate": 0.003, + "loss": 4.025, + "step": 15716 + }, + { + "epoch": 0.15717, + "grad_norm": 0.6631869673728943, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 15717 + }, + { + "epoch": 0.15718, + "grad_norm": 0.8022869825363159, + "learning_rate": 0.003, + "loss": 4.007, + "step": 15718 + }, + { + "epoch": 0.15719, + "grad_norm": 0.8688627481460571, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 15719 + }, + { + "epoch": 0.1572, + "grad_norm": 0.8488649725914001, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 15720 + }, + { + "epoch": 0.15721, + "grad_norm": 0.7200070023536682, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 15721 + }, + { + "epoch": 0.15722, + "grad_norm": 0.6765528321266174, + "learning_rate": 0.003, + "loss": 3.992, + "step": 15722 + }, + { + "epoch": 0.15723, + "grad_norm": 0.7749486565589905, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 15723 + }, + { + "epoch": 0.15724, + "grad_norm": 0.9206483364105225, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 15724 + }, + { + "epoch": 0.15725, + "grad_norm": 0.7665677070617676, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 15725 + }, + { + "epoch": 0.15726, + "grad_norm": 0.6814460754394531, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 15726 + }, + { + "epoch": 0.15727, + "grad_norm": 0.6091095805168152, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 15727 + }, + { + "epoch": 0.15728, + "grad_norm": 0.6103879809379578, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 15728 + }, + { + "epoch": 0.15729, + "grad_norm": 0.6739366054534912, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 15729 + }, + { + "epoch": 0.1573, + "grad_norm": 0.738776445388794, + "learning_rate": 0.003, + "loss": 4.013, + "step": 15730 + }, + { + "epoch": 0.15731, + "grad_norm": 0.7757909893989563, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 15731 + }, + { + "epoch": 0.15732, + "grad_norm": 0.998759388923645, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 15732 + }, + { + "epoch": 0.15733, + "grad_norm": 1.298292875289917, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 15733 + }, + { + "epoch": 0.15734, + "grad_norm": 0.7509281039237976, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 15734 + }, + { + "epoch": 0.15735, + "grad_norm": 0.7110625505447388, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 15735 + }, + { + "epoch": 0.15736, + "grad_norm": 0.7740097641944885, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 15736 + }, + { + "epoch": 0.15737, + "grad_norm": 0.7516487836837769, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 15737 + }, + { + "epoch": 0.15738, + "grad_norm": 0.6357292532920837, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 15738 + }, + { + "epoch": 0.15739, + "grad_norm": 0.5393524765968323, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 15739 + }, + { + "epoch": 0.1574, + "grad_norm": 0.6245474219322205, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 15740 + }, + { + "epoch": 0.15741, + "grad_norm": 0.9935439825057983, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 15741 + }, + { + "epoch": 0.15742, + "grad_norm": 1.4136745929718018, + "learning_rate": 0.003, + "loss": 4.007, + "step": 15742 + }, + { + "epoch": 0.15743, + "grad_norm": 0.5433087944984436, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 15743 + }, + { + "epoch": 0.15744, + "grad_norm": 0.6950448751449585, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 15744 + }, + { + "epoch": 0.15745, + "grad_norm": 0.832676112651825, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 15745 + }, + { + "epoch": 0.15746, + "grad_norm": 0.8238332271575928, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 15746 + }, + { + "epoch": 0.15747, + "grad_norm": 0.8855404257774353, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 15747 + }, + { + "epoch": 0.15748, + "grad_norm": 0.8645646572113037, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 15748 + }, + { + "epoch": 0.15749, + "grad_norm": 0.9085519909858704, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 15749 + }, + { + "epoch": 0.1575, + "grad_norm": 1.0006437301635742, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 15750 + }, + { + "epoch": 0.15751, + "grad_norm": 1.1213877201080322, + "learning_rate": 0.003, + "loss": 4.011, + "step": 15751 + }, + { + "epoch": 0.15752, + "grad_norm": 1.1313105821609497, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 15752 + }, + { + "epoch": 0.15753, + "grad_norm": 0.9690887928009033, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 15753 + }, + { + "epoch": 0.15754, + "grad_norm": 0.9049549102783203, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 15754 + }, + { + "epoch": 0.15755, + "grad_norm": 0.8756823539733887, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 15755 + }, + { + "epoch": 0.15756, + "grad_norm": 0.7845616936683655, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 15756 + }, + { + "epoch": 0.15757, + "grad_norm": 0.7886598110198975, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 15757 + }, + { + "epoch": 0.15758, + "grad_norm": 0.8714204430580139, + "learning_rate": 0.003, + "loss": 4.03, + "step": 15758 + }, + { + "epoch": 0.15759, + "grad_norm": 1.0030721426010132, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 15759 + }, + { + "epoch": 0.1576, + "grad_norm": 1.263763189315796, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 15760 + }, + { + "epoch": 0.15761, + "grad_norm": 0.9564282894134521, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 15761 + }, + { + "epoch": 0.15762, + "grad_norm": 0.9213220477104187, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 15762 + }, + { + "epoch": 0.15763, + "grad_norm": 0.9312294721603394, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 15763 + }, + { + "epoch": 0.15764, + "grad_norm": 0.7343323826789856, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 15764 + }, + { + "epoch": 0.15765, + "grad_norm": 0.7667588591575623, + "learning_rate": 0.003, + "loss": 4.024, + "step": 15765 + }, + { + "epoch": 0.15766, + "grad_norm": 0.7190549373626709, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 15766 + }, + { + "epoch": 0.15767, + "grad_norm": 0.7208852767944336, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 15767 + }, + { + "epoch": 0.15768, + "grad_norm": 0.7332645058631897, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 15768 + }, + { + "epoch": 0.15769, + "grad_norm": 0.825160801410675, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 15769 + }, + { + "epoch": 0.1577, + "grad_norm": 0.7511894702911377, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 15770 + }, + { + "epoch": 0.15771, + "grad_norm": 0.8381373882293701, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 15771 + }, + { + "epoch": 0.15772, + "grad_norm": 1.0264532566070557, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 15772 + }, + { + "epoch": 0.15773, + "grad_norm": 0.9846251010894775, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 15773 + }, + { + "epoch": 0.15774, + "grad_norm": 0.8334699869155884, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 15774 + }, + { + "epoch": 0.15775, + "grad_norm": 0.6779347062110901, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 15775 + }, + { + "epoch": 0.15776, + "grad_norm": 0.6254589557647705, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 15776 + }, + { + "epoch": 0.15777, + "grad_norm": 0.679844856262207, + "learning_rate": 0.003, + "loss": 3.991, + "step": 15777 + }, + { + "epoch": 0.15778, + "grad_norm": 0.7129006385803223, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 15778 + }, + { + "epoch": 0.15779, + "grad_norm": 0.7573176026344299, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 15779 + }, + { + "epoch": 0.1578, + "grad_norm": 0.7286023497581482, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 15780 + }, + { + "epoch": 0.15781, + "grad_norm": 0.7774902582168579, + "learning_rate": 0.003, + "loss": 3.9493, + "step": 15781 + }, + { + "epoch": 0.15782, + "grad_norm": 0.8229116201400757, + "learning_rate": 0.003, + "loss": 4.022, + "step": 15782 + }, + { + "epoch": 0.15783, + "grad_norm": 1.1103692054748535, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 15783 + }, + { + "epoch": 0.15784, + "grad_norm": 0.8943964242935181, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 15784 + }, + { + "epoch": 0.15785, + "grad_norm": 0.8067341446876526, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 15785 + }, + { + "epoch": 0.15786, + "grad_norm": 0.7550935745239258, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 15786 + }, + { + "epoch": 0.15787, + "grad_norm": 0.6463281512260437, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 15787 + }, + { + "epoch": 0.15788, + "grad_norm": 0.6048409938812256, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 15788 + }, + { + "epoch": 0.15789, + "grad_norm": 0.586916983127594, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 15789 + }, + { + "epoch": 0.1579, + "grad_norm": 0.606907308101654, + "learning_rate": 0.003, + "loss": 4.003, + "step": 15790 + }, + { + "epoch": 0.15791, + "grad_norm": 0.7055268883705139, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 15791 + }, + { + "epoch": 0.15792, + "grad_norm": 0.8679223656654358, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 15792 + }, + { + "epoch": 0.15793, + "grad_norm": 1.0493624210357666, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 15793 + }, + { + "epoch": 0.15794, + "grad_norm": 1.0888004302978516, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 15794 + }, + { + "epoch": 0.15795, + "grad_norm": 0.9175351858139038, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 15795 + }, + { + "epoch": 0.15796, + "grad_norm": 0.8576933145523071, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 15796 + }, + { + "epoch": 0.15797, + "grad_norm": 0.7691969871520996, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 15797 + }, + { + "epoch": 0.15798, + "grad_norm": 0.7251110672950745, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 15798 + }, + { + "epoch": 0.15799, + "grad_norm": 0.7915005683898926, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 15799 + }, + { + "epoch": 0.158, + "grad_norm": 0.9329673051834106, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 15800 + }, + { + "epoch": 0.15801, + "grad_norm": 0.9955234527587891, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 15801 + }, + { + "epoch": 0.15802, + "grad_norm": 1.0439205169677734, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 15802 + }, + { + "epoch": 0.15803, + "grad_norm": 0.9350157380104065, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 15803 + }, + { + "epoch": 0.15804, + "grad_norm": 0.9399392604827881, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 15804 + }, + { + "epoch": 0.15805, + "grad_norm": 1.160086750984192, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 15805 + }, + { + "epoch": 0.15806, + "grad_norm": 1.1166666746139526, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 15806 + }, + { + "epoch": 0.15807, + "grad_norm": 0.7115695476531982, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 15807 + }, + { + "epoch": 0.15808, + "grad_norm": 0.5912373065948486, + "learning_rate": 0.003, + "loss": 3.995, + "step": 15808 + }, + { + "epoch": 0.15809, + "grad_norm": 0.6347526907920837, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 15809 + }, + { + "epoch": 0.1581, + "grad_norm": 0.7153865694999695, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 15810 + }, + { + "epoch": 0.15811, + "grad_norm": 0.8052754998207092, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 15811 + }, + { + "epoch": 0.15812, + "grad_norm": 0.8784002661705017, + "learning_rate": 0.003, + "loss": 4.002, + "step": 15812 + }, + { + "epoch": 0.15813, + "grad_norm": 0.9777888059616089, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 15813 + }, + { + "epoch": 0.15814, + "grad_norm": 1.0682368278503418, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 15814 + }, + { + "epoch": 0.15815, + "grad_norm": 1.0067962408065796, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 15815 + }, + { + "epoch": 0.15816, + "grad_norm": 0.9310998916625977, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 15816 + }, + { + "epoch": 0.15817, + "grad_norm": 0.8633456826210022, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 15817 + }, + { + "epoch": 0.15818, + "grad_norm": 0.8136230111122131, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 15818 + }, + { + "epoch": 0.15819, + "grad_norm": 0.6792727708816528, + "learning_rate": 0.003, + "loss": 3.97, + "step": 15819 + }, + { + "epoch": 0.1582, + "grad_norm": 0.6035311818122864, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 15820 + }, + { + "epoch": 0.15821, + "grad_norm": 0.6503746509552002, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 15821 + }, + { + "epoch": 0.15822, + "grad_norm": 0.6729138493537903, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 15822 + }, + { + "epoch": 0.15823, + "grad_norm": 0.7227945923805237, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 15823 + }, + { + "epoch": 0.15824, + "grad_norm": 0.846311628818512, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 15824 + }, + { + "epoch": 0.15825, + "grad_norm": 1.0390722751617432, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 15825 + }, + { + "epoch": 0.15826, + "grad_norm": 1.2302485704421997, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 15826 + }, + { + "epoch": 0.15827, + "grad_norm": 0.8671656847000122, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 15827 + }, + { + "epoch": 0.15828, + "grad_norm": 0.9332629442214966, + "learning_rate": 0.003, + "loss": 3.997, + "step": 15828 + }, + { + "epoch": 0.15829, + "grad_norm": 0.970983624458313, + "learning_rate": 0.003, + "loss": 4.023, + "step": 15829 + }, + { + "epoch": 0.1583, + "grad_norm": 1.1105905771255493, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 15830 + }, + { + "epoch": 0.15831, + "grad_norm": 1.189861536026001, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 15831 + }, + { + "epoch": 0.15832, + "grad_norm": 0.7376668453216553, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 15832 + }, + { + "epoch": 0.15833, + "grad_norm": 0.6104328036308289, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 15833 + }, + { + "epoch": 0.15834, + "grad_norm": 0.6534768342971802, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 15834 + }, + { + "epoch": 0.15835, + "grad_norm": 0.6755269169807434, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 15835 + }, + { + "epoch": 0.15836, + "grad_norm": 0.6206930875778198, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 15836 + }, + { + "epoch": 0.15837, + "grad_norm": 0.5505279898643494, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 15837 + }, + { + "epoch": 0.15838, + "grad_norm": 0.5139374136924744, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 15838 + }, + { + "epoch": 0.15839, + "grad_norm": 0.5601384043693542, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 15839 + }, + { + "epoch": 0.1584, + "grad_norm": 0.6331465244293213, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 15840 + }, + { + "epoch": 0.15841, + "grad_norm": 0.7213279604911804, + "learning_rate": 0.003, + "loss": 4.016, + "step": 15841 + }, + { + "epoch": 0.15842, + "grad_norm": 0.7569580674171448, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 15842 + }, + { + "epoch": 0.15843, + "grad_norm": 0.9473897814750671, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 15843 + }, + { + "epoch": 0.15844, + "grad_norm": 1.1790201663970947, + "learning_rate": 0.003, + "loss": 4.017, + "step": 15844 + }, + { + "epoch": 0.15845, + "grad_norm": 0.8182543516159058, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 15845 + }, + { + "epoch": 0.15846, + "grad_norm": 0.7447865009307861, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 15846 + }, + { + "epoch": 0.15847, + "grad_norm": 0.8887273669242859, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 15847 + }, + { + "epoch": 0.15848, + "grad_norm": 1.0807547569274902, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 15848 + }, + { + "epoch": 0.15849, + "grad_norm": 0.9198710918426514, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 15849 + }, + { + "epoch": 0.1585, + "grad_norm": 0.796409010887146, + "learning_rate": 0.003, + "loss": 4.034, + "step": 15850 + }, + { + "epoch": 0.15851, + "grad_norm": 0.7506511211395264, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 15851 + }, + { + "epoch": 0.15852, + "grad_norm": 0.6153442859649658, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 15852 + }, + { + "epoch": 0.15853, + "grad_norm": 0.637804388999939, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 15853 + }, + { + "epoch": 0.15854, + "grad_norm": 0.6735568642616272, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 15854 + }, + { + "epoch": 0.15855, + "grad_norm": 0.7883623838424683, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 15855 + }, + { + "epoch": 0.15856, + "grad_norm": 0.9529778957366943, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 15856 + }, + { + "epoch": 0.15857, + "grad_norm": 0.9971096515655518, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 15857 + }, + { + "epoch": 0.15858, + "grad_norm": 1.0070446729660034, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 15858 + }, + { + "epoch": 0.15859, + "grad_norm": 0.981824517250061, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 15859 + }, + { + "epoch": 0.1586, + "grad_norm": 0.8740189671516418, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 15860 + }, + { + "epoch": 0.15861, + "grad_norm": 0.9106821417808533, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 15861 + }, + { + "epoch": 0.15862, + "grad_norm": 0.8984867930412292, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 15862 + }, + { + "epoch": 0.15863, + "grad_norm": 0.8848163485527039, + "learning_rate": 0.003, + "loss": 3.992, + "step": 15863 + }, + { + "epoch": 0.15864, + "grad_norm": 0.9364362359046936, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 15864 + }, + { + "epoch": 0.15865, + "grad_norm": 1.0542372465133667, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 15865 + }, + { + "epoch": 0.15866, + "grad_norm": 1.015795350074768, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 15866 + }, + { + "epoch": 0.15867, + "grad_norm": 1.0119582414627075, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 15867 + }, + { + "epoch": 0.15868, + "grad_norm": 1.1160649061203003, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 15868 + }, + { + "epoch": 0.15869, + "grad_norm": 0.9358794093132019, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 15869 + }, + { + "epoch": 0.1587, + "grad_norm": 0.9446738958358765, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 15870 + }, + { + "epoch": 0.15871, + "grad_norm": 1.2070125341415405, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 15871 + }, + { + "epoch": 0.15872, + "grad_norm": 0.9497233033180237, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 15872 + }, + { + "epoch": 0.15873, + "grad_norm": 0.9464455842971802, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 15873 + }, + { + "epoch": 0.15874, + "grad_norm": 0.9484769701957703, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 15874 + }, + { + "epoch": 0.15875, + "grad_norm": 1.0746848583221436, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 15875 + }, + { + "epoch": 0.15876, + "grad_norm": 1.0386066436767578, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 15876 + }, + { + "epoch": 0.15877, + "grad_norm": 0.9693458080291748, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 15877 + }, + { + "epoch": 0.15878, + "grad_norm": 0.8267617225646973, + "learning_rate": 0.003, + "loss": 4.009, + "step": 15878 + }, + { + "epoch": 0.15879, + "grad_norm": 0.811514139175415, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 15879 + }, + { + "epoch": 0.1588, + "grad_norm": 0.7325130701065063, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 15880 + }, + { + "epoch": 0.15881, + "grad_norm": 0.8104916214942932, + "learning_rate": 0.003, + "loss": 4.004, + "step": 15881 + }, + { + "epoch": 0.15882, + "grad_norm": 0.9261886477470398, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 15882 + }, + { + "epoch": 0.15883, + "grad_norm": 1.1093589067459106, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 15883 + }, + { + "epoch": 0.15884, + "grad_norm": 0.7875494956970215, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 15884 + }, + { + "epoch": 0.15885, + "grad_norm": 0.6347299814224243, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 15885 + }, + { + "epoch": 0.15886, + "grad_norm": 0.619236409664154, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 15886 + }, + { + "epoch": 0.15887, + "grad_norm": 0.6201211810112, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 15887 + }, + { + "epoch": 0.15888, + "grad_norm": 0.6558188199996948, + "learning_rate": 0.003, + "loss": 4.001, + "step": 15888 + }, + { + "epoch": 0.15889, + "grad_norm": 0.6662975549697876, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 15889 + }, + { + "epoch": 0.1589, + "grad_norm": 0.6249191761016846, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 15890 + }, + { + "epoch": 0.15891, + "grad_norm": 0.6488112807273865, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 15891 + }, + { + "epoch": 0.15892, + "grad_norm": 0.5491712093353271, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 15892 + }, + { + "epoch": 0.15893, + "grad_norm": 0.5083919763565063, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 15893 + }, + { + "epoch": 0.15894, + "grad_norm": 0.5192022323608398, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 15894 + }, + { + "epoch": 0.15895, + "grad_norm": 0.5186970829963684, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 15895 + }, + { + "epoch": 0.15896, + "grad_norm": 0.6246152520179749, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 15896 + }, + { + "epoch": 0.15897, + "grad_norm": 0.788753092288971, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 15897 + }, + { + "epoch": 0.15898, + "grad_norm": 0.9366752505302429, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 15898 + }, + { + "epoch": 0.15899, + "grad_norm": 1.1852505207061768, + "learning_rate": 0.003, + "loss": 3.976, + "step": 15899 + }, + { + "epoch": 0.159, + "grad_norm": 0.8160002827644348, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 15900 + }, + { + "epoch": 0.15901, + "grad_norm": 0.7241095900535583, + "learning_rate": 0.003, + "loss": 4.013, + "step": 15901 + }, + { + "epoch": 0.15902, + "grad_norm": 0.7413921356201172, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 15902 + }, + { + "epoch": 0.15903, + "grad_norm": 0.7482760548591614, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 15903 + }, + { + "epoch": 0.15904, + "grad_norm": 0.7704154253005981, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 15904 + }, + { + "epoch": 0.15905, + "grad_norm": 0.8344385027885437, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 15905 + }, + { + "epoch": 0.15906, + "grad_norm": 0.7981237173080444, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 15906 + }, + { + "epoch": 0.15907, + "grad_norm": 0.7326487898826599, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 15907 + }, + { + "epoch": 0.15908, + "grad_norm": 0.7611872553825378, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 15908 + }, + { + "epoch": 0.15909, + "grad_norm": 0.9336971044540405, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 15909 + }, + { + "epoch": 0.1591, + "grad_norm": 1.031105637550354, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 15910 + }, + { + "epoch": 0.15911, + "grad_norm": 0.9963960647583008, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 15911 + }, + { + "epoch": 0.15912, + "grad_norm": 0.822158932685852, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 15912 + }, + { + "epoch": 0.15913, + "grad_norm": 0.7185021042823792, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 15913 + }, + { + "epoch": 0.15914, + "grad_norm": 0.6723746061325073, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 15914 + }, + { + "epoch": 0.15915, + "grad_norm": 0.6805190443992615, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 15915 + }, + { + "epoch": 0.15916, + "grad_norm": 0.7104006409645081, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 15916 + }, + { + "epoch": 0.15917, + "grad_norm": 0.8186883926391602, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 15917 + }, + { + "epoch": 0.15918, + "grad_norm": 0.9097923636436462, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 15918 + }, + { + "epoch": 0.15919, + "grad_norm": 0.849427342414856, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 15919 + }, + { + "epoch": 0.1592, + "grad_norm": 0.8371151685714722, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 15920 + }, + { + "epoch": 0.15921, + "grad_norm": 0.8601378798484802, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 15921 + }, + { + "epoch": 0.15922, + "grad_norm": 0.8923529982566833, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 15922 + }, + { + "epoch": 0.15923, + "grad_norm": 1.1591925621032715, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 15923 + }, + { + "epoch": 0.15924, + "grad_norm": 1.07288658618927, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 15924 + }, + { + "epoch": 0.15925, + "grad_norm": 0.9240902066230774, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 15925 + }, + { + "epoch": 0.15926, + "grad_norm": 0.9564006924629211, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 15926 + }, + { + "epoch": 0.15927, + "grad_norm": 0.9875341057777405, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 15927 + }, + { + "epoch": 0.15928, + "grad_norm": 0.8872766494750977, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 15928 + }, + { + "epoch": 0.15929, + "grad_norm": 0.7050520777702332, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 15929 + }, + { + "epoch": 0.1593, + "grad_norm": 0.675094485282898, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 15930 + }, + { + "epoch": 0.15931, + "grad_norm": 0.6591386795043945, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 15931 + }, + { + "epoch": 0.15932, + "grad_norm": 0.6734553575515747, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 15932 + }, + { + "epoch": 0.15933, + "grad_norm": 0.8183740377426147, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 15933 + }, + { + "epoch": 0.15934, + "grad_norm": 1.0227583646774292, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 15934 + }, + { + "epoch": 0.15935, + "grad_norm": 1.2014424800872803, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 15935 + }, + { + "epoch": 0.15936, + "grad_norm": 0.6601992845535278, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 15936 + }, + { + "epoch": 0.15937, + "grad_norm": 0.5800651907920837, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 15937 + }, + { + "epoch": 0.15938, + "grad_norm": 0.7739246487617493, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 15938 + }, + { + "epoch": 0.15939, + "grad_norm": 0.7693057656288147, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 15939 + }, + { + "epoch": 0.1594, + "grad_norm": 0.7467857599258423, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 15940 + }, + { + "epoch": 0.15941, + "grad_norm": 0.7646198868751526, + "learning_rate": 0.003, + "loss": 3.987, + "step": 15941 + }, + { + "epoch": 0.15942, + "grad_norm": 0.7261171936988831, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 15942 + }, + { + "epoch": 0.15943, + "grad_norm": 0.7164345383644104, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 15943 + }, + { + "epoch": 0.15944, + "grad_norm": 0.7011243104934692, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 15944 + }, + { + "epoch": 0.15945, + "grad_norm": 0.8210936784744263, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 15945 + }, + { + "epoch": 0.15946, + "grad_norm": 0.8095575571060181, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 15946 + }, + { + "epoch": 0.15947, + "grad_norm": 0.674490749835968, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 15947 + }, + { + "epoch": 0.15948, + "grad_norm": 0.6978493928909302, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 15948 + }, + { + "epoch": 0.15949, + "grad_norm": 0.7934913039207458, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 15949 + }, + { + "epoch": 0.1595, + "grad_norm": 0.9713555574417114, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 15950 + }, + { + "epoch": 0.15951, + "grad_norm": 1.0996856689453125, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 15951 + }, + { + "epoch": 0.15952, + "grad_norm": 0.8714984655380249, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 15952 + }, + { + "epoch": 0.15953, + "grad_norm": 0.8879398107528687, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 15953 + }, + { + "epoch": 0.15954, + "grad_norm": 1.0821666717529297, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 15954 + }, + { + "epoch": 0.15955, + "grad_norm": 0.8716177344322205, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 15955 + }, + { + "epoch": 0.15956, + "grad_norm": 0.8655083179473877, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 15956 + }, + { + "epoch": 0.15957, + "grad_norm": 0.902826726436615, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 15957 + }, + { + "epoch": 0.15958, + "grad_norm": 0.8519112467765808, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 15958 + }, + { + "epoch": 0.15959, + "grad_norm": 0.7480038404464722, + "learning_rate": 0.003, + "loss": 4.027, + "step": 15959 + }, + { + "epoch": 0.1596, + "grad_norm": 0.8103182911872864, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 15960 + }, + { + "epoch": 0.15961, + "grad_norm": 0.9224309921264648, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 15961 + }, + { + "epoch": 0.15962, + "grad_norm": 0.8068740367889404, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 15962 + }, + { + "epoch": 0.15963, + "grad_norm": 0.9216637015342712, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 15963 + }, + { + "epoch": 0.15964, + "grad_norm": 1.3789762258529663, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 15964 + }, + { + "epoch": 0.15965, + "grad_norm": 0.8822488784790039, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 15965 + }, + { + "epoch": 0.15966, + "grad_norm": 0.852074921131134, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 15966 + }, + { + "epoch": 0.15967, + "grad_norm": 0.842533528804779, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 15967 + }, + { + "epoch": 0.15968, + "grad_norm": 0.8331539034843445, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 15968 + }, + { + "epoch": 0.15969, + "grad_norm": 0.6840645670890808, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 15969 + }, + { + "epoch": 0.1597, + "grad_norm": 0.712909460067749, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 15970 + }, + { + "epoch": 0.15971, + "grad_norm": 0.7712053060531616, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 15971 + }, + { + "epoch": 0.15972, + "grad_norm": 0.7960266470909119, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 15972 + }, + { + "epoch": 0.15973, + "grad_norm": 0.8820586204528809, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 15973 + }, + { + "epoch": 0.15974, + "grad_norm": 1.0755890607833862, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 15974 + }, + { + "epoch": 0.15975, + "grad_norm": 0.9829635620117188, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 15975 + }, + { + "epoch": 0.15976, + "grad_norm": 0.9537001848220825, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 15976 + }, + { + "epoch": 0.15977, + "grad_norm": 0.8291124105453491, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 15977 + }, + { + "epoch": 0.15978, + "grad_norm": 0.7226710319519043, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 15978 + }, + { + "epoch": 0.15979, + "grad_norm": 0.6324989795684814, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 15979 + }, + { + "epoch": 0.1598, + "grad_norm": 0.6973319053649902, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 15980 + }, + { + "epoch": 0.15981, + "grad_norm": 0.7571434378623962, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 15981 + }, + { + "epoch": 0.15982, + "grad_norm": 0.7698507905006409, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 15982 + }, + { + "epoch": 0.15983, + "grad_norm": 0.8050060272216797, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 15983 + }, + { + "epoch": 0.15984, + "grad_norm": 1.0299527645111084, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 15984 + }, + { + "epoch": 0.15985, + "grad_norm": 1.2539191246032715, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 15985 + }, + { + "epoch": 0.15986, + "grad_norm": 0.7665183544158936, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 15986 + }, + { + "epoch": 0.15987, + "grad_norm": 0.6669110655784607, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 15987 + }, + { + "epoch": 0.15988, + "grad_norm": 0.7086806297302246, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 15988 + }, + { + "epoch": 0.15989, + "grad_norm": 0.7054063677787781, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 15989 + }, + { + "epoch": 0.1599, + "grad_norm": 0.7590455412864685, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 15990 + }, + { + "epoch": 0.15991, + "grad_norm": 0.753837525844574, + "learning_rate": 0.003, + "loss": 4.031, + "step": 15991 + }, + { + "epoch": 0.15992, + "grad_norm": 0.8459493517875671, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 15992 + }, + { + "epoch": 0.15993, + "grad_norm": 1.016892433166504, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 15993 + }, + { + "epoch": 0.15994, + "grad_norm": 1.1503039598464966, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 15994 + }, + { + "epoch": 0.15995, + "grad_norm": 0.8415044546127319, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 15995 + }, + { + "epoch": 0.15996, + "grad_norm": 0.806657075881958, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 15996 + }, + { + "epoch": 0.15997, + "grad_norm": 0.7540598511695862, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 15997 + }, + { + "epoch": 0.15998, + "grad_norm": 0.6794475317001343, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 15998 + }, + { + "epoch": 0.15999, + "grad_norm": 0.7693151831626892, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 15999 + }, + { + "epoch": 0.16, + "grad_norm": 0.9738240242004395, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 16000 + }, + { + "epoch": 0.16001, + "grad_norm": 1.1086915731430054, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 16001 + }, + { + "epoch": 0.16002, + "grad_norm": 0.867268443107605, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 16002 + }, + { + "epoch": 0.16003, + "grad_norm": 0.8266265392303467, + "learning_rate": 0.003, + "loss": 4.043, + "step": 16003 + }, + { + "epoch": 0.16004, + "grad_norm": 0.7528146505355835, + "learning_rate": 0.003, + "loss": 4.006, + "step": 16004 + }, + { + "epoch": 0.16005, + "grad_norm": 0.8397828340530396, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 16005 + }, + { + "epoch": 0.16006, + "grad_norm": 0.9907081127166748, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 16006 + }, + { + "epoch": 0.16007, + "grad_norm": 1.0469940900802612, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 16007 + }, + { + "epoch": 0.16008, + "grad_norm": 1.0257529020309448, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 16008 + }, + { + "epoch": 0.16009, + "grad_norm": 1.0527772903442383, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 16009 + }, + { + "epoch": 0.1601, + "grad_norm": 0.8561283349990845, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 16010 + }, + { + "epoch": 0.16011, + "grad_norm": 0.8803778886795044, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 16011 + }, + { + "epoch": 0.16012, + "grad_norm": 1.0238463878631592, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 16012 + }, + { + "epoch": 0.16013, + "grad_norm": 0.9067773818969727, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 16013 + }, + { + "epoch": 0.16014, + "grad_norm": 0.9412308931350708, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 16014 + }, + { + "epoch": 0.16015, + "grad_norm": 0.956563413143158, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 16015 + }, + { + "epoch": 0.16016, + "grad_norm": 1.1569287776947021, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 16016 + }, + { + "epoch": 0.16017, + "grad_norm": 0.8871514797210693, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 16017 + }, + { + "epoch": 0.16018, + "grad_norm": 0.8447285890579224, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 16018 + }, + { + "epoch": 0.16019, + "grad_norm": 0.9038500785827637, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 16019 + }, + { + "epoch": 0.1602, + "grad_norm": 0.8824516534805298, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 16020 + }, + { + "epoch": 0.16021, + "grad_norm": 0.9204214811325073, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 16021 + }, + { + "epoch": 0.16022, + "grad_norm": 0.9645723104476929, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 16022 + }, + { + "epoch": 0.16023, + "grad_norm": 0.8048455119132996, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 16023 + }, + { + "epoch": 0.16024, + "grad_norm": 0.7482070326805115, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 16024 + }, + { + "epoch": 0.16025, + "grad_norm": 0.9110950231552124, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 16025 + }, + { + "epoch": 0.16026, + "grad_norm": 1.1130304336547852, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 16026 + }, + { + "epoch": 0.16027, + "grad_norm": 0.8323398232460022, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 16027 + }, + { + "epoch": 0.16028, + "grad_norm": 0.7672797441482544, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 16028 + }, + { + "epoch": 0.16029, + "grad_norm": 0.7083603143692017, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 16029 + }, + { + "epoch": 0.1603, + "grad_norm": 0.6222543716430664, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 16030 + }, + { + "epoch": 0.16031, + "grad_norm": 0.5898983478546143, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 16031 + }, + { + "epoch": 0.16032, + "grad_norm": 0.6412619948387146, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 16032 + }, + { + "epoch": 0.16033, + "grad_norm": 0.6066755056381226, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 16033 + }, + { + "epoch": 0.16034, + "grad_norm": 0.5793968439102173, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 16034 + }, + { + "epoch": 0.16035, + "grad_norm": 0.5370006561279297, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 16035 + }, + { + "epoch": 0.16036, + "grad_norm": 0.44992175698280334, + "learning_rate": 0.003, + "loss": 3.989, + "step": 16036 + }, + { + "epoch": 0.16037, + "grad_norm": 0.6282742023468018, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 16037 + }, + { + "epoch": 0.16038, + "grad_norm": 0.9777520298957825, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 16038 + }, + { + "epoch": 0.16039, + "grad_norm": 1.5354535579681396, + "learning_rate": 0.003, + "loss": 4.0, + "step": 16039 + }, + { + "epoch": 0.1604, + "grad_norm": 0.4473283588886261, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 16040 + }, + { + "epoch": 0.16041, + "grad_norm": 0.9890369772911072, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 16041 + }, + { + "epoch": 0.16042, + "grad_norm": 1.2385085821151733, + "learning_rate": 0.003, + "loss": 4.033, + "step": 16042 + }, + { + "epoch": 0.16043, + "grad_norm": 0.6593718528747559, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 16043 + }, + { + "epoch": 0.16044, + "grad_norm": 0.744037389755249, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 16044 + }, + { + "epoch": 0.16045, + "grad_norm": 0.7767122387886047, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 16045 + }, + { + "epoch": 0.16046, + "grad_norm": 0.8072572946548462, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 16046 + }, + { + "epoch": 0.16047, + "grad_norm": 0.8254396915435791, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 16047 + }, + { + "epoch": 0.16048, + "grad_norm": 0.7795100212097168, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 16048 + }, + { + "epoch": 0.16049, + "grad_norm": 0.7616690397262573, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 16049 + }, + { + "epoch": 0.1605, + "grad_norm": 0.8673222064971924, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 16050 + }, + { + "epoch": 0.16051, + "grad_norm": 0.8658856153488159, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 16051 + }, + { + "epoch": 0.16052, + "grad_norm": 0.7555408477783203, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 16052 + }, + { + "epoch": 0.16053, + "grad_norm": 0.7539330720901489, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 16053 + }, + { + "epoch": 0.16054, + "grad_norm": 0.8005216121673584, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 16054 + }, + { + "epoch": 0.16055, + "grad_norm": 0.7881186604499817, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 16055 + }, + { + "epoch": 0.16056, + "grad_norm": 0.7905475497245789, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 16056 + }, + { + "epoch": 0.16057, + "grad_norm": 0.7403662204742432, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 16057 + }, + { + "epoch": 0.16058, + "grad_norm": 0.9307411313056946, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 16058 + }, + { + "epoch": 0.16059, + "grad_norm": 1.0022494792938232, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 16059 + }, + { + "epoch": 0.1606, + "grad_norm": 0.9395740628242493, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 16060 + }, + { + "epoch": 0.16061, + "grad_norm": 0.8792284727096558, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 16061 + }, + { + "epoch": 0.16062, + "grad_norm": 0.879179835319519, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 16062 + }, + { + "epoch": 0.16063, + "grad_norm": 1.2204158306121826, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 16063 + }, + { + "epoch": 0.16064, + "grad_norm": 1.0658458471298218, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 16064 + }, + { + "epoch": 0.16065, + "grad_norm": 0.7353061437606812, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 16065 + }, + { + "epoch": 0.16066, + "grad_norm": 0.5892605781555176, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 16066 + }, + { + "epoch": 0.16067, + "grad_norm": 0.5376728773117065, + "learning_rate": 0.003, + "loss": 4.029, + "step": 16067 + }, + { + "epoch": 0.16068, + "grad_norm": 0.5359166264533997, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 16068 + }, + { + "epoch": 0.16069, + "grad_norm": 0.6350721120834351, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 16069 + }, + { + "epoch": 0.1607, + "grad_norm": 0.8995853066444397, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 16070 + }, + { + "epoch": 0.16071, + "grad_norm": 1.21028733253479, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 16071 + }, + { + "epoch": 0.16072, + "grad_norm": 0.6996341347694397, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 16072 + }, + { + "epoch": 0.16073, + "grad_norm": 0.6284819841384888, + "learning_rate": 0.003, + "loss": 4.01, + "step": 16073 + }, + { + "epoch": 0.16074, + "grad_norm": 0.8318029046058655, + "learning_rate": 0.003, + "loss": 4.005, + "step": 16074 + }, + { + "epoch": 0.16075, + "grad_norm": 0.9144317507743835, + "learning_rate": 0.003, + "loss": 4.027, + "step": 16075 + }, + { + "epoch": 0.16076, + "grad_norm": 0.935553252696991, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 16076 + }, + { + "epoch": 0.16077, + "grad_norm": 0.8554702997207642, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 16077 + }, + { + "epoch": 0.16078, + "grad_norm": 0.7261465787887573, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 16078 + }, + { + "epoch": 0.16079, + "grad_norm": 0.6737211346626282, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 16079 + }, + { + "epoch": 0.1608, + "grad_norm": 0.6472561955451965, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 16080 + }, + { + "epoch": 0.16081, + "grad_norm": 0.7049079537391663, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 16081 + }, + { + "epoch": 0.16082, + "grad_norm": 0.8032575845718384, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 16082 + }, + { + "epoch": 0.16083, + "grad_norm": 0.8006021976470947, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 16083 + }, + { + "epoch": 0.16084, + "grad_norm": 0.5820706486701965, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 16084 + }, + { + "epoch": 0.16085, + "grad_norm": 0.5111738443374634, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 16085 + }, + { + "epoch": 0.16086, + "grad_norm": 0.6005224585533142, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 16086 + }, + { + "epoch": 0.16087, + "grad_norm": 0.8387770056724548, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 16087 + }, + { + "epoch": 0.16088, + "grad_norm": 1.1707104444503784, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 16088 + }, + { + "epoch": 0.16089, + "grad_norm": 0.8848904371261597, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 16089 + }, + { + "epoch": 0.1609, + "grad_norm": 0.76739102602005, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 16090 + }, + { + "epoch": 0.16091, + "grad_norm": 0.8427786827087402, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 16091 + }, + { + "epoch": 0.16092, + "grad_norm": 1.0178136825561523, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 16092 + }, + { + "epoch": 0.16093, + "grad_norm": 1.1659822463989258, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 16093 + }, + { + "epoch": 0.16094, + "grad_norm": 1.002729892730713, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 16094 + }, + { + "epoch": 0.16095, + "grad_norm": 0.9544771909713745, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 16095 + }, + { + "epoch": 0.16096, + "grad_norm": 1.047904372215271, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 16096 + }, + { + "epoch": 0.16097, + "grad_norm": 1.0617247819900513, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 16097 + }, + { + "epoch": 0.16098, + "grad_norm": 0.9110809564590454, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 16098 + }, + { + "epoch": 0.16099, + "grad_norm": 0.942016065120697, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 16099 + }, + { + "epoch": 0.161, + "grad_norm": 1.09463369846344, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 16100 + }, + { + "epoch": 0.16101, + "grad_norm": 0.9544133543968201, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 16101 + }, + { + "epoch": 0.16102, + "grad_norm": 0.9932712316513062, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 16102 + }, + { + "epoch": 0.16103, + "grad_norm": 0.9824539422988892, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 16103 + }, + { + "epoch": 0.16104, + "grad_norm": 0.9473547339439392, + "learning_rate": 0.003, + "loss": 4.017, + "step": 16104 + }, + { + "epoch": 0.16105, + "grad_norm": 0.8918814659118652, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 16105 + }, + { + "epoch": 0.16106, + "grad_norm": 0.9244639277458191, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 16106 + }, + { + "epoch": 0.16107, + "grad_norm": 0.8934754133224487, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 16107 + }, + { + "epoch": 0.16108, + "grad_norm": 0.9675315618515015, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 16108 + }, + { + "epoch": 0.16109, + "grad_norm": 1.0206494331359863, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 16109 + }, + { + "epoch": 0.1611, + "grad_norm": 0.8940903544425964, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 16110 + }, + { + "epoch": 0.16111, + "grad_norm": 0.9600245952606201, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 16111 + }, + { + "epoch": 0.16112, + "grad_norm": 0.8340402245521545, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 16112 + }, + { + "epoch": 0.16113, + "grad_norm": 0.693452775478363, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 16113 + }, + { + "epoch": 0.16114, + "grad_norm": 0.6798891425132751, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 16114 + }, + { + "epoch": 0.16115, + "grad_norm": 0.6488301157951355, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 16115 + }, + { + "epoch": 0.16116, + "grad_norm": 0.6416028141975403, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 16116 + }, + { + "epoch": 0.16117, + "grad_norm": 0.7598521113395691, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 16117 + }, + { + "epoch": 0.16118, + "grad_norm": 1.1467598676681519, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 16118 + }, + { + "epoch": 0.16119, + "grad_norm": 1.098025918006897, + "learning_rate": 0.003, + "loss": 4.02, + "step": 16119 + }, + { + "epoch": 0.1612, + "grad_norm": 0.8007514476776123, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 16120 + }, + { + "epoch": 0.16121, + "grad_norm": 0.6835646629333496, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 16121 + }, + { + "epoch": 0.16122, + "grad_norm": 0.7399167418479919, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 16122 + }, + { + "epoch": 0.16123, + "grad_norm": 0.7732407450675964, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 16123 + }, + { + "epoch": 0.16124, + "grad_norm": 0.6835975050926208, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 16124 + }, + { + "epoch": 0.16125, + "grad_norm": 0.5874419808387756, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 16125 + }, + { + "epoch": 0.16126, + "grad_norm": 0.6312301754951477, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 16126 + }, + { + "epoch": 0.16127, + "grad_norm": 0.6397377252578735, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 16127 + }, + { + "epoch": 0.16128, + "grad_norm": 0.7009602785110474, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 16128 + }, + { + "epoch": 0.16129, + "grad_norm": 0.6710821390151978, + "learning_rate": 0.003, + "loss": 4.014, + "step": 16129 + }, + { + "epoch": 0.1613, + "grad_norm": 0.7803601622581482, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 16130 + }, + { + "epoch": 0.16131, + "grad_norm": 1.0563775300979614, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 16131 + }, + { + "epoch": 0.16132, + "grad_norm": 1.0355607271194458, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 16132 + }, + { + "epoch": 0.16133, + "grad_norm": 0.7016539573669434, + "learning_rate": 0.003, + "loss": 3.998, + "step": 16133 + }, + { + "epoch": 0.16134, + "grad_norm": 0.5129038691520691, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 16134 + }, + { + "epoch": 0.16135, + "grad_norm": 0.5346699357032776, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 16135 + }, + { + "epoch": 0.16136, + "grad_norm": 0.6307773590087891, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 16136 + }, + { + "epoch": 0.16137, + "grad_norm": 0.7481401562690735, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 16137 + }, + { + "epoch": 0.16138, + "grad_norm": 0.7668987512588501, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 16138 + }, + { + "epoch": 0.16139, + "grad_norm": 0.7831903100013733, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 16139 + }, + { + "epoch": 0.1614, + "grad_norm": 0.9852398037910461, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16140 + }, + { + "epoch": 0.16141, + "grad_norm": 1.1628477573394775, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 16141 + }, + { + "epoch": 0.16142, + "grad_norm": 0.8994024991989136, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 16142 + }, + { + "epoch": 0.16143, + "grad_norm": 0.778788149356842, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 16143 + }, + { + "epoch": 0.16144, + "grad_norm": 0.8439193964004517, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 16144 + }, + { + "epoch": 0.16145, + "grad_norm": 0.7844627499580383, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 16145 + }, + { + "epoch": 0.16146, + "grad_norm": 0.873829185962677, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 16146 + }, + { + "epoch": 0.16147, + "grad_norm": 1.0035282373428345, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 16147 + }, + { + "epoch": 0.16148, + "grad_norm": 0.9474618434906006, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 16148 + }, + { + "epoch": 0.16149, + "grad_norm": 0.9962616562843323, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 16149 + }, + { + "epoch": 0.1615, + "grad_norm": 1.0283039808273315, + "learning_rate": 0.003, + "loss": 4.03, + "step": 16150 + }, + { + "epoch": 0.16151, + "grad_norm": 1.0023417472839355, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 16151 + }, + { + "epoch": 0.16152, + "grad_norm": 0.9786602854728699, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 16152 + }, + { + "epoch": 0.16153, + "grad_norm": 0.9277042746543884, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 16153 + }, + { + "epoch": 0.16154, + "grad_norm": 0.8046708703041077, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 16154 + }, + { + "epoch": 0.16155, + "grad_norm": 0.7589371204376221, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 16155 + }, + { + "epoch": 0.16156, + "grad_norm": 0.8702610731124878, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 16156 + }, + { + "epoch": 0.16157, + "grad_norm": 0.9417670369148254, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16157 + }, + { + "epoch": 0.16158, + "grad_norm": 0.9658819437026978, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 16158 + }, + { + "epoch": 0.16159, + "grad_norm": 0.9872148633003235, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 16159 + }, + { + "epoch": 0.1616, + "grad_norm": 0.8776973485946655, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 16160 + }, + { + "epoch": 0.16161, + "grad_norm": 0.8566110134124756, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 16161 + }, + { + "epoch": 0.16162, + "grad_norm": 0.9043013453483582, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 16162 + }, + { + "epoch": 0.16163, + "grad_norm": 0.8688957691192627, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 16163 + }, + { + "epoch": 0.16164, + "grad_norm": 0.8106199502944946, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 16164 + }, + { + "epoch": 0.16165, + "grad_norm": 0.8550721406936646, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 16165 + }, + { + "epoch": 0.16166, + "grad_norm": 0.9837195873260498, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 16166 + }, + { + "epoch": 0.16167, + "grad_norm": 1.1086894273757935, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 16167 + }, + { + "epoch": 0.16168, + "grad_norm": 0.8541070222854614, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 16168 + }, + { + "epoch": 0.16169, + "grad_norm": 0.6221586465835571, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 16169 + }, + { + "epoch": 0.1617, + "grad_norm": 0.6150177717208862, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 16170 + }, + { + "epoch": 0.16171, + "grad_norm": 0.6094152927398682, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 16171 + }, + { + "epoch": 0.16172, + "grad_norm": 0.5822041630744934, + "learning_rate": 0.003, + "loss": 4.008, + "step": 16172 + }, + { + "epoch": 0.16173, + "grad_norm": 0.6120102405548096, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 16173 + }, + { + "epoch": 0.16174, + "grad_norm": 0.6903099417686462, + "learning_rate": 0.003, + "loss": 4.022, + "step": 16174 + }, + { + "epoch": 0.16175, + "grad_norm": 0.6956382393836975, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 16175 + }, + { + "epoch": 0.16176, + "grad_norm": 0.6636068224906921, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16176 + }, + { + "epoch": 0.16177, + "grad_norm": 0.675407886505127, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 16177 + }, + { + "epoch": 0.16178, + "grad_norm": 0.9279387593269348, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 16178 + }, + { + "epoch": 0.16179, + "grad_norm": 1.267047643661499, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 16179 + }, + { + "epoch": 0.1618, + "grad_norm": 0.8936948180198669, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 16180 + }, + { + "epoch": 0.16181, + "grad_norm": 0.9198494553565979, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 16181 + }, + { + "epoch": 0.16182, + "grad_norm": 0.9184507131576538, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 16182 + }, + { + "epoch": 0.16183, + "grad_norm": 0.8332173824310303, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 16183 + }, + { + "epoch": 0.16184, + "grad_norm": 0.8280366659164429, + "learning_rate": 0.003, + "loss": 3.994, + "step": 16184 + }, + { + "epoch": 0.16185, + "grad_norm": 0.8992508053779602, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 16185 + }, + { + "epoch": 0.16186, + "grad_norm": 0.960120677947998, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 16186 + }, + { + "epoch": 0.16187, + "grad_norm": 0.903139591217041, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 16187 + }, + { + "epoch": 0.16188, + "grad_norm": 0.8598777055740356, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 16188 + }, + { + "epoch": 0.16189, + "grad_norm": 0.796144962310791, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 16189 + }, + { + "epoch": 0.1619, + "grad_norm": 0.8381752967834473, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 16190 + }, + { + "epoch": 0.16191, + "grad_norm": 0.802821695804596, + "learning_rate": 0.003, + "loss": 4.012, + "step": 16191 + }, + { + "epoch": 0.16192, + "grad_norm": 0.8790308237075806, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 16192 + }, + { + "epoch": 0.16193, + "grad_norm": 1.0374481678009033, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 16193 + }, + { + "epoch": 0.16194, + "grad_norm": 0.9121679663658142, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 16194 + }, + { + "epoch": 0.16195, + "grad_norm": 0.8900026082992554, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 16195 + }, + { + "epoch": 0.16196, + "grad_norm": 1.0195919275283813, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 16196 + }, + { + "epoch": 0.16197, + "grad_norm": 1.005298376083374, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 16197 + }, + { + "epoch": 0.16198, + "grad_norm": 0.8747978210449219, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 16198 + }, + { + "epoch": 0.16199, + "grad_norm": 0.9359422326087952, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 16199 + }, + { + "epoch": 0.162, + "grad_norm": 0.8763812184333801, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 16200 + }, + { + "epoch": 0.16201, + "grad_norm": 0.7494961023330688, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 16201 + }, + { + "epoch": 0.16202, + "grad_norm": 0.7147844433784485, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 16202 + }, + { + "epoch": 0.16203, + "grad_norm": 0.7433773875236511, + "learning_rate": 0.003, + "loss": 3.999, + "step": 16203 + }, + { + "epoch": 0.16204, + "grad_norm": 0.9177911281585693, + "learning_rate": 0.003, + "loss": 4.038, + "step": 16204 + }, + { + "epoch": 0.16205, + "grad_norm": 1.0803254842758179, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 16205 + }, + { + "epoch": 0.16206, + "grad_norm": 1.1402027606964111, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 16206 + }, + { + "epoch": 0.16207, + "grad_norm": 0.8355861306190491, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 16207 + }, + { + "epoch": 0.16208, + "grad_norm": 0.6533427834510803, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 16208 + }, + { + "epoch": 0.16209, + "grad_norm": 0.6891869902610779, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 16209 + }, + { + "epoch": 0.1621, + "grad_norm": 0.7308468818664551, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 16210 + }, + { + "epoch": 0.16211, + "grad_norm": 0.7914701700210571, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 16211 + }, + { + "epoch": 0.16212, + "grad_norm": 0.839141845703125, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 16212 + }, + { + "epoch": 0.16213, + "grad_norm": 0.888842761516571, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 16213 + }, + { + "epoch": 0.16214, + "grad_norm": 0.9561465382575989, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 16214 + }, + { + "epoch": 0.16215, + "grad_norm": 0.8719993829727173, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 16215 + }, + { + "epoch": 0.16216, + "grad_norm": 0.8226259350776672, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 16216 + }, + { + "epoch": 0.16217, + "grad_norm": 0.7822552919387817, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 16217 + }, + { + "epoch": 0.16218, + "grad_norm": 0.7415698766708374, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 16218 + }, + { + "epoch": 0.16219, + "grad_norm": 0.7511895298957825, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 16219 + }, + { + "epoch": 0.1622, + "grad_norm": 0.8057510852813721, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 16220 + }, + { + "epoch": 0.16221, + "grad_norm": 0.7829014658927917, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 16221 + }, + { + "epoch": 0.16222, + "grad_norm": 0.7896765470504761, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 16222 + }, + { + "epoch": 0.16223, + "grad_norm": 0.7292823195457458, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 16223 + }, + { + "epoch": 0.16224, + "grad_norm": 0.7166668772697449, + "learning_rate": 0.003, + "loss": 3.996, + "step": 16224 + }, + { + "epoch": 0.16225, + "grad_norm": 0.7180719971656799, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 16225 + }, + { + "epoch": 0.16226, + "grad_norm": 0.8296489119529724, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 16226 + }, + { + "epoch": 0.16227, + "grad_norm": 1.0246996879577637, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 16227 + }, + { + "epoch": 0.16228, + "grad_norm": 1.2064777612686157, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 16228 + }, + { + "epoch": 0.16229, + "grad_norm": 0.7835697531700134, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 16229 + }, + { + "epoch": 0.1623, + "grad_norm": 0.7163366079330444, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 16230 + }, + { + "epoch": 0.16231, + "grad_norm": 0.8635886907577515, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 16231 + }, + { + "epoch": 0.16232, + "grad_norm": 0.9135334491729736, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 16232 + }, + { + "epoch": 0.16233, + "grad_norm": 1.0937687158584595, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 16233 + }, + { + "epoch": 0.16234, + "grad_norm": 0.9035110473632812, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 16234 + }, + { + "epoch": 0.16235, + "grad_norm": 0.8154218196868896, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16235 + }, + { + "epoch": 0.16236, + "grad_norm": 0.717538058757782, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 16236 + }, + { + "epoch": 0.16237, + "grad_norm": 0.8207351565361023, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 16237 + }, + { + "epoch": 0.16238, + "grad_norm": 0.9286295771598816, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 16238 + }, + { + "epoch": 0.16239, + "grad_norm": 0.8026152849197388, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 16239 + }, + { + "epoch": 0.1624, + "grad_norm": 0.7241533398628235, + "learning_rate": 0.003, + "loss": 4.002, + "step": 16240 + }, + { + "epoch": 0.16241, + "grad_norm": 0.7309879064559937, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 16241 + }, + { + "epoch": 0.16242, + "grad_norm": 0.8363540172576904, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 16242 + }, + { + "epoch": 0.16243, + "grad_norm": 0.8413676023483276, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 16243 + }, + { + "epoch": 0.16244, + "grad_norm": 0.9204793572425842, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 16244 + }, + { + "epoch": 0.16245, + "grad_norm": 1.1601897478103638, + "learning_rate": 0.003, + "loss": 4.003, + "step": 16245 + }, + { + "epoch": 0.16246, + "grad_norm": 0.9563637375831604, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 16246 + }, + { + "epoch": 0.16247, + "grad_norm": 0.9307214617729187, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 16247 + }, + { + "epoch": 0.16248, + "grad_norm": 1.002216100692749, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 16248 + }, + { + "epoch": 0.16249, + "grad_norm": 1.1238837242126465, + "learning_rate": 0.003, + "loss": 4.022, + "step": 16249 + }, + { + "epoch": 0.1625, + "grad_norm": 0.8523120880126953, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 16250 + }, + { + "epoch": 0.16251, + "grad_norm": 0.7996931076049805, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 16251 + }, + { + "epoch": 0.16252, + "grad_norm": 0.8345876932144165, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 16252 + }, + { + "epoch": 0.16253, + "grad_norm": 0.9343923926353455, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 16253 + }, + { + "epoch": 0.16254, + "grad_norm": 0.8420714735984802, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 16254 + }, + { + "epoch": 0.16255, + "grad_norm": 0.6601258516311646, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 16255 + }, + { + "epoch": 0.16256, + "grad_norm": 0.66865074634552, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 16256 + }, + { + "epoch": 0.16257, + "grad_norm": 0.7216544151306152, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 16257 + }, + { + "epoch": 0.16258, + "grad_norm": 0.7510634064674377, + "learning_rate": 0.003, + "loss": 4.01, + "step": 16258 + }, + { + "epoch": 0.16259, + "grad_norm": 1.0844184160232544, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 16259 + }, + { + "epoch": 0.1626, + "grad_norm": 1.138572096824646, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 16260 + }, + { + "epoch": 0.16261, + "grad_norm": 0.8396537899971008, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 16261 + }, + { + "epoch": 0.16262, + "grad_norm": 0.7281916737556458, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 16262 + }, + { + "epoch": 0.16263, + "grad_norm": 0.6730375289916992, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 16263 + }, + { + "epoch": 0.16264, + "grad_norm": 0.6375163197517395, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 16264 + }, + { + "epoch": 0.16265, + "grad_norm": 0.6660574674606323, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 16265 + }, + { + "epoch": 0.16266, + "grad_norm": 0.7589208483695984, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 16266 + }, + { + "epoch": 0.16267, + "grad_norm": 0.9495667219161987, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 16267 + }, + { + "epoch": 0.16268, + "grad_norm": 1.158148169517517, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 16268 + }, + { + "epoch": 0.16269, + "grad_norm": 0.7837066650390625, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 16269 + }, + { + "epoch": 0.1627, + "grad_norm": 0.7256262898445129, + "learning_rate": 0.003, + "loss": 4.02, + "step": 16270 + }, + { + "epoch": 0.16271, + "grad_norm": 0.8901933431625366, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 16271 + }, + { + "epoch": 0.16272, + "grad_norm": 0.9590361714363098, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 16272 + }, + { + "epoch": 0.16273, + "grad_norm": 1.023372769355774, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 16273 + }, + { + "epoch": 0.16274, + "grad_norm": 0.9545578360557556, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 16274 + }, + { + "epoch": 0.16275, + "grad_norm": 0.7929362654685974, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 16275 + }, + { + "epoch": 0.16276, + "grad_norm": 0.6679825186729431, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 16276 + }, + { + "epoch": 0.16277, + "grad_norm": 0.58383709192276, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 16277 + }, + { + "epoch": 0.16278, + "grad_norm": 0.7074006199836731, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 16278 + }, + { + "epoch": 0.16279, + "grad_norm": 0.774600625038147, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 16279 + }, + { + "epoch": 0.1628, + "grad_norm": 0.6628326177597046, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 16280 + }, + { + "epoch": 0.16281, + "grad_norm": 0.6247139573097229, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 16281 + }, + { + "epoch": 0.16282, + "grad_norm": 0.5879446864128113, + "learning_rate": 0.003, + "loss": 3.979, + "step": 16282 + }, + { + "epoch": 0.16283, + "grad_norm": 0.6044084429740906, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 16283 + }, + { + "epoch": 0.16284, + "grad_norm": 0.6138390898704529, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 16284 + }, + { + "epoch": 0.16285, + "grad_norm": 0.6722159385681152, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 16285 + }, + { + "epoch": 0.16286, + "grad_norm": 0.7046880722045898, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 16286 + }, + { + "epoch": 0.16287, + "grad_norm": 0.7353987097740173, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 16287 + }, + { + "epoch": 0.16288, + "grad_norm": 0.9673696756362915, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 16288 + }, + { + "epoch": 0.16289, + "grad_norm": 1.1569815874099731, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 16289 + }, + { + "epoch": 0.1629, + "grad_norm": 0.7916101217269897, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 16290 + }, + { + "epoch": 0.16291, + "grad_norm": 0.8873964548110962, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 16291 + }, + { + "epoch": 0.16292, + "grad_norm": 0.9866530299186707, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 16292 + }, + { + "epoch": 0.16293, + "grad_norm": 1.2426506280899048, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16293 + }, + { + "epoch": 0.16294, + "grad_norm": 0.674839973449707, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 16294 + }, + { + "epoch": 0.16295, + "grad_norm": 0.697393536567688, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 16295 + }, + { + "epoch": 0.16296, + "grad_norm": 0.8670492768287659, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 16296 + }, + { + "epoch": 0.16297, + "grad_norm": 1.480337381362915, + "learning_rate": 0.003, + "loss": 4.021, + "step": 16297 + }, + { + "epoch": 0.16298, + "grad_norm": 0.5470685362815857, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 16298 + }, + { + "epoch": 0.16299, + "grad_norm": 0.7625333070755005, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 16299 + }, + { + "epoch": 0.163, + "grad_norm": 0.8565612435340881, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16300 + }, + { + "epoch": 0.16301, + "grad_norm": 0.8694688677787781, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 16301 + }, + { + "epoch": 0.16302, + "grad_norm": 0.8741586804389954, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 16302 + }, + { + "epoch": 0.16303, + "grad_norm": 0.9137376546859741, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 16303 + }, + { + "epoch": 0.16304, + "grad_norm": 1.0089272260665894, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 16304 + }, + { + "epoch": 0.16305, + "grad_norm": 1.0015945434570312, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 16305 + }, + { + "epoch": 0.16306, + "grad_norm": 1.0515681505203247, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 16306 + }, + { + "epoch": 0.16307, + "grad_norm": 1.0251307487487793, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 16307 + }, + { + "epoch": 0.16308, + "grad_norm": 0.9413036704063416, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 16308 + }, + { + "epoch": 0.16309, + "grad_norm": 1.1088650226593018, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 16309 + }, + { + "epoch": 0.1631, + "grad_norm": 0.9975844025611877, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 16310 + }, + { + "epoch": 0.16311, + "grad_norm": 1.0978097915649414, + "learning_rate": 0.003, + "loss": 4.033, + "step": 16311 + }, + { + "epoch": 0.16312, + "grad_norm": 0.9465668201446533, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 16312 + }, + { + "epoch": 0.16313, + "grad_norm": 0.9554963111877441, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 16313 + }, + { + "epoch": 0.16314, + "grad_norm": 0.9088976979255676, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 16314 + }, + { + "epoch": 0.16315, + "grad_norm": 0.9877479076385498, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 16315 + }, + { + "epoch": 0.16316, + "grad_norm": 0.7756491899490356, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 16316 + }, + { + "epoch": 0.16317, + "grad_norm": 0.705436110496521, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 16317 + }, + { + "epoch": 0.16318, + "grad_norm": 0.7293632626533508, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 16318 + }, + { + "epoch": 0.16319, + "grad_norm": 0.6516294479370117, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 16319 + }, + { + "epoch": 0.1632, + "grad_norm": 0.6582109928131104, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 16320 + }, + { + "epoch": 0.16321, + "grad_norm": 0.7615460157394409, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 16321 + }, + { + "epoch": 0.16322, + "grad_norm": 0.9899001121520996, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 16322 + }, + { + "epoch": 0.16323, + "grad_norm": 1.1688170433044434, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 16323 + }, + { + "epoch": 0.16324, + "grad_norm": 1.1293865442276, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 16324 + }, + { + "epoch": 0.16325, + "grad_norm": 0.7544232606887817, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 16325 + }, + { + "epoch": 0.16326, + "grad_norm": 0.5888630151748657, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 16326 + }, + { + "epoch": 0.16327, + "grad_norm": 0.7046893835067749, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 16327 + }, + { + "epoch": 0.16328, + "grad_norm": 0.7926900386810303, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 16328 + }, + { + "epoch": 0.16329, + "grad_norm": 0.9033036231994629, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 16329 + }, + { + "epoch": 0.1633, + "grad_norm": 0.8790760040283203, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 16330 + }, + { + "epoch": 0.16331, + "grad_norm": 0.8309141397476196, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 16331 + }, + { + "epoch": 0.16332, + "grad_norm": 0.8339505195617676, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 16332 + }, + { + "epoch": 0.16333, + "grad_norm": 0.8661202192306519, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 16333 + }, + { + "epoch": 0.16334, + "grad_norm": 0.7818444967269897, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 16334 + }, + { + "epoch": 0.16335, + "grad_norm": 0.7368418574333191, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 16335 + }, + { + "epoch": 0.16336, + "grad_norm": 0.6508287191390991, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 16336 + }, + { + "epoch": 0.16337, + "grad_norm": 0.7051012516021729, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 16337 + }, + { + "epoch": 0.16338, + "grad_norm": 0.7545062303543091, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 16338 + }, + { + "epoch": 0.16339, + "grad_norm": 0.8058009147644043, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 16339 + }, + { + "epoch": 0.1634, + "grad_norm": 0.9874696135520935, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 16340 + }, + { + "epoch": 0.16341, + "grad_norm": 0.9253709316253662, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 16341 + }, + { + "epoch": 0.16342, + "grad_norm": 0.6806486248970032, + "learning_rate": 0.003, + "loss": 3.991, + "step": 16342 + }, + { + "epoch": 0.16343, + "grad_norm": 0.6025413274765015, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 16343 + }, + { + "epoch": 0.16344, + "grad_norm": 0.6132237315177917, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 16344 + }, + { + "epoch": 0.16345, + "grad_norm": 0.6393827795982361, + "learning_rate": 0.003, + "loss": 4.013, + "step": 16345 + }, + { + "epoch": 0.16346, + "grad_norm": 0.615208625793457, + "learning_rate": 0.003, + "loss": 4.006, + "step": 16346 + }, + { + "epoch": 0.16347, + "grad_norm": 0.7005764842033386, + "learning_rate": 0.003, + "loss": 4.027, + "step": 16347 + }, + { + "epoch": 0.16348, + "grad_norm": 0.7764422297477722, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 16348 + }, + { + "epoch": 0.16349, + "grad_norm": 0.7720112800598145, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 16349 + }, + { + "epoch": 0.1635, + "grad_norm": 0.838747501373291, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 16350 + }, + { + "epoch": 0.16351, + "grad_norm": 0.8589238524436951, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 16351 + }, + { + "epoch": 0.16352, + "grad_norm": 1.0065206289291382, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 16352 + }, + { + "epoch": 0.16353, + "grad_norm": 1.0837733745574951, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 16353 + }, + { + "epoch": 0.16354, + "grad_norm": 0.788565456867218, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 16354 + }, + { + "epoch": 0.16355, + "grad_norm": 0.6954969167709351, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 16355 + }, + { + "epoch": 0.16356, + "grad_norm": 0.7750489115715027, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 16356 + }, + { + "epoch": 0.16357, + "grad_norm": 0.9297155141830444, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 16357 + }, + { + "epoch": 0.16358, + "grad_norm": 0.9632990956306458, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 16358 + }, + { + "epoch": 0.16359, + "grad_norm": 0.8834934830665588, + "learning_rate": 0.003, + "loss": 4.014, + "step": 16359 + }, + { + "epoch": 0.1636, + "grad_norm": 0.9617258906364441, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 16360 + }, + { + "epoch": 0.16361, + "grad_norm": 0.9705206155776978, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 16361 + }, + { + "epoch": 0.16362, + "grad_norm": 0.8674737215042114, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 16362 + }, + { + "epoch": 0.16363, + "grad_norm": 0.8213301301002502, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 16363 + }, + { + "epoch": 0.16364, + "grad_norm": 0.8834892511367798, + "learning_rate": 0.003, + "loss": 4.031, + "step": 16364 + }, + { + "epoch": 0.16365, + "grad_norm": 0.8041810393333435, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 16365 + }, + { + "epoch": 0.16366, + "grad_norm": 0.6862834692001343, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 16366 + }, + { + "epoch": 0.16367, + "grad_norm": 0.7433396577835083, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 16367 + }, + { + "epoch": 0.16368, + "grad_norm": 0.8846827149391174, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 16368 + }, + { + "epoch": 0.16369, + "grad_norm": 1.0452004671096802, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 16369 + }, + { + "epoch": 0.1637, + "grad_norm": 1.139094352722168, + "learning_rate": 0.003, + "loss": 4.006, + "step": 16370 + }, + { + "epoch": 0.16371, + "grad_norm": 0.9088010787963867, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 16371 + }, + { + "epoch": 0.16372, + "grad_norm": 0.8271872997283936, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 16372 + }, + { + "epoch": 0.16373, + "grad_norm": 0.7686737179756165, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 16373 + }, + { + "epoch": 0.16374, + "grad_norm": 0.6929585337638855, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 16374 + }, + { + "epoch": 0.16375, + "grad_norm": 0.8070968985557556, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 16375 + }, + { + "epoch": 0.16376, + "grad_norm": 0.7533002495765686, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 16376 + }, + { + "epoch": 0.16377, + "grad_norm": 0.7526688575744629, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 16377 + }, + { + "epoch": 0.16378, + "grad_norm": 0.8268585801124573, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 16378 + }, + { + "epoch": 0.16379, + "grad_norm": 1.0098788738250732, + "learning_rate": 0.003, + "loss": 4.003, + "step": 16379 + }, + { + "epoch": 0.1638, + "grad_norm": 1.2174149751663208, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 16380 + }, + { + "epoch": 0.16381, + "grad_norm": 0.6912074685096741, + "learning_rate": 0.003, + "loss": 4.001, + "step": 16381 + }, + { + "epoch": 0.16382, + "grad_norm": 0.7847248315811157, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 16382 + }, + { + "epoch": 0.16383, + "grad_norm": 0.8508479595184326, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 16383 + }, + { + "epoch": 0.16384, + "grad_norm": 0.7754083275794983, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 16384 + }, + { + "epoch": 0.16385, + "grad_norm": 0.7800403237342834, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 16385 + }, + { + "epoch": 0.16386, + "grad_norm": 0.7742740511894226, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 16386 + }, + { + "epoch": 0.16387, + "grad_norm": 0.8291677832603455, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 16387 + }, + { + "epoch": 0.16388, + "grad_norm": 0.8471203446388245, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 16388 + }, + { + "epoch": 0.16389, + "grad_norm": 0.7955463528633118, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 16389 + }, + { + "epoch": 0.1639, + "grad_norm": 0.7575664520263672, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 16390 + }, + { + "epoch": 0.16391, + "grad_norm": 0.6484594941139221, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 16391 + }, + { + "epoch": 0.16392, + "grad_norm": 0.6634478569030762, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 16392 + }, + { + "epoch": 0.16393, + "grad_norm": 0.5433070659637451, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 16393 + }, + { + "epoch": 0.16394, + "grad_norm": 0.49226635694503784, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 16394 + }, + { + "epoch": 0.16395, + "grad_norm": 0.5242682695388794, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 16395 + }, + { + "epoch": 0.16396, + "grad_norm": 0.4910825490951538, + "learning_rate": 0.003, + "loss": 3.98, + "step": 16396 + }, + { + "epoch": 0.16397, + "grad_norm": 0.6202237010002136, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 16397 + }, + { + "epoch": 0.16398, + "grad_norm": 0.8250946998596191, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 16398 + }, + { + "epoch": 0.16399, + "grad_norm": 1.031864047050476, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 16399 + }, + { + "epoch": 0.164, + "grad_norm": 1.2869422435760498, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 16400 + }, + { + "epoch": 0.16401, + "grad_norm": 0.6132083535194397, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 16401 + }, + { + "epoch": 0.16402, + "grad_norm": 0.6066889762878418, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 16402 + }, + { + "epoch": 0.16403, + "grad_norm": 0.7775403261184692, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 16403 + }, + { + "epoch": 0.16404, + "grad_norm": 0.8699544668197632, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 16404 + }, + { + "epoch": 0.16405, + "grad_norm": 0.9804795384407043, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 16405 + }, + { + "epoch": 0.16406, + "grad_norm": 0.8409727811813354, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 16406 + }, + { + "epoch": 0.16407, + "grad_norm": 0.6656890511512756, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 16407 + }, + { + "epoch": 0.16408, + "grad_norm": 0.8278241157531738, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 16408 + }, + { + "epoch": 0.16409, + "grad_norm": 1.022950291633606, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 16409 + }, + { + "epoch": 0.1641, + "grad_norm": 1.1753534078598022, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 16410 + }, + { + "epoch": 0.16411, + "grad_norm": 0.8887650370597839, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 16411 + }, + { + "epoch": 0.16412, + "grad_norm": 1.007218599319458, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 16412 + }, + { + "epoch": 0.16413, + "grad_norm": 1.1239910125732422, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 16413 + }, + { + "epoch": 0.16414, + "grad_norm": 0.863362193107605, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 16414 + }, + { + "epoch": 0.16415, + "grad_norm": 0.9276494979858398, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 16415 + }, + { + "epoch": 0.16416, + "grad_norm": 1.1005187034606934, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 16416 + }, + { + "epoch": 0.16417, + "grad_norm": 0.9491989016532898, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 16417 + }, + { + "epoch": 0.16418, + "grad_norm": 0.9228178262710571, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 16418 + }, + { + "epoch": 0.16419, + "grad_norm": 0.9737118482589722, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 16419 + }, + { + "epoch": 0.1642, + "grad_norm": 0.9858505725860596, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 16420 + }, + { + "epoch": 0.16421, + "grad_norm": 1.0404865741729736, + "learning_rate": 0.003, + "loss": 4.023, + "step": 16421 + }, + { + "epoch": 0.16422, + "grad_norm": 1.0032793283462524, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 16422 + }, + { + "epoch": 0.16423, + "grad_norm": 1.0093377828598022, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 16423 + }, + { + "epoch": 0.16424, + "grad_norm": 1.0854952335357666, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 16424 + }, + { + "epoch": 0.16425, + "grad_norm": 0.910970151424408, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 16425 + }, + { + "epoch": 0.16426, + "grad_norm": 0.8588374257087708, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 16426 + }, + { + "epoch": 0.16427, + "grad_norm": 0.9977859854698181, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 16427 + }, + { + "epoch": 0.16428, + "grad_norm": 1.202165126800537, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 16428 + }, + { + "epoch": 0.16429, + "grad_norm": 0.7841586470603943, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16429 + }, + { + "epoch": 0.1643, + "grad_norm": 0.8664776682853699, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 16430 + }, + { + "epoch": 0.16431, + "grad_norm": 0.7448684573173523, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 16431 + }, + { + "epoch": 0.16432, + "grad_norm": 0.8166662454605103, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 16432 + }, + { + "epoch": 0.16433, + "grad_norm": 0.8164058327674866, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 16433 + }, + { + "epoch": 0.16434, + "grad_norm": 0.7713918685913086, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 16434 + }, + { + "epoch": 0.16435, + "grad_norm": 0.7069947719573975, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 16435 + }, + { + "epoch": 0.16436, + "grad_norm": 0.6676192879676819, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 16436 + }, + { + "epoch": 0.16437, + "grad_norm": 0.7678779363632202, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 16437 + }, + { + "epoch": 0.16438, + "grad_norm": 0.7208015322685242, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 16438 + }, + { + "epoch": 0.16439, + "grad_norm": 0.7070453763008118, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 16439 + }, + { + "epoch": 0.1644, + "grad_norm": 0.9220455884933472, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 16440 + }, + { + "epoch": 0.16441, + "grad_norm": 1.2636245489120483, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 16441 + }, + { + "epoch": 0.16442, + "grad_norm": 0.678617537021637, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 16442 + }, + { + "epoch": 0.16443, + "grad_norm": 0.49778828024864197, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 16443 + }, + { + "epoch": 0.16444, + "grad_norm": 0.5420036315917969, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 16444 + }, + { + "epoch": 0.16445, + "grad_norm": 0.727878212928772, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 16445 + }, + { + "epoch": 0.16446, + "grad_norm": 0.7359219789505005, + "learning_rate": 0.003, + "loss": 3.993, + "step": 16446 + }, + { + "epoch": 0.16447, + "grad_norm": 0.5902706980705261, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 16447 + }, + { + "epoch": 0.16448, + "grad_norm": 0.5670353770256042, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 16448 + }, + { + "epoch": 0.16449, + "grad_norm": 0.6856237053871155, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 16449 + }, + { + "epoch": 0.1645, + "grad_norm": 0.8142359852790833, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 16450 + }, + { + "epoch": 0.16451, + "grad_norm": 0.8532720804214478, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 16451 + }, + { + "epoch": 0.16452, + "grad_norm": 1.0314480066299438, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 16452 + }, + { + "epoch": 0.16453, + "grad_norm": 1.0708210468292236, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 16453 + }, + { + "epoch": 0.16454, + "grad_norm": 0.8132475018501282, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 16454 + }, + { + "epoch": 0.16455, + "grad_norm": 0.7674928307533264, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 16455 + }, + { + "epoch": 0.16456, + "grad_norm": 0.8469036221504211, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 16456 + }, + { + "epoch": 0.16457, + "grad_norm": 0.852376401424408, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 16457 + }, + { + "epoch": 0.16458, + "grad_norm": 0.8676538467407227, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 16458 + }, + { + "epoch": 0.16459, + "grad_norm": 1.0059013366699219, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 16459 + }, + { + "epoch": 0.1646, + "grad_norm": 1.1781777143478394, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 16460 + }, + { + "epoch": 0.16461, + "grad_norm": 0.795784056186676, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 16461 + }, + { + "epoch": 0.16462, + "grad_norm": 0.6251187920570374, + "learning_rate": 0.003, + "loss": 3.991, + "step": 16462 + }, + { + "epoch": 0.16463, + "grad_norm": 0.5951520204544067, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 16463 + }, + { + "epoch": 0.16464, + "grad_norm": 0.6764260530471802, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 16464 + }, + { + "epoch": 0.16465, + "grad_norm": 0.7524860501289368, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 16465 + }, + { + "epoch": 0.16466, + "grad_norm": 0.9263510704040527, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 16466 + }, + { + "epoch": 0.16467, + "grad_norm": 1.0180227756500244, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 16467 + }, + { + "epoch": 0.16468, + "grad_norm": 0.9685873985290527, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 16468 + }, + { + "epoch": 0.16469, + "grad_norm": 1.026856780052185, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 16469 + }, + { + "epoch": 0.1647, + "grad_norm": 0.895588755607605, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 16470 + }, + { + "epoch": 0.16471, + "grad_norm": 0.8129714727401733, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 16471 + }, + { + "epoch": 0.16472, + "grad_norm": 1.013070821762085, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 16472 + }, + { + "epoch": 0.16473, + "grad_norm": 1.2826343774795532, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 16473 + }, + { + "epoch": 0.16474, + "grad_norm": 0.7825457453727722, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 16474 + }, + { + "epoch": 0.16475, + "grad_norm": 0.6030428409576416, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 16475 + }, + { + "epoch": 0.16476, + "grad_norm": 0.6580950617790222, + "learning_rate": 0.003, + "loss": 4.017, + "step": 16476 + }, + { + "epoch": 0.16477, + "grad_norm": 0.6692718267440796, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 16477 + }, + { + "epoch": 0.16478, + "grad_norm": 0.7654966711997986, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 16478 + }, + { + "epoch": 0.16479, + "grad_norm": 0.7372002601623535, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 16479 + }, + { + "epoch": 0.1648, + "grad_norm": 0.7242112755775452, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 16480 + }, + { + "epoch": 0.16481, + "grad_norm": 0.8388112783432007, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 16481 + }, + { + "epoch": 0.16482, + "grad_norm": 0.8586564660072327, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 16482 + }, + { + "epoch": 0.16483, + "grad_norm": 0.9531363844871521, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 16483 + }, + { + "epoch": 0.16484, + "grad_norm": 1.023642897605896, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 16484 + }, + { + "epoch": 0.16485, + "grad_norm": 0.9483336806297302, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 16485 + }, + { + "epoch": 0.16486, + "grad_norm": 0.8429510593414307, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 16486 + }, + { + "epoch": 0.16487, + "grad_norm": 0.8054298162460327, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 16487 + }, + { + "epoch": 0.16488, + "grad_norm": 1.0099833011627197, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 16488 + }, + { + "epoch": 0.16489, + "grad_norm": 1.0440032482147217, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 16489 + }, + { + "epoch": 0.1649, + "grad_norm": 0.9935981035232544, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 16490 + }, + { + "epoch": 0.16491, + "grad_norm": 0.8320918083190918, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 16491 + }, + { + "epoch": 0.16492, + "grad_norm": 0.6321850419044495, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 16492 + }, + { + "epoch": 0.16493, + "grad_norm": 0.5548147559165955, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 16493 + }, + { + "epoch": 0.16494, + "grad_norm": 0.6225996017456055, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 16494 + }, + { + "epoch": 0.16495, + "grad_norm": 0.7261919975280762, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 16495 + }, + { + "epoch": 0.16496, + "grad_norm": 0.774777352809906, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 16496 + }, + { + "epoch": 0.16497, + "grad_norm": 0.6230810284614563, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 16497 + }, + { + "epoch": 0.16498, + "grad_norm": 0.5928381085395813, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 16498 + }, + { + "epoch": 0.16499, + "grad_norm": 0.6339703798294067, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 16499 + }, + { + "epoch": 0.165, + "grad_norm": 0.593946099281311, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 16500 + }, + { + "epoch": 0.16501, + "grad_norm": 0.7014836072921753, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 16501 + }, + { + "epoch": 0.16502, + "grad_norm": 0.9443208575248718, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 16502 + }, + { + "epoch": 0.16503, + "grad_norm": 1.3297762870788574, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 16503 + }, + { + "epoch": 0.16504, + "grad_norm": 0.6993222236633301, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 16504 + }, + { + "epoch": 0.16505, + "grad_norm": 0.7262223958969116, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 16505 + }, + { + "epoch": 0.16506, + "grad_norm": 0.7761648893356323, + "learning_rate": 0.003, + "loss": 4.033, + "step": 16506 + }, + { + "epoch": 0.16507, + "grad_norm": 0.6623278260231018, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 16507 + }, + { + "epoch": 0.16508, + "grad_norm": 0.6675810813903809, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 16508 + }, + { + "epoch": 0.16509, + "grad_norm": 0.6640301942825317, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 16509 + }, + { + "epoch": 0.1651, + "grad_norm": 0.7328985929489136, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 16510 + }, + { + "epoch": 0.16511, + "grad_norm": 0.861212968826294, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 16511 + }, + { + "epoch": 0.16512, + "grad_norm": 0.9499975442886353, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 16512 + }, + { + "epoch": 0.16513, + "grad_norm": 1.1305769681930542, + "learning_rate": 0.003, + "loss": 4.018, + "step": 16513 + }, + { + "epoch": 0.16514, + "grad_norm": 0.7863883376121521, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 16514 + }, + { + "epoch": 0.16515, + "grad_norm": 0.7641639113426208, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 16515 + }, + { + "epoch": 0.16516, + "grad_norm": 0.8611436486244202, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 16516 + }, + { + "epoch": 0.16517, + "grad_norm": 0.9765501022338867, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 16517 + }, + { + "epoch": 0.16518, + "grad_norm": 0.8455554246902466, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 16518 + }, + { + "epoch": 0.16519, + "grad_norm": 0.692939043045044, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 16519 + }, + { + "epoch": 0.1652, + "grad_norm": 0.7636265754699707, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 16520 + }, + { + "epoch": 0.16521, + "grad_norm": 0.7926337122917175, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 16521 + }, + { + "epoch": 0.16522, + "grad_norm": 0.7739710211753845, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 16522 + }, + { + "epoch": 0.16523, + "grad_norm": 0.8571917414665222, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 16523 + }, + { + "epoch": 0.16524, + "grad_norm": 0.9024523496627808, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 16524 + }, + { + "epoch": 0.16525, + "grad_norm": 0.8396733999252319, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 16525 + }, + { + "epoch": 0.16526, + "grad_norm": 0.9109805226325989, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 16526 + }, + { + "epoch": 0.16527, + "grad_norm": 0.8133218884468079, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 16527 + }, + { + "epoch": 0.16528, + "grad_norm": 0.7340258955955505, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 16528 + }, + { + "epoch": 0.16529, + "grad_norm": 0.6597772836685181, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 16529 + }, + { + "epoch": 0.1653, + "grad_norm": 0.7231074571609497, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 16530 + }, + { + "epoch": 0.16531, + "grad_norm": 0.704673707485199, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 16531 + }, + { + "epoch": 0.16532, + "grad_norm": 0.6683825850486755, + "learning_rate": 0.003, + "loss": 4.011, + "step": 16532 + }, + { + "epoch": 0.16533, + "grad_norm": 0.7330564260482788, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 16533 + }, + { + "epoch": 0.16534, + "grad_norm": 0.8109389543533325, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 16534 + }, + { + "epoch": 0.16535, + "grad_norm": 0.9890372157096863, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 16535 + }, + { + "epoch": 0.16536, + "grad_norm": 1.5045017004013062, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 16536 + }, + { + "epoch": 0.16537, + "grad_norm": 0.7912871241569519, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 16537 + }, + { + "epoch": 0.16538, + "grad_norm": 0.8432285189628601, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 16538 + }, + { + "epoch": 0.16539, + "grad_norm": 1.0448075532913208, + "learning_rate": 0.003, + "loss": 4.004, + "step": 16539 + }, + { + "epoch": 0.1654, + "grad_norm": 0.9978594779968262, + "learning_rate": 0.003, + "loss": 4.021, + "step": 16540 + }, + { + "epoch": 0.16541, + "grad_norm": 0.9974130988121033, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 16541 + }, + { + "epoch": 0.16542, + "grad_norm": 1.2907527685165405, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 16542 + }, + { + "epoch": 0.16543, + "grad_norm": 0.981609046459198, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 16543 + }, + { + "epoch": 0.16544, + "grad_norm": 1.1068936586380005, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 16544 + }, + { + "epoch": 0.16545, + "grad_norm": 1.0436385869979858, + "learning_rate": 0.003, + "loss": 4.036, + "step": 16545 + }, + { + "epoch": 0.16546, + "grad_norm": 0.9572142958641052, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 16546 + }, + { + "epoch": 0.16547, + "grad_norm": 1.1582400798797607, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 16547 + }, + { + "epoch": 0.16548, + "grad_norm": 1.0990976095199585, + "learning_rate": 0.003, + "loss": 4.004, + "step": 16548 + }, + { + "epoch": 0.16549, + "grad_norm": 0.7857990264892578, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 16549 + }, + { + "epoch": 0.1655, + "grad_norm": 0.8027048110961914, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 16550 + }, + { + "epoch": 0.16551, + "grad_norm": 0.73116534948349, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 16551 + }, + { + "epoch": 0.16552, + "grad_norm": 0.6601966619491577, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 16552 + }, + { + "epoch": 0.16553, + "grad_norm": 0.6313268542289734, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 16553 + }, + { + "epoch": 0.16554, + "grad_norm": 0.82774418592453, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 16554 + }, + { + "epoch": 0.16555, + "grad_norm": 1.204330563545227, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 16555 + }, + { + "epoch": 0.16556, + "grad_norm": 0.6298305988311768, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 16556 + }, + { + "epoch": 0.16557, + "grad_norm": 0.6001787185668945, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 16557 + }, + { + "epoch": 0.16558, + "grad_norm": 0.665256142616272, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 16558 + }, + { + "epoch": 0.16559, + "grad_norm": 0.7350388765335083, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16559 + }, + { + "epoch": 0.1656, + "grad_norm": 0.9374911785125732, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 16560 + }, + { + "epoch": 0.16561, + "grad_norm": 0.9782078266143799, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 16561 + }, + { + "epoch": 0.16562, + "grad_norm": 0.7695868015289307, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 16562 + }, + { + "epoch": 0.16563, + "grad_norm": 0.7391797304153442, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 16563 + }, + { + "epoch": 0.16564, + "grad_norm": 0.7675747275352478, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 16564 + }, + { + "epoch": 0.16565, + "grad_norm": 0.7832668423652649, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 16565 + }, + { + "epoch": 0.16566, + "grad_norm": 0.6910805702209473, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 16566 + }, + { + "epoch": 0.16567, + "grad_norm": 0.7149823307991028, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 16567 + }, + { + "epoch": 0.16568, + "grad_norm": 0.6877647638320923, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 16568 + }, + { + "epoch": 0.16569, + "grad_norm": 0.7742563486099243, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 16569 + }, + { + "epoch": 0.1657, + "grad_norm": 0.8138877153396606, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 16570 + }, + { + "epoch": 0.16571, + "grad_norm": 0.7903016805648804, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 16571 + }, + { + "epoch": 0.16572, + "grad_norm": 0.861447811126709, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 16572 + }, + { + "epoch": 0.16573, + "grad_norm": 0.8637712001800537, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 16573 + }, + { + "epoch": 0.16574, + "grad_norm": 0.8199882507324219, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 16574 + }, + { + "epoch": 0.16575, + "grad_norm": 0.7627727389335632, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 16575 + }, + { + "epoch": 0.16576, + "grad_norm": 0.9434555768966675, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 16576 + }, + { + "epoch": 0.16577, + "grad_norm": 1.17056143283844, + "learning_rate": 0.003, + "loss": 4.019, + "step": 16577 + }, + { + "epoch": 0.16578, + "grad_norm": 0.8861210346221924, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 16578 + }, + { + "epoch": 0.16579, + "grad_norm": 0.977642834186554, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 16579 + }, + { + "epoch": 0.1658, + "grad_norm": 1.097947120666504, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 16580 + }, + { + "epoch": 0.16581, + "grad_norm": 0.8989746570587158, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 16581 + }, + { + "epoch": 0.16582, + "grad_norm": 0.8438913822174072, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 16582 + }, + { + "epoch": 0.16583, + "grad_norm": 1.0367250442504883, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 16583 + }, + { + "epoch": 0.16584, + "grad_norm": 0.9683894515037537, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 16584 + }, + { + "epoch": 0.16585, + "grad_norm": 0.8536483645439148, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 16585 + }, + { + "epoch": 0.16586, + "grad_norm": 0.8029671907424927, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 16586 + }, + { + "epoch": 0.16587, + "grad_norm": 0.7941018342971802, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 16587 + }, + { + "epoch": 0.16588, + "grad_norm": 0.7110337615013123, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 16588 + }, + { + "epoch": 0.16589, + "grad_norm": 0.6337109804153442, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 16589 + }, + { + "epoch": 0.1659, + "grad_norm": 0.5970135927200317, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 16590 + }, + { + "epoch": 0.16591, + "grad_norm": 0.7086123824119568, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 16591 + }, + { + "epoch": 0.16592, + "grad_norm": 0.7931826114654541, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 16592 + }, + { + "epoch": 0.16593, + "grad_norm": 0.9041086435317993, + "learning_rate": 0.003, + "loss": 4.005, + "step": 16593 + }, + { + "epoch": 0.16594, + "grad_norm": 0.8759629726409912, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 16594 + }, + { + "epoch": 0.16595, + "grad_norm": 0.7272791862487793, + "learning_rate": 0.003, + "loss": 4.025, + "step": 16595 + }, + { + "epoch": 0.16596, + "grad_norm": 0.6055447459220886, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 16596 + }, + { + "epoch": 0.16597, + "grad_norm": 0.65510493516922, + "learning_rate": 0.003, + "loss": 4.001, + "step": 16597 + }, + { + "epoch": 0.16598, + "grad_norm": 0.7917929887771606, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 16598 + }, + { + "epoch": 0.16599, + "grad_norm": 1.0322705507278442, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 16599 + }, + { + "epoch": 0.166, + "grad_norm": 1.1653863191604614, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 16600 + }, + { + "epoch": 0.16601, + "grad_norm": 0.7524556517601013, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 16601 + }, + { + "epoch": 0.16602, + "grad_norm": 0.7310243248939514, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 16602 + }, + { + "epoch": 0.16603, + "grad_norm": 0.7169476747512817, + "learning_rate": 0.003, + "loss": 4.033, + "step": 16603 + }, + { + "epoch": 0.16604, + "grad_norm": 0.7632569670677185, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 16604 + }, + { + "epoch": 0.16605, + "grad_norm": 0.7261447906494141, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 16605 + }, + { + "epoch": 0.16606, + "grad_norm": 0.7755059599876404, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 16606 + }, + { + "epoch": 0.16607, + "grad_norm": 0.7412046790122986, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 16607 + }, + { + "epoch": 0.16608, + "grad_norm": 0.9550292491912842, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 16608 + }, + { + "epoch": 0.16609, + "grad_norm": 1.1417499780654907, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 16609 + }, + { + "epoch": 0.1661, + "grad_norm": 0.9677286744117737, + "learning_rate": 0.003, + "loss": 3.979, + "step": 16610 + }, + { + "epoch": 0.16611, + "grad_norm": 0.9584037065505981, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 16611 + }, + { + "epoch": 0.16612, + "grad_norm": 0.834425687789917, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 16612 + }, + { + "epoch": 0.16613, + "grad_norm": 0.8633244037628174, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 16613 + }, + { + "epoch": 0.16614, + "grad_norm": 0.9459644556045532, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 16614 + }, + { + "epoch": 0.16615, + "grad_norm": 1.3256431818008423, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 16615 + }, + { + "epoch": 0.16616, + "grad_norm": 0.7476480603218079, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 16616 + }, + { + "epoch": 0.16617, + "grad_norm": 0.7218948006629944, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 16617 + }, + { + "epoch": 0.16618, + "grad_norm": 0.9353631734848022, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 16618 + }, + { + "epoch": 0.16619, + "grad_norm": 0.949628472328186, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 16619 + }, + { + "epoch": 0.1662, + "grad_norm": 1.0367956161499023, + "learning_rate": 0.003, + "loss": 4.019, + "step": 16620 + }, + { + "epoch": 0.16621, + "grad_norm": 1.2297966480255127, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 16621 + }, + { + "epoch": 0.16622, + "grad_norm": 0.6927123069763184, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 16622 + }, + { + "epoch": 0.16623, + "grad_norm": 0.6782436966896057, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 16623 + }, + { + "epoch": 0.16624, + "grad_norm": 0.645195722579956, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16624 + }, + { + "epoch": 0.16625, + "grad_norm": 0.6785486340522766, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 16625 + }, + { + "epoch": 0.16626, + "grad_norm": 0.8116571307182312, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 16626 + }, + { + "epoch": 0.16627, + "grad_norm": 0.9605870842933655, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 16627 + }, + { + "epoch": 0.16628, + "grad_norm": 1.026713490486145, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 16628 + }, + { + "epoch": 0.16629, + "grad_norm": 0.8040949106216431, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 16629 + }, + { + "epoch": 0.1663, + "grad_norm": 0.6801572442054749, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 16630 + }, + { + "epoch": 0.16631, + "grad_norm": 0.7101470232009888, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 16631 + }, + { + "epoch": 0.16632, + "grad_norm": 0.78169846534729, + "learning_rate": 0.003, + "loss": 4.003, + "step": 16632 + }, + { + "epoch": 0.16633, + "grad_norm": 0.7497729659080505, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 16633 + }, + { + "epoch": 0.16634, + "grad_norm": 0.7496538758277893, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 16634 + }, + { + "epoch": 0.16635, + "grad_norm": 0.7790054082870483, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 16635 + }, + { + "epoch": 0.16636, + "grad_norm": 0.8766295313835144, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 16636 + }, + { + "epoch": 0.16637, + "grad_norm": 0.895172119140625, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 16637 + }, + { + "epoch": 0.16638, + "grad_norm": 0.9995051026344299, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 16638 + }, + { + "epoch": 0.16639, + "grad_norm": 0.9869672656059265, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 16639 + }, + { + "epoch": 0.1664, + "grad_norm": 0.9185425043106079, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 16640 + }, + { + "epoch": 0.16641, + "grad_norm": 0.9075988531112671, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 16641 + }, + { + "epoch": 0.16642, + "grad_norm": 0.8588910698890686, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 16642 + }, + { + "epoch": 0.16643, + "grad_norm": 0.8271640539169312, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 16643 + }, + { + "epoch": 0.16644, + "grad_norm": 0.7942617535591125, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 16644 + }, + { + "epoch": 0.16645, + "grad_norm": 0.8908455967903137, + "learning_rate": 0.003, + "loss": 3.998, + "step": 16645 + }, + { + "epoch": 0.16646, + "grad_norm": 0.8527781963348389, + "learning_rate": 0.003, + "loss": 4.005, + "step": 16646 + }, + { + "epoch": 0.16647, + "grad_norm": 0.9275540113449097, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 16647 + }, + { + "epoch": 0.16648, + "grad_norm": 0.8698004484176636, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 16648 + }, + { + "epoch": 0.16649, + "grad_norm": 0.7929522395133972, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 16649 + }, + { + "epoch": 0.1665, + "grad_norm": 0.7279219031333923, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 16650 + }, + { + "epoch": 0.16651, + "grad_norm": 0.6361696720123291, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 16651 + }, + { + "epoch": 0.16652, + "grad_norm": 0.7483872175216675, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 16652 + }, + { + "epoch": 0.16653, + "grad_norm": 0.8242842555046082, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 16653 + }, + { + "epoch": 0.16654, + "grad_norm": 0.7325317859649658, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 16654 + }, + { + "epoch": 0.16655, + "grad_norm": 0.6815661787986755, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 16655 + }, + { + "epoch": 0.16656, + "grad_norm": 0.8499429821968079, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 16656 + }, + { + "epoch": 0.16657, + "grad_norm": 1.0942624807357788, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 16657 + }, + { + "epoch": 0.16658, + "grad_norm": 1.0901638269424438, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 16658 + }, + { + "epoch": 0.16659, + "grad_norm": 0.8472113013267517, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 16659 + }, + { + "epoch": 0.1666, + "grad_norm": 0.6041252613067627, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 16660 + }, + { + "epoch": 0.16661, + "grad_norm": 0.8856112957000732, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 16661 + }, + { + "epoch": 0.16662, + "grad_norm": 1.2002215385437012, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 16662 + }, + { + "epoch": 0.16663, + "grad_norm": 0.8653046488761902, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 16663 + }, + { + "epoch": 0.16664, + "grad_norm": 0.8390600085258484, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 16664 + }, + { + "epoch": 0.16665, + "grad_norm": 0.8085476160049438, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 16665 + }, + { + "epoch": 0.16666, + "grad_norm": 0.8118124604225159, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 16666 + }, + { + "epoch": 0.16667, + "grad_norm": 0.8476701378822327, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 16667 + }, + { + "epoch": 0.16668, + "grad_norm": 0.6915823817253113, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 16668 + }, + { + "epoch": 0.16669, + "grad_norm": 0.6230525374412537, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 16669 + }, + { + "epoch": 0.1667, + "grad_norm": 0.6813180446624756, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 16670 + }, + { + "epoch": 0.16671, + "grad_norm": 0.7478421926498413, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 16671 + }, + { + "epoch": 0.16672, + "grad_norm": 0.7379879951477051, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 16672 + }, + { + "epoch": 0.16673, + "grad_norm": 0.9299353361129761, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 16673 + }, + { + "epoch": 0.16674, + "grad_norm": 1.1519417762756348, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 16674 + }, + { + "epoch": 0.16675, + "grad_norm": 0.934518575668335, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 16675 + }, + { + "epoch": 0.16676, + "grad_norm": 0.9274440407752991, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 16676 + }, + { + "epoch": 0.16677, + "grad_norm": 0.9019926190376282, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 16677 + }, + { + "epoch": 0.16678, + "grad_norm": 0.9301829934120178, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 16678 + }, + { + "epoch": 0.16679, + "grad_norm": 1.0301803350448608, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 16679 + }, + { + "epoch": 0.1668, + "grad_norm": 0.9349686503410339, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 16680 + }, + { + "epoch": 0.16681, + "grad_norm": 1.0394617319107056, + "learning_rate": 0.003, + "loss": 4.007, + "step": 16681 + }, + { + "epoch": 0.16682, + "grad_norm": 1.0207794904708862, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 16682 + }, + { + "epoch": 0.16683, + "grad_norm": 0.8925315737724304, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 16683 + }, + { + "epoch": 0.16684, + "grad_norm": 0.7841940522193909, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 16684 + }, + { + "epoch": 0.16685, + "grad_norm": 0.7427353262901306, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 16685 + }, + { + "epoch": 0.16686, + "grad_norm": 0.6968230605125427, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 16686 + }, + { + "epoch": 0.16687, + "grad_norm": 0.5778324007987976, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 16687 + }, + { + "epoch": 0.16688, + "grad_norm": 0.6392512321472168, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 16688 + }, + { + "epoch": 0.16689, + "grad_norm": 0.7645626664161682, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 16689 + }, + { + "epoch": 0.1669, + "grad_norm": 0.8958572745323181, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 16690 + }, + { + "epoch": 0.16691, + "grad_norm": 0.8632878065109253, + "learning_rate": 0.003, + "loss": 4.02, + "step": 16691 + }, + { + "epoch": 0.16692, + "grad_norm": 0.75998854637146, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 16692 + }, + { + "epoch": 0.16693, + "grad_norm": 0.7607892155647278, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 16693 + }, + { + "epoch": 0.16694, + "grad_norm": 0.8183885216712952, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 16694 + }, + { + "epoch": 0.16695, + "grad_norm": 0.8750672936439514, + "learning_rate": 0.003, + "loss": 4.024, + "step": 16695 + }, + { + "epoch": 0.16696, + "grad_norm": 0.9015554189682007, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 16696 + }, + { + "epoch": 0.16697, + "grad_norm": 1.065500259399414, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 16697 + }, + { + "epoch": 0.16698, + "grad_norm": 1.045125961303711, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 16698 + }, + { + "epoch": 0.16699, + "grad_norm": 0.9421795606613159, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 16699 + }, + { + "epoch": 0.167, + "grad_norm": 0.8270700573921204, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 16700 + }, + { + "epoch": 0.16701, + "grad_norm": 0.771172821521759, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 16701 + }, + { + "epoch": 0.16702, + "grad_norm": 0.7797114849090576, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 16702 + }, + { + "epoch": 0.16703, + "grad_norm": 0.7795373797416687, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 16703 + }, + { + "epoch": 0.16704, + "grad_norm": 0.8451706171035767, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 16704 + }, + { + "epoch": 0.16705, + "grad_norm": 0.8600301146507263, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 16705 + }, + { + "epoch": 0.16706, + "grad_norm": 0.8787464499473572, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 16706 + }, + { + "epoch": 0.16707, + "grad_norm": 0.8668013215065002, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 16707 + }, + { + "epoch": 0.16708, + "grad_norm": 0.8679298758506775, + "learning_rate": 0.003, + "loss": 3.994, + "step": 16708 + }, + { + "epoch": 0.16709, + "grad_norm": 0.8530462980270386, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 16709 + }, + { + "epoch": 0.1671, + "grad_norm": 0.8643242716789246, + "learning_rate": 0.003, + "loss": 3.991, + "step": 16710 + }, + { + "epoch": 0.16711, + "grad_norm": 0.8478008508682251, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 16711 + }, + { + "epoch": 0.16712, + "grad_norm": 0.7355970144271851, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 16712 + }, + { + "epoch": 0.16713, + "grad_norm": 0.6986602544784546, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 16713 + }, + { + "epoch": 0.16714, + "grad_norm": 0.7060433030128479, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 16714 + }, + { + "epoch": 0.16715, + "grad_norm": 0.7944496273994446, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 16715 + }, + { + "epoch": 0.16716, + "grad_norm": 0.8400862812995911, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 16716 + }, + { + "epoch": 0.16717, + "grad_norm": 0.7830513715744019, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 16717 + }, + { + "epoch": 0.16718, + "grad_norm": 0.8651089072227478, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 16718 + }, + { + "epoch": 0.16719, + "grad_norm": 0.9211392402648926, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 16719 + }, + { + "epoch": 0.1672, + "grad_norm": 0.8560593724250793, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 16720 + }, + { + "epoch": 0.16721, + "grad_norm": 0.7442037463188171, + "learning_rate": 0.003, + "loss": 4.002, + "step": 16721 + }, + { + "epoch": 0.16722, + "grad_norm": 0.783574104309082, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 16722 + }, + { + "epoch": 0.16723, + "grad_norm": 0.8946051001548767, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 16723 + }, + { + "epoch": 0.16724, + "grad_norm": 0.8144287467002869, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 16724 + }, + { + "epoch": 0.16725, + "grad_norm": 0.6272501349449158, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 16725 + }, + { + "epoch": 0.16726, + "grad_norm": 0.6548783183097839, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 16726 + }, + { + "epoch": 0.16727, + "grad_norm": 0.7995438575744629, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 16727 + }, + { + "epoch": 0.16728, + "grad_norm": 0.9519756436347961, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 16728 + }, + { + "epoch": 0.16729, + "grad_norm": 1.064491868019104, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 16729 + }, + { + "epoch": 0.1673, + "grad_norm": 0.819176197052002, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 16730 + }, + { + "epoch": 0.16731, + "grad_norm": 0.6019748449325562, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 16731 + }, + { + "epoch": 0.16732, + "grad_norm": 0.7550683617591858, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 16732 + }, + { + "epoch": 0.16733, + "grad_norm": 0.9951023459434509, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 16733 + }, + { + "epoch": 0.16734, + "grad_norm": 1.3040541410446167, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 16734 + }, + { + "epoch": 0.16735, + "grad_norm": 0.8052306175231934, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 16735 + }, + { + "epoch": 0.16736, + "grad_norm": 0.7533053159713745, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 16736 + }, + { + "epoch": 0.16737, + "grad_norm": 0.7331545948982239, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 16737 + }, + { + "epoch": 0.16738, + "grad_norm": 0.7126925587654114, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 16738 + }, + { + "epoch": 0.16739, + "grad_norm": 0.707188606262207, + "learning_rate": 0.003, + "loss": 4.004, + "step": 16739 + }, + { + "epoch": 0.1674, + "grad_norm": 0.6312726736068726, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 16740 + }, + { + "epoch": 0.16741, + "grad_norm": 0.631080150604248, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 16741 + }, + { + "epoch": 0.16742, + "grad_norm": 0.6698132157325745, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 16742 + }, + { + "epoch": 0.16743, + "grad_norm": 0.8329685926437378, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 16743 + }, + { + "epoch": 0.16744, + "grad_norm": 1.090051293373108, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 16744 + }, + { + "epoch": 0.16745, + "grad_norm": 1.2471741437911987, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 16745 + }, + { + "epoch": 0.16746, + "grad_norm": 0.7032690048217773, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 16746 + }, + { + "epoch": 0.16747, + "grad_norm": 0.5997713804244995, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 16747 + }, + { + "epoch": 0.16748, + "grad_norm": 0.7260429263114929, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 16748 + }, + { + "epoch": 0.16749, + "grad_norm": 0.8784670829772949, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 16749 + }, + { + "epoch": 0.1675, + "grad_norm": 0.969279944896698, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 16750 + }, + { + "epoch": 0.16751, + "grad_norm": 0.830579936504364, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 16751 + }, + { + "epoch": 0.16752, + "grad_norm": 0.7464983463287354, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 16752 + }, + { + "epoch": 0.16753, + "grad_norm": 0.7157581448554993, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 16753 + }, + { + "epoch": 0.16754, + "grad_norm": 0.8800062537193298, + "learning_rate": 0.003, + "loss": 4.024, + "step": 16754 + }, + { + "epoch": 0.16755, + "grad_norm": 0.9574975371360779, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 16755 + }, + { + "epoch": 0.16756, + "grad_norm": 0.9339281916618347, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 16756 + }, + { + "epoch": 0.16757, + "grad_norm": 1.0285277366638184, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 16757 + }, + { + "epoch": 0.16758, + "grad_norm": 1.0120329856872559, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 16758 + }, + { + "epoch": 0.16759, + "grad_norm": 0.9101712107658386, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 16759 + }, + { + "epoch": 0.1676, + "grad_norm": 1.0395931005477905, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 16760 + }, + { + "epoch": 0.16761, + "grad_norm": 1.0496395826339722, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 16761 + }, + { + "epoch": 0.16762, + "grad_norm": 1.0535210371017456, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 16762 + }, + { + "epoch": 0.16763, + "grad_norm": 1.2106959819793701, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 16763 + }, + { + "epoch": 0.16764, + "grad_norm": 0.9122406840324402, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 16764 + }, + { + "epoch": 0.16765, + "grad_norm": 0.9126542210578918, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 16765 + }, + { + "epoch": 0.16766, + "grad_norm": 0.9665969014167786, + "learning_rate": 0.003, + "loss": 4.062, + "step": 16766 + }, + { + "epoch": 0.16767, + "grad_norm": 0.9059301614761353, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 16767 + }, + { + "epoch": 0.16768, + "grad_norm": 0.8744325637817383, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 16768 + }, + { + "epoch": 0.16769, + "grad_norm": 0.9320008754730225, + "learning_rate": 0.003, + "loss": 4.02, + "step": 16769 + }, + { + "epoch": 0.1677, + "grad_norm": 0.8519328236579895, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 16770 + }, + { + "epoch": 0.16771, + "grad_norm": 0.6243749856948853, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 16771 + }, + { + "epoch": 0.16772, + "grad_norm": 0.702466607093811, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 16772 + }, + { + "epoch": 0.16773, + "grad_norm": 0.7545448541641235, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 16773 + }, + { + "epoch": 0.16774, + "grad_norm": 0.8205435872077942, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 16774 + }, + { + "epoch": 0.16775, + "grad_norm": 1.0039379596710205, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 16775 + }, + { + "epoch": 0.16776, + "grad_norm": 1.189281702041626, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 16776 + }, + { + "epoch": 0.16777, + "grad_norm": 0.6821681261062622, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 16777 + }, + { + "epoch": 0.16778, + "grad_norm": 0.6866998076438904, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 16778 + }, + { + "epoch": 0.16779, + "grad_norm": 0.813926637172699, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 16779 + }, + { + "epoch": 0.1678, + "grad_norm": 0.9408845901489258, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 16780 + }, + { + "epoch": 0.16781, + "grad_norm": 0.8010312914848328, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 16781 + }, + { + "epoch": 0.16782, + "grad_norm": 0.6520599126815796, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 16782 + }, + { + "epoch": 0.16783, + "grad_norm": 0.5568283200263977, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 16783 + }, + { + "epoch": 0.16784, + "grad_norm": 0.5546129941940308, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 16784 + }, + { + "epoch": 0.16785, + "grad_norm": 0.5724064707756042, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 16785 + }, + { + "epoch": 0.16786, + "grad_norm": 0.5485352873802185, + "learning_rate": 0.003, + "loss": 4.028, + "step": 16786 + }, + { + "epoch": 0.16787, + "grad_norm": 0.6238598227500916, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 16787 + }, + { + "epoch": 0.16788, + "grad_norm": 0.8130132555961609, + "learning_rate": 0.003, + "loss": 4.033, + "step": 16788 + }, + { + "epoch": 0.16789, + "grad_norm": 0.9928077459335327, + "learning_rate": 0.003, + "loss": 4.021, + "step": 16789 + }, + { + "epoch": 0.1679, + "grad_norm": 1.0398330688476562, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 16790 + }, + { + "epoch": 0.16791, + "grad_norm": 0.8814020752906799, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 16791 + }, + { + "epoch": 0.16792, + "grad_norm": 0.8732738494873047, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 16792 + }, + { + "epoch": 0.16793, + "grad_norm": 0.848568856716156, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 16793 + }, + { + "epoch": 0.16794, + "grad_norm": 0.9220749735832214, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 16794 + }, + { + "epoch": 0.16795, + "grad_norm": 1.1375749111175537, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 16795 + }, + { + "epoch": 0.16796, + "grad_norm": 0.8748759031295776, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 16796 + }, + { + "epoch": 0.16797, + "grad_norm": 0.7760801315307617, + "learning_rate": 0.003, + "loss": 4.018, + "step": 16797 + }, + { + "epoch": 0.16798, + "grad_norm": 0.7592577338218689, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 16798 + }, + { + "epoch": 0.16799, + "grad_norm": 0.634211003780365, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 16799 + }, + { + "epoch": 0.168, + "grad_norm": 0.5644124150276184, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 16800 + }, + { + "epoch": 0.16801, + "grad_norm": 0.5643258690834045, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 16801 + }, + { + "epoch": 0.16802, + "grad_norm": 0.6898741126060486, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 16802 + }, + { + "epoch": 0.16803, + "grad_norm": 0.7783620357513428, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 16803 + }, + { + "epoch": 0.16804, + "grad_norm": 0.9071454405784607, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 16804 + }, + { + "epoch": 0.16805, + "grad_norm": 1.0606682300567627, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 16805 + }, + { + "epoch": 0.16806, + "grad_norm": 0.7918750643730164, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 16806 + }, + { + "epoch": 0.16807, + "grad_norm": 0.5339258909225464, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 16807 + }, + { + "epoch": 0.16808, + "grad_norm": 0.5802764296531677, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 16808 + }, + { + "epoch": 0.16809, + "grad_norm": 0.613307774066925, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 16809 + }, + { + "epoch": 0.1681, + "grad_norm": 0.6131017804145813, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 16810 + }, + { + "epoch": 0.16811, + "grad_norm": 0.6089868545532227, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 16811 + }, + { + "epoch": 0.16812, + "grad_norm": 0.6131261587142944, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 16812 + }, + { + "epoch": 0.16813, + "grad_norm": 0.6756606698036194, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 16813 + }, + { + "epoch": 0.16814, + "grad_norm": 0.8853565454483032, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 16814 + }, + { + "epoch": 0.16815, + "grad_norm": 1.149666428565979, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 16815 + }, + { + "epoch": 0.16816, + "grad_norm": 1.1497527360916138, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 16816 + }, + { + "epoch": 0.16817, + "grad_norm": 0.889690637588501, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 16817 + }, + { + "epoch": 0.16818, + "grad_norm": 0.818009078502655, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 16818 + }, + { + "epoch": 0.16819, + "grad_norm": 0.9402246475219727, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 16819 + }, + { + "epoch": 0.1682, + "grad_norm": 1.109734058380127, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 16820 + }, + { + "epoch": 0.16821, + "grad_norm": 1.1249537467956543, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 16821 + }, + { + "epoch": 0.16822, + "grad_norm": 0.8454878926277161, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 16822 + }, + { + "epoch": 0.16823, + "grad_norm": 0.7891453504562378, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 16823 + }, + { + "epoch": 0.16824, + "grad_norm": 0.9035543203353882, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 16824 + }, + { + "epoch": 0.16825, + "grad_norm": 1.0534626245498657, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 16825 + }, + { + "epoch": 0.16826, + "grad_norm": 0.8612683415412903, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16826 + }, + { + "epoch": 0.16827, + "grad_norm": 0.8174473643302917, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 16827 + }, + { + "epoch": 0.16828, + "grad_norm": 0.9313732385635376, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 16828 + }, + { + "epoch": 0.16829, + "grad_norm": 0.9340553283691406, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 16829 + }, + { + "epoch": 0.1683, + "grad_norm": 0.9243738651275635, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 16830 + }, + { + "epoch": 0.16831, + "grad_norm": 0.9216597080230713, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 16831 + }, + { + "epoch": 0.16832, + "grad_norm": 0.8400397300720215, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 16832 + }, + { + "epoch": 0.16833, + "grad_norm": 0.7118893265724182, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 16833 + }, + { + "epoch": 0.16834, + "grad_norm": 0.7463409900665283, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 16834 + }, + { + "epoch": 0.16835, + "grad_norm": 0.798879086971283, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 16835 + }, + { + "epoch": 0.16836, + "grad_norm": 0.7490203976631165, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 16836 + }, + { + "epoch": 0.16837, + "grad_norm": 0.7311861515045166, + "learning_rate": 0.003, + "loss": 4.006, + "step": 16837 + }, + { + "epoch": 0.16838, + "grad_norm": 0.7826172113418579, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 16838 + }, + { + "epoch": 0.16839, + "grad_norm": 0.7967138886451721, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 16839 + }, + { + "epoch": 0.1684, + "grad_norm": 0.8535910248756409, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 16840 + }, + { + "epoch": 0.16841, + "grad_norm": 0.9957929253578186, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 16841 + }, + { + "epoch": 0.16842, + "grad_norm": 1.0661824941635132, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 16842 + }, + { + "epoch": 0.16843, + "grad_norm": 0.8872292637825012, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 16843 + }, + { + "epoch": 0.16844, + "grad_norm": 0.641571581363678, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 16844 + }, + { + "epoch": 0.16845, + "grad_norm": 0.6273783445358276, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 16845 + }, + { + "epoch": 0.16846, + "grad_norm": 0.738833487033844, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 16846 + }, + { + "epoch": 0.16847, + "grad_norm": 0.9297849535942078, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 16847 + }, + { + "epoch": 0.16848, + "grad_norm": 1.1222742795944214, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 16848 + }, + { + "epoch": 0.16849, + "grad_norm": 0.9273313283920288, + "learning_rate": 0.003, + "loss": 3.992, + "step": 16849 + }, + { + "epoch": 0.1685, + "grad_norm": 0.8675814867019653, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 16850 + }, + { + "epoch": 0.16851, + "grad_norm": 0.8588464856147766, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 16851 + }, + { + "epoch": 0.16852, + "grad_norm": 0.8039618730545044, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 16852 + }, + { + "epoch": 0.16853, + "grad_norm": 0.8666275143623352, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 16853 + }, + { + "epoch": 0.16854, + "grad_norm": 0.8161850571632385, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 16854 + }, + { + "epoch": 0.16855, + "grad_norm": 0.6379879713058472, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 16855 + }, + { + "epoch": 0.16856, + "grad_norm": 0.7508741021156311, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 16856 + }, + { + "epoch": 0.16857, + "grad_norm": 0.7264812588691711, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 16857 + }, + { + "epoch": 0.16858, + "grad_norm": 0.6979163885116577, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 16858 + }, + { + "epoch": 0.16859, + "grad_norm": 0.7951305508613586, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 16859 + }, + { + "epoch": 0.1686, + "grad_norm": 0.9455273151397705, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 16860 + }, + { + "epoch": 0.16861, + "grad_norm": 0.9702208042144775, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16861 + }, + { + "epoch": 0.16862, + "grad_norm": 0.8496451377868652, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 16862 + }, + { + "epoch": 0.16863, + "grad_norm": 0.9306349754333496, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 16863 + }, + { + "epoch": 0.16864, + "grad_norm": 0.916183590888977, + "learning_rate": 0.003, + "loss": 4.007, + "step": 16864 + }, + { + "epoch": 0.16865, + "grad_norm": 0.8136289715766907, + "learning_rate": 0.003, + "loss": 3.98, + "step": 16865 + }, + { + "epoch": 0.16866, + "grad_norm": 0.7357578873634338, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 16866 + }, + { + "epoch": 0.16867, + "grad_norm": 0.6378408074378967, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 16867 + }, + { + "epoch": 0.16868, + "grad_norm": 0.614768385887146, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 16868 + }, + { + "epoch": 0.16869, + "grad_norm": 0.5955229997634888, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 16869 + }, + { + "epoch": 0.1687, + "grad_norm": 0.5942284464836121, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 16870 + }, + { + "epoch": 0.16871, + "grad_norm": 0.6663064956665039, + "learning_rate": 0.003, + "loss": 4.018, + "step": 16871 + }, + { + "epoch": 0.16872, + "grad_norm": 0.7257936000823975, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 16872 + }, + { + "epoch": 0.16873, + "grad_norm": 0.7455652952194214, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 16873 + }, + { + "epoch": 0.16874, + "grad_norm": 0.8603283762931824, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 16874 + }, + { + "epoch": 0.16875, + "grad_norm": 1.2169922590255737, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 16875 + }, + { + "epoch": 0.16876, + "grad_norm": 0.9302839040756226, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 16876 + }, + { + "epoch": 0.16877, + "grad_norm": 0.9746444821357727, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 16877 + }, + { + "epoch": 0.16878, + "grad_norm": 1.0206416845321655, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 16878 + }, + { + "epoch": 0.16879, + "grad_norm": 1.0273840427398682, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 16879 + }, + { + "epoch": 0.1688, + "grad_norm": 0.8442237973213196, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 16880 + }, + { + "epoch": 0.16881, + "grad_norm": 0.6956433057785034, + "learning_rate": 0.003, + "loss": 3.993, + "step": 16881 + }, + { + "epoch": 0.16882, + "grad_norm": 0.7793794870376587, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 16882 + }, + { + "epoch": 0.16883, + "grad_norm": 0.8479170799255371, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 16883 + }, + { + "epoch": 0.16884, + "grad_norm": 0.8433241248130798, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 16884 + }, + { + "epoch": 0.16885, + "grad_norm": 0.7989548444747925, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 16885 + }, + { + "epoch": 0.16886, + "grad_norm": 0.8560846447944641, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 16886 + }, + { + "epoch": 0.16887, + "grad_norm": 0.8832461833953857, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 16887 + }, + { + "epoch": 0.16888, + "grad_norm": 0.9571264982223511, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 16888 + }, + { + "epoch": 0.16889, + "grad_norm": 0.9143347144126892, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 16889 + }, + { + "epoch": 0.1689, + "grad_norm": 0.8070544600486755, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 16890 + }, + { + "epoch": 0.16891, + "grad_norm": 0.7932570576667786, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 16891 + }, + { + "epoch": 0.16892, + "grad_norm": 0.7867026925086975, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 16892 + }, + { + "epoch": 0.16893, + "grad_norm": 0.7658373117446899, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 16893 + }, + { + "epoch": 0.16894, + "grad_norm": 0.7852148413658142, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 16894 + }, + { + "epoch": 0.16895, + "grad_norm": 0.7895588278770447, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 16895 + }, + { + "epoch": 0.16896, + "grad_norm": 0.8559044599533081, + "learning_rate": 0.003, + "loss": 4.006, + "step": 16896 + }, + { + "epoch": 0.16897, + "grad_norm": 1.0634825229644775, + "learning_rate": 0.003, + "loss": 3.997, + "step": 16897 + }, + { + "epoch": 0.16898, + "grad_norm": 0.9409705996513367, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 16898 + }, + { + "epoch": 0.16899, + "grad_norm": 0.8026694655418396, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 16899 + }, + { + "epoch": 0.169, + "grad_norm": 0.7315845489501953, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 16900 + }, + { + "epoch": 0.16901, + "grad_norm": 0.7260497808456421, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 16901 + }, + { + "epoch": 0.16902, + "grad_norm": 0.6549625396728516, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 16902 + }, + { + "epoch": 0.16903, + "grad_norm": 0.607406735420227, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 16903 + }, + { + "epoch": 0.16904, + "grad_norm": 0.6448104381561279, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 16904 + }, + { + "epoch": 0.16905, + "grad_norm": 0.8571560978889465, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 16905 + }, + { + "epoch": 0.16906, + "grad_norm": 0.9251641035079956, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 16906 + }, + { + "epoch": 0.16907, + "grad_norm": 0.9457172751426697, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 16907 + }, + { + "epoch": 0.16908, + "grad_norm": 0.8236315250396729, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 16908 + }, + { + "epoch": 0.16909, + "grad_norm": 0.8186244368553162, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 16909 + }, + { + "epoch": 0.1691, + "grad_norm": 0.8274247646331787, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 16910 + }, + { + "epoch": 0.16911, + "grad_norm": 0.9622073173522949, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 16911 + }, + { + "epoch": 0.16912, + "grad_norm": 1.0782620906829834, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 16912 + }, + { + "epoch": 0.16913, + "grad_norm": 1.0125151872634888, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 16913 + }, + { + "epoch": 0.16914, + "grad_norm": 0.798346757888794, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 16914 + }, + { + "epoch": 0.16915, + "grad_norm": 0.672568142414093, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 16915 + }, + { + "epoch": 0.16916, + "grad_norm": 0.6295229196548462, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 16916 + }, + { + "epoch": 0.16917, + "grad_norm": 0.6785711050033569, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 16917 + }, + { + "epoch": 0.16918, + "grad_norm": 0.7268692851066589, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 16918 + }, + { + "epoch": 0.16919, + "grad_norm": 0.7384181022644043, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 16919 + }, + { + "epoch": 0.1692, + "grad_norm": 0.9648261070251465, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 16920 + }, + { + "epoch": 0.16921, + "grad_norm": 1.593048334121704, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 16921 + }, + { + "epoch": 0.16922, + "grad_norm": 0.5928442478179932, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 16922 + }, + { + "epoch": 0.16923, + "grad_norm": 0.7853564620018005, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 16923 + }, + { + "epoch": 0.16924, + "grad_norm": 1.0473350286483765, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 16924 + }, + { + "epoch": 0.16925, + "grad_norm": 1.0236759185791016, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 16925 + }, + { + "epoch": 0.16926, + "grad_norm": 0.8573715090751648, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 16926 + }, + { + "epoch": 0.16927, + "grad_norm": 0.7441908717155457, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 16927 + }, + { + "epoch": 0.16928, + "grad_norm": 0.8400618433952332, + "learning_rate": 0.003, + "loss": 4.018, + "step": 16928 + }, + { + "epoch": 0.16929, + "grad_norm": 1.0212396383285522, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 16929 + }, + { + "epoch": 0.1693, + "grad_norm": 1.0390324592590332, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 16930 + }, + { + "epoch": 0.16931, + "grad_norm": 0.9738713502883911, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 16931 + }, + { + "epoch": 0.16932, + "grad_norm": 1.1137603521347046, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 16932 + }, + { + "epoch": 0.16933, + "grad_norm": 1.0305066108703613, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 16933 + }, + { + "epoch": 0.16934, + "grad_norm": 0.9579649567604065, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 16934 + }, + { + "epoch": 0.16935, + "grad_norm": 0.7607799768447876, + "learning_rate": 0.003, + "loss": 4.029, + "step": 16935 + }, + { + "epoch": 0.16936, + "grad_norm": 0.7850809693336487, + "learning_rate": 0.003, + "loss": 3.99, + "step": 16936 + }, + { + "epoch": 0.16937, + "grad_norm": 0.8200135231018066, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 16937 + }, + { + "epoch": 0.16938, + "grad_norm": 0.7746044397354126, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 16938 + }, + { + "epoch": 0.16939, + "grad_norm": 0.8466827273368835, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 16939 + }, + { + "epoch": 0.1694, + "grad_norm": 0.7908685207366943, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 16940 + }, + { + "epoch": 0.16941, + "grad_norm": 0.6856575608253479, + "learning_rate": 0.003, + "loss": 4.022, + "step": 16941 + }, + { + "epoch": 0.16942, + "grad_norm": 0.6925505995750427, + "learning_rate": 0.003, + "loss": 4.008, + "step": 16942 + }, + { + "epoch": 0.16943, + "grad_norm": 0.8004992008209229, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 16943 + }, + { + "epoch": 0.16944, + "grad_norm": 0.822459876537323, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 16944 + }, + { + "epoch": 0.16945, + "grad_norm": 0.8093191981315613, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 16945 + }, + { + "epoch": 0.16946, + "grad_norm": 0.8074596524238586, + "learning_rate": 0.003, + "loss": 4.019, + "step": 16946 + }, + { + "epoch": 0.16947, + "grad_norm": 0.9427773356437683, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 16947 + }, + { + "epoch": 0.16948, + "grad_norm": 1.1139631271362305, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 16948 + }, + { + "epoch": 0.16949, + "grad_norm": 0.9527763724327087, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 16949 + }, + { + "epoch": 0.1695, + "grad_norm": 0.7660547494888306, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 16950 + }, + { + "epoch": 0.16951, + "grad_norm": 0.6911223530769348, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 16951 + }, + { + "epoch": 0.16952, + "grad_norm": 0.6668634414672852, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 16952 + }, + { + "epoch": 0.16953, + "grad_norm": 0.687681257724762, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 16953 + }, + { + "epoch": 0.16954, + "grad_norm": 0.6563870310783386, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 16954 + }, + { + "epoch": 0.16955, + "grad_norm": 0.6711004972457886, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 16955 + }, + { + "epoch": 0.16956, + "grad_norm": 0.7026411294937134, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 16956 + }, + { + "epoch": 0.16957, + "grad_norm": 0.704535186290741, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 16957 + }, + { + "epoch": 0.16958, + "grad_norm": 0.6693637371063232, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 16958 + }, + { + "epoch": 0.16959, + "grad_norm": 0.6984379887580872, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 16959 + }, + { + "epoch": 0.1696, + "grad_norm": 0.750443696975708, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 16960 + }, + { + "epoch": 0.16961, + "grad_norm": 0.8894288539886475, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 16961 + }, + { + "epoch": 0.16962, + "grad_norm": 1.128329873085022, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 16962 + }, + { + "epoch": 0.16963, + "grad_norm": 0.7519640922546387, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 16963 + }, + { + "epoch": 0.16964, + "grad_norm": 0.552846372127533, + "learning_rate": 0.003, + "loss": 4.016, + "step": 16964 + }, + { + "epoch": 0.16965, + "grad_norm": 0.5697802901268005, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 16965 + }, + { + "epoch": 0.16966, + "grad_norm": 0.6609813570976257, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 16966 + }, + { + "epoch": 0.16967, + "grad_norm": 0.7198853492736816, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 16967 + }, + { + "epoch": 0.16968, + "grad_norm": 0.7277693748474121, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 16968 + }, + { + "epoch": 0.16969, + "grad_norm": 0.7942861914634705, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 16969 + }, + { + "epoch": 0.1697, + "grad_norm": 0.7987948656082153, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 16970 + }, + { + "epoch": 0.16971, + "grad_norm": 0.8817940354347229, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 16971 + }, + { + "epoch": 0.16972, + "grad_norm": 1.0301092863082886, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 16972 + }, + { + "epoch": 0.16973, + "grad_norm": 0.9825281500816345, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 16973 + }, + { + "epoch": 0.16974, + "grad_norm": 0.9258411526679993, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 16974 + }, + { + "epoch": 0.16975, + "grad_norm": 1.0217163562774658, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 16975 + }, + { + "epoch": 0.16976, + "grad_norm": 1.1775392293930054, + "learning_rate": 0.003, + "loss": 3.977, + "step": 16976 + }, + { + "epoch": 0.16977, + "grad_norm": 0.9654213786125183, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 16977 + }, + { + "epoch": 0.16978, + "grad_norm": 1.1151385307312012, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 16978 + }, + { + "epoch": 0.16979, + "grad_norm": 1.012771487236023, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 16979 + }, + { + "epoch": 0.1698, + "grad_norm": 0.8385891914367676, + "learning_rate": 0.003, + "loss": 4.004, + "step": 16980 + }, + { + "epoch": 0.16981, + "grad_norm": 0.7704838514328003, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 16981 + }, + { + "epoch": 0.16982, + "grad_norm": 0.9659813046455383, + "learning_rate": 0.003, + "loss": 4.033, + "step": 16982 + }, + { + "epoch": 0.16983, + "grad_norm": 1.3794572353363037, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 16983 + }, + { + "epoch": 0.16984, + "grad_norm": 0.6425350904464722, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 16984 + }, + { + "epoch": 0.16985, + "grad_norm": 0.7135833501815796, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 16985 + }, + { + "epoch": 0.16986, + "grad_norm": 0.7565526962280273, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 16986 + }, + { + "epoch": 0.16987, + "grad_norm": 0.7284208536148071, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 16987 + }, + { + "epoch": 0.16988, + "grad_norm": 0.7384746074676514, + "learning_rate": 0.003, + "loss": 4.004, + "step": 16988 + }, + { + "epoch": 0.16989, + "grad_norm": 0.7987564206123352, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 16989 + }, + { + "epoch": 0.1699, + "grad_norm": 0.6784647107124329, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 16990 + }, + { + "epoch": 0.16991, + "grad_norm": 0.6576104760169983, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 16991 + }, + { + "epoch": 0.16992, + "grad_norm": 0.818426251411438, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 16992 + }, + { + "epoch": 0.16993, + "grad_norm": 1.1293405294418335, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 16993 + }, + { + "epoch": 0.16994, + "grad_norm": 1.1304106712341309, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 16994 + }, + { + "epoch": 0.16995, + "grad_norm": 0.7580990791320801, + "learning_rate": 0.003, + "loss": 3.973, + "step": 16995 + }, + { + "epoch": 0.16996, + "grad_norm": 0.6400741934776306, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 16996 + }, + { + "epoch": 0.16997, + "grad_norm": 0.6320247650146484, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 16997 + }, + { + "epoch": 0.16998, + "grad_norm": 0.6405655145645142, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 16998 + }, + { + "epoch": 0.16999, + "grad_norm": 0.7054705023765564, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 16999 + }, + { + "epoch": 0.17, + "grad_norm": 0.7184433937072754, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 17000 + }, + { + "epoch": 0.17001, + "grad_norm": 0.7353405952453613, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 17001 + }, + { + "epoch": 0.17002, + "grad_norm": 0.7217193841934204, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 17002 + }, + { + "epoch": 0.17003, + "grad_norm": 0.9274918437004089, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 17003 + }, + { + "epoch": 0.17004, + "grad_norm": 1.2057504653930664, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 17004 + }, + { + "epoch": 0.17005, + "grad_norm": 0.794136643409729, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 17005 + }, + { + "epoch": 0.17006, + "grad_norm": 0.6768636107444763, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 17006 + }, + { + "epoch": 0.17007, + "grad_norm": 0.725312352180481, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 17007 + }, + { + "epoch": 0.17008, + "grad_norm": 0.7155160903930664, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 17008 + }, + { + "epoch": 0.17009, + "grad_norm": 0.6952875852584839, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 17009 + }, + { + "epoch": 0.1701, + "grad_norm": 0.7479856014251709, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 17010 + }, + { + "epoch": 0.17011, + "grad_norm": 0.8017983436584473, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 17011 + }, + { + "epoch": 0.17012, + "grad_norm": 1.0040581226348877, + "learning_rate": 0.003, + "loss": 4.033, + "step": 17012 + }, + { + "epoch": 0.17013, + "grad_norm": 1.1289502382278442, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 17013 + }, + { + "epoch": 0.17014, + "grad_norm": 0.9964608550071716, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 17014 + }, + { + "epoch": 0.17015, + "grad_norm": 1.200086236000061, + "learning_rate": 0.003, + "loss": 4.015, + "step": 17015 + }, + { + "epoch": 0.17016, + "grad_norm": 0.8935628533363342, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 17016 + }, + { + "epoch": 0.17017, + "grad_norm": 0.835453450679779, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 17017 + }, + { + "epoch": 0.17018, + "grad_norm": 0.9155903458595276, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 17018 + }, + { + "epoch": 0.17019, + "grad_norm": 0.8293262720108032, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 17019 + }, + { + "epoch": 0.1702, + "grad_norm": 0.7839001417160034, + "learning_rate": 0.003, + "loss": 4.016, + "step": 17020 + }, + { + "epoch": 0.17021, + "grad_norm": 0.833850622177124, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 17021 + }, + { + "epoch": 0.17022, + "grad_norm": 0.7615661025047302, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 17022 + }, + { + "epoch": 0.17023, + "grad_norm": 0.7861036062240601, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 17023 + }, + { + "epoch": 0.17024, + "grad_norm": 0.9979937672615051, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 17024 + }, + { + "epoch": 0.17025, + "grad_norm": 1.267797589302063, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 17025 + }, + { + "epoch": 0.17026, + "grad_norm": 1.0182993412017822, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 17026 + }, + { + "epoch": 0.17027, + "grad_norm": 0.9555907845497131, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 17027 + }, + { + "epoch": 0.17028, + "grad_norm": 1.3939357995986938, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 17028 + }, + { + "epoch": 0.17029, + "grad_norm": 1.0221747159957886, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 17029 + }, + { + "epoch": 0.1703, + "grad_norm": 0.7883939146995544, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 17030 + }, + { + "epoch": 0.17031, + "grad_norm": 0.8329399228096008, + "learning_rate": 0.003, + "loss": 4.036, + "step": 17031 + }, + { + "epoch": 0.17032, + "grad_norm": 0.9606727957725525, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 17032 + }, + { + "epoch": 0.17033, + "grad_norm": 0.8926258683204651, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 17033 + }, + { + "epoch": 0.17034, + "grad_norm": 0.837275505065918, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 17034 + }, + { + "epoch": 0.17035, + "grad_norm": 0.8217374682426453, + "learning_rate": 0.003, + "loss": 3.988, + "step": 17035 + }, + { + "epoch": 0.17036, + "grad_norm": 0.7080724239349365, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 17036 + }, + { + "epoch": 0.17037, + "grad_norm": 0.759881317615509, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 17037 + }, + { + "epoch": 0.17038, + "grad_norm": 0.6957381963729858, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 17038 + }, + { + "epoch": 0.17039, + "grad_norm": 0.6571491956710815, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 17039 + }, + { + "epoch": 0.1704, + "grad_norm": 0.5505919456481934, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 17040 + }, + { + "epoch": 0.17041, + "grad_norm": 0.5832685232162476, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 17041 + }, + { + "epoch": 0.17042, + "grad_norm": 0.6317238807678223, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 17042 + }, + { + "epoch": 0.17043, + "grad_norm": 0.8054482936859131, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 17043 + }, + { + "epoch": 0.17044, + "grad_norm": 1.0918316841125488, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 17044 + }, + { + "epoch": 0.17045, + "grad_norm": 1.0211206674575806, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 17045 + }, + { + "epoch": 0.17046, + "grad_norm": 0.8288164734840393, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 17046 + }, + { + "epoch": 0.17047, + "grad_norm": 0.6238090991973877, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 17047 + }, + { + "epoch": 0.17048, + "grad_norm": 0.6770544052124023, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 17048 + }, + { + "epoch": 0.17049, + "grad_norm": 0.8049353957176208, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 17049 + }, + { + "epoch": 0.1705, + "grad_norm": 0.8502638339996338, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 17050 + }, + { + "epoch": 0.17051, + "grad_norm": 0.8071856498718262, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 17051 + }, + { + "epoch": 0.17052, + "grad_norm": 0.7418029308319092, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 17052 + }, + { + "epoch": 0.17053, + "grad_norm": 0.883597195148468, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 17053 + }, + { + "epoch": 0.17054, + "grad_norm": 1.0723153352737427, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 17054 + }, + { + "epoch": 0.17055, + "grad_norm": 1.1307390928268433, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 17055 + }, + { + "epoch": 0.17056, + "grad_norm": 0.7959531545639038, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 17056 + }, + { + "epoch": 0.17057, + "grad_norm": 0.7517292499542236, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 17057 + }, + { + "epoch": 0.17058, + "grad_norm": 0.8400816917419434, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 17058 + }, + { + "epoch": 0.17059, + "grad_norm": 0.8377096652984619, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 17059 + }, + { + "epoch": 0.1706, + "grad_norm": 0.8193461298942566, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 17060 + }, + { + "epoch": 0.17061, + "grad_norm": 0.845766007900238, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 17061 + }, + { + "epoch": 0.17062, + "grad_norm": 0.8220046162605286, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 17062 + }, + { + "epoch": 0.17063, + "grad_norm": 0.6880900859832764, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 17063 + }, + { + "epoch": 0.17064, + "grad_norm": 0.6444327235221863, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 17064 + }, + { + "epoch": 0.17065, + "grad_norm": 0.7091438174247742, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 17065 + }, + { + "epoch": 0.17066, + "grad_norm": 0.744178831577301, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 17066 + }, + { + "epoch": 0.17067, + "grad_norm": 0.7807066440582275, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 17067 + }, + { + "epoch": 0.17068, + "grad_norm": 0.7357064485549927, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 17068 + }, + { + "epoch": 0.17069, + "grad_norm": 0.8833433985710144, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 17069 + }, + { + "epoch": 0.1707, + "grad_norm": 1.200100064277649, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 17070 + }, + { + "epoch": 0.17071, + "grad_norm": 0.8609374165534973, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 17071 + }, + { + "epoch": 0.17072, + "grad_norm": 0.7197414636611938, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 17072 + }, + { + "epoch": 0.17073, + "grad_norm": 0.6767376065254211, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 17073 + }, + { + "epoch": 0.17074, + "grad_norm": 0.7874352335929871, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 17074 + }, + { + "epoch": 0.17075, + "grad_norm": 0.8052144646644592, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 17075 + }, + { + "epoch": 0.17076, + "grad_norm": 0.8144032955169678, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 17076 + }, + { + "epoch": 0.17077, + "grad_norm": 0.9464995861053467, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 17077 + }, + { + "epoch": 0.17078, + "grad_norm": 1.0766133069992065, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 17078 + }, + { + "epoch": 0.17079, + "grad_norm": 1.0682998895645142, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 17079 + }, + { + "epoch": 0.1708, + "grad_norm": 1.1437551975250244, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 17080 + }, + { + "epoch": 0.17081, + "grad_norm": 0.9291083812713623, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 17081 + }, + { + "epoch": 0.17082, + "grad_norm": 0.7274229526519775, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 17082 + }, + { + "epoch": 0.17083, + "grad_norm": 0.7047369480133057, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 17083 + }, + { + "epoch": 0.17084, + "grad_norm": 0.7894013524055481, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 17084 + }, + { + "epoch": 0.17085, + "grad_norm": 0.9227026700973511, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 17085 + }, + { + "epoch": 0.17086, + "grad_norm": 0.8737115263938904, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 17086 + }, + { + "epoch": 0.17087, + "grad_norm": 0.8501720428466797, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 17087 + }, + { + "epoch": 0.17088, + "grad_norm": 0.8258364796638489, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 17088 + }, + { + "epoch": 0.17089, + "grad_norm": 0.8701712489128113, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 17089 + }, + { + "epoch": 0.1709, + "grad_norm": 0.8529579639434814, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 17090 + }, + { + "epoch": 0.17091, + "grad_norm": 0.876214325428009, + "learning_rate": 0.003, + "loss": 4.045, + "step": 17091 + }, + { + "epoch": 0.17092, + "grad_norm": 0.9395101070404053, + "learning_rate": 0.003, + "loss": 3.987, + "step": 17092 + }, + { + "epoch": 0.17093, + "grad_norm": 1.09381103515625, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 17093 + }, + { + "epoch": 0.17094, + "grad_norm": 0.8975692391395569, + "learning_rate": 0.003, + "loss": 4.023, + "step": 17094 + }, + { + "epoch": 0.17095, + "grad_norm": 1.1432256698608398, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 17095 + }, + { + "epoch": 0.17096, + "grad_norm": 0.9737615585327148, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 17096 + }, + { + "epoch": 0.17097, + "grad_norm": 0.8718317747116089, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 17097 + }, + { + "epoch": 0.17098, + "grad_norm": 0.8652920126914978, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 17098 + }, + { + "epoch": 0.17099, + "grad_norm": 0.9636838436126709, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 17099 + }, + { + "epoch": 0.171, + "grad_norm": 1.0163567066192627, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 17100 + }, + { + "epoch": 0.17101, + "grad_norm": 0.9739014506340027, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 17101 + }, + { + "epoch": 0.17102, + "grad_norm": 0.8297356963157654, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 17102 + }, + { + "epoch": 0.17103, + "grad_norm": 0.7012573480606079, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 17103 + }, + { + "epoch": 0.17104, + "grad_norm": 0.6135175228118896, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 17104 + }, + { + "epoch": 0.17105, + "grad_norm": 0.7124812602996826, + "learning_rate": 0.003, + "loss": 3.988, + "step": 17105 + }, + { + "epoch": 0.17106, + "grad_norm": 0.8142275214195251, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 17106 + }, + { + "epoch": 0.17107, + "grad_norm": 0.9401968717575073, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 17107 + }, + { + "epoch": 0.17108, + "grad_norm": 0.9679330587387085, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 17108 + }, + { + "epoch": 0.17109, + "grad_norm": 0.8321598172187805, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 17109 + }, + { + "epoch": 0.1711, + "grad_norm": 0.675540566444397, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 17110 + }, + { + "epoch": 0.17111, + "grad_norm": 0.7044729590415955, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 17111 + }, + { + "epoch": 0.17112, + "grad_norm": 0.8573277592658997, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 17112 + }, + { + "epoch": 0.17113, + "grad_norm": 0.9914045333862305, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 17113 + }, + { + "epoch": 0.17114, + "grad_norm": 1.0339431762695312, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 17114 + }, + { + "epoch": 0.17115, + "grad_norm": 0.9043312668800354, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 17115 + }, + { + "epoch": 0.17116, + "grad_norm": 0.9394837021827698, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 17116 + }, + { + "epoch": 0.17117, + "grad_norm": 1.0119091272354126, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 17117 + }, + { + "epoch": 0.17118, + "grad_norm": 0.8556259274482727, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 17118 + }, + { + "epoch": 0.17119, + "grad_norm": 0.7036623358726501, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 17119 + }, + { + "epoch": 0.1712, + "grad_norm": 0.7242422699928284, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 17120 + }, + { + "epoch": 0.17121, + "grad_norm": 0.8174219131469727, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 17121 + }, + { + "epoch": 0.17122, + "grad_norm": 0.9132465124130249, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 17122 + }, + { + "epoch": 0.17123, + "grad_norm": 0.9040917158126831, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 17123 + }, + { + "epoch": 0.17124, + "grad_norm": 0.8150893449783325, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 17124 + }, + { + "epoch": 0.17125, + "grad_norm": 0.7183358073234558, + "learning_rate": 0.003, + "loss": 4.036, + "step": 17125 + }, + { + "epoch": 0.17126, + "grad_norm": 0.7179135680198669, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 17126 + }, + { + "epoch": 0.17127, + "grad_norm": 0.7805479168891907, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 17127 + }, + { + "epoch": 0.17128, + "grad_norm": 0.8239583969116211, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 17128 + }, + { + "epoch": 0.17129, + "grad_norm": 0.8165447115898132, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 17129 + }, + { + "epoch": 0.1713, + "grad_norm": 0.7133782505989075, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 17130 + }, + { + "epoch": 0.17131, + "grad_norm": 0.6737720370292664, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 17131 + }, + { + "epoch": 0.17132, + "grad_norm": 0.65351402759552, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 17132 + }, + { + "epoch": 0.17133, + "grad_norm": 0.7062671780586243, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 17133 + }, + { + "epoch": 0.17134, + "grad_norm": 0.7021439075469971, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 17134 + }, + { + "epoch": 0.17135, + "grad_norm": 0.6557730436325073, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 17135 + }, + { + "epoch": 0.17136, + "grad_norm": 0.6355788707733154, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 17136 + }, + { + "epoch": 0.17137, + "grad_norm": 0.6955825686454773, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 17137 + }, + { + "epoch": 0.17138, + "grad_norm": 0.7713178992271423, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 17138 + }, + { + "epoch": 0.17139, + "grad_norm": 0.8570329546928406, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 17139 + }, + { + "epoch": 0.1714, + "grad_norm": 1.0380743741989136, + "learning_rate": 0.003, + "loss": 4.028, + "step": 17140 + }, + { + "epoch": 0.17141, + "grad_norm": 1.0252989530563354, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 17141 + }, + { + "epoch": 0.17142, + "grad_norm": 0.8437224626541138, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 17142 + }, + { + "epoch": 0.17143, + "grad_norm": 0.7631940245628357, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 17143 + }, + { + "epoch": 0.17144, + "grad_norm": 0.8039262890815735, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 17144 + }, + { + "epoch": 0.17145, + "grad_norm": 0.9135908484458923, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 17145 + }, + { + "epoch": 0.17146, + "grad_norm": 0.8931128978729248, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 17146 + }, + { + "epoch": 0.17147, + "grad_norm": 0.9946384429931641, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 17147 + }, + { + "epoch": 0.17148, + "grad_norm": 1.0222218036651611, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 17148 + }, + { + "epoch": 0.17149, + "grad_norm": 0.7616196274757385, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 17149 + }, + { + "epoch": 0.1715, + "grad_norm": 0.7758917212486267, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 17150 + }, + { + "epoch": 0.17151, + "grad_norm": 0.7636691331863403, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 17151 + }, + { + "epoch": 0.17152, + "grad_norm": 0.8937175273895264, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 17152 + }, + { + "epoch": 0.17153, + "grad_norm": 0.8333864212036133, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 17153 + }, + { + "epoch": 0.17154, + "grad_norm": 0.6168978214263916, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 17154 + }, + { + "epoch": 0.17155, + "grad_norm": 0.6731457710266113, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 17155 + }, + { + "epoch": 0.17156, + "grad_norm": 0.8194024562835693, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 17156 + }, + { + "epoch": 0.17157, + "grad_norm": 0.8781210780143738, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 17157 + }, + { + "epoch": 0.17158, + "grad_norm": 0.9894959926605225, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 17158 + }, + { + "epoch": 0.17159, + "grad_norm": 1.1849749088287354, + "learning_rate": 0.003, + "loss": 4.033, + "step": 17159 + }, + { + "epoch": 0.1716, + "grad_norm": 1.0248125791549683, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 17160 + }, + { + "epoch": 0.17161, + "grad_norm": 0.9380013942718506, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 17161 + }, + { + "epoch": 0.17162, + "grad_norm": 0.9853338003158569, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 17162 + }, + { + "epoch": 0.17163, + "grad_norm": 1.119477391242981, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 17163 + }, + { + "epoch": 0.17164, + "grad_norm": 1.0069806575775146, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 17164 + }, + { + "epoch": 0.17165, + "grad_norm": 1.0723400115966797, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 17165 + }, + { + "epoch": 0.17166, + "grad_norm": 1.010719656944275, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 17166 + }, + { + "epoch": 0.17167, + "grad_norm": 0.8696228861808777, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 17167 + }, + { + "epoch": 0.17168, + "grad_norm": 0.7739221453666687, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 17168 + }, + { + "epoch": 0.17169, + "grad_norm": 0.7838475704193115, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 17169 + }, + { + "epoch": 0.1717, + "grad_norm": 0.7348381280899048, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 17170 + }, + { + "epoch": 0.17171, + "grad_norm": 0.801749050617218, + "learning_rate": 0.003, + "loss": 4.017, + "step": 17171 + }, + { + "epoch": 0.17172, + "grad_norm": 0.8654886484146118, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 17172 + }, + { + "epoch": 0.17173, + "grad_norm": 0.8426519632339478, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 17173 + }, + { + "epoch": 0.17174, + "grad_norm": 0.754588782787323, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 17174 + }, + { + "epoch": 0.17175, + "grad_norm": 0.7090778350830078, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 17175 + }, + { + "epoch": 0.17176, + "grad_norm": 0.6298714280128479, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 17176 + }, + { + "epoch": 0.17177, + "grad_norm": 0.5926030278205872, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 17177 + }, + { + "epoch": 0.17178, + "grad_norm": 0.6813340783119202, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 17178 + }, + { + "epoch": 0.17179, + "grad_norm": 0.9781045913696289, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 17179 + }, + { + "epoch": 0.1718, + "grad_norm": 1.287935733795166, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 17180 + }, + { + "epoch": 0.17181, + "grad_norm": 0.6834552884101868, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 17181 + }, + { + "epoch": 0.17182, + "grad_norm": 0.9143573641777039, + "learning_rate": 0.003, + "loss": 3.981, + "step": 17182 + }, + { + "epoch": 0.17183, + "grad_norm": 1.3454179763793945, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 17183 + }, + { + "epoch": 0.17184, + "grad_norm": 0.5757114291191101, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 17184 + }, + { + "epoch": 0.17185, + "grad_norm": 0.6277640461921692, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 17185 + }, + { + "epoch": 0.17186, + "grad_norm": 0.7600275874137878, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 17186 + }, + { + "epoch": 0.17187, + "grad_norm": 0.7353819608688354, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 17187 + }, + { + "epoch": 0.17188, + "grad_norm": 0.6280838847160339, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 17188 + }, + { + "epoch": 0.17189, + "grad_norm": 0.6756056547164917, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 17189 + }, + { + "epoch": 0.1719, + "grad_norm": 0.7582512497901917, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 17190 + }, + { + "epoch": 0.17191, + "grad_norm": 0.9030353426933289, + "learning_rate": 0.003, + "loss": 4.032, + "step": 17191 + }, + { + "epoch": 0.17192, + "grad_norm": 0.9921842217445374, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 17192 + }, + { + "epoch": 0.17193, + "grad_norm": 0.9811756014823914, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 17193 + }, + { + "epoch": 0.17194, + "grad_norm": 0.9288318753242493, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 17194 + }, + { + "epoch": 0.17195, + "grad_norm": 1.0272241830825806, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 17195 + }, + { + "epoch": 0.17196, + "grad_norm": 0.9172542095184326, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 17196 + }, + { + "epoch": 0.17197, + "grad_norm": 0.8295038342475891, + "learning_rate": 0.003, + "loss": 4.0, + "step": 17197 + }, + { + "epoch": 0.17198, + "grad_norm": 0.8457723259925842, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 17198 + }, + { + "epoch": 0.17199, + "grad_norm": 0.9383748173713684, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 17199 + }, + { + "epoch": 0.172, + "grad_norm": 0.847207248210907, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 17200 + }, + { + "epoch": 0.17201, + "grad_norm": 0.6900160312652588, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 17201 + }, + { + "epoch": 0.17202, + "grad_norm": 0.8649606704711914, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 17202 + }, + { + "epoch": 0.17203, + "grad_norm": 1.037705421447754, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 17203 + }, + { + "epoch": 0.17204, + "grad_norm": 1.084540605545044, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 17204 + }, + { + "epoch": 0.17205, + "grad_norm": 0.826554000377655, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 17205 + }, + { + "epoch": 0.17206, + "grad_norm": 0.9185588359832764, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 17206 + }, + { + "epoch": 0.17207, + "grad_norm": 0.9008485078811646, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 17207 + }, + { + "epoch": 0.17208, + "grad_norm": 0.8682863116264343, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 17208 + }, + { + "epoch": 0.17209, + "grad_norm": 0.895900309085846, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 17209 + }, + { + "epoch": 0.1721, + "grad_norm": 0.7852314114570618, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 17210 + }, + { + "epoch": 0.17211, + "grad_norm": 0.9408220648765564, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 17211 + }, + { + "epoch": 0.17212, + "grad_norm": 1.2642772197723389, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 17212 + }, + { + "epoch": 0.17213, + "grad_norm": 0.8733190298080444, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 17213 + }, + { + "epoch": 0.17214, + "grad_norm": 0.7148447632789612, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 17214 + }, + { + "epoch": 0.17215, + "grad_norm": 0.6363114714622498, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 17215 + }, + { + "epoch": 0.17216, + "grad_norm": 0.6281545758247375, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 17216 + }, + { + "epoch": 0.17217, + "grad_norm": 0.6153284311294556, + "learning_rate": 0.003, + "loss": 3.965, + "step": 17217 + }, + { + "epoch": 0.17218, + "grad_norm": 0.6405128240585327, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 17218 + }, + { + "epoch": 0.17219, + "grad_norm": 0.7058283090591431, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 17219 + }, + { + "epoch": 0.1722, + "grad_norm": 0.7759990096092224, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 17220 + }, + { + "epoch": 0.17221, + "grad_norm": 0.8714685440063477, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 17221 + }, + { + "epoch": 0.17222, + "grad_norm": 0.8724744319915771, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 17222 + }, + { + "epoch": 0.17223, + "grad_norm": 0.7614827752113342, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 17223 + }, + { + "epoch": 0.17224, + "grad_norm": 0.692354142665863, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 17224 + }, + { + "epoch": 0.17225, + "grad_norm": 0.7642661929130554, + "learning_rate": 0.003, + "loss": 4.0, + "step": 17225 + }, + { + "epoch": 0.17226, + "grad_norm": 0.732671320438385, + "learning_rate": 0.003, + "loss": 4.03, + "step": 17226 + }, + { + "epoch": 0.17227, + "grad_norm": 0.7117012143135071, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 17227 + }, + { + "epoch": 0.17228, + "grad_norm": 0.8455045819282532, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 17228 + }, + { + "epoch": 0.17229, + "grad_norm": 0.9891947507858276, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 17229 + }, + { + "epoch": 0.1723, + "grad_norm": 1.20091712474823, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 17230 + }, + { + "epoch": 0.17231, + "grad_norm": 0.8701549768447876, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 17231 + }, + { + "epoch": 0.17232, + "grad_norm": 0.7704989314079285, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 17232 + }, + { + "epoch": 0.17233, + "grad_norm": 0.7895516157150269, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 17233 + }, + { + "epoch": 0.17234, + "grad_norm": 0.7732393145561218, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 17234 + }, + { + "epoch": 0.17235, + "grad_norm": 0.7781100273132324, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 17235 + }, + { + "epoch": 0.17236, + "grad_norm": 0.7803873419761658, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 17236 + }, + { + "epoch": 0.17237, + "grad_norm": 0.8356990814208984, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 17237 + }, + { + "epoch": 0.17238, + "grad_norm": 0.7919490933418274, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 17238 + }, + { + "epoch": 0.17239, + "grad_norm": 0.9274510145187378, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 17239 + }, + { + "epoch": 0.1724, + "grad_norm": 1.0621308088302612, + "learning_rate": 0.003, + "loss": 3.952, + "step": 17240 + }, + { + "epoch": 0.17241, + "grad_norm": 0.7842041254043579, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 17241 + }, + { + "epoch": 0.17242, + "grad_norm": 0.6512811779975891, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 17242 + }, + { + "epoch": 0.17243, + "grad_norm": 0.6548734307289124, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 17243 + }, + { + "epoch": 0.17244, + "grad_norm": 0.7729583382606506, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 17244 + }, + { + "epoch": 0.17245, + "grad_norm": 1.075770378112793, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 17245 + }, + { + "epoch": 0.17246, + "grad_norm": 0.8497618436813354, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 17246 + }, + { + "epoch": 0.17247, + "grad_norm": 0.7210960388183594, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 17247 + }, + { + "epoch": 0.17248, + "grad_norm": 0.7670256495475769, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 17248 + }, + { + "epoch": 0.17249, + "grad_norm": 0.7562484741210938, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 17249 + }, + { + "epoch": 0.1725, + "grad_norm": 0.6801670789718628, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 17250 + }, + { + "epoch": 0.17251, + "grad_norm": 0.7317301034927368, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 17251 + }, + { + "epoch": 0.17252, + "grad_norm": 0.733239471912384, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 17252 + }, + { + "epoch": 0.17253, + "grad_norm": 0.8458531498908997, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 17253 + }, + { + "epoch": 0.17254, + "grad_norm": 1.0262222290039062, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 17254 + }, + { + "epoch": 0.17255, + "grad_norm": 1.158956527709961, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 17255 + }, + { + "epoch": 0.17256, + "grad_norm": 0.7779101729393005, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 17256 + }, + { + "epoch": 0.17257, + "grad_norm": 0.7064163088798523, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 17257 + }, + { + "epoch": 0.17258, + "grad_norm": 0.7622659802436829, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 17258 + }, + { + "epoch": 0.17259, + "grad_norm": 0.9194077849388123, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 17259 + }, + { + "epoch": 0.1726, + "grad_norm": 0.8892446756362915, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 17260 + }, + { + "epoch": 0.17261, + "grad_norm": 0.8862422108650208, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 17261 + }, + { + "epoch": 0.17262, + "grad_norm": 0.937934935092926, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 17262 + }, + { + "epoch": 0.17263, + "grad_norm": 1.2220501899719238, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 17263 + }, + { + "epoch": 0.17264, + "grad_norm": 0.7869848608970642, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 17264 + }, + { + "epoch": 0.17265, + "grad_norm": 0.7467491030693054, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 17265 + }, + { + "epoch": 0.17266, + "grad_norm": 0.6699659824371338, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 17266 + }, + { + "epoch": 0.17267, + "grad_norm": 0.6560962200164795, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 17267 + }, + { + "epoch": 0.17268, + "grad_norm": 0.6503502726554871, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 17268 + }, + { + "epoch": 0.17269, + "grad_norm": 0.7429957985877991, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 17269 + }, + { + "epoch": 0.1727, + "grad_norm": 0.7676070332527161, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 17270 + }, + { + "epoch": 0.17271, + "grad_norm": 0.9311237335205078, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 17271 + }, + { + "epoch": 0.17272, + "grad_norm": 0.8896855711936951, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 17272 + }, + { + "epoch": 0.17273, + "grad_norm": 0.8906723856925964, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 17273 + }, + { + "epoch": 0.17274, + "grad_norm": 0.9185470342636108, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 17274 + }, + { + "epoch": 0.17275, + "grad_norm": 0.9359113574028015, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 17275 + }, + { + "epoch": 0.17276, + "grad_norm": 1.127929449081421, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 17276 + }, + { + "epoch": 0.17277, + "grad_norm": 0.961293637752533, + "learning_rate": 0.003, + "loss": 4.018, + "step": 17277 + }, + { + "epoch": 0.17278, + "grad_norm": 0.8514905571937561, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 17278 + }, + { + "epoch": 0.17279, + "grad_norm": 0.8976634740829468, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 17279 + }, + { + "epoch": 0.1728, + "grad_norm": 1.0153040885925293, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 17280 + }, + { + "epoch": 0.17281, + "grad_norm": 1.1854093074798584, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 17281 + }, + { + "epoch": 0.17282, + "grad_norm": 0.6242632865905762, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 17282 + }, + { + "epoch": 0.17283, + "grad_norm": 0.7148314118385315, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 17283 + }, + { + "epoch": 0.17284, + "grad_norm": 0.5994605422019958, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 17284 + }, + { + "epoch": 0.17285, + "grad_norm": 0.5930424928665161, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 17285 + }, + { + "epoch": 0.17286, + "grad_norm": 0.6516755223274231, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 17286 + }, + { + "epoch": 0.17287, + "grad_norm": 0.5659880638122559, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 17287 + }, + { + "epoch": 0.17288, + "grad_norm": 0.5813016295433044, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 17288 + }, + { + "epoch": 0.17289, + "grad_norm": 0.6589696407318115, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 17289 + }, + { + "epoch": 0.1729, + "grad_norm": 0.6921236515045166, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 17290 + }, + { + "epoch": 0.17291, + "grad_norm": 0.6518563628196716, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 17291 + }, + { + "epoch": 0.17292, + "grad_norm": 0.638307511806488, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 17292 + }, + { + "epoch": 0.17293, + "grad_norm": 0.7391517162322998, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 17293 + }, + { + "epoch": 0.17294, + "grad_norm": 0.8546097874641418, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 17294 + }, + { + "epoch": 0.17295, + "grad_norm": 0.9340639710426331, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 17295 + }, + { + "epoch": 0.17296, + "grad_norm": 0.9694673418998718, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 17296 + }, + { + "epoch": 0.17297, + "grad_norm": 1.064598560333252, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 17297 + }, + { + "epoch": 0.17298, + "grad_norm": 0.958608865737915, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 17298 + }, + { + "epoch": 0.17299, + "grad_norm": 0.8235632181167603, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 17299 + }, + { + "epoch": 0.173, + "grad_norm": 0.9273446202278137, + "learning_rate": 0.003, + "loss": 4.029, + "step": 17300 + }, + { + "epoch": 0.17301, + "grad_norm": 1.0050894021987915, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 17301 + }, + { + "epoch": 0.17302, + "grad_norm": 1.0514724254608154, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 17302 + }, + { + "epoch": 0.17303, + "grad_norm": 1.0046824216842651, + "learning_rate": 0.003, + "loss": 4.014, + "step": 17303 + }, + { + "epoch": 0.17304, + "grad_norm": 1.0234330892562866, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 17304 + }, + { + "epoch": 0.17305, + "grad_norm": 1.0548959970474243, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 17305 + }, + { + "epoch": 0.17306, + "grad_norm": 1.0699856281280518, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 17306 + }, + { + "epoch": 0.17307, + "grad_norm": 0.8867344260215759, + "learning_rate": 0.003, + "loss": 4.014, + "step": 17307 + }, + { + "epoch": 0.17308, + "grad_norm": 0.8486331105232239, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 17308 + }, + { + "epoch": 0.17309, + "grad_norm": 0.85123211145401, + "learning_rate": 0.003, + "loss": 4.041, + "step": 17309 + }, + { + "epoch": 0.1731, + "grad_norm": 1.0190993547439575, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 17310 + }, + { + "epoch": 0.17311, + "grad_norm": 1.1102086305618286, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 17311 + }, + { + "epoch": 0.17312, + "grad_norm": 0.920134425163269, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 17312 + }, + { + "epoch": 0.17313, + "grad_norm": 0.8725000023841858, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 17313 + }, + { + "epoch": 0.17314, + "grad_norm": 0.7170385122299194, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 17314 + }, + { + "epoch": 0.17315, + "grad_norm": 0.7331812977790833, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 17315 + }, + { + "epoch": 0.17316, + "grad_norm": 0.9502736926078796, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 17316 + }, + { + "epoch": 0.17317, + "grad_norm": 0.9615503549575806, + "learning_rate": 0.003, + "loss": 4.019, + "step": 17317 + }, + { + "epoch": 0.17318, + "grad_norm": 0.8293330669403076, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 17318 + }, + { + "epoch": 0.17319, + "grad_norm": 0.7895271182060242, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 17319 + }, + { + "epoch": 0.1732, + "grad_norm": 0.9117076396942139, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 17320 + }, + { + "epoch": 0.17321, + "grad_norm": 0.9139329791069031, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 17321 + }, + { + "epoch": 0.17322, + "grad_norm": 1.0657029151916504, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 17322 + }, + { + "epoch": 0.17323, + "grad_norm": 1.171359896659851, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 17323 + }, + { + "epoch": 0.17324, + "grad_norm": 0.7427003979682922, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 17324 + }, + { + "epoch": 0.17325, + "grad_norm": 0.5160406231880188, + "learning_rate": 0.003, + "loss": 3.98, + "step": 17325 + }, + { + "epoch": 0.17326, + "grad_norm": 0.7402018308639526, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 17326 + }, + { + "epoch": 0.17327, + "grad_norm": 0.8790965676307678, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 17327 + }, + { + "epoch": 0.17328, + "grad_norm": 0.928956151008606, + "learning_rate": 0.003, + "loss": 3.988, + "step": 17328 + }, + { + "epoch": 0.17329, + "grad_norm": 0.8769863843917847, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 17329 + }, + { + "epoch": 0.1733, + "grad_norm": 0.7440482378005981, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 17330 + }, + { + "epoch": 0.17331, + "grad_norm": 0.5764461755752563, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 17331 + }, + { + "epoch": 0.17332, + "grad_norm": 0.5576710104942322, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 17332 + }, + { + "epoch": 0.17333, + "grad_norm": 0.695499837398529, + "learning_rate": 0.003, + "loss": 4.004, + "step": 17333 + }, + { + "epoch": 0.17334, + "grad_norm": 0.7192192077636719, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 17334 + }, + { + "epoch": 0.17335, + "grad_norm": 0.6879513263702393, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 17335 + }, + { + "epoch": 0.17336, + "grad_norm": 0.6746930480003357, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 17336 + }, + { + "epoch": 0.17337, + "grad_norm": 0.7167201042175293, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 17337 + }, + { + "epoch": 0.17338, + "grad_norm": 0.7446791529655457, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 17338 + }, + { + "epoch": 0.17339, + "grad_norm": 0.7925171256065369, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 17339 + }, + { + "epoch": 0.1734, + "grad_norm": 0.9689291715621948, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 17340 + }, + { + "epoch": 0.17341, + "grad_norm": 1.0888067483901978, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 17341 + }, + { + "epoch": 0.17342, + "grad_norm": 0.6682828068733215, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 17342 + }, + { + "epoch": 0.17343, + "grad_norm": 0.66913902759552, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 17343 + }, + { + "epoch": 0.17344, + "grad_norm": 0.6523056030273438, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 17344 + }, + { + "epoch": 0.17345, + "grad_norm": 0.5623226165771484, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 17345 + }, + { + "epoch": 0.17346, + "grad_norm": 0.583284854888916, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 17346 + }, + { + "epoch": 0.17347, + "grad_norm": 0.6016397476196289, + "learning_rate": 0.003, + "loss": 3.977, + "step": 17347 + }, + { + "epoch": 0.17348, + "grad_norm": 0.589728832244873, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 17348 + }, + { + "epoch": 0.17349, + "grad_norm": 0.5602003931999207, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 17349 + }, + { + "epoch": 0.1735, + "grad_norm": 0.6237813234329224, + "learning_rate": 0.003, + "loss": 3.984, + "step": 17350 + }, + { + "epoch": 0.17351, + "grad_norm": 0.5999209880828857, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 17351 + }, + { + "epoch": 0.17352, + "grad_norm": 0.652887225151062, + "learning_rate": 0.003, + "loss": 3.991, + "step": 17352 + }, + { + "epoch": 0.17353, + "grad_norm": 0.7345176935195923, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 17353 + }, + { + "epoch": 0.17354, + "grad_norm": 0.7107290029525757, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 17354 + }, + { + "epoch": 0.17355, + "grad_norm": 0.7638047337532043, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 17355 + }, + { + "epoch": 0.17356, + "grad_norm": 0.683320939540863, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 17356 + }, + { + "epoch": 0.17357, + "grad_norm": 0.8579681515693665, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 17357 + }, + { + "epoch": 0.17358, + "grad_norm": 1.1549464464187622, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 17358 + }, + { + "epoch": 0.17359, + "grad_norm": 1.3736201524734497, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 17359 + }, + { + "epoch": 0.1736, + "grad_norm": 0.847314178943634, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 17360 + }, + { + "epoch": 0.17361, + "grad_norm": 0.7125725150108337, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 17361 + }, + { + "epoch": 0.17362, + "grad_norm": 0.6976441144943237, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 17362 + }, + { + "epoch": 0.17363, + "grad_norm": 0.6987001895904541, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 17363 + }, + { + "epoch": 0.17364, + "grad_norm": 0.6975489258766174, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 17364 + }, + { + "epoch": 0.17365, + "grad_norm": 0.8965944647789001, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 17365 + }, + { + "epoch": 0.17366, + "grad_norm": 1.063056230545044, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 17366 + }, + { + "epoch": 0.17367, + "grad_norm": 1.0388587713241577, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 17367 + }, + { + "epoch": 0.17368, + "grad_norm": 0.911588728427887, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 17368 + }, + { + "epoch": 0.17369, + "grad_norm": 0.8996477723121643, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 17369 + }, + { + "epoch": 0.1737, + "grad_norm": 0.9115211367607117, + "learning_rate": 0.003, + "loss": 4.025, + "step": 17370 + }, + { + "epoch": 0.17371, + "grad_norm": 0.8117915391921997, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 17371 + }, + { + "epoch": 0.17372, + "grad_norm": 1.026132345199585, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 17372 + }, + { + "epoch": 0.17373, + "grad_norm": 1.2777268886566162, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 17373 + }, + { + "epoch": 0.17374, + "grad_norm": 0.8393291234970093, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 17374 + }, + { + "epoch": 0.17375, + "grad_norm": 0.724292516708374, + "learning_rate": 0.003, + "loss": 4.009, + "step": 17375 + }, + { + "epoch": 0.17376, + "grad_norm": 0.6945507526397705, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 17376 + }, + { + "epoch": 0.17377, + "grad_norm": 0.693756639957428, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 17377 + }, + { + "epoch": 0.17378, + "grad_norm": 0.8418163061141968, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 17378 + }, + { + "epoch": 0.17379, + "grad_norm": 0.6874514222145081, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 17379 + }, + { + "epoch": 0.1738, + "grad_norm": 0.6427291035652161, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 17380 + }, + { + "epoch": 0.17381, + "grad_norm": 0.6788190007209778, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 17381 + }, + { + "epoch": 0.17382, + "grad_norm": 0.7309727668762207, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 17382 + }, + { + "epoch": 0.17383, + "grad_norm": 0.8250257968902588, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 17383 + }, + { + "epoch": 0.17384, + "grad_norm": 1.0060784816741943, + "learning_rate": 0.003, + "loss": 4.003, + "step": 17384 + }, + { + "epoch": 0.17385, + "grad_norm": 1.2069662809371948, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 17385 + }, + { + "epoch": 0.17386, + "grad_norm": 0.9430493712425232, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 17386 + }, + { + "epoch": 0.17387, + "grad_norm": 0.9377607107162476, + "learning_rate": 0.003, + "loss": 4.009, + "step": 17387 + }, + { + "epoch": 0.17388, + "grad_norm": 0.9175955057144165, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 17388 + }, + { + "epoch": 0.17389, + "grad_norm": 0.9722581505775452, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 17389 + }, + { + "epoch": 0.1739, + "grad_norm": 0.8540912866592407, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 17390 + }, + { + "epoch": 0.17391, + "grad_norm": 0.8366203308105469, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 17391 + }, + { + "epoch": 0.17392, + "grad_norm": 0.7991741299629211, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 17392 + }, + { + "epoch": 0.17393, + "grad_norm": 0.7675210237503052, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 17393 + }, + { + "epoch": 0.17394, + "grad_norm": 0.9029843807220459, + "learning_rate": 0.003, + "loss": 4.017, + "step": 17394 + }, + { + "epoch": 0.17395, + "grad_norm": 0.9093694090843201, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 17395 + }, + { + "epoch": 0.17396, + "grad_norm": 1.017406940460205, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 17396 + }, + { + "epoch": 0.17397, + "grad_norm": 1.1229156255722046, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 17397 + }, + { + "epoch": 0.17398, + "grad_norm": 1.0047029256820679, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 17398 + }, + { + "epoch": 0.17399, + "grad_norm": 0.9616352319717407, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 17399 + }, + { + "epoch": 0.174, + "grad_norm": 0.7757841348648071, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 17400 + }, + { + "epoch": 0.17401, + "grad_norm": 0.8163595795631409, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 17401 + }, + { + "epoch": 0.17402, + "grad_norm": 0.7642260789871216, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 17402 + }, + { + "epoch": 0.17403, + "grad_norm": 0.7993416786193848, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 17403 + }, + { + "epoch": 0.17404, + "grad_norm": 0.7023702263832092, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 17404 + }, + { + "epoch": 0.17405, + "grad_norm": 0.6713720560073853, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 17405 + }, + { + "epoch": 0.17406, + "grad_norm": 0.7091724872589111, + "learning_rate": 0.003, + "loss": 4.01, + "step": 17406 + }, + { + "epoch": 0.17407, + "grad_norm": 0.6860936880111694, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 17407 + }, + { + "epoch": 0.17408, + "grad_norm": 0.6985510587692261, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 17408 + }, + { + "epoch": 0.17409, + "grad_norm": 0.753868043422699, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 17409 + }, + { + "epoch": 0.1741, + "grad_norm": 0.9177746176719666, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 17410 + }, + { + "epoch": 0.17411, + "grad_norm": 1.244633436203003, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 17411 + }, + { + "epoch": 0.17412, + "grad_norm": 1.0975927114486694, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 17412 + }, + { + "epoch": 0.17413, + "grad_norm": 1.0281498432159424, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 17413 + }, + { + "epoch": 0.17414, + "grad_norm": 1.0181832313537598, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 17414 + }, + { + "epoch": 0.17415, + "grad_norm": 0.9445314407348633, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 17415 + }, + { + "epoch": 0.17416, + "grad_norm": 0.8794584274291992, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 17416 + }, + { + "epoch": 0.17417, + "grad_norm": 0.839174211025238, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 17417 + }, + { + "epoch": 0.17418, + "grad_norm": 0.7632862329483032, + "learning_rate": 0.003, + "loss": 3.995, + "step": 17418 + }, + { + "epoch": 0.17419, + "grad_norm": 0.7480762004852295, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 17419 + }, + { + "epoch": 0.1742, + "grad_norm": 0.9292443990707397, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 17420 + }, + { + "epoch": 0.17421, + "grad_norm": 1.1653343439102173, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 17421 + }, + { + "epoch": 0.17422, + "grad_norm": 0.8291074633598328, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 17422 + }, + { + "epoch": 0.17423, + "grad_norm": 0.6278573870658875, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 17423 + }, + { + "epoch": 0.17424, + "grad_norm": 0.5656075477600098, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 17424 + }, + { + "epoch": 0.17425, + "grad_norm": 0.6245643496513367, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 17425 + }, + { + "epoch": 0.17426, + "grad_norm": 0.5158340930938721, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 17426 + }, + { + "epoch": 0.17427, + "grad_norm": 0.5346347689628601, + "learning_rate": 0.003, + "loss": 3.993, + "step": 17427 + }, + { + "epoch": 0.17428, + "grad_norm": 0.6253666281700134, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 17428 + }, + { + "epoch": 0.17429, + "grad_norm": 0.6902015209197998, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 17429 + }, + { + "epoch": 0.1743, + "grad_norm": 0.7939674854278564, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 17430 + }, + { + "epoch": 0.17431, + "grad_norm": 0.9390441179275513, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 17431 + }, + { + "epoch": 0.17432, + "grad_norm": 1.0236916542053223, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 17432 + }, + { + "epoch": 0.17433, + "grad_norm": 0.9324219822883606, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 17433 + }, + { + "epoch": 0.17434, + "grad_norm": 0.8530586957931519, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 17434 + }, + { + "epoch": 0.17435, + "grad_norm": 0.8744862079620361, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 17435 + }, + { + "epoch": 0.17436, + "grad_norm": 0.999214768409729, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 17436 + }, + { + "epoch": 0.17437, + "grad_norm": 1.099568247795105, + "learning_rate": 0.003, + "loss": 4.006, + "step": 17437 + }, + { + "epoch": 0.17438, + "grad_norm": 0.8788972496986389, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 17438 + }, + { + "epoch": 0.17439, + "grad_norm": 0.7022820711135864, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 17439 + }, + { + "epoch": 0.1744, + "grad_norm": 0.6724662780761719, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 17440 + }, + { + "epoch": 0.17441, + "grad_norm": 0.7439224123954773, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 17441 + }, + { + "epoch": 0.17442, + "grad_norm": 0.7784821391105652, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 17442 + }, + { + "epoch": 0.17443, + "grad_norm": 0.7862952947616577, + "learning_rate": 0.003, + "loss": 4.01, + "step": 17443 + }, + { + "epoch": 0.17444, + "grad_norm": 0.9709135293960571, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 17444 + }, + { + "epoch": 0.17445, + "grad_norm": 1.1998265981674194, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 17445 + }, + { + "epoch": 0.17446, + "grad_norm": 0.9091138243675232, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 17446 + }, + { + "epoch": 0.17447, + "grad_norm": 0.9684601426124573, + "learning_rate": 0.003, + "loss": 3.994, + "step": 17447 + }, + { + "epoch": 0.17448, + "grad_norm": 1.0080211162567139, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 17448 + }, + { + "epoch": 0.17449, + "grad_norm": 0.9079150557518005, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 17449 + }, + { + "epoch": 0.1745, + "grad_norm": 0.6905058026313782, + "learning_rate": 0.003, + "loss": 3.999, + "step": 17450 + }, + { + "epoch": 0.17451, + "grad_norm": 0.632532000541687, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 17451 + }, + { + "epoch": 0.17452, + "grad_norm": 0.7154324054718018, + "learning_rate": 0.003, + "loss": 4.005, + "step": 17452 + }, + { + "epoch": 0.17453, + "grad_norm": 0.7559896111488342, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 17453 + }, + { + "epoch": 0.17454, + "grad_norm": 0.7690410017967224, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 17454 + }, + { + "epoch": 0.17455, + "grad_norm": 0.7744832038879395, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 17455 + }, + { + "epoch": 0.17456, + "grad_norm": 0.787226676940918, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 17456 + }, + { + "epoch": 0.17457, + "grad_norm": 0.8192123770713806, + "learning_rate": 0.003, + "loss": 4.011, + "step": 17457 + }, + { + "epoch": 0.17458, + "grad_norm": 0.8754841089248657, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 17458 + }, + { + "epoch": 0.17459, + "grad_norm": 0.8357239365577698, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 17459 + }, + { + "epoch": 0.1746, + "grad_norm": 0.9402945637702942, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 17460 + }, + { + "epoch": 0.17461, + "grad_norm": 1.1382694244384766, + "learning_rate": 0.003, + "loss": 4.006, + "step": 17461 + }, + { + "epoch": 0.17462, + "grad_norm": 0.9957150816917419, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 17462 + }, + { + "epoch": 0.17463, + "grad_norm": 0.9740569591522217, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 17463 + }, + { + "epoch": 0.17464, + "grad_norm": 0.8861000537872314, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 17464 + }, + { + "epoch": 0.17465, + "grad_norm": 0.7433106303215027, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 17465 + }, + { + "epoch": 0.17466, + "grad_norm": 0.6907755732536316, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 17466 + }, + { + "epoch": 0.17467, + "grad_norm": 0.7396727204322815, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 17467 + }, + { + "epoch": 0.17468, + "grad_norm": 0.7633123993873596, + "learning_rate": 0.003, + "loss": 3.998, + "step": 17468 + }, + { + "epoch": 0.17469, + "grad_norm": 0.6752055883407593, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 17469 + }, + { + "epoch": 0.1747, + "grad_norm": 0.6542142033576965, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 17470 + }, + { + "epoch": 0.17471, + "grad_norm": 0.724885106086731, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 17471 + }, + { + "epoch": 0.17472, + "grad_norm": 0.7976656556129456, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 17472 + }, + { + "epoch": 0.17473, + "grad_norm": 0.8885031342506409, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 17473 + }, + { + "epoch": 0.17474, + "grad_norm": 1.2055965662002563, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 17474 + }, + { + "epoch": 0.17475, + "grad_norm": 1.0816218852996826, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 17475 + }, + { + "epoch": 0.17476, + "grad_norm": 0.8149133920669556, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 17476 + }, + { + "epoch": 0.17477, + "grad_norm": 0.7656036615371704, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 17477 + }, + { + "epoch": 0.17478, + "grad_norm": 0.7564748525619507, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 17478 + }, + { + "epoch": 0.17479, + "grad_norm": 0.7418794631958008, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 17479 + }, + { + "epoch": 0.1748, + "grad_norm": 0.6966302394866943, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 17480 + }, + { + "epoch": 0.17481, + "grad_norm": 0.6780658960342407, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 17481 + }, + { + "epoch": 0.17482, + "grad_norm": 0.7791665196418762, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 17482 + }, + { + "epoch": 0.17483, + "grad_norm": 0.7765859365463257, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 17483 + }, + { + "epoch": 0.17484, + "grad_norm": 0.9068479537963867, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 17484 + }, + { + "epoch": 0.17485, + "grad_norm": 1.1002610921859741, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 17485 + }, + { + "epoch": 0.17486, + "grad_norm": 0.99592524766922, + "learning_rate": 0.003, + "loss": 4.036, + "step": 17486 + }, + { + "epoch": 0.17487, + "grad_norm": 1.0522546768188477, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 17487 + }, + { + "epoch": 0.17488, + "grad_norm": 0.7485376000404358, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 17488 + }, + { + "epoch": 0.17489, + "grad_norm": 0.6797248125076294, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 17489 + }, + { + "epoch": 0.1749, + "grad_norm": 0.7047320604324341, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 17490 + }, + { + "epoch": 0.17491, + "grad_norm": 0.8675941824913025, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 17491 + }, + { + "epoch": 0.17492, + "grad_norm": 0.8862576484680176, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 17492 + }, + { + "epoch": 0.17493, + "grad_norm": 0.9569544792175293, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 17493 + }, + { + "epoch": 0.17494, + "grad_norm": 0.8425646424293518, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 17494 + }, + { + "epoch": 0.17495, + "grad_norm": 0.785433828830719, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 17495 + }, + { + "epoch": 0.17496, + "grad_norm": 0.7538527846336365, + "learning_rate": 0.003, + "loss": 3.987, + "step": 17496 + }, + { + "epoch": 0.17497, + "grad_norm": 0.7240882515907288, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 17497 + }, + { + "epoch": 0.17498, + "grad_norm": 0.7209701538085938, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 17498 + }, + { + "epoch": 0.17499, + "grad_norm": 0.7353964447975159, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 17499 + }, + { + "epoch": 0.175, + "grad_norm": 0.8131001591682434, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 17500 + }, + { + "epoch": 0.17501, + "grad_norm": 1.0257822275161743, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 17501 + }, + { + "epoch": 0.17502, + "grad_norm": 1.1355948448181152, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 17502 + }, + { + "epoch": 0.17503, + "grad_norm": 0.8676596879959106, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 17503 + }, + { + "epoch": 0.17504, + "grad_norm": 0.7113892436027527, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 17504 + }, + { + "epoch": 0.17505, + "grad_norm": 0.7276456952095032, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 17505 + }, + { + "epoch": 0.17506, + "grad_norm": 0.8023135662078857, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 17506 + }, + { + "epoch": 0.17507, + "grad_norm": 0.8918072581291199, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 17507 + }, + { + "epoch": 0.17508, + "grad_norm": 0.9726418256759644, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 17508 + }, + { + "epoch": 0.17509, + "grad_norm": 1.2349975109100342, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 17509 + }, + { + "epoch": 0.1751, + "grad_norm": 0.7286527156829834, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 17510 + }, + { + "epoch": 0.17511, + "grad_norm": 0.7270987033843994, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 17511 + }, + { + "epoch": 0.17512, + "grad_norm": 0.9057313203811646, + "learning_rate": 0.003, + "loss": 4.024, + "step": 17512 + }, + { + "epoch": 0.17513, + "grad_norm": 1.2127231359481812, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 17513 + }, + { + "epoch": 0.17514, + "grad_norm": 1.0335619449615479, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 17514 + }, + { + "epoch": 0.17515, + "grad_norm": 1.0899924039840698, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 17515 + }, + { + "epoch": 0.17516, + "grad_norm": 0.788173496723175, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 17516 + }, + { + "epoch": 0.17517, + "grad_norm": 0.7215850949287415, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 17517 + }, + { + "epoch": 0.17518, + "grad_norm": 0.6537896394729614, + "learning_rate": 0.003, + "loss": 4.008, + "step": 17518 + }, + { + "epoch": 0.17519, + "grad_norm": 0.7272935509681702, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 17519 + }, + { + "epoch": 0.1752, + "grad_norm": 0.8016128540039062, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 17520 + }, + { + "epoch": 0.17521, + "grad_norm": 0.9136723279953003, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 17521 + }, + { + "epoch": 0.17522, + "grad_norm": 1.1162629127502441, + "learning_rate": 0.003, + "loss": 3.999, + "step": 17522 + }, + { + "epoch": 0.17523, + "grad_norm": 1.1937209367752075, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 17523 + }, + { + "epoch": 0.17524, + "grad_norm": 0.7823119163513184, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 17524 + }, + { + "epoch": 0.17525, + "grad_norm": 0.5235046148300171, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 17525 + }, + { + "epoch": 0.17526, + "grad_norm": 0.6116799712181091, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 17526 + }, + { + "epoch": 0.17527, + "grad_norm": 0.5505747199058533, + "learning_rate": 0.003, + "loss": 3.969, + "step": 17527 + }, + { + "epoch": 0.17528, + "grad_norm": 0.6316537857055664, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 17528 + }, + { + "epoch": 0.17529, + "grad_norm": 0.7600823044776917, + "learning_rate": 0.003, + "loss": 3.996, + "step": 17529 + }, + { + "epoch": 0.1753, + "grad_norm": 0.9958966374397278, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 17530 + }, + { + "epoch": 0.17531, + "grad_norm": 1.2995574474334717, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 17531 + }, + { + "epoch": 0.17532, + "grad_norm": 0.7968376874923706, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 17532 + }, + { + "epoch": 0.17533, + "grad_norm": 0.7426732778549194, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 17533 + }, + { + "epoch": 0.17534, + "grad_norm": 0.6420416831970215, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 17534 + }, + { + "epoch": 0.17535, + "grad_norm": 0.6831434369087219, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 17535 + }, + { + "epoch": 0.17536, + "grad_norm": 0.7888033986091614, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 17536 + }, + { + "epoch": 0.17537, + "grad_norm": 0.7804220914840698, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 17537 + }, + { + "epoch": 0.17538, + "grad_norm": 0.7175537347793579, + "learning_rate": 0.003, + "loss": 4.026, + "step": 17538 + }, + { + "epoch": 0.17539, + "grad_norm": 0.867682933807373, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 17539 + }, + { + "epoch": 0.1754, + "grad_norm": 0.8793272376060486, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 17540 + }, + { + "epoch": 0.17541, + "grad_norm": 0.8773455023765564, + "learning_rate": 0.003, + "loss": 4.021, + "step": 17541 + }, + { + "epoch": 0.17542, + "grad_norm": 0.8339686393737793, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 17542 + }, + { + "epoch": 0.17543, + "grad_norm": 0.8761228919029236, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 17543 + }, + { + "epoch": 0.17544, + "grad_norm": 0.9914465546607971, + "learning_rate": 0.003, + "loss": 4.014, + "step": 17544 + }, + { + "epoch": 0.17545, + "grad_norm": 1.115749478340149, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 17545 + }, + { + "epoch": 0.17546, + "grad_norm": 0.8497521877288818, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 17546 + }, + { + "epoch": 0.17547, + "grad_norm": 0.8092749118804932, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 17547 + }, + { + "epoch": 0.17548, + "grad_norm": 0.732430636882782, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 17548 + }, + { + "epoch": 0.17549, + "grad_norm": 0.6795975565910339, + "learning_rate": 0.003, + "loss": 3.985, + "step": 17549 + }, + { + "epoch": 0.1755, + "grad_norm": 0.5819090008735657, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 17550 + }, + { + "epoch": 0.17551, + "grad_norm": 0.6748693585395813, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 17551 + }, + { + "epoch": 0.17552, + "grad_norm": 0.9079766273498535, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 17552 + }, + { + "epoch": 0.17553, + "grad_norm": 1.239787220954895, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 17553 + }, + { + "epoch": 0.17554, + "grad_norm": 0.8258132934570312, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 17554 + }, + { + "epoch": 0.17555, + "grad_norm": 0.7910948395729065, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 17555 + }, + { + "epoch": 0.17556, + "grad_norm": 1.1124839782714844, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 17556 + }, + { + "epoch": 0.17557, + "grad_norm": 1.1162309646606445, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 17557 + }, + { + "epoch": 0.17558, + "grad_norm": 0.9350470304489136, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 17558 + }, + { + "epoch": 0.17559, + "grad_norm": 0.8715490102767944, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 17559 + }, + { + "epoch": 0.1756, + "grad_norm": 0.7804958820343018, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 17560 + }, + { + "epoch": 0.17561, + "grad_norm": 0.8452410697937012, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 17561 + }, + { + "epoch": 0.17562, + "grad_norm": 0.81874680519104, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 17562 + }, + { + "epoch": 0.17563, + "grad_norm": 0.7408578991889954, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 17563 + }, + { + "epoch": 0.17564, + "grad_norm": 0.6901379227638245, + "learning_rate": 0.003, + "loss": 3.999, + "step": 17564 + }, + { + "epoch": 0.17565, + "grad_norm": 0.6317659616470337, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 17565 + }, + { + "epoch": 0.17566, + "grad_norm": 0.7332043647766113, + "learning_rate": 0.003, + "loss": 3.977, + "step": 17566 + }, + { + "epoch": 0.17567, + "grad_norm": 0.9280351400375366, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 17567 + }, + { + "epoch": 0.17568, + "grad_norm": 0.961087703704834, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 17568 + }, + { + "epoch": 0.17569, + "grad_norm": 1.0218905210494995, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 17569 + }, + { + "epoch": 0.1757, + "grad_norm": 1.013644814491272, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 17570 + }, + { + "epoch": 0.17571, + "grad_norm": 0.866844654083252, + "learning_rate": 0.003, + "loss": 4.006, + "step": 17571 + }, + { + "epoch": 0.17572, + "grad_norm": 0.9111921787261963, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 17572 + }, + { + "epoch": 0.17573, + "grad_norm": 0.7767279148101807, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 17573 + }, + { + "epoch": 0.17574, + "grad_norm": 0.6741935014724731, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 17574 + }, + { + "epoch": 0.17575, + "grad_norm": 0.7317989468574524, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 17575 + }, + { + "epoch": 0.17576, + "grad_norm": 0.9040318727493286, + "learning_rate": 0.003, + "loss": 4.022, + "step": 17576 + }, + { + "epoch": 0.17577, + "grad_norm": 1.078898549079895, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 17577 + }, + { + "epoch": 0.17578, + "grad_norm": 0.9824024438858032, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 17578 + }, + { + "epoch": 0.17579, + "grad_norm": 1.0256359577178955, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 17579 + }, + { + "epoch": 0.1758, + "grad_norm": 0.9958242177963257, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 17580 + }, + { + "epoch": 0.17581, + "grad_norm": 0.9190477728843689, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 17581 + }, + { + "epoch": 0.17582, + "grad_norm": 0.7948879599571228, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 17582 + }, + { + "epoch": 0.17583, + "grad_norm": 0.7672145366668701, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 17583 + }, + { + "epoch": 0.17584, + "grad_norm": 0.7990474700927734, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 17584 + }, + { + "epoch": 0.17585, + "grad_norm": 0.8452971577644348, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 17585 + }, + { + "epoch": 0.17586, + "grad_norm": 0.8646570444107056, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 17586 + }, + { + "epoch": 0.17587, + "grad_norm": 0.8637148141860962, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 17587 + }, + { + "epoch": 0.17588, + "grad_norm": 0.8240643739700317, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 17588 + }, + { + "epoch": 0.17589, + "grad_norm": 0.9793906211853027, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 17589 + }, + { + "epoch": 0.1759, + "grad_norm": 1.1525877714157104, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 17590 + }, + { + "epoch": 0.17591, + "grad_norm": 0.8227669596672058, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 17591 + }, + { + "epoch": 0.17592, + "grad_norm": 0.6809818744659424, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 17592 + }, + { + "epoch": 0.17593, + "grad_norm": 0.559374213218689, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 17593 + }, + { + "epoch": 0.17594, + "grad_norm": 0.5239890217781067, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 17594 + }, + { + "epoch": 0.17595, + "grad_norm": 0.5396626591682434, + "learning_rate": 0.003, + "loss": 3.991, + "step": 17595 + }, + { + "epoch": 0.17596, + "grad_norm": 0.6020597219467163, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 17596 + }, + { + "epoch": 0.17597, + "grad_norm": 0.6247676610946655, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 17597 + }, + { + "epoch": 0.17598, + "grad_norm": 0.7158964276313782, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 17598 + }, + { + "epoch": 0.17599, + "grad_norm": 0.8355786800384521, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 17599 + }, + { + "epoch": 0.176, + "grad_norm": 1.1567732095718384, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 17600 + }, + { + "epoch": 0.17601, + "grad_norm": 1.0251920223236084, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 17601 + }, + { + "epoch": 0.17602, + "grad_norm": 0.8615152835845947, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 17602 + }, + { + "epoch": 0.17603, + "grad_norm": 0.7872258424758911, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 17603 + }, + { + "epoch": 0.17604, + "grad_norm": 0.7728003859519958, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 17604 + }, + { + "epoch": 0.17605, + "grad_norm": 0.9679343104362488, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 17605 + }, + { + "epoch": 0.17606, + "grad_norm": 0.9719768762588501, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 17606 + }, + { + "epoch": 0.17607, + "grad_norm": 1.1245415210723877, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 17607 + }, + { + "epoch": 0.17608, + "grad_norm": 0.9123148322105408, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 17608 + }, + { + "epoch": 0.17609, + "grad_norm": 0.8522974252700806, + "learning_rate": 0.003, + "loss": 3.997, + "step": 17609 + }, + { + "epoch": 0.1761, + "grad_norm": 0.9533160328865051, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 17610 + }, + { + "epoch": 0.17611, + "grad_norm": 1.060482382774353, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 17611 + }, + { + "epoch": 0.17612, + "grad_norm": 1.0258617401123047, + "learning_rate": 0.003, + "loss": 3.951, + "step": 17612 + }, + { + "epoch": 0.17613, + "grad_norm": 0.9553532600402832, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 17613 + }, + { + "epoch": 0.17614, + "grad_norm": 0.9101959466934204, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 17614 + }, + { + "epoch": 0.17615, + "grad_norm": 0.8651487827301025, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 17615 + }, + { + "epoch": 0.17616, + "grad_norm": 0.8503139019012451, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 17616 + }, + { + "epoch": 0.17617, + "grad_norm": 0.8145707845687866, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 17617 + }, + { + "epoch": 0.17618, + "grad_norm": 0.8305964469909668, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 17618 + }, + { + "epoch": 0.17619, + "grad_norm": 0.784797728061676, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 17619 + }, + { + "epoch": 0.1762, + "grad_norm": 0.7057626247406006, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 17620 + }, + { + "epoch": 0.17621, + "grad_norm": 0.722081184387207, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 17621 + }, + { + "epoch": 0.17622, + "grad_norm": 0.7251399755477905, + "learning_rate": 0.003, + "loss": 4.01, + "step": 17622 + }, + { + "epoch": 0.17623, + "grad_norm": 0.8381239175796509, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 17623 + }, + { + "epoch": 0.17624, + "grad_norm": 0.9010316729545593, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 17624 + }, + { + "epoch": 0.17625, + "grad_norm": 0.8895271420478821, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 17625 + }, + { + "epoch": 0.17626, + "grad_norm": 1.0438815355300903, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 17626 + }, + { + "epoch": 0.17627, + "grad_norm": 1.0572185516357422, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 17627 + }, + { + "epoch": 0.17628, + "grad_norm": 0.9122909307479858, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 17628 + }, + { + "epoch": 0.17629, + "grad_norm": 0.8782704472541809, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 17629 + }, + { + "epoch": 0.1763, + "grad_norm": 0.7416739463806152, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 17630 + }, + { + "epoch": 0.17631, + "grad_norm": 0.6938516497612, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 17631 + }, + { + "epoch": 0.17632, + "grad_norm": 0.732664167881012, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 17632 + }, + { + "epoch": 0.17633, + "grad_norm": 0.7271563410758972, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 17633 + }, + { + "epoch": 0.17634, + "grad_norm": 0.6744954586029053, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 17634 + }, + { + "epoch": 0.17635, + "grad_norm": 0.6522406339645386, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 17635 + }, + { + "epoch": 0.17636, + "grad_norm": 0.7126379013061523, + "learning_rate": 0.003, + "loss": 3.999, + "step": 17636 + }, + { + "epoch": 0.17637, + "grad_norm": 0.8472570180892944, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 17637 + }, + { + "epoch": 0.17638, + "grad_norm": 1.177600383758545, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 17638 + }, + { + "epoch": 0.17639, + "grad_norm": 0.9043630361557007, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 17639 + }, + { + "epoch": 0.1764, + "grad_norm": 0.6743602752685547, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 17640 + }, + { + "epoch": 0.17641, + "grad_norm": 0.540255606174469, + "learning_rate": 0.003, + "loss": 3.991, + "step": 17641 + }, + { + "epoch": 0.17642, + "grad_norm": 0.5593344569206238, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 17642 + }, + { + "epoch": 0.17643, + "grad_norm": 0.6615741848945618, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 17643 + }, + { + "epoch": 0.17644, + "grad_norm": 0.8354815244674683, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 17644 + }, + { + "epoch": 0.17645, + "grad_norm": 0.9365418553352356, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 17645 + }, + { + "epoch": 0.17646, + "grad_norm": 0.91257643699646, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 17646 + }, + { + "epoch": 0.17647, + "grad_norm": 0.8841412663459778, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 17647 + }, + { + "epoch": 0.17648, + "grad_norm": 0.8749768733978271, + "learning_rate": 0.003, + "loss": 4.006, + "step": 17648 + }, + { + "epoch": 0.17649, + "grad_norm": 1.000697135925293, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 17649 + }, + { + "epoch": 0.1765, + "grad_norm": 1.0813151597976685, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 17650 + }, + { + "epoch": 0.17651, + "grad_norm": 0.8042722344398499, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 17651 + }, + { + "epoch": 0.17652, + "grad_norm": 0.7402278780937195, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 17652 + }, + { + "epoch": 0.17653, + "grad_norm": 0.7824121117591858, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 17653 + }, + { + "epoch": 0.17654, + "grad_norm": 0.7480524778366089, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 17654 + }, + { + "epoch": 0.17655, + "grad_norm": 0.6704444289207458, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 17655 + }, + { + "epoch": 0.17656, + "grad_norm": 0.8114060163497925, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 17656 + }, + { + "epoch": 0.17657, + "grad_norm": 0.8143280744552612, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 17657 + }, + { + "epoch": 0.17658, + "grad_norm": 0.8782177567481995, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 17658 + }, + { + "epoch": 0.17659, + "grad_norm": 0.9757961630821228, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 17659 + }, + { + "epoch": 0.1766, + "grad_norm": 0.8597797751426697, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 17660 + }, + { + "epoch": 0.17661, + "grad_norm": 0.8613868355751038, + "learning_rate": 0.003, + "loss": 4.01, + "step": 17661 + }, + { + "epoch": 0.17662, + "grad_norm": 1.0099040269851685, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 17662 + }, + { + "epoch": 0.17663, + "grad_norm": 0.9791454076766968, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 17663 + }, + { + "epoch": 0.17664, + "grad_norm": 0.9757301807403564, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 17664 + }, + { + "epoch": 0.17665, + "grad_norm": 1.0718437433242798, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 17665 + }, + { + "epoch": 0.17666, + "grad_norm": 1.065012812614441, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 17666 + }, + { + "epoch": 0.17667, + "grad_norm": 1.106485366821289, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 17667 + }, + { + "epoch": 0.17668, + "grad_norm": 0.8106964826583862, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 17668 + }, + { + "epoch": 0.17669, + "grad_norm": 0.6266219019889832, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 17669 + }, + { + "epoch": 0.1767, + "grad_norm": 0.6873462796211243, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 17670 + }, + { + "epoch": 0.17671, + "grad_norm": 0.9363231062889099, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 17671 + }, + { + "epoch": 0.17672, + "grad_norm": 1.082198977470398, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 17672 + }, + { + "epoch": 0.17673, + "grad_norm": 0.9996940493583679, + "learning_rate": 0.003, + "loss": 3.99, + "step": 17673 + }, + { + "epoch": 0.17674, + "grad_norm": 0.9465346932411194, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 17674 + }, + { + "epoch": 0.17675, + "grad_norm": 0.9348333477973938, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 17675 + }, + { + "epoch": 0.17676, + "grad_norm": 0.8782925009727478, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 17676 + }, + { + "epoch": 0.17677, + "grad_norm": 0.7545113563537598, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 17677 + }, + { + "epoch": 0.17678, + "grad_norm": 0.7324469089508057, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 17678 + }, + { + "epoch": 0.17679, + "grad_norm": 0.6657153964042664, + "learning_rate": 0.003, + "loss": 3.986, + "step": 17679 + }, + { + "epoch": 0.1768, + "grad_norm": 0.6143937706947327, + "learning_rate": 0.003, + "loss": 3.972, + "step": 17680 + }, + { + "epoch": 0.17681, + "grad_norm": 0.5606423020362854, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 17681 + }, + { + "epoch": 0.17682, + "grad_norm": 0.521074652671814, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 17682 + }, + { + "epoch": 0.17683, + "grad_norm": 0.5197908878326416, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 17683 + }, + { + "epoch": 0.17684, + "grad_norm": 0.5607931017875671, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 17684 + }, + { + "epoch": 0.17685, + "grad_norm": 0.6494010090827942, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 17685 + }, + { + "epoch": 0.17686, + "grad_norm": 0.9178204536437988, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 17686 + }, + { + "epoch": 0.17687, + "grad_norm": 1.1997172832489014, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 17687 + }, + { + "epoch": 0.17688, + "grad_norm": 0.9303160309791565, + "learning_rate": 0.003, + "loss": 3.983, + "step": 17688 + }, + { + "epoch": 0.17689, + "grad_norm": 0.7332910299301147, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 17689 + }, + { + "epoch": 0.1769, + "grad_norm": 0.6390018463134766, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 17690 + }, + { + "epoch": 0.17691, + "grad_norm": 0.6789406538009644, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 17691 + }, + { + "epoch": 0.17692, + "grad_norm": 0.76705402135849, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 17692 + }, + { + "epoch": 0.17693, + "grad_norm": 0.8636718392372131, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 17693 + }, + { + "epoch": 0.17694, + "grad_norm": 0.8098905682563782, + "learning_rate": 0.003, + "loss": 3.983, + "step": 17694 + }, + { + "epoch": 0.17695, + "grad_norm": 0.8161396384239197, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 17695 + }, + { + "epoch": 0.17696, + "grad_norm": 0.9196518063545227, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 17696 + }, + { + "epoch": 0.17697, + "grad_norm": 1.0340486764907837, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 17697 + }, + { + "epoch": 0.17698, + "grad_norm": 0.9755218029022217, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 17698 + }, + { + "epoch": 0.17699, + "grad_norm": 1.0434486865997314, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 17699 + }, + { + "epoch": 0.177, + "grad_norm": 0.9362537264823914, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 17700 + }, + { + "epoch": 0.17701, + "grad_norm": 0.9179715514183044, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 17701 + }, + { + "epoch": 0.17702, + "grad_norm": 0.8468559384346008, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 17702 + }, + { + "epoch": 0.17703, + "grad_norm": 0.9051650762557983, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 17703 + }, + { + "epoch": 0.17704, + "grad_norm": 0.9996410608291626, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 17704 + }, + { + "epoch": 0.17705, + "grad_norm": 0.9762026071548462, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 17705 + }, + { + "epoch": 0.17706, + "grad_norm": 0.8613827228546143, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 17706 + }, + { + "epoch": 0.17707, + "grad_norm": 0.8776628971099854, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 17707 + }, + { + "epoch": 0.17708, + "grad_norm": 1.010719895362854, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 17708 + }, + { + "epoch": 0.17709, + "grad_norm": 1.096747875213623, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 17709 + }, + { + "epoch": 0.1771, + "grad_norm": 1.0291926860809326, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 17710 + }, + { + "epoch": 0.17711, + "grad_norm": 1.0713984966278076, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 17711 + }, + { + "epoch": 0.17712, + "grad_norm": 0.9482304453849792, + "learning_rate": 0.003, + "loss": 3.978, + "step": 17712 + }, + { + "epoch": 0.17713, + "grad_norm": 0.9021075963973999, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 17713 + }, + { + "epoch": 0.17714, + "grad_norm": 0.9854664206504822, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 17714 + }, + { + "epoch": 0.17715, + "grad_norm": 1.0893782377243042, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 17715 + }, + { + "epoch": 0.17716, + "grad_norm": 0.9051546454429626, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 17716 + }, + { + "epoch": 0.17717, + "grad_norm": 0.9802340865135193, + "learning_rate": 0.003, + "loss": 4.041, + "step": 17717 + }, + { + "epoch": 0.17718, + "grad_norm": 0.7992175817489624, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 17718 + }, + { + "epoch": 0.17719, + "grad_norm": 0.6974555253982544, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 17719 + }, + { + "epoch": 0.1772, + "grad_norm": 0.6435659527778625, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 17720 + }, + { + "epoch": 0.17721, + "grad_norm": 0.6219088435173035, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 17721 + }, + { + "epoch": 0.17722, + "grad_norm": 0.5757900476455688, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 17722 + }, + { + "epoch": 0.17723, + "grad_norm": 0.5606616735458374, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 17723 + }, + { + "epoch": 0.17724, + "grad_norm": 0.6454625725746155, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 17724 + }, + { + "epoch": 0.17725, + "grad_norm": 0.8007955551147461, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 17725 + }, + { + "epoch": 0.17726, + "grad_norm": 1.0168530941009521, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 17726 + }, + { + "epoch": 0.17727, + "grad_norm": 1.1221123933792114, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 17727 + }, + { + "epoch": 0.17728, + "grad_norm": 0.8694456219673157, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 17728 + }, + { + "epoch": 0.17729, + "grad_norm": 0.6848198771476746, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 17729 + }, + { + "epoch": 0.1773, + "grad_norm": 0.6626553535461426, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 17730 + }, + { + "epoch": 0.17731, + "grad_norm": 0.9850336909294128, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 17731 + }, + { + "epoch": 0.17732, + "grad_norm": 1.075743317604065, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 17732 + }, + { + "epoch": 0.17733, + "grad_norm": 0.8826678991317749, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 17733 + }, + { + "epoch": 0.17734, + "grad_norm": 0.829764187335968, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 17734 + }, + { + "epoch": 0.17735, + "grad_norm": 0.62376469373703, + "learning_rate": 0.003, + "loss": 3.998, + "step": 17735 + }, + { + "epoch": 0.17736, + "grad_norm": 0.6764979362487793, + "learning_rate": 0.003, + "loss": 4.007, + "step": 17736 + }, + { + "epoch": 0.17737, + "grad_norm": 0.594389021396637, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 17737 + }, + { + "epoch": 0.17738, + "grad_norm": 0.6828976273536682, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 17738 + }, + { + "epoch": 0.17739, + "grad_norm": 0.7793011665344238, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 17739 + }, + { + "epoch": 0.1774, + "grad_norm": 0.8825653791427612, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 17740 + }, + { + "epoch": 0.17741, + "grad_norm": 0.895940363407135, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 17741 + }, + { + "epoch": 0.17742, + "grad_norm": 0.8027079105377197, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 17742 + }, + { + "epoch": 0.17743, + "grad_norm": 0.7605303525924683, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 17743 + }, + { + "epoch": 0.17744, + "grad_norm": 0.736718475818634, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 17744 + }, + { + "epoch": 0.17745, + "grad_norm": 0.7389856576919556, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 17745 + }, + { + "epoch": 0.17746, + "grad_norm": 0.8453066945075989, + "learning_rate": 0.003, + "loss": 4.014, + "step": 17746 + }, + { + "epoch": 0.17747, + "grad_norm": 1.043473482131958, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 17747 + }, + { + "epoch": 0.17748, + "grad_norm": 1.101148247718811, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 17748 + }, + { + "epoch": 0.17749, + "grad_norm": 0.9800677299499512, + "learning_rate": 0.003, + "loss": 3.982, + "step": 17749 + }, + { + "epoch": 0.1775, + "grad_norm": 1.0185678005218506, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 17750 + }, + { + "epoch": 0.17751, + "grad_norm": 1.0174145698547363, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 17751 + }, + { + "epoch": 0.17752, + "grad_norm": 0.8327456116676331, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 17752 + }, + { + "epoch": 0.17753, + "grad_norm": 0.7818776369094849, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 17753 + }, + { + "epoch": 0.17754, + "grad_norm": 0.8434064388275146, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 17754 + }, + { + "epoch": 0.17755, + "grad_norm": 0.8647765517234802, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 17755 + }, + { + "epoch": 0.17756, + "grad_norm": 0.831998884677887, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 17756 + }, + { + "epoch": 0.17757, + "grad_norm": 0.7533503174781799, + "learning_rate": 0.003, + "loss": 4.018, + "step": 17757 + }, + { + "epoch": 0.17758, + "grad_norm": 0.6916507482528687, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 17758 + }, + { + "epoch": 0.17759, + "grad_norm": 0.5789796710014343, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 17759 + }, + { + "epoch": 0.1776, + "grad_norm": 0.5916829705238342, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 17760 + }, + { + "epoch": 0.17761, + "grad_norm": 0.5936658978462219, + "learning_rate": 0.003, + "loss": 4.001, + "step": 17761 + }, + { + "epoch": 0.17762, + "grad_norm": 0.6573385000228882, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 17762 + }, + { + "epoch": 0.17763, + "grad_norm": 0.7643323540687561, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 17763 + }, + { + "epoch": 0.17764, + "grad_norm": 1.0655878782272339, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 17764 + }, + { + "epoch": 0.17765, + "grad_norm": 1.1750019788742065, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 17765 + }, + { + "epoch": 0.17766, + "grad_norm": 0.6205437779426575, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 17766 + }, + { + "epoch": 0.17767, + "grad_norm": 0.6816059947013855, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 17767 + }, + { + "epoch": 0.17768, + "grad_norm": 0.936151385307312, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 17768 + }, + { + "epoch": 0.17769, + "grad_norm": 0.905048131942749, + "learning_rate": 0.003, + "loss": 4.016, + "step": 17769 + }, + { + "epoch": 0.1777, + "grad_norm": 0.9736447334289551, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 17770 + }, + { + "epoch": 0.17771, + "grad_norm": 0.9936445951461792, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 17771 + }, + { + "epoch": 0.17772, + "grad_norm": 0.8724110722541809, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 17772 + }, + { + "epoch": 0.17773, + "grad_norm": 0.8918551802635193, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 17773 + }, + { + "epoch": 0.17774, + "grad_norm": 1.0295112133026123, + "learning_rate": 0.003, + "loss": 4.016, + "step": 17774 + }, + { + "epoch": 0.17775, + "grad_norm": 1.0086634159088135, + "learning_rate": 0.003, + "loss": 3.976, + "step": 17775 + }, + { + "epoch": 0.17776, + "grad_norm": 0.9048023819923401, + "learning_rate": 0.003, + "loss": 4.012, + "step": 17776 + }, + { + "epoch": 0.17777, + "grad_norm": 0.7482655048370361, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 17777 + }, + { + "epoch": 0.17778, + "grad_norm": 0.7453811168670654, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 17778 + }, + { + "epoch": 0.17779, + "grad_norm": 0.7985710501670837, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 17779 + }, + { + "epoch": 0.1778, + "grad_norm": 0.8785742521286011, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 17780 + }, + { + "epoch": 0.17781, + "grad_norm": 0.9537194967269897, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 17781 + }, + { + "epoch": 0.17782, + "grad_norm": 0.8183233141899109, + "learning_rate": 0.003, + "loss": 4.012, + "step": 17782 + }, + { + "epoch": 0.17783, + "grad_norm": 0.7182443737983704, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 17783 + }, + { + "epoch": 0.17784, + "grad_norm": 0.7940345406532288, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 17784 + }, + { + "epoch": 0.17785, + "grad_norm": 0.6857940554618835, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 17785 + }, + { + "epoch": 0.17786, + "grad_norm": 0.7146942019462585, + "learning_rate": 0.003, + "loss": 4.022, + "step": 17786 + }, + { + "epoch": 0.17787, + "grad_norm": 0.8008255362510681, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 17787 + }, + { + "epoch": 0.17788, + "grad_norm": 0.9737480282783508, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 17788 + }, + { + "epoch": 0.17789, + "grad_norm": 1.1552413702011108, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 17789 + }, + { + "epoch": 0.1779, + "grad_norm": 0.997272253036499, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 17790 + }, + { + "epoch": 0.17791, + "grad_norm": 0.931243896484375, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 17791 + }, + { + "epoch": 0.17792, + "grad_norm": 0.8716450929641724, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 17792 + }, + { + "epoch": 0.17793, + "grad_norm": 0.8943480253219604, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 17793 + }, + { + "epoch": 0.17794, + "grad_norm": 0.9397181868553162, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 17794 + }, + { + "epoch": 0.17795, + "grad_norm": 0.7583839893341064, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 17795 + }, + { + "epoch": 0.17796, + "grad_norm": 0.6900312304496765, + "learning_rate": 0.003, + "loss": 3.983, + "step": 17796 + }, + { + "epoch": 0.17797, + "grad_norm": 0.8090738654136658, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 17797 + }, + { + "epoch": 0.17798, + "grad_norm": 0.8548142910003662, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 17798 + }, + { + "epoch": 0.17799, + "grad_norm": 0.8663538694381714, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 17799 + }, + { + "epoch": 0.178, + "grad_norm": 0.9889453649520874, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 17800 + }, + { + "epoch": 0.17801, + "grad_norm": 1.1211903095245361, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 17801 + }, + { + "epoch": 0.17802, + "grad_norm": 0.9584589004516602, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 17802 + }, + { + "epoch": 0.17803, + "grad_norm": 1.0522129535675049, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 17803 + }, + { + "epoch": 0.17804, + "grad_norm": 0.9472935795783997, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 17804 + }, + { + "epoch": 0.17805, + "grad_norm": 0.9455471038818359, + "learning_rate": 0.003, + "loss": 4.022, + "step": 17805 + }, + { + "epoch": 0.17806, + "grad_norm": 0.8628377914428711, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 17806 + }, + { + "epoch": 0.17807, + "grad_norm": 0.7941399812698364, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 17807 + }, + { + "epoch": 0.17808, + "grad_norm": 0.8009958863258362, + "learning_rate": 0.003, + "loss": 3.978, + "step": 17808 + }, + { + "epoch": 0.17809, + "grad_norm": 0.8584080338478088, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 17809 + }, + { + "epoch": 0.1781, + "grad_norm": 0.7621896862983704, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 17810 + }, + { + "epoch": 0.17811, + "grad_norm": 0.7111849188804626, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 17811 + }, + { + "epoch": 0.17812, + "grad_norm": 0.7301157712936401, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 17812 + }, + { + "epoch": 0.17813, + "grad_norm": 0.7553697228431702, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 17813 + }, + { + "epoch": 0.17814, + "grad_norm": 0.9111964106559753, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 17814 + }, + { + "epoch": 0.17815, + "grad_norm": 1.1202391386032104, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 17815 + }, + { + "epoch": 0.17816, + "grad_norm": 1.1454354524612427, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 17816 + }, + { + "epoch": 0.17817, + "grad_norm": 0.9268150329589844, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 17817 + }, + { + "epoch": 0.17818, + "grad_norm": 0.9023136496543884, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 17818 + }, + { + "epoch": 0.17819, + "grad_norm": 0.8537625670433044, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 17819 + }, + { + "epoch": 0.1782, + "grad_norm": 0.8082426190376282, + "learning_rate": 0.003, + "loss": 4.032, + "step": 17820 + }, + { + "epoch": 0.17821, + "grad_norm": 0.753790020942688, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 17821 + }, + { + "epoch": 0.17822, + "grad_norm": 0.6044871807098389, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 17822 + }, + { + "epoch": 0.17823, + "grad_norm": 0.5598384737968445, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 17823 + }, + { + "epoch": 0.17824, + "grad_norm": 0.5412915945053101, + "learning_rate": 0.003, + "loss": 3.98, + "step": 17824 + }, + { + "epoch": 0.17825, + "grad_norm": 0.4867744445800781, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 17825 + }, + { + "epoch": 0.17826, + "grad_norm": 0.5870778560638428, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 17826 + }, + { + "epoch": 0.17827, + "grad_norm": 0.8643435835838318, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 17827 + }, + { + "epoch": 0.17828, + "grad_norm": 1.3006690740585327, + "learning_rate": 0.003, + "loss": 4.007, + "step": 17828 + }, + { + "epoch": 0.17829, + "grad_norm": 0.7963849902153015, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 17829 + }, + { + "epoch": 0.1783, + "grad_norm": 0.6619006991386414, + "learning_rate": 0.003, + "loss": 3.958, + "step": 17830 + }, + { + "epoch": 0.17831, + "grad_norm": 0.6422700881958008, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 17831 + }, + { + "epoch": 0.17832, + "grad_norm": 0.5084357261657715, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 17832 + }, + { + "epoch": 0.17833, + "grad_norm": 0.5677739977836609, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 17833 + }, + { + "epoch": 0.17834, + "grad_norm": 0.5317157506942749, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 17834 + }, + { + "epoch": 0.17835, + "grad_norm": 0.5522458553314209, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 17835 + }, + { + "epoch": 0.17836, + "grad_norm": 0.6171836853027344, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 17836 + }, + { + "epoch": 0.17837, + "grad_norm": 0.7240206003189087, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 17837 + }, + { + "epoch": 0.17838, + "grad_norm": 0.9882445335388184, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 17838 + }, + { + "epoch": 0.17839, + "grad_norm": 1.2277603149414062, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 17839 + }, + { + "epoch": 0.1784, + "grad_norm": 0.5247293710708618, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 17840 + }, + { + "epoch": 0.17841, + "grad_norm": 0.6750937104225159, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 17841 + }, + { + "epoch": 0.17842, + "grad_norm": 0.897888720035553, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 17842 + }, + { + "epoch": 0.17843, + "grad_norm": 0.9377698302268982, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 17843 + }, + { + "epoch": 0.17844, + "grad_norm": 0.974871039390564, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 17844 + }, + { + "epoch": 0.17845, + "grad_norm": 0.8839754462242126, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 17845 + }, + { + "epoch": 0.17846, + "grad_norm": 0.772807240486145, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 17846 + }, + { + "epoch": 0.17847, + "grad_norm": 0.715358316898346, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 17847 + }, + { + "epoch": 0.17848, + "grad_norm": 0.8007568120956421, + "learning_rate": 0.003, + "loss": 4.001, + "step": 17848 + }, + { + "epoch": 0.17849, + "grad_norm": 0.9028872847557068, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 17849 + }, + { + "epoch": 0.1785, + "grad_norm": 0.976070761680603, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 17850 + }, + { + "epoch": 0.17851, + "grad_norm": 1.1514437198638916, + "learning_rate": 0.003, + "loss": 4.032, + "step": 17851 + }, + { + "epoch": 0.17852, + "grad_norm": 0.7972047328948975, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 17852 + }, + { + "epoch": 0.17853, + "grad_norm": 0.6677350401878357, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 17853 + }, + { + "epoch": 0.17854, + "grad_norm": 0.8710585832595825, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 17854 + }, + { + "epoch": 0.17855, + "grad_norm": 1.0941059589385986, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 17855 + }, + { + "epoch": 0.17856, + "grad_norm": 0.9320690631866455, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 17856 + }, + { + "epoch": 0.17857, + "grad_norm": 1.0536068677902222, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 17857 + }, + { + "epoch": 0.17858, + "grad_norm": 0.9712250828742981, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 17858 + }, + { + "epoch": 0.17859, + "grad_norm": 0.8574034571647644, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 17859 + }, + { + "epoch": 0.1786, + "grad_norm": 0.8647499084472656, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 17860 + }, + { + "epoch": 0.17861, + "grad_norm": 0.7013272047042847, + "learning_rate": 0.003, + "loss": 4.037, + "step": 17861 + }, + { + "epoch": 0.17862, + "grad_norm": 0.746783971786499, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 17862 + }, + { + "epoch": 0.17863, + "grad_norm": 0.8463944792747498, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 17863 + }, + { + "epoch": 0.17864, + "grad_norm": 0.998805582523346, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 17864 + }, + { + "epoch": 0.17865, + "grad_norm": 0.8989658951759338, + "learning_rate": 0.003, + "loss": 3.989, + "step": 17865 + }, + { + "epoch": 0.17866, + "grad_norm": 0.8039606213569641, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 17866 + }, + { + "epoch": 0.17867, + "grad_norm": 0.8249936699867249, + "learning_rate": 0.003, + "loss": 4.017, + "step": 17867 + }, + { + "epoch": 0.17868, + "grad_norm": 0.9631441235542297, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 17868 + }, + { + "epoch": 0.17869, + "grad_norm": 1.0489217042922974, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 17869 + }, + { + "epoch": 0.1787, + "grad_norm": 1.0523375272750854, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 17870 + }, + { + "epoch": 0.17871, + "grad_norm": 1.0117212533950806, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 17871 + }, + { + "epoch": 0.17872, + "grad_norm": 0.9836920499801636, + "learning_rate": 0.003, + "loss": 3.999, + "step": 17872 + }, + { + "epoch": 0.17873, + "grad_norm": 0.9232009649276733, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 17873 + }, + { + "epoch": 0.17874, + "grad_norm": 0.9044313430786133, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 17874 + }, + { + "epoch": 0.17875, + "grad_norm": 0.9934422373771667, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 17875 + }, + { + "epoch": 0.17876, + "grad_norm": 0.9941718578338623, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 17876 + }, + { + "epoch": 0.17877, + "grad_norm": 0.8592278957366943, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 17877 + }, + { + "epoch": 0.17878, + "grad_norm": 0.774692714214325, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 17878 + }, + { + "epoch": 0.17879, + "grad_norm": 0.8519881367683411, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 17879 + }, + { + "epoch": 0.1788, + "grad_norm": 0.853733479976654, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 17880 + }, + { + "epoch": 0.17881, + "grad_norm": 0.8926007151603699, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 17881 + }, + { + "epoch": 0.17882, + "grad_norm": 1.009367823600769, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 17882 + }, + { + "epoch": 0.17883, + "grad_norm": 1.2331897020339966, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 17883 + }, + { + "epoch": 0.17884, + "grad_norm": 0.7124277949333191, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 17884 + }, + { + "epoch": 0.17885, + "grad_norm": 0.6617794632911682, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 17885 + }, + { + "epoch": 0.17886, + "grad_norm": 0.6539735794067383, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 17886 + }, + { + "epoch": 0.17887, + "grad_norm": 0.658829391002655, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 17887 + }, + { + "epoch": 0.17888, + "grad_norm": 0.621078610420227, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 17888 + }, + { + "epoch": 0.17889, + "grad_norm": 0.6504934430122375, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 17889 + }, + { + "epoch": 0.1789, + "grad_norm": 0.7506583333015442, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 17890 + }, + { + "epoch": 0.17891, + "grad_norm": 0.850108802318573, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 17891 + }, + { + "epoch": 0.17892, + "grad_norm": 0.8623435497283936, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 17892 + }, + { + "epoch": 0.17893, + "grad_norm": 0.9529052376747131, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 17893 + }, + { + "epoch": 0.17894, + "grad_norm": 1.041895866394043, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 17894 + }, + { + "epoch": 0.17895, + "grad_norm": 0.9011391401290894, + "learning_rate": 0.003, + "loss": 4.035, + "step": 17895 + }, + { + "epoch": 0.17896, + "grad_norm": 0.7943750619888306, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 17896 + }, + { + "epoch": 0.17897, + "grad_norm": 0.7554261088371277, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 17897 + }, + { + "epoch": 0.17898, + "grad_norm": 0.7515415549278259, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 17898 + }, + { + "epoch": 0.17899, + "grad_norm": 0.7488419413566589, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 17899 + }, + { + "epoch": 0.179, + "grad_norm": 0.6361393928527832, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 17900 + }, + { + "epoch": 0.17901, + "grad_norm": 0.6696352958679199, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 17901 + }, + { + "epoch": 0.17902, + "grad_norm": 0.6741111874580383, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 17902 + }, + { + "epoch": 0.17903, + "grad_norm": 0.7678008079528809, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 17903 + }, + { + "epoch": 0.17904, + "grad_norm": 0.9445978403091431, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 17904 + }, + { + "epoch": 0.17905, + "grad_norm": 1.1481324434280396, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 17905 + }, + { + "epoch": 0.17906, + "grad_norm": 0.8214532136917114, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 17906 + }, + { + "epoch": 0.17907, + "grad_norm": 0.7733437418937683, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 17907 + }, + { + "epoch": 0.17908, + "grad_norm": 0.6816405653953552, + "learning_rate": 0.003, + "loss": 4.012, + "step": 17908 + }, + { + "epoch": 0.17909, + "grad_norm": 0.6628265976905823, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 17909 + }, + { + "epoch": 0.1791, + "grad_norm": 0.69963538646698, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 17910 + }, + { + "epoch": 0.17911, + "grad_norm": 0.7206888794898987, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 17911 + }, + { + "epoch": 0.17912, + "grad_norm": 0.6892799139022827, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 17912 + }, + { + "epoch": 0.17913, + "grad_norm": 0.6421831250190735, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 17913 + }, + { + "epoch": 0.17914, + "grad_norm": 0.793516993522644, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 17914 + }, + { + "epoch": 0.17915, + "grad_norm": 0.6671417355537415, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 17915 + }, + { + "epoch": 0.17916, + "grad_norm": 0.6116956472396851, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 17916 + }, + { + "epoch": 0.17917, + "grad_norm": 0.7654872536659241, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 17917 + }, + { + "epoch": 0.17918, + "grad_norm": 1.0274306535720825, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 17918 + }, + { + "epoch": 0.17919, + "grad_norm": 1.1414265632629395, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 17919 + }, + { + "epoch": 0.1792, + "grad_norm": 0.8894082903862, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 17920 + }, + { + "epoch": 0.17921, + "grad_norm": 0.8943668007850647, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 17921 + }, + { + "epoch": 0.17922, + "grad_norm": 0.8579666018486023, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 17922 + }, + { + "epoch": 0.17923, + "grad_norm": 0.9284310936927795, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 17923 + }, + { + "epoch": 0.17924, + "grad_norm": 1.0336041450500488, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 17924 + }, + { + "epoch": 0.17925, + "grad_norm": 1.0972650051116943, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 17925 + }, + { + "epoch": 0.17926, + "grad_norm": 0.9823690056800842, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 17926 + }, + { + "epoch": 0.17927, + "grad_norm": 1.0724045038223267, + "learning_rate": 0.003, + "loss": 4.025, + "step": 17927 + }, + { + "epoch": 0.17928, + "grad_norm": 0.9181076288223267, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 17928 + }, + { + "epoch": 0.17929, + "grad_norm": 0.8142296075820923, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 17929 + }, + { + "epoch": 0.1793, + "grad_norm": 0.7954555749893188, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 17930 + }, + { + "epoch": 0.17931, + "grad_norm": 0.8270518779754639, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 17931 + }, + { + "epoch": 0.17932, + "grad_norm": 0.8402372002601624, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 17932 + }, + { + "epoch": 0.17933, + "grad_norm": 1.0668981075286865, + "learning_rate": 0.003, + "loss": 3.998, + "step": 17933 + }, + { + "epoch": 0.17934, + "grad_norm": 0.9972578287124634, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 17934 + }, + { + "epoch": 0.17935, + "grad_norm": 0.9289863705635071, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 17935 + }, + { + "epoch": 0.17936, + "grad_norm": 0.8743131160736084, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 17936 + }, + { + "epoch": 0.17937, + "grad_norm": 0.848335325717926, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 17937 + }, + { + "epoch": 0.17938, + "grad_norm": 0.9841867089271545, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 17938 + }, + { + "epoch": 0.17939, + "grad_norm": 1.2921292781829834, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 17939 + }, + { + "epoch": 0.1794, + "grad_norm": 0.7182594537734985, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 17940 + }, + { + "epoch": 0.17941, + "grad_norm": 0.5401883721351624, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 17941 + }, + { + "epoch": 0.17942, + "grad_norm": 0.6777501106262207, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 17942 + }, + { + "epoch": 0.17943, + "grad_norm": 0.7719506621360779, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 17943 + }, + { + "epoch": 0.17944, + "grad_norm": 0.9649561643600464, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 17944 + }, + { + "epoch": 0.17945, + "grad_norm": 1.1592811346054077, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 17945 + }, + { + "epoch": 0.17946, + "grad_norm": 0.7516922950744629, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 17946 + }, + { + "epoch": 0.17947, + "grad_norm": 0.6371436715126038, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 17947 + }, + { + "epoch": 0.17948, + "grad_norm": 0.6841728091239929, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 17948 + }, + { + "epoch": 0.17949, + "grad_norm": 0.7224218249320984, + "learning_rate": 0.003, + "loss": 4.006, + "step": 17949 + }, + { + "epoch": 0.1795, + "grad_norm": 0.7025465965270996, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 17950 + }, + { + "epoch": 0.17951, + "grad_norm": 0.7663628458976746, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 17951 + }, + { + "epoch": 0.17952, + "grad_norm": 0.7727324366569519, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 17952 + }, + { + "epoch": 0.17953, + "grad_norm": 0.750734269618988, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 17953 + }, + { + "epoch": 0.17954, + "grad_norm": 0.7506475448608398, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 17954 + }, + { + "epoch": 0.17955, + "grad_norm": 0.7635833024978638, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 17955 + }, + { + "epoch": 0.17956, + "grad_norm": 0.7833163738250732, + "learning_rate": 0.003, + "loss": 3.981, + "step": 17956 + }, + { + "epoch": 0.17957, + "grad_norm": 0.7768687605857849, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 17957 + }, + { + "epoch": 0.17958, + "grad_norm": 0.8088453412055969, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 17958 + }, + { + "epoch": 0.17959, + "grad_norm": 0.8046279549598694, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 17959 + }, + { + "epoch": 0.1796, + "grad_norm": 0.886318564414978, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 17960 + }, + { + "epoch": 0.17961, + "grad_norm": 1.0145245790481567, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 17961 + }, + { + "epoch": 0.17962, + "grad_norm": 1.001228928565979, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 17962 + }, + { + "epoch": 0.17963, + "grad_norm": 0.8505359292030334, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 17963 + }, + { + "epoch": 0.17964, + "grad_norm": 0.8970061540603638, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 17964 + }, + { + "epoch": 0.17965, + "grad_norm": 0.9306725263595581, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 17965 + }, + { + "epoch": 0.17966, + "grad_norm": 1.0478090047836304, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 17966 + }, + { + "epoch": 0.17967, + "grad_norm": 1.050741195678711, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 17967 + }, + { + "epoch": 0.17968, + "grad_norm": 0.9878668785095215, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 17968 + }, + { + "epoch": 0.17969, + "grad_norm": 1.103552222251892, + "learning_rate": 0.003, + "loss": 4.013, + "step": 17969 + }, + { + "epoch": 0.1797, + "grad_norm": 1.0772671699523926, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 17970 + }, + { + "epoch": 0.17971, + "grad_norm": 0.7761634588241577, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 17971 + }, + { + "epoch": 0.17972, + "grad_norm": 0.9355222582817078, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 17972 + }, + { + "epoch": 0.17973, + "grad_norm": 1.0587787628173828, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 17973 + }, + { + "epoch": 0.17974, + "grad_norm": 1.0033667087554932, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 17974 + }, + { + "epoch": 0.17975, + "grad_norm": 0.9196949601173401, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 17975 + }, + { + "epoch": 0.17976, + "grad_norm": 0.8534526824951172, + "learning_rate": 0.003, + "loss": 4.001, + "step": 17976 + }, + { + "epoch": 0.17977, + "grad_norm": 0.9210339188575745, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 17977 + }, + { + "epoch": 0.17978, + "grad_norm": 0.9203281402587891, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 17978 + }, + { + "epoch": 0.17979, + "grad_norm": 1.0668666362762451, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 17979 + }, + { + "epoch": 0.1798, + "grad_norm": 1.1201149225234985, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 17980 + }, + { + "epoch": 0.17981, + "grad_norm": 0.7923021912574768, + "learning_rate": 0.003, + "loss": 3.939, + "step": 17981 + }, + { + "epoch": 0.17982, + "grad_norm": 0.5935437679290771, + "learning_rate": 0.003, + "loss": 4.011, + "step": 17982 + }, + { + "epoch": 0.17983, + "grad_norm": 0.5846838355064392, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 17983 + }, + { + "epoch": 0.17984, + "grad_norm": 0.43343862891197205, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 17984 + }, + { + "epoch": 0.17985, + "grad_norm": 0.5557079315185547, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 17985 + }, + { + "epoch": 0.17986, + "grad_norm": 0.5190477967262268, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 17986 + }, + { + "epoch": 0.17987, + "grad_norm": 0.5365965962409973, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 17987 + }, + { + "epoch": 0.17988, + "grad_norm": 0.5814077258110046, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 17988 + }, + { + "epoch": 0.17989, + "grad_norm": 0.6688368320465088, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 17989 + }, + { + "epoch": 0.1799, + "grad_norm": 0.8527092933654785, + "learning_rate": 0.003, + "loss": 4.028, + "step": 17990 + }, + { + "epoch": 0.17991, + "grad_norm": 1.0361605882644653, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 17991 + }, + { + "epoch": 0.17992, + "grad_norm": 1.0612369775772095, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 17992 + }, + { + "epoch": 0.17993, + "grad_norm": 0.8605156540870667, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 17993 + }, + { + "epoch": 0.17994, + "grad_norm": 0.6299520134925842, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 17994 + }, + { + "epoch": 0.17995, + "grad_norm": 0.6558441519737244, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 17995 + }, + { + "epoch": 0.17996, + "grad_norm": 0.646214485168457, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 17996 + }, + { + "epoch": 0.17997, + "grad_norm": 0.685188353061676, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 17997 + }, + { + "epoch": 0.17998, + "grad_norm": 0.7944631576538086, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 17998 + }, + { + "epoch": 0.17999, + "grad_norm": 0.9150182604789734, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 17999 + }, + { + "epoch": 0.18, + "grad_norm": 1.0325008630752563, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 18000 + }, + { + "epoch": 0.18001, + "grad_norm": 0.9318733215332031, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 18001 + }, + { + "epoch": 0.18002, + "grad_norm": 0.8521453738212585, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 18002 + }, + { + "epoch": 0.18003, + "grad_norm": 0.7938467860221863, + "learning_rate": 0.003, + "loss": 4.003, + "step": 18003 + }, + { + "epoch": 0.18004, + "grad_norm": 0.9001320004463196, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 18004 + }, + { + "epoch": 0.18005, + "grad_norm": 0.9790751338005066, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 18005 + }, + { + "epoch": 0.18006, + "grad_norm": 1.0301158428192139, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 18006 + }, + { + "epoch": 0.18007, + "grad_norm": 1.0088856220245361, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 18007 + }, + { + "epoch": 0.18008, + "grad_norm": 0.986687958240509, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 18008 + }, + { + "epoch": 0.18009, + "grad_norm": 0.918765127658844, + "learning_rate": 0.003, + "loss": 3.982, + "step": 18009 + }, + { + "epoch": 0.1801, + "grad_norm": 0.8073814511299133, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 18010 + }, + { + "epoch": 0.18011, + "grad_norm": 1.0299934148788452, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 18011 + }, + { + "epoch": 0.18012, + "grad_norm": 1.068831205368042, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 18012 + }, + { + "epoch": 0.18013, + "grad_norm": 1.015882968902588, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 18013 + }, + { + "epoch": 0.18014, + "grad_norm": 0.8612130880355835, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 18014 + }, + { + "epoch": 0.18015, + "grad_norm": 0.8178862929344177, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 18015 + }, + { + "epoch": 0.18016, + "grad_norm": 0.9154928922653198, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 18016 + }, + { + "epoch": 0.18017, + "grad_norm": 1.0131648778915405, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 18017 + }, + { + "epoch": 0.18018, + "grad_norm": 0.968364417552948, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 18018 + }, + { + "epoch": 0.18019, + "grad_norm": 1.172088623046875, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 18019 + }, + { + "epoch": 0.1802, + "grad_norm": 0.922442615032196, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 18020 + }, + { + "epoch": 0.18021, + "grad_norm": 0.8483938574790955, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 18021 + }, + { + "epoch": 0.18022, + "grad_norm": 0.8102627992630005, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 18022 + }, + { + "epoch": 0.18023, + "grad_norm": 0.5910297632217407, + "learning_rate": 0.003, + "loss": 3.987, + "step": 18023 + }, + { + "epoch": 0.18024, + "grad_norm": 0.6116005778312683, + "learning_rate": 0.003, + "loss": 3.993, + "step": 18024 + }, + { + "epoch": 0.18025, + "grad_norm": 0.5381456017494202, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 18025 + }, + { + "epoch": 0.18026, + "grad_norm": 0.5271363854408264, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 18026 + }, + { + "epoch": 0.18027, + "grad_norm": 0.49178123474121094, + "learning_rate": 0.003, + "loss": 3.995, + "step": 18027 + }, + { + "epoch": 0.18028, + "grad_norm": 0.5047873258590698, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 18028 + }, + { + "epoch": 0.18029, + "grad_norm": 0.5563161373138428, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 18029 + }, + { + "epoch": 0.1803, + "grad_norm": 0.6534048318862915, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 18030 + }, + { + "epoch": 0.18031, + "grad_norm": 0.7242395877838135, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 18031 + }, + { + "epoch": 0.18032, + "grad_norm": 0.7478570938110352, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 18032 + }, + { + "epoch": 0.18033, + "grad_norm": 0.8335338830947876, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 18033 + }, + { + "epoch": 0.18034, + "grad_norm": 0.7908081412315369, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 18034 + }, + { + "epoch": 0.18035, + "grad_norm": 0.7753511071205139, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 18035 + }, + { + "epoch": 0.18036, + "grad_norm": 0.9649012088775635, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 18036 + }, + { + "epoch": 0.18037, + "grad_norm": 1.1775555610656738, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 18037 + }, + { + "epoch": 0.18038, + "grad_norm": 0.8960131406784058, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 18038 + }, + { + "epoch": 0.18039, + "grad_norm": 0.8114643692970276, + "learning_rate": 0.003, + "loss": 3.998, + "step": 18039 + }, + { + "epoch": 0.1804, + "grad_norm": 0.8493918776512146, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 18040 + }, + { + "epoch": 0.18041, + "grad_norm": 0.8604472279548645, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 18041 + }, + { + "epoch": 0.18042, + "grad_norm": 0.9029722213745117, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 18042 + }, + { + "epoch": 0.18043, + "grad_norm": 0.9265522360801697, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 18043 + }, + { + "epoch": 0.18044, + "grad_norm": 0.9179112911224365, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 18044 + }, + { + "epoch": 0.18045, + "grad_norm": 1.0061774253845215, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 18045 + }, + { + "epoch": 0.18046, + "grad_norm": 0.8425312638282776, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 18046 + }, + { + "epoch": 0.18047, + "grad_norm": 0.7756473422050476, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 18047 + }, + { + "epoch": 0.18048, + "grad_norm": 0.814687192440033, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 18048 + }, + { + "epoch": 0.18049, + "grad_norm": 0.8933665752410889, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 18049 + }, + { + "epoch": 0.1805, + "grad_norm": 0.9201660752296448, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 18050 + }, + { + "epoch": 0.18051, + "grad_norm": 0.955101728439331, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 18051 + }, + { + "epoch": 0.18052, + "grad_norm": 0.8565345406532288, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 18052 + }, + { + "epoch": 0.18053, + "grad_norm": 0.8543877601623535, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 18053 + }, + { + "epoch": 0.18054, + "grad_norm": 0.8501734733581543, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 18054 + }, + { + "epoch": 0.18055, + "grad_norm": 0.8169307708740234, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 18055 + }, + { + "epoch": 0.18056, + "grad_norm": 0.7411114573478699, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 18056 + }, + { + "epoch": 0.18057, + "grad_norm": 0.7389774918556213, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 18057 + }, + { + "epoch": 0.18058, + "grad_norm": 0.8780634999275208, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 18058 + }, + { + "epoch": 0.18059, + "grad_norm": 0.9774113893508911, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 18059 + }, + { + "epoch": 0.1806, + "grad_norm": 1.1417102813720703, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 18060 + }, + { + "epoch": 0.18061, + "grad_norm": 0.9026036262512207, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 18061 + }, + { + "epoch": 0.18062, + "grad_norm": 1.0317175388336182, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 18062 + }, + { + "epoch": 0.18063, + "grad_norm": 0.9787594676017761, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 18063 + }, + { + "epoch": 0.18064, + "grad_norm": 0.8721433877944946, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 18064 + }, + { + "epoch": 0.18065, + "grad_norm": 0.8833760023117065, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 18065 + }, + { + "epoch": 0.18066, + "grad_norm": 0.8670377731323242, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 18066 + }, + { + "epoch": 0.18067, + "grad_norm": 0.8080344200134277, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 18067 + }, + { + "epoch": 0.18068, + "grad_norm": 0.9190772175788879, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 18068 + }, + { + "epoch": 0.18069, + "grad_norm": 0.9776795506477356, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 18069 + }, + { + "epoch": 0.1807, + "grad_norm": 0.9771139621734619, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 18070 + }, + { + "epoch": 0.18071, + "grad_norm": 1.0068480968475342, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 18071 + }, + { + "epoch": 0.18072, + "grad_norm": 0.9160780310630798, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 18072 + }, + { + "epoch": 0.18073, + "grad_norm": 0.7548110485076904, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 18073 + }, + { + "epoch": 0.18074, + "grad_norm": 0.6938346028327942, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 18074 + }, + { + "epoch": 0.18075, + "grad_norm": 0.6407344937324524, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 18075 + }, + { + "epoch": 0.18076, + "grad_norm": 0.5780854225158691, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 18076 + }, + { + "epoch": 0.18077, + "grad_norm": 0.6665111184120178, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 18077 + }, + { + "epoch": 0.18078, + "grad_norm": 0.7108888030052185, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 18078 + }, + { + "epoch": 0.18079, + "grad_norm": 0.7258763313293457, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 18079 + }, + { + "epoch": 0.1808, + "grad_norm": 0.7347949147224426, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 18080 + }, + { + "epoch": 0.18081, + "grad_norm": 0.6717410087585449, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 18081 + }, + { + "epoch": 0.18082, + "grad_norm": 0.5788361430168152, + "learning_rate": 0.003, + "loss": 3.995, + "step": 18082 + }, + { + "epoch": 0.18083, + "grad_norm": 0.5914314389228821, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 18083 + }, + { + "epoch": 0.18084, + "grad_norm": 0.6312091946601868, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 18084 + }, + { + "epoch": 0.18085, + "grad_norm": 0.6804765462875366, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 18085 + }, + { + "epoch": 0.18086, + "grad_norm": 0.6745128035545349, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 18086 + }, + { + "epoch": 0.18087, + "grad_norm": 0.6256589293479919, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 18087 + }, + { + "epoch": 0.18088, + "grad_norm": 0.5829675793647766, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 18088 + }, + { + "epoch": 0.18089, + "grad_norm": 0.7264362573623657, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 18089 + }, + { + "epoch": 0.1809, + "grad_norm": 1.0110256671905518, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 18090 + }, + { + "epoch": 0.18091, + "grad_norm": 1.305204153060913, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 18091 + }, + { + "epoch": 0.18092, + "grad_norm": 0.9151637554168701, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 18092 + }, + { + "epoch": 0.18093, + "grad_norm": 0.8778868913650513, + "learning_rate": 0.003, + "loss": 4.026, + "step": 18093 + }, + { + "epoch": 0.18094, + "grad_norm": 0.7514395713806152, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 18094 + }, + { + "epoch": 0.18095, + "grad_norm": 0.9953643679618835, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 18095 + }, + { + "epoch": 0.18096, + "grad_norm": 1.349854826927185, + "learning_rate": 0.003, + "loss": 3.999, + "step": 18096 + }, + { + "epoch": 0.18097, + "grad_norm": 0.6634512543678284, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 18097 + }, + { + "epoch": 0.18098, + "grad_norm": 0.746289849281311, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 18098 + }, + { + "epoch": 0.18099, + "grad_norm": 0.7250983715057373, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 18099 + }, + { + "epoch": 0.181, + "grad_norm": 0.723181962966919, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 18100 + }, + { + "epoch": 0.18101, + "grad_norm": 0.7149516940116882, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 18101 + }, + { + "epoch": 0.18102, + "grad_norm": 0.8280004262924194, + "learning_rate": 0.003, + "loss": 4.005, + "step": 18102 + }, + { + "epoch": 0.18103, + "grad_norm": 0.8888851404190063, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 18103 + }, + { + "epoch": 0.18104, + "grad_norm": 0.9927442669868469, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 18104 + }, + { + "epoch": 0.18105, + "grad_norm": 1.3739959001541138, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 18105 + }, + { + "epoch": 0.18106, + "grad_norm": 0.5654383897781372, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 18106 + }, + { + "epoch": 0.18107, + "grad_norm": 0.6091672778129578, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 18107 + }, + { + "epoch": 0.18108, + "grad_norm": 0.6785678267478943, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 18108 + }, + { + "epoch": 0.18109, + "grad_norm": 0.7807140946388245, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 18109 + }, + { + "epoch": 0.1811, + "grad_norm": 0.95677250623703, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 18110 + }, + { + "epoch": 0.18111, + "grad_norm": 1.0871576070785522, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 18111 + }, + { + "epoch": 0.18112, + "grad_norm": 0.8533523082733154, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 18112 + }, + { + "epoch": 0.18113, + "grad_norm": 0.7580128908157349, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 18113 + }, + { + "epoch": 0.18114, + "grad_norm": 0.8775519132614136, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 18114 + }, + { + "epoch": 0.18115, + "grad_norm": 0.9075474739074707, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 18115 + }, + { + "epoch": 0.18116, + "grad_norm": 0.8368702530860901, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 18116 + }, + { + "epoch": 0.18117, + "grad_norm": 0.9043243527412415, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 18117 + }, + { + "epoch": 0.18118, + "grad_norm": 0.9974503517150879, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 18118 + }, + { + "epoch": 0.18119, + "grad_norm": 0.9757204055786133, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 18119 + }, + { + "epoch": 0.1812, + "grad_norm": 1.045078158378601, + "learning_rate": 0.003, + "loss": 4.009, + "step": 18120 + }, + { + "epoch": 0.18121, + "grad_norm": 1.0194590091705322, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 18121 + }, + { + "epoch": 0.18122, + "grad_norm": 0.9902092218399048, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 18122 + }, + { + "epoch": 0.18123, + "grad_norm": 0.964513897895813, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 18123 + }, + { + "epoch": 0.18124, + "grad_norm": 0.9905357360839844, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 18124 + }, + { + "epoch": 0.18125, + "grad_norm": 1.0209734439849854, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 18125 + }, + { + "epoch": 0.18126, + "grad_norm": 0.9319140315055847, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 18126 + }, + { + "epoch": 0.18127, + "grad_norm": 0.9061564207077026, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 18127 + }, + { + "epoch": 0.18128, + "grad_norm": 0.8320205807685852, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 18128 + }, + { + "epoch": 0.18129, + "grad_norm": 0.77691650390625, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 18129 + }, + { + "epoch": 0.1813, + "grad_norm": 0.6979191303253174, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 18130 + }, + { + "epoch": 0.18131, + "grad_norm": 0.8580296635627747, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 18131 + }, + { + "epoch": 0.18132, + "grad_norm": 1.3209208250045776, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 18132 + }, + { + "epoch": 0.18133, + "grad_norm": 0.8732667565345764, + "learning_rate": 0.003, + "loss": 4.012, + "step": 18133 + }, + { + "epoch": 0.18134, + "grad_norm": 0.7871861457824707, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 18134 + }, + { + "epoch": 0.18135, + "grad_norm": 0.7526431083679199, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 18135 + }, + { + "epoch": 0.18136, + "grad_norm": 0.604056179523468, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 18136 + }, + { + "epoch": 0.18137, + "grad_norm": 0.5123692154884338, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 18137 + }, + { + "epoch": 0.18138, + "grad_norm": 0.5805838108062744, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 18138 + }, + { + "epoch": 0.18139, + "grad_norm": 0.7583757042884827, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 18139 + }, + { + "epoch": 0.1814, + "grad_norm": 1.0473604202270508, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 18140 + }, + { + "epoch": 0.18141, + "grad_norm": 1.088240385055542, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 18141 + }, + { + "epoch": 0.18142, + "grad_norm": 0.8162732720375061, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 18142 + }, + { + "epoch": 0.18143, + "grad_norm": 0.6132739782333374, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 18143 + }, + { + "epoch": 0.18144, + "grad_norm": 0.5057741403579712, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 18144 + }, + { + "epoch": 0.18145, + "grad_norm": 0.6503153443336487, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 18145 + }, + { + "epoch": 0.18146, + "grad_norm": 0.7001628279685974, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 18146 + }, + { + "epoch": 0.18147, + "grad_norm": 0.6886793375015259, + "learning_rate": 0.003, + "loss": 3.996, + "step": 18147 + }, + { + "epoch": 0.18148, + "grad_norm": 0.6724410057067871, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 18148 + }, + { + "epoch": 0.18149, + "grad_norm": 0.7682032585144043, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 18149 + }, + { + "epoch": 0.1815, + "grad_norm": 0.9040877819061279, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 18150 + }, + { + "epoch": 0.18151, + "grad_norm": 1.0435184240341187, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 18151 + }, + { + "epoch": 0.18152, + "grad_norm": 0.9554518461227417, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 18152 + }, + { + "epoch": 0.18153, + "grad_norm": 0.8426952362060547, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 18153 + }, + { + "epoch": 0.18154, + "grad_norm": 0.7699375152587891, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 18154 + }, + { + "epoch": 0.18155, + "grad_norm": 0.7254806756973267, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 18155 + }, + { + "epoch": 0.18156, + "grad_norm": 0.6597497463226318, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 18156 + }, + { + "epoch": 0.18157, + "grad_norm": 0.8115432262420654, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 18157 + }, + { + "epoch": 0.18158, + "grad_norm": 0.9796165823936462, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 18158 + }, + { + "epoch": 0.18159, + "grad_norm": 0.8970003724098206, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 18159 + }, + { + "epoch": 0.1816, + "grad_norm": 0.902718722820282, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 18160 + }, + { + "epoch": 0.18161, + "grad_norm": 0.9734975695610046, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 18161 + }, + { + "epoch": 0.18162, + "grad_norm": 0.8031430840492249, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 18162 + }, + { + "epoch": 0.18163, + "grad_norm": 0.7273565530776978, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 18163 + }, + { + "epoch": 0.18164, + "grad_norm": 0.6953762769699097, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 18164 + }, + { + "epoch": 0.18165, + "grad_norm": 0.7612891793251038, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 18165 + }, + { + "epoch": 0.18166, + "grad_norm": 0.8077008724212646, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 18166 + }, + { + "epoch": 0.18167, + "grad_norm": 0.9646387696266174, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 18167 + }, + { + "epoch": 0.18168, + "grad_norm": 1.1055657863616943, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 18168 + }, + { + "epoch": 0.18169, + "grad_norm": 0.9168370366096497, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 18169 + }, + { + "epoch": 0.1817, + "grad_norm": 0.914101243019104, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 18170 + }, + { + "epoch": 0.18171, + "grad_norm": 1.0902481079101562, + "learning_rate": 0.003, + "loss": 4.022, + "step": 18171 + }, + { + "epoch": 0.18172, + "grad_norm": 0.7929781675338745, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 18172 + }, + { + "epoch": 0.18173, + "grad_norm": 0.6111283898353577, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 18173 + }, + { + "epoch": 0.18174, + "grad_norm": 0.5835784077644348, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 18174 + }, + { + "epoch": 0.18175, + "grad_norm": 0.823890209197998, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 18175 + }, + { + "epoch": 0.18176, + "grad_norm": 0.9427134990692139, + "learning_rate": 0.003, + "loss": 4.011, + "step": 18176 + }, + { + "epoch": 0.18177, + "grad_norm": 0.7927230596542358, + "learning_rate": 0.003, + "loss": 4.015, + "step": 18177 + }, + { + "epoch": 0.18178, + "grad_norm": 0.8689695000648499, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 18178 + }, + { + "epoch": 0.18179, + "grad_norm": 1.0723934173583984, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 18179 + }, + { + "epoch": 0.1818, + "grad_norm": 1.0629585981369019, + "learning_rate": 0.003, + "loss": 4.006, + "step": 18180 + }, + { + "epoch": 0.18181, + "grad_norm": 0.8775352239608765, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 18181 + }, + { + "epoch": 0.18182, + "grad_norm": 0.9303977489471436, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 18182 + }, + { + "epoch": 0.18183, + "grad_norm": 1.0644172430038452, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 18183 + }, + { + "epoch": 0.18184, + "grad_norm": 1.0199772119522095, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 18184 + }, + { + "epoch": 0.18185, + "grad_norm": 0.963274359703064, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 18185 + }, + { + "epoch": 0.18186, + "grad_norm": 0.9673733115196228, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 18186 + }, + { + "epoch": 0.18187, + "grad_norm": 1.017067313194275, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 18187 + }, + { + "epoch": 0.18188, + "grad_norm": 0.920941174030304, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 18188 + }, + { + "epoch": 0.18189, + "grad_norm": 0.8725190162658691, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 18189 + }, + { + "epoch": 0.1819, + "grad_norm": 0.944081723690033, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 18190 + }, + { + "epoch": 0.18191, + "grad_norm": 0.9849206805229187, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 18191 + }, + { + "epoch": 0.18192, + "grad_norm": 1.0786023139953613, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 18192 + }, + { + "epoch": 0.18193, + "grad_norm": 1.1245348453521729, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 18193 + }, + { + "epoch": 0.18194, + "grad_norm": 0.9885661005973816, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 18194 + }, + { + "epoch": 0.18195, + "grad_norm": 1.0349547863006592, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 18195 + }, + { + "epoch": 0.18196, + "grad_norm": 0.8013774156570435, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 18196 + }, + { + "epoch": 0.18197, + "grad_norm": 0.6548697352409363, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 18197 + }, + { + "epoch": 0.18198, + "grad_norm": 0.6461350321769714, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 18198 + }, + { + "epoch": 0.18199, + "grad_norm": 0.6670629978179932, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 18199 + }, + { + "epoch": 0.182, + "grad_norm": 0.6904429793357849, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 18200 + }, + { + "epoch": 0.18201, + "grad_norm": 0.7718185186386108, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 18201 + }, + { + "epoch": 0.18202, + "grad_norm": 0.8350328207015991, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 18202 + }, + { + "epoch": 0.18203, + "grad_norm": 0.9042280316352844, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 18203 + }, + { + "epoch": 0.18204, + "grad_norm": 0.9488075971603394, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 18204 + }, + { + "epoch": 0.18205, + "grad_norm": 0.9429847002029419, + "learning_rate": 0.003, + "loss": 3.976, + "step": 18205 + }, + { + "epoch": 0.18206, + "grad_norm": 0.782609224319458, + "learning_rate": 0.003, + "loss": 3.986, + "step": 18206 + }, + { + "epoch": 0.18207, + "grad_norm": 0.6234448552131653, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 18207 + }, + { + "epoch": 0.18208, + "grad_norm": 0.547356128692627, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 18208 + }, + { + "epoch": 0.18209, + "grad_norm": 0.6364834308624268, + "learning_rate": 0.003, + "loss": 3.986, + "step": 18209 + }, + { + "epoch": 0.1821, + "grad_norm": 0.6573218703269958, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 18210 + }, + { + "epoch": 0.18211, + "grad_norm": 0.726592481136322, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 18211 + }, + { + "epoch": 0.18212, + "grad_norm": 0.7832172513008118, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 18212 + }, + { + "epoch": 0.18213, + "grad_norm": 0.8910562992095947, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 18213 + }, + { + "epoch": 0.18214, + "grad_norm": 0.8994468450546265, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 18214 + }, + { + "epoch": 0.18215, + "grad_norm": 0.7811818718910217, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 18215 + }, + { + "epoch": 0.18216, + "grad_norm": 0.849797248840332, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 18216 + }, + { + "epoch": 0.18217, + "grad_norm": 0.9664706587791443, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 18217 + }, + { + "epoch": 0.18218, + "grad_norm": 1.0398706197738647, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 18218 + }, + { + "epoch": 0.18219, + "grad_norm": 0.838563084602356, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 18219 + }, + { + "epoch": 0.1822, + "grad_norm": 0.7078306078910828, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 18220 + }, + { + "epoch": 0.18221, + "grad_norm": 0.7152316570281982, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 18221 + }, + { + "epoch": 0.18222, + "grad_norm": 0.7453733086585999, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 18222 + }, + { + "epoch": 0.18223, + "grad_norm": 0.7840257287025452, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 18223 + }, + { + "epoch": 0.18224, + "grad_norm": 0.8685861825942993, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 18224 + }, + { + "epoch": 0.18225, + "grad_norm": 1.0737313032150269, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 18225 + }, + { + "epoch": 0.18226, + "grad_norm": 1.0292390584945679, + "learning_rate": 0.003, + "loss": 4.023, + "step": 18226 + }, + { + "epoch": 0.18227, + "grad_norm": 0.7910824418067932, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 18227 + }, + { + "epoch": 0.18228, + "grad_norm": 0.6173461675643921, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 18228 + }, + { + "epoch": 0.18229, + "grad_norm": 0.6397079229354858, + "learning_rate": 0.003, + "loss": 3.982, + "step": 18229 + }, + { + "epoch": 0.1823, + "grad_norm": 0.6707412600517273, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 18230 + }, + { + "epoch": 0.18231, + "grad_norm": 0.7674610018730164, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 18231 + }, + { + "epoch": 0.18232, + "grad_norm": 0.851983904838562, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 18232 + }, + { + "epoch": 0.18233, + "grad_norm": 1.0599827766418457, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 18233 + }, + { + "epoch": 0.18234, + "grad_norm": 1.0854276418685913, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 18234 + }, + { + "epoch": 0.18235, + "grad_norm": 0.8180278539657593, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 18235 + }, + { + "epoch": 0.18236, + "grad_norm": 0.7597306966781616, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 18236 + }, + { + "epoch": 0.18237, + "grad_norm": 0.7044374346733093, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 18237 + }, + { + "epoch": 0.18238, + "grad_norm": 0.8133132457733154, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 18238 + }, + { + "epoch": 0.18239, + "grad_norm": 1.0175007581710815, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 18239 + }, + { + "epoch": 0.1824, + "grad_norm": 1.062023401260376, + "learning_rate": 0.003, + "loss": 3.972, + "step": 18240 + }, + { + "epoch": 0.18241, + "grad_norm": 0.6800371408462524, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 18241 + }, + { + "epoch": 0.18242, + "grad_norm": 0.635016679763794, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 18242 + }, + { + "epoch": 0.18243, + "grad_norm": 0.7265605330467224, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 18243 + }, + { + "epoch": 0.18244, + "grad_norm": 0.7801223397254944, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 18244 + }, + { + "epoch": 0.18245, + "grad_norm": 0.781179666519165, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 18245 + }, + { + "epoch": 0.18246, + "grad_norm": 0.7372574210166931, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 18246 + }, + { + "epoch": 0.18247, + "grad_norm": 0.7139351963996887, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 18247 + }, + { + "epoch": 0.18248, + "grad_norm": 0.7166587114334106, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 18248 + }, + { + "epoch": 0.18249, + "grad_norm": 0.8879257440567017, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 18249 + }, + { + "epoch": 0.1825, + "grad_norm": 1.0072455406188965, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 18250 + }, + { + "epoch": 0.18251, + "grad_norm": 1.1003729104995728, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 18251 + }, + { + "epoch": 0.18252, + "grad_norm": 0.8618923425674438, + "learning_rate": 0.003, + "loss": 4.038, + "step": 18252 + }, + { + "epoch": 0.18253, + "grad_norm": 0.8108559846878052, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 18253 + }, + { + "epoch": 0.18254, + "grad_norm": 0.7113040685653687, + "learning_rate": 0.003, + "loss": 4.011, + "step": 18254 + }, + { + "epoch": 0.18255, + "grad_norm": 0.8643912672996521, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 18255 + }, + { + "epoch": 0.18256, + "grad_norm": 1.0000271797180176, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 18256 + }, + { + "epoch": 0.18257, + "grad_norm": 1.3140206336975098, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 18257 + }, + { + "epoch": 0.18258, + "grad_norm": 0.747545599937439, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 18258 + }, + { + "epoch": 0.18259, + "grad_norm": 0.7669658064842224, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 18259 + }, + { + "epoch": 0.1826, + "grad_norm": 0.7207746505737305, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 18260 + }, + { + "epoch": 0.18261, + "grad_norm": 0.7719569802284241, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 18261 + }, + { + "epoch": 0.18262, + "grad_norm": 0.7602952122688293, + "learning_rate": 0.003, + "loss": 3.999, + "step": 18262 + }, + { + "epoch": 0.18263, + "grad_norm": 0.7253813743591309, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 18263 + }, + { + "epoch": 0.18264, + "grad_norm": 0.6444506645202637, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 18264 + }, + { + "epoch": 0.18265, + "grad_norm": 0.6988993287086487, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 18265 + }, + { + "epoch": 0.18266, + "grad_norm": 0.7080886960029602, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 18266 + }, + { + "epoch": 0.18267, + "grad_norm": 0.7752986550331116, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 18267 + }, + { + "epoch": 0.18268, + "grad_norm": 0.9107593297958374, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 18268 + }, + { + "epoch": 0.18269, + "grad_norm": 1.1095401048660278, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 18269 + }, + { + "epoch": 0.1827, + "grad_norm": 0.9966729283332825, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 18270 + }, + { + "epoch": 0.18271, + "grad_norm": 0.93555748462677, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 18271 + }, + { + "epoch": 0.18272, + "grad_norm": 0.8204525709152222, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 18272 + }, + { + "epoch": 0.18273, + "grad_norm": 0.863254725933075, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 18273 + }, + { + "epoch": 0.18274, + "grad_norm": 0.7821072936058044, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 18274 + }, + { + "epoch": 0.18275, + "grad_norm": 0.8419560194015503, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 18275 + }, + { + "epoch": 0.18276, + "grad_norm": 0.8162816762924194, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 18276 + }, + { + "epoch": 0.18277, + "grad_norm": 0.8232567310333252, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 18277 + }, + { + "epoch": 0.18278, + "grad_norm": 0.8190836310386658, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 18278 + }, + { + "epoch": 0.18279, + "grad_norm": 0.7407211065292358, + "learning_rate": 0.003, + "loss": 4.013, + "step": 18279 + }, + { + "epoch": 0.1828, + "grad_norm": 0.7084978222846985, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 18280 + }, + { + "epoch": 0.18281, + "grad_norm": 0.8958796858787537, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 18281 + }, + { + "epoch": 0.18282, + "grad_norm": 1.1421873569488525, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 18282 + }, + { + "epoch": 0.18283, + "grad_norm": 1.0158741474151611, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 18283 + }, + { + "epoch": 0.18284, + "grad_norm": 1.1559165716171265, + "learning_rate": 0.003, + "loss": 4.042, + "step": 18284 + }, + { + "epoch": 0.18285, + "grad_norm": 0.8005707859992981, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 18285 + }, + { + "epoch": 0.18286, + "grad_norm": 0.8458976149559021, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 18286 + }, + { + "epoch": 0.18287, + "grad_norm": 0.8141799569129944, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 18287 + }, + { + "epoch": 0.18288, + "grad_norm": 0.8615123629570007, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 18288 + }, + { + "epoch": 0.18289, + "grad_norm": 0.8300474882125854, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 18289 + }, + { + "epoch": 0.1829, + "grad_norm": 0.8556902408599854, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 18290 + }, + { + "epoch": 0.18291, + "grad_norm": 0.8768019676208496, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 18291 + }, + { + "epoch": 0.18292, + "grad_norm": 0.8756127953529358, + "learning_rate": 0.003, + "loss": 4.028, + "step": 18292 + }, + { + "epoch": 0.18293, + "grad_norm": 0.8733711838722229, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 18293 + }, + { + "epoch": 0.18294, + "grad_norm": 0.8341532349586487, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 18294 + }, + { + "epoch": 0.18295, + "grad_norm": 0.8398283123970032, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 18295 + }, + { + "epoch": 0.18296, + "grad_norm": 0.8624936938285828, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 18296 + }, + { + "epoch": 0.18297, + "grad_norm": 0.6987881064414978, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 18297 + }, + { + "epoch": 0.18298, + "grad_norm": 0.628121018409729, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 18298 + }, + { + "epoch": 0.18299, + "grad_norm": 0.7382155060768127, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 18299 + }, + { + "epoch": 0.183, + "grad_norm": 0.7305464148521423, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 18300 + }, + { + "epoch": 0.18301, + "grad_norm": 0.8464676737785339, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 18301 + }, + { + "epoch": 0.18302, + "grad_norm": 0.9060691595077515, + "learning_rate": 0.003, + "loss": 4.029, + "step": 18302 + }, + { + "epoch": 0.18303, + "grad_norm": 1.0593427419662476, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 18303 + }, + { + "epoch": 0.18304, + "grad_norm": 1.2884917259216309, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 18304 + }, + { + "epoch": 0.18305, + "grad_norm": 0.8706284165382385, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 18305 + }, + { + "epoch": 0.18306, + "grad_norm": 0.8215622305870056, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 18306 + }, + { + "epoch": 0.18307, + "grad_norm": 0.8193975687026978, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 18307 + }, + { + "epoch": 0.18308, + "grad_norm": 0.7148175835609436, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 18308 + }, + { + "epoch": 0.18309, + "grad_norm": 0.7559876441955566, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 18309 + }, + { + "epoch": 0.1831, + "grad_norm": 0.8626182079315186, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 18310 + }, + { + "epoch": 0.18311, + "grad_norm": 0.8986365795135498, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 18311 + }, + { + "epoch": 0.18312, + "grad_norm": 1.1372230052947998, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 18312 + }, + { + "epoch": 0.18313, + "grad_norm": 0.8173007965087891, + "learning_rate": 0.003, + "loss": 3.9305, + "step": 18313 + }, + { + "epoch": 0.18314, + "grad_norm": 0.7488594651222229, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 18314 + }, + { + "epoch": 0.18315, + "grad_norm": 0.7553419470787048, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 18315 + }, + { + "epoch": 0.18316, + "grad_norm": 0.5957514047622681, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 18316 + }, + { + "epoch": 0.18317, + "grad_norm": 0.5239382982254028, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 18317 + }, + { + "epoch": 0.18318, + "grad_norm": 0.5718421339988708, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 18318 + }, + { + "epoch": 0.18319, + "grad_norm": 0.6599254608154297, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 18319 + }, + { + "epoch": 0.1832, + "grad_norm": 0.6726394891738892, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 18320 + }, + { + "epoch": 0.18321, + "grad_norm": 0.7308098077774048, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 18321 + }, + { + "epoch": 0.18322, + "grad_norm": 0.7821747064590454, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 18322 + }, + { + "epoch": 0.18323, + "grad_norm": 0.7732888460159302, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 18323 + }, + { + "epoch": 0.18324, + "grad_norm": 0.6457021832466125, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 18324 + }, + { + "epoch": 0.18325, + "grad_norm": 0.6706361770629883, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 18325 + }, + { + "epoch": 0.18326, + "grad_norm": 0.7337856888771057, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 18326 + }, + { + "epoch": 0.18327, + "grad_norm": 0.7500945329666138, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 18327 + }, + { + "epoch": 0.18328, + "grad_norm": 0.9249593019485474, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 18328 + }, + { + "epoch": 0.18329, + "grad_norm": 1.1571935415267944, + "learning_rate": 0.003, + "loss": 3.977, + "step": 18329 + }, + { + "epoch": 0.1833, + "grad_norm": 0.886870265007019, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 18330 + }, + { + "epoch": 0.18331, + "grad_norm": 0.7807754874229431, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 18331 + }, + { + "epoch": 0.18332, + "grad_norm": 0.7621500492095947, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 18332 + }, + { + "epoch": 0.18333, + "grad_norm": 0.8085874915122986, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 18333 + }, + { + "epoch": 0.18334, + "grad_norm": 0.7710543870925903, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 18334 + }, + { + "epoch": 0.18335, + "grad_norm": 0.7949119210243225, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 18335 + }, + { + "epoch": 0.18336, + "grad_norm": 1.0128542184829712, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 18336 + }, + { + "epoch": 0.18337, + "grad_norm": 1.1886122226715088, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 18337 + }, + { + "epoch": 0.18338, + "grad_norm": 0.817167341709137, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 18338 + }, + { + "epoch": 0.18339, + "grad_norm": 0.7243970036506653, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 18339 + }, + { + "epoch": 0.1834, + "grad_norm": 0.6925933957099915, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 18340 + }, + { + "epoch": 0.18341, + "grad_norm": 0.6670451760292053, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 18341 + }, + { + "epoch": 0.18342, + "grad_norm": 0.7994633913040161, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 18342 + }, + { + "epoch": 0.18343, + "grad_norm": 0.9264425039291382, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 18343 + }, + { + "epoch": 0.18344, + "grad_norm": 1.0609327554702759, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 18344 + }, + { + "epoch": 0.18345, + "grad_norm": 1.0058785676956177, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 18345 + }, + { + "epoch": 0.18346, + "grad_norm": 1.009697675704956, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 18346 + }, + { + "epoch": 0.18347, + "grad_norm": 0.9138298034667969, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 18347 + }, + { + "epoch": 0.18348, + "grad_norm": 0.8665712475776672, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 18348 + }, + { + "epoch": 0.18349, + "grad_norm": 0.9085323810577393, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 18349 + }, + { + "epoch": 0.1835, + "grad_norm": 0.8555454611778259, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 18350 + }, + { + "epoch": 0.18351, + "grad_norm": 0.8092401027679443, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 18351 + }, + { + "epoch": 0.18352, + "grad_norm": 0.7945643663406372, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 18352 + }, + { + "epoch": 0.18353, + "grad_norm": 0.9307811856269836, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 18353 + }, + { + "epoch": 0.18354, + "grad_norm": 0.9853703379631042, + "learning_rate": 0.003, + "loss": 4.013, + "step": 18354 + }, + { + "epoch": 0.18355, + "grad_norm": 1.050830364227295, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 18355 + }, + { + "epoch": 0.18356, + "grad_norm": 0.8668641448020935, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 18356 + }, + { + "epoch": 0.18357, + "grad_norm": 0.7032343149185181, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 18357 + }, + { + "epoch": 0.18358, + "grad_norm": 0.6883024573326111, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 18358 + }, + { + "epoch": 0.18359, + "grad_norm": 0.7748553156852722, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 18359 + }, + { + "epoch": 0.1836, + "grad_norm": 0.9895306825637817, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 18360 + }, + { + "epoch": 0.18361, + "grad_norm": 1.1863415241241455, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 18361 + }, + { + "epoch": 0.18362, + "grad_norm": 0.914741039276123, + "learning_rate": 0.003, + "loss": 3.999, + "step": 18362 + }, + { + "epoch": 0.18363, + "grad_norm": 0.8926665186882019, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 18363 + }, + { + "epoch": 0.18364, + "grad_norm": 0.9242895245552063, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 18364 + }, + { + "epoch": 0.18365, + "grad_norm": 0.8577793836593628, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 18365 + }, + { + "epoch": 0.18366, + "grad_norm": 0.8163846731185913, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 18366 + }, + { + "epoch": 0.18367, + "grad_norm": 0.7950718402862549, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 18367 + }, + { + "epoch": 0.18368, + "grad_norm": 0.8748380541801453, + "learning_rate": 0.003, + "loss": 3.985, + "step": 18368 + }, + { + "epoch": 0.18369, + "grad_norm": 0.9596766829490662, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 18369 + }, + { + "epoch": 0.1837, + "grad_norm": 0.9462250471115112, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 18370 + }, + { + "epoch": 0.18371, + "grad_norm": 0.954529345035553, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 18371 + }, + { + "epoch": 0.18372, + "grad_norm": 0.8737909197807312, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 18372 + }, + { + "epoch": 0.18373, + "grad_norm": 0.8716749548912048, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 18373 + }, + { + "epoch": 0.18374, + "grad_norm": 0.8134250044822693, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 18374 + }, + { + "epoch": 0.18375, + "grad_norm": 0.9541025757789612, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 18375 + }, + { + "epoch": 0.18376, + "grad_norm": 0.9908921718597412, + "learning_rate": 0.003, + "loss": 4.036, + "step": 18376 + }, + { + "epoch": 0.18377, + "grad_norm": 1.1104575395584106, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 18377 + }, + { + "epoch": 0.18378, + "grad_norm": 0.9030351042747498, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 18378 + }, + { + "epoch": 0.18379, + "grad_norm": 0.851387619972229, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 18379 + }, + { + "epoch": 0.1838, + "grad_norm": 0.8044515252113342, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 18380 + }, + { + "epoch": 0.18381, + "grad_norm": 0.7882006764411926, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 18381 + }, + { + "epoch": 0.18382, + "grad_norm": 0.6093977093696594, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 18382 + }, + { + "epoch": 0.18383, + "grad_norm": 0.52236008644104, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 18383 + }, + { + "epoch": 0.18384, + "grad_norm": 0.47218433022499084, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 18384 + }, + { + "epoch": 0.18385, + "grad_norm": 0.5042617917060852, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 18385 + }, + { + "epoch": 0.18386, + "grad_norm": 0.5072876811027527, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 18386 + }, + { + "epoch": 0.18387, + "grad_norm": 0.5613670349121094, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 18387 + }, + { + "epoch": 0.18388, + "grad_norm": 0.5694621205329895, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 18388 + }, + { + "epoch": 0.18389, + "grad_norm": 0.6198713183403015, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 18389 + }, + { + "epoch": 0.1839, + "grad_norm": 0.694005012512207, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 18390 + }, + { + "epoch": 0.18391, + "grad_norm": 0.781154215335846, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 18391 + }, + { + "epoch": 0.18392, + "grad_norm": 0.8123658895492554, + "learning_rate": 0.003, + "loss": 3.97, + "step": 18392 + }, + { + "epoch": 0.18393, + "grad_norm": 0.9070790410041809, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 18393 + }, + { + "epoch": 0.18394, + "grad_norm": 1.081210970878601, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 18394 + }, + { + "epoch": 0.18395, + "grad_norm": 1.0967609882354736, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 18395 + }, + { + "epoch": 0.18396, + "grad_norm": 1.050869345664978, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 18396 + }, + { + "epoch": 0.18397, + "grad_norm": 1.079639196395874, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 18397 + }, + { + "epoch": 0.18398, + "grad_norm": 0.7882301807403564, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 18398 + }, + { + "epoch": 0.18399, + "grad_norm": 0.63349848985672, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 18399 + }, + { + "epoch": 0.184, + "grad_norm": 0.8485949635505676, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 18400 + }, + { + "epoch": 0.18401, + "grad_norm": 0.8544250130653381, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 18401 + }, + { + "epoch": 0.18402, + "grad_norm": 0.7469441890716553, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 18402 + }, + { + "epoch": 0.18403, + "grad_norm": 0.9119636416435242, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 18403 + }, + { + "epoch": 0.18404, + "grad_norm": 1.0456124544143677, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 18404 + }, + { + "epoch": 0.18405, + "grad_norm": 1.1618139743804932, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 18405 + }, + { + "epoch": 0.18406, + "grad_norm": 0.9747986793518066, + "learning_rate": 0.003, + "loss": 4.016, + "step": 18406 + }, + { + "epoch": 0.18407, + "grad_norm": 0.9599617123603821, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 18407 + }, + { + "epoch": 0.18408, + "grad_norm": 1.067495346069336, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 18408 + }, + { + "epoch": 0.18409, + "grad_norm": 0.8262394666671753, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 18409 + }, + { + "epoch": 0.1841, + "grad_norm": 0.7933745384216309, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 18410 + }, + { + "epoch": 0.18411, + "grad_norm": 0.8479217886924744, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 18411 + }, + { + "epoch": 0.18412, + "grad_norm": 0.855431079864502, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 18412 + }, + { + "epoch": 0.18413, + "grad_norm": 0.8933448791503906, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 18413 + }, + { + "epoch": 0.18414, + "grad_norm": 0.9668148756027222, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 18414 + }, + { + "epoch": 0.18415, + "grad_norm": 1.121019959449768, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 18415 + }, + { + "epoch": 0.18416, + "grad_norm": 0.9086769819259644, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 18416 + }, + { + "epoch": 0.18417, + "grad_norm": 0.9084094166755676, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 18417 + }, + { + "epoch": 0.18418, + "grad_norm": 0.8724955916404724, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 18418 + }, + { + "epoch": 0.18419, + "grad_norm": 0.7465152144432068, + "learning_rate": 0.003, + "loss": 3.99, + "step": 18419 + }, + { + "epoch": 0.1842, + "grad_norm": 0.768031895160675, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 18420 + }, + { + "epoch": 0.18421, + "grad_norm": 0.8390229940414429, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 18421 + }, + { + "epoch": 0.18422, + "grad_norm": 0.8336851596832275, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 18422 + }, + { + "epoch": 0.18423, + "grad_norm": 0.7277620434761047, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 18423 + }, + { + "epoch": 0.18424, + "grad_norm": 0.7897154688835144, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 18424 + }, + { + "epoch": 0.18425, + "grad_norm": 1.0583573579788208, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 18425 + }, + { + "epoch": 0.18426, + "grad_norm": 1.0781868696212769, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 18426 + }, + { + "epoch": 0.18427, + "grad_norm": 0.8976330161094666, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 18427 + }, + { + "epoch": 0.18428, + "grad_norm": 0.7958218455314636, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 18428 + }, + { + "epoch": 0.18429, + "grad_norm": 0.6814959645271301, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 18429 + }, + { + "epoch": 0.1843, + "grad_norm": 0.6394864916801453, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 18430 + }, + { + "epoch": 0.18431, + "grad_norm": 0.685542643070221, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 18431 + }, + { + "epoch": 0.18432, + "grad_norm": 0.8013325929641724, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 18432 + }, + { + "epoch": 0.18433, + "grad_norm": 0.8234718441963196, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 18433 + }, + { + "epoch": 0.18434, + "grad_norm": 0.7327991724014282, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 18434 + }, + { + "epoch": 0.18435, + "grad_norm": 0.6646810173988342, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 18435 + }, + { + "epoch": 0.18436, + "grad_norm": 0.7256752848625183, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 18436 + }, + { + "epoch": 0.18437, + "grad_norm": 0.8754481673240662, + "learning_rate": 0.003, + "loss": 4.026, + "step": 18437 + }, + { + "epoch": 0.18438, + "grad_norm": 0.9004335999488831, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 18438 + }, + { + "epoch": 0.18439, + "grad_norm": 0.8405164480209351, + "learning_rate": 0.003, + "loss": 3.979, + "step": 18439 + }, + { + "epoch": 0.1844, + "grad_norm": 0.8154463768005371, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 18440 + }, + { + "epoch": 0.18441, + "grad_norm": 0.9057301878929138, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 18441 + }, + { + "epoch": 0.18442, + "grad_norm": 1.1145095825195312, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 18442 + }, + { + "epoch": 0.18443, + "grad_norm": 0.9151156544685364, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 18443 + }, + { + "epoch": 0.18444, + "grad_norm": 0.9163838028907776, + "learning_rate": 0.003, + "loss": 4.01, + "step": 18444 + }, + { + "epoch": 0.18445, + "grad_norm": 1.0952199697494507, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 18445 + }, + { + "epoch": 0.18446, + "grad_norm": 0.7924932837486267, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 18446 + }, + { + "epoch": 0.18447, + "grad_norm": 0.7247111201286316, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 18447 + }, + { + "epoch": 0.18448, + "grad_norm": 0.7970155477523804, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 18448 + }, + { + "epoch": 0.18449, + "grad_norm": 0.759702742099762, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 18449 + }, + { + "epoch": 0.1845, + "grad_norm": 0.7928878664970398, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 18450 + }, + { + "epoch": 0.18451, + "grad_norm": 0.7595303058624268, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 18451 + }, + { + "epoch": 0.18452, + "grad_norm": 0.7852331399917603, + "learning_rate": 0.003, + "loss": 3.992, + "step": 18452 + }, + { + "epoch": 0.18453, + "grad_norm": 0.9528810977935791, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 18453 + }, + { + "epoch": 0.18454, + "grad_norm": 1.0351529121398926, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 18454 + }, + { + "epoch": 0.18455, + "grad_norm": 0.801206111907959, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 18455 + }, + { + "epoch": 0.18456, + "grad_norm": 0.7646304965019226, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 18456 + }, + { + "epoch": 0.18457, + "grad_norm": 0.7039245963096619, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 18457 + }, + { + "epoch": 0.18458, + "grad_norm": 0.6283014416694641, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 18458 + }, + { + "epoch": 0.18459, + "grad_norm": 0.6567989587783813, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 18459 + }, + { + "epoch": 0.1846, + "grad_norm": 0.9653926491737366, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 18460 + }, + { + "epoch": 0.18461, + "grad_norm": 1.556738257408142, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 18461 + }, + { + "epoch": 0.18462, + "grad_norm": 0.5955246686935425, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 18462 + }, + { + "epoch": 0.18463, + "grad_norm": 0.7615717649459839, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 18463 + }, + { + "epoch": 0.18464, + "grad_norm": 0.8574117422103882, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 18464 + }, + { + "epoch": 0.18465, + "grad_norm": 1.0039912462234497, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 18465 + }, + { + "epoch": 0.18466, + "grad_norm": 1.3914353847503662, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 18466 + }, + { + "epoch": 0.18467, + "grad_norm": 0.5892895460128784, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 18467 + }, + { + "epoch": 0.18468, + "grad_norm": 0.8511627316474915, + "learning_rate": 0.003, + "loss": 4.019, + "step": 18468 + }, + { + "epoch": 0.18469, + "grad_norm": 0.9471988677978516, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 18469 + }, + { + "epoch": 0.1847, + "grad_norm": 1.0093071460723877, + "learning_rate": 0.003, + "loss": 3.979, + "step": 18470 + }, + { + "epoch": 0.18471, + "grad_norm": 1.2240053415298462, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 18471 + }, + { + "epoch": 0.18472, + "grad_norm": 0.7324399948120117, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 18472 + }, + { + "epoch": 0.18473, + "grad_norm": 0.6687946915626526, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 18473 + }, + { + "epoch": 0.18474, + "grad_norm": 0.8151763081550598, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 18474 + }, + { + "epoch": 0.18475, + "grad_norm": 0.8017345666885376, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 18475 + }, + { + "epoch": 0.18476, + "grad_norm": 0.8550369739532471, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 18476 + }, + { + "epoch": 0.18477, + "grad_norm": 1.0418061017990112, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 18477 + }, + { + "epoch": 0.18478, + "grad_norm": 0.9475833773612976, + "learning_rate": 0.003, + "loss": 4.016, + "step": 18478 + }, + { + "epoch": 0.18479, + "grad_norm": 1.0286717414855957, + "learning_rate": 0.003, + "loss": 4.009, + "step": 18479 + }, + { + "epoch": 0.1848, + "grad_norm": 1.016998291015625, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 18480 + }, + { + "epoch": 0.18481, + "grad_norm": 1.0216963291168213, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 18481 + }, + { + "epoch": 0.18482, + "grad_norm": 0.9507943391799927, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 18482 + }, + { + "epoch": 0.18483, + "grad_norm": 0.8995311260223389, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 18483 + }, + { + "epoch": 0.18484, + "grad_norm": 0.7644209265708923, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 18484 + }, + { + "epoch": 0.18485, + "grad_norm": 0.9264317154884338, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 18485 + }, + { + "epoch": 0.18486, + "grad_norm": 0.9133664965629578, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 18486 + }, + { + "epoch": 0.18487, + "grad_norm": 0.9255028963088989, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 18487 + }, + { + "epoch": 0.18488, + "grad_norm": 0.9390118718147278, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 18488 + }, + { + "epoch": 0.18489, + "grad_norm": 1.0404934883117676, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 18489 + }, + { + "epoch": 0.1849, + "grad_norm": 0.8677457571029663, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 18490 + }, + { + "epoch": 0.18491, + "grad_norm": 0.7368670105934143, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 18491 + }, + { + "epoch": 0.18492, + "grad_norm": 0.6713873744010925, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 18492 + }, + { + "epoch": 0.18493, + "grad_norm": 0.6910095810890198, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 18493 + }, + { + "epoch": 0.18494, + "grad_norm": 0.6025274991989136, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 18494 + }, + { + "epoch": 0.18495, + "grad_norm": 0.6360000371932983, + "learning_rate": 0.003, + "loss": 3.994, + "step": 18495 + }, + { + "epoch": 0.18496, + "grad_norm": 0.7263234853744507, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 18496 + }, + { + "epoch": 0.18497, + "grad_norm": 0.7698788046836853, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 18497 + }, + { + "epoch": 0.18498, + "grad_norm": 0.9591576457023621, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 18498 + }, + { + "epoch": 0.18499, + "grad_norm": 1.0750406980514526, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 18499 + }, + { + "epoch": 0.185, + "grad_norm": 0.9537850022315979, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 18500 + }, + { + "epoch": 0.18501, + "grad_norm": 0.8107244372367859, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 18501 + }, + { + "epoch": 0.18502, + "grad_norm": 0.5002862215042114, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 18502 + }, + { + "epoch": 0.18503, + "grad_norm": 0.5681113600730896, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 18503 + }, + { + "epoch": 0.18504, + "grad_norm": 0.5782869458198547, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 18504 + }, + { + "epoch": 0.18505, + "grad_norm": 0.6032190322875977, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 18505 + }, + { + "epoch": 0.18506, + "grad_norm": 0.759202778339386, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 18506 + }, + { + "epoch": 0.18507, + "grad_norm": 0.9024019837379456, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 18507 + }, + { + "epoch": 0.18508, + "grad_norm": 0.9643970727920532, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 18508 + }, + { + "epoch": 0.18509, + "grad_norm": 1.0283045768737793, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 18509 + }, + { + "epoch": 0.1851, + "grad_norm": 0.8149793744087219, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 18510 + }, + { + "epoch": 0.18511, + "grad_norm": 0.6232828497886658, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 18511 + }, + { + "epoch": 0.18512, + "grad_norm": 0.6650475859642029, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 18512 + }, + { + "epoch": 0.18513, + "grad_norm": 0.6731092929840088, + "learning_rate": 0.003, + "loss": 4.003, + "step": 18513 + }, + { + "epoch": 0.18514, + "grad_norm": 0.7437854409217834, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 18514 + }, + { + "epoch": 0.18515, + "grad_norm": 0.7701714038848877, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 18515 + }, + { + "epoch": 0.18516, + "grad_norm": 0.6693366765975952, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 18516 + }, + { + "epoch": 0.18517, + "grad_norm": 0.7216953635215759, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 18517 + }, + { + "epoch": 0.18518, + "grad_norm": 0.7274855375289917, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 18518 + }, + { + "epoch": 0.18519, + "grad_norm": 0.7271580696105957, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 18519 + }, + { + "epoch": 0.1852, + "grad_norm": 0.7157036662101746, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 18520 + }, + { + "epoch": 0.18521, + "grad_norm": 0.6716972589492798, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 18521 + }, + { + "epoch": 0.18522, + "grad_norm": 0.634327232837677, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 18522 + }, + { + "epoch": 0.18523, + "grad_norm": 0.756804883480072, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 18523 + }, + { + "epoch": 0.18524, + "grad_norm": 0.978564441204071, + "learning_rate": 0.003, + "loss": 4.021, + "step": 18524 + }, + { + "epoch": 0.18525, + "grad_norm": 0.9920554757118225, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 18525 + }, + { + "epoch": 0.18526, + "grad_norm": 0.9714943170547485, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 18526 + }, + { + "epoch": 0.18527, + "grad_norm": 0.8749993443489075, + "learning_rate": 0.003, + "loss": 4.011, + "step": 18527 + }, + { + "epoch": 0.18528, + "grad_norm": 0.9511075615882874, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 18528 + }, + { + "epoch": 0.18529, + "grad_norm": 1.1085911989212036, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 18529 + }, + { + "epoch": 0.1853, + "grad_norm": 0.9463942050933838, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 18530 + }, + { + "epoch": 0.18531, + "grad_norm": 0.9208247065544128, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 18531 + }, + { + "epoch": 0.18532, + "grad_norm": 0.9658119678497314, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 18532 + }, + { + "epoch": 0.18533, + "grad_norm": 1.1414157152175903, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 18533 + }, + { + "epoch": 0.18534, + "grad_norm": 0.9629420042037964, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 18534 + }, + { + "epoch": 0.18535, + "grad_norm": 0.9314190149307251, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 18535 + }, + { + "epoch": 0.18536, + "grad_norm": 0.9639930129051208, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 18536 + }, + { + "epoch": 0.18537, + "grad_norm": 0.9416884183883667, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 18537 + }, + { + "epoch": 0.18538, + "grad_norm": 0.8365035653114319, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 18538 + }, + { + "epoch": 0.18539, + "grad_norm": 0.908024787902832, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 18539 + }, + { + "epoch": 0.1854, + "grad_norm": 0.9564443826675415, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 18540 + }, + { + "epoch": 0.18541, + "grad_norm": 1.1441367864608765, + "learning_rate": 0.003, + "loss": 3.998, + "step": 18541 + }, + { + "epoch": 0.18542, + "grad_norm": 0.7297008037567139, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 18542 + }, + { + "epoch": 0.18543, + "grad_norm": 0.6466062068939209, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 18543 + }, + { + "epoch": 0.18544, + "grad_norm": 0.8631340265274048, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 18544 + }, + { + "epoch": 0.18545, + "grad_norm": 0.8763319253921509, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 18545 + }, + { + "epoch": 0.18546, + "grad_norm": 0.8306936025619507, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 18546 + }, + { + "epoch": 0.18547, + "grad_norm": 0.7532835006713867, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 18547 + }, + { + "epoch": 0.18548, + "grad_norm": 0.6749123930931091, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 18548 + }, + { + "epoch": 0.18549, + "grad_norm": 0.7876031398773193, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 18549 + }, + { + "epoch": 0.1855, + "grad_norm": 0.8753268718719482, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 18550 + }, + { + "epoch": 0.18551, + "grad_norm": 1.1003737449645996, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 18551 + }, + { + "epoch": 0.18552, + "grad_norm": 0.9563855528831482, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 18552 + }, + { + "epoch": 0.18553, + "grad_norm": 0.9888499975204468, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 18553 + }, + { + "epoch": 0.18554, + "grad_norm": 0.88606196641922, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 18554 + }, + { + "epoch": 0.18555, + "grad_norm": 0.6885478496551514, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 18555 + }, + { + "epoch": 0.18556, + "grad_norm": 0.7235888838768005, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 18556 + }, + { + "epoch": 0.18557, + "grad_norm": 0.7943369150161743, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 18557 + }, + { + "epoch": 0.18558, + "grad_norm": 0.8080901503562927, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 18558 + }, + { + "epoch": 0.18559, + "grad_norm": 0.6975395679473877, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 18559 + }, + { + "epoch": 0.1856, + "grad_norm": 0.624427855014801, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 18560 + }, + { + "epoch": 0.18561, + "grad_norm": 0.5854123830795288, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 18561 + }, + { + "epoch": 0.18562, + "grad_norm": 0.731335461139679, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 18562 + }, + { + "epoch": 0.18563, + "grad_norm": 0.8278244733810425, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 18563 + }, + { + "epoch": 0.18564, + "grad_norm": 0.8942627906799316, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 18564 + }, + { + "epoch": 0.18565, + "grad_norm": 1.1789724826812744, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 18565 + }, + { + "epoch": 0.18566, + "grad_norm": 0.8848671913146973, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 18566 + }, + { + "epoch": 0.18567, + "grad_norm": 0.6374181509017944, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 18567 + }, + { + "epoch": 0.18568, + "grad_norm": 0.49829450249671936, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 18568 + }, + { + "epoch": 0.18569, + "grad_norm": 0.5674188137054443, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 18569 + }, + { + "epoch": 0.1857, + "grad_norm": 0.6359483003616333, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 18570 + }, + { + "epoch": 0.18571, + "grad_norm": 0.6313909888267517, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 18571 + }, + { + "epoch": 0.18572, + "grad_norm": 0.6858205199241638, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 18572 + }, + { + "epoch": 0.18573, + "grad_norm": 0.8699098825454712, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 18573 + }, + { + "epoch": 0.18574, + "grad_norm": 1.1460634469985962, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 18574 + }, + { + "epoch": 0.18575, + "grad_norm": 0.8282952308654785, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 18575 + }, + { + "epoch": 0.18576, + "grad_norm": 0.8061177730560303, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 18576 + }, + { + "epoch": 0.18577, + "grad_norm": 0.8096114993095398, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 18577 + }, + { + "epoch": 0.18578, + "grad_norm": 0.764124870300293, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 18578 + }, + { + "epoch": 0.18579, + "grad_norm": 0.7354459762573242, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 18579 + }, + { + "epoch": 0.1858, + "grad_norm": 0.917849063873291, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 18580 + }, + { + "epoch": 0.18581, + "grad_norm": 1.1212433576583862, + "learning_rate": 0.003, + "loss": 3.973, + "step": 18581 + }, + { + "epoch": 0.18582, + "grad_norm": 1.0859284400939941, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 18582 + }, + { + "epoch": 0.18583, + "grad_norm": 0.7617725133895874, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 18583 + }, + { + "epoch": 0.18584, + "grad_norm": 0.6631203293800354, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 18584 + }, + { + "epoch": 0.18585, + "grad_norm": 0.7645079493522644, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 18585 + }, + { + "epoch": 0.18586, + "grad_norm": 0.7495100498199463, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 18586 + }, + { + "epoch": 0.18587, + "grad_norm": 0.810477077960968, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 18587 + }, + { + "epoch": 0.18588, + "grad_norm": 0.8579960465431213, + "learning_rate": 0.003, + "loss": 3.987, + "step": 18588 + }, + { + "epoch": 0.18589, + "grad_norm": 0.8831098675727844, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 18589 + }, + { + "epoch": 0.1859, + "grad_norm": 0.8370000720024109, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 18590 + }, + { + "epoch": 0.18591, + "grad_norm": 0.8281912803649902, + "learning_rate": 0.003, + "loss": 3.973, + "step": 18591 + }, + { + "epoch": 0.18592, + "grad_norm": 0.926630437374115, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 18592 + }, + { + "epoch": 0.18593, + "grad_norm": 0.9408486485481262, + "learning_rate": 0.003, + "loss": 4.025, + "step": 18593 + }, + { + "epoch": 0.18594, + "grad_norm": 1.0552456378936768, + "learning_rate": 0.003, + "loss": 4.02, + "step": 18594 + }, + { + "epoch": 0.18595, + "grad_norm": 1.0246379375457764, + "learning_rate": 0.003, + "loss": 4.009, + "step": 18595 + }, + { + "epoch": 0.18596, + "grad_norm": 0.9521511793136597, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 18596 + }, + { + "epoch": 0.18597, + "grad_norm": 0.8722307682037354, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 18597 + }, + { + "epoch": 0.18598, + "grad_norm": 0.93885338306427, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 18598 + }, + { + "epoch": 0.18599, + "grad_norm": 1.083472728729248, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 18599 + }, + { + "epoch": 0.186, + "grad_norm": 1.032609462738037, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 18600 + }, + { + "epoch": 0.18601, + "grad_norm": 0.8510088324546814, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 18601 + }, + { + "epoch": 0.18602, + "grad_norm": 0.8725912570953369, + "learning_rate": 0.003, + "loss": 3.981, + "step": 18602 + }, + { + "epoch": 0.18603, + "grad_norm": 1.2701548337936401, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 18603 + }, + { + "epoch": 0.18604, + "grad_norm": 1.0352787971496582, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 18604 + }, + { + "epoch": 0.18605, + "grad_norm": 0.9110256433486938, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 18605 + }, + { + "epoch": 0.18606, + "grad_norm": 0.8785024285316467, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 18606 + }, + { + "epoch": 0.18607, + "grad_norm": 0.7071681022644043, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 18607 + }, + { + "epoch": 0.18608, + "grad_norm": 0.6964113116264343, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 18608 + }, + { + "epoch": 0.18609, + "grad_norm": 0.654489278793335, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 18609 + }, + { + "epoch": 0.1861, + "grad_norm": 0.6638596653938293, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 18610 + }, + { + "epoch": 0.18611, + "grad_norm": 0.6475112438201904, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 18611 + }, + { + "epoch": 0.18612, + "grad_norm": 0.6679487824440002, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 18612 + }, + { + "epoch": 0.18613, + "grad_norm": 0.7296201586723328, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 18613 + }, + { + "epoch": 0.18614, + "grad_norm": 0.7591899037361145, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 18614 + }, + { + "epoch": 0.18615, + "grad_norm": 0.8773738145828247, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 18615 + }, + { + "epoch": 0.18616, + "grad_norm": 0.9669382572174072, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 18616 + }, + { + "epoch": 0.18617, + "grad_norm": 1.0774128437042236, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 18617 + }, + { + "epoch": 0.18618, + "grad_norm": 0.9947457909584045, + "learning_rate": 0.003, + "loss": 4.021, + "step": 18618 + }, + { + "epoch": 0.18619, + "grad_norm": 1.021847128868103, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 18619 + }, + { + "epoch": 0.1862, + "grad_norm": 0.8634188175201416, + "learning_rate": 0.003, + "loss": 4.025, + "step": 18620 + }, + { + "epoch": 0.18621, + "grad_norm": 0.8271879553794861, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 18621 + }, + { + "epoch": 0.18622, + "grad_norm": 0.871752142906189, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 18622 + }, + { + "epoch": 0.18623, + "grad_norm": 0.7900510430335999, + "learning_rate": 0.003, + "loss": 3.997, + "step": 18623 + }, + { + "epoch": 0.18624, + "grad_norm": 0.7515827417373657, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 18624 + }, + { + "epoch": 0.18625, + "grad_norm": 0.6824390888214111, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 18625 + }, + { + "epoch": 0.18626, + "grad_norm": 0.7836955785751343, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 18626 + }, + { + "epoch": 0.18627, + "grad_norm": 0.9825575947761536, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 18627 + }, + { + "epoch": 0.18628, + "grad_norm": 1.022945761680603, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 18628 + }, + { + "epoch": 0.18629, + "grad_norm": 1.023468255996704, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 18629 + }, + { + "epoch": 0.1863, + "grad_norm": 0.9927089810371399, + "learning_rate": 0.003, + "loss": 4.001, + "step": 18630 + }, + { + "epoch": 0.18631, + "grad_norm": 0.9182254076004028, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 18631 + }, + { + "epoch": 0.18632, + "grad_norm": 0.9914144277572632, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 18632 + }, + { + "epoch": 0.18633, + "grad_norm": 1.050477385520935, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 18633 + }, + { + "epoch": 0.18634, + "grad_norm": 0.7706889510154724, + "learning_rate": 0.003, + "loss": 4.011, + "step": 18634 + }, + { + "epoch": 0.18635, + "grad_norm": 0.6501280069351196, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 18635 + }, + { + "epoch": 0.18636, + "grad_norm": 0.5715956687927246, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 18636 + }, + { + "epoch": 0.18637, + "grad_norm": 0.538364589214325, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 18637 + }, + { + "epoch": 0.18638, + "grad_norm": 0.556035578250885, + "learning_rate": 0.003, + "loss": 4.006, + "step": 18638 + }, + { + "epoch": 0.18639, + "grad_norm": 0.5113933086395264, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 18639 + }, + { + "epoch": 0.1864, + "grad_norm": 0.4925510585308075, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 18640 + }, + { + "epoch": 0.18641, + "grad_norm": 0.4664783179759979, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 18641 + }, + { + "epoch": 0.18642, + "grad_norm": 0.5348178744316101, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 18642 + }, + { + "epoch": 0.18643, + "grad_norm": 0.6688624620437622, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 18643 + }, + { + "epoch": 0.18644, + "grad_norm": 0.7310923337936401, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 18644 + }, + { + "epoch": 0.18645, + "grad_norm": 0.8424361348152161, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 18645 + }, + { + "epoch": 0.18646, + "grad_norm": 0.922663152217865, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 18646 + }, + { + "epoch": 0.18647, + "grad_norm": 0.9829749464988708, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 18647 + }, + { + "epoch": 0.18648, + "grad_norm": 1.1478034257888794, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 18648 + }, + { + "epoch": 0.18649, + "grad_norm": 0.9126496315002441, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 18649 + }, + { + "epoch": 0.1865, + "grad_norm": 0.9763796925544739, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 18650 + }, + { + "epoch": 0.18651, + "grad_norm": 1.0723530054092407, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 18651 + }, + { + "epoch": 0.18652, + "grad_norm": 0.7517349123954773, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 18652 + }, + { + "epoch": 0.18653, + "grad_norm": 0.6144676804542542, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 18653 + }, + { + "epoch": 0.18654, + "grad_norm": 0.6148786544799805, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 18654 + }, + { + "epoch": 0.18655, + "grad_norm": 0.8592705726623535, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 18655 + }, + { + "epoch": 0.18656, + "grad_norm": 1.053800344467163, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 18656 + }, + { + "epoch": 0.18657, + "grad_norm": 0.9127708673477173, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 18657 + }, + { + "epoch": 0.18658, + "grad_norm": 0.7173557281494141, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 18658 + }, + { + "epoch": 0.18659, + "grad_norm": 0.6722642779350281, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 18659 + }, + { + "epoch": 0.1866, + "grad_norm": 0.7511748671531677, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 18660 + }, + { + "epoch": 0.18661, + "grad_norm": 0.8174583315849304, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 18661 + }, + { + "epoch": 0.18662, + "grad_norm": 0.9325646162033081, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 18662 + }, + { + "epoch": 0.18663, + "grad_norm": 0.9656684398651123, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 18663 + }, + { + "epoch": 0.18664, + "grad_norm": 0.9210911989212036, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 18664 + }, + { + "epoch": 0.18665, + "grad_norm": 1.0134801864624023, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 18665 + }, + { + "epoch": 0.18666, + "grad_norm": 1.0047428607940674, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 18666 + }, + { + "epoch": 0.18667, + "grad_norm": 0.9935968518257141, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 18667 + }, + { + "epoch": 0.18668, + "grad_norm": 1.0019229650497437, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 18668 + }, + { + "epoch": 0.18669, + "grad_norm": 1.104162335395813, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 18669 + }, + { + "epoch": 0.1867, + "grad_norm": 0.913928747177124, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 18670 + }, + { + "epoch": 0.18671, + "grad_norm": 0.9824994206428528, + "learning_rate": 0.003, + "loss": 4.022, + "step": 18671 + }, + { + "epoch": 0.18672, + "grad_norm": 1.0232406854629517, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 18672 + }, + { + "epoch": 0.18673, + "grad_norm": 1.027138113975525, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 18673 + }, + { + "epoch": 0.18674, + "grad_norm": 1.1209120750427246, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 18674 + }, + { + "epoch": 0.18675, + "grad_norm": 0.9312422275543213, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 18675 + }, + { + "epoch": 0.18676, + "grad_norm": 0.7721902132034302, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 18676 + }, + { + "epoch": 0.18677, + "grad_norm": 0.8749040961265564, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 18677 + }, + { + "epoch": 0.18678, + "grad_norm": 0.8137281537055969, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 18678 + }, + { + "epoch": 0.18679, + "grad_norm": 0.7488441467285156, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 18679 + }, + { + "epoch": 0.1868, + "grad_norm": 0.6902167797088623, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 18680 + }, + { + "epoch": 0.18681, + "grad_norm": 0.6958896517753601, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 18681 + }, + { + "epoch": 0.18682, + "grad_norm": 0.7173922657966614, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 18682 + }, + { + "epoch": 0.18683, + "grad_norm": 0.7283394932746887, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 18683 + }, + { + "epoch": 0.18684, + "grad_norm": 0.7280787825584412, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 18684 + }, + { + "epoch": 0.18685, + "grad_norm": 0.7102197408676147, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 18685 + }, + { + "epoch": 0.18686, + "grad_norm": 0.7416349649429321, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 18686 + }, + { + "epoch": 0.18687, + "grad_norm": 0.7413874268531799, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 18687 + }, + { + "epoch": 0.18688, + "grad_norm": 0.7215458154678345, + "learning_rate": 0.003, + "loss": 3.992, + "step": 18688 + }, + { + "epoch": 0.18689, + "grad_norm": 0.6806061863899231, + "learning_rate": 0.003, + "loss": 3.967, + "step": 18689 + }, + { + "epoch": 0.1869, + "grad_norm": 0.8180581331253052, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 18690 + }, + { + "epoch": 0.18691, + "grad_norm": 1.1742573976516724, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 18691 + }, + { + "epoch": 0.18692, + "grad_norm": 1.1151024103164673, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 18692 + }, + { + "epoch": 0.18693, + "grad_norm": 0.6225581169128418, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 18693 + }, + { + "epoch": 0.18694, + "grad_norm": 0.5789169073104858, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 18694 + }, + { + "epoch": 0.18695, + "grad_norm": 0.8254528045654297, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 18695 + }, + { + "epoch": 0.18696, + "grad_norm": 1.1789621114730835, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 18696 + }, + { + "epoch": 0.18697, + "grad_norm": 1.0584477186203003, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 18697 + }, + { + "epoch": 0.18698, + "grad_norm": 0.8069787621498108, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 18698 + }, + { + "epoch": 0.18699, + "grad_norm": 0.7804241180419922, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 18699 + }, + { + "epoch": 0.187, + "grad_norm": 0.7091956734657288, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 18700 + }, + { + "epoch": 0.18701, + "grad_norm": 0.6648358106613159, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 18701 + }, + { + "epoch": 0.18702, + "grad_norm": 0.6405989527702332, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 18702 + }, + { + "epoch": 0.18703, + "grad_norm": 0.6682977080345154, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 18703 + }, + { + "epoch": 0.18704, + "grad_norm": 0.7034135460853577, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 18704 + }, + { + "epoch": 0.18705, + "grad_norm": 0.7313877940177917, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 18705 + }, + { + "epoch": 0.18706, + "grad_norm": 0.7735674977302551, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 18706 + }, + { + "epoch": 0.18707, + "grad_norm": 0.8979125618934631, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 18707 + }, + { + "epoch": 0.18708, + "grad_norm": 0.9722280502319336, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 18708 + }, + { + "epoch": 0.18709, + "grad_norm": 1.004054307937622, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 18709 + }, + { + "epoch": 0.1871, + "grad_norm": 0.8962096571922302, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 18710 + }, + { + "epoch": 0.18711, + "grad_norm": 0.7256183624267578, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 18711 + }, + { + "epoch": 0.18712, + "grad_norm": 0.7422738075256348, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 18712 + }, + { + "epoch": 0.18713, + "grad_norm": 0.7318476438522339, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 18713 + }, + { + "epoch": 0.18714, + "grad_norm": 0.9459980726242065, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 18714 + }, + { + "epoch": 0.18715, + "grad_norm": 1.1251165866851807, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 18715 + }, + { + "epoch": 0.18716, + "grad_norm": 0.9163543581962585, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 18716 + }, + { + "epoch": 0.18717, + "grad_norm": 1.009063720703125, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 18717 + }, + { + "epoch": 0.18718, + "grad_norm": 1.2291123867034912, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 18718 + }, + { + "epoch": 0.18719, + "grad_norm": 0.8537262678146362, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 18719 + }, + { + "epoch": 0.1872, + "grad_norm": 0.675536036491394, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 18720 + }, + { + "epoch": 0.18721, + "grad_norm": 0.6935260891914368, + "learning_rate": 0.003, + "loss": 3.996, + "step": 18721 + }, + { + "epoch": 0.18722, + "grad_norm": 0.7502979040145874, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 18722 + }, + { + "epoch": 0.18723, + "grad_norm": 0.8143985867500305, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 18723 + }, + { + "epoch": 0.18724, + "grad_norm": 0.8948759436607361, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 18724 + }, + { + "epoch": 0.18725, + "grad_norm": 0.8538362979888916, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 18725 + }, + { + "epoch": 0.18726, + "grad_norm": 1.0155415534973145, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 18726 + }, + { + "epoch": 0.18727, + "grad_norm": 1.1329224109649658, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 18727 + }, + { + "epoch": 0.18728, + "grad_norm": 0.8613154292106628, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 18728 + }, + { + "epoch": 0.18729, + "grad_norm": 0.8102912306785583, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 18729 + }, + { + "epoch": 0.1873, + "grad_norm": 0.720447301864624, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 18730 + }, + { + "epoch": 0.18731, + "grad_norm": 0.7747088670730591, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 18731 + }, + { + "epoch": 0.18732, + "grad_norm": 0.8366368412971497, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 18732 + }, + { + "epoch": 0.18733, + "grad_norm": 1.0016186237335205, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 18733 + }, + { + "epoch": 0.18734, + "grad_norm": 1.1299272775650024, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 18734 + }, + { + "epoch": 0.18735, + "grad_norm": 0.7979495525360107, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 18735 + }, + { + "epoch": 0.18736, + "grad_norm": 0.7674964666366577, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 18736 + }, + { + "epoch": 0.18737, + "grad_norm": 0.8326266407966614, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 18737 + }, + { + "epoch": 0.18738, + "grad_norm": 1.0282922983169556, + "learning_rate": 0.003, + "loss": 3.992, + "step": 18738 + }, + { + "epoch": 0.18739, + "grad_norm": 1.325468897819519, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 18739 + }, + { + "epoch": 0.1874, + "grad_norm": 0.7673645615577698, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 18740 + }, + { + "epoch": 0.18741, + "grad_norm": 0.6764647960662842, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 18741 + }, + { + "epoch": 0.18742, + "grad_norm": 0.8184046149253845, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 18742 + }, + { + "epoch": 0.18743, + "grad_norm": 0.7239495515823364, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 18743 + }, + { + "epoch": 0.18744, + "grad_norm": 0.6987858414649963, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 18744 + }, + { + "epoch": 0.18745, + "grad_norm": 0.8750888109207153, + "learning_rate": 0.003, + "loss": 3.987, + "step": 18745 + }, + { + "epoch": 0.18746, + "grad_norm": 0.8903581500053406, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 18746 + }, + { + "epoch": 0.18747, + "grad_norm": 0.8389039635658264, + "learning_rate": 0.003, + "loss": 4.028, + "step": 18747 + }, + { + "epoch": 0.18748, + "grad_norm": 0.8459964394569397, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 18748 + }, + { + "epoch": 0.18749, + "grad_norm": 0.8524193167686462, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 18749 + }, + { + "epoch": 0.1875, + "grad_norm": 0.8610021471977234, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 18750 + }, + { + "epoch": 0.18751, + "grad_norm": 0.9103726744651794, + "learning_rate": 0.003, + "loss": 4.003, + "step": 18751 + }, + { + "epoch": 0.18752, + "grad_norm": 0.8229556679725647, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 18752 + }, + { + "epoch": 0.18753, + "grad_norm": 0.8255165219306946, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 18753 + }, + { + "epoch": 0.18754, + "grad_norm": 0.7489179372787476, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 18754 + }, + { + "epoch": 0.18755, + "grad_norm": 0.8602535128593445, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 18755 + }, + { + "epoch": 0.18756, + "grad_norm": 0.8368017673492432, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 18756 + }, + { + "epoch": 0.18757, + "grad_norm": 0.7005070447921753, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 18757 + }, + { + "epoch": 0.18758, + "grad_norm": 0.7822446823120117, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 18758 + }, + { + "epoch": 0.18759, + "grad_norm": 0.775924026966095, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 18759 + }, + { + "epoch": 0.1876, + "grad_norm": 0.7569155693054199, + "learning_rate": 0.003, + "loss": 3.982, + "step": 18760 + }, + { + "epoch": 0.18761, + "grad_norm": 0.7290402054786682, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 18761 + }, + { + "epoch": 0.18762, + "grad_norm": 0.7267779111862183, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 18762 + }, + { + "epoch": 0.18763, + "grad_norm": 0.6764037013053894, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 18763 + }, + { + "epoch": 0.18764, + "grad_norm": 0.7469682693481445, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 18764 + }, + { + "epoch": 0.18765, + "grad_norm": 0.9405340552330017, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 18765 + }, + { + "epoch": 0.18766, + "grad_norm": 1.1264852285385132, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 18766 + }, + { + "epoch": 0.18767, + "grad_norm": 1.2135547399520874, + "learning_rate": 0.003, + "loss": 4.02, + "step": 18767 + }, + { + "epoch": 0.18768, + "grad_norm": 0.7948119044303894, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 18768 + }, + { + "epoch": 0.18769, + "grad_norm": 0.8168572783470154, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 18769 + }, + { + "epoch": 0.1877, + "grad_norm": 0.8725208044052124, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 18770 + }, + { + "epoch": 0.18771, + "grad_norm": 0.886518120765686, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 18771 + }, + { + "epoch": 0.18772, + "grad_norm": 0.8297489881515503, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 18772 + }, + { + "epoch": 0.18773, + "grad_norm": 0.7882468700408936, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 18773 + }, + { + "epoch": 0.18774, + "grad_norm": 0.7615990042686462, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 18774 + }, + { + "epoch": 0.18775, + "grad_norm": 0.7042850255966187, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 18775 + }, + { + "epoch": 0.18776, + "grad_norm": 0.6992802619934082, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 18776 + }, + { + "epoch": 0.18777, + "grad_norm": 0.7362062931060791, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 18777 + }, + { + "epoch": 0.18778, + "grad_norm": 0.991096556186676, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 18778 + }, + { + "epoch": 0.18779, + "grad_norm": 1.3604698181152344, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 18779 + }, + { + "epoch": 0.1878, + "grad_norm": 0.6573266386985779, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 18780 + }, + { + "epoch": 0.18781, + "grad_norm": 0.6684979796409607, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 18781 + }, + { + "epoch": 0.18782, + "grad_norm": 0.6320002675056458, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 18782 + }, + { + "epoch": 0.18783, + "grad_norm": 0.675640881061554, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 18783 + }, + { + "epoch": 0.18784, + "grad_norm": 0.670727014541626, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 18784 + }, + { + "epoch": 0.18785, + "grad_norm": 0.6947329640388489, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 18785 + }, + { + "epoch": 0.18786, + "grad_norm": 0.813109815120697, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 18786 + }, + { + "epoch": 0.18787, + "grad_norm": 0.9462290406227112, + "learning_rate": 0.003, + "loss": 3.997, + "step": 18787 + }, + { + "epoch": 0.18788, + "grad_norm": 1.2447432279586792, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 18788 + }, + { + "epoch": 0.18789, + "grad_norm": 0.8113349676132202, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 18789 + }, + { + "epoch": 0.1879, + "grad_norm": 0.6997334361076355, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 18790 + }, + { + "epoch": 0.18791, + "grad_norm": 0.7395490407943726, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 18791 + }, + { + "epoch": 0.18792, + "grad_norm": 0.653052031993866, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 18792 + }, + { + "epoch": 0.18793, + "grad_norm": 0.6461042761802673, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 18793 + }, + { + "epoch": 0.18794, + "grad_norm": 0.8122109770774841, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 18794 + }, + { + "epoch": 0.18795, + "grad_norm": 1.1163297891616821, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 18795 + }, + { + "epoch": 0.18796, + "grad_norm": 1.1696635484695435, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 18796 + }, + { + "epoch": 0.18797, + "grad_norm": 0.8114674091339111, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 18797 + }, + { + "epoch": 0.18798, + "grad_norm": 0.7650184035301208, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 18798 + }, + { + "epoch": 0.18799, + "grad_norm": 0.8254711031913757, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 18799 + }, + { + "epoch": 0.188, + "grad_norm": 0.9358893632888794, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 18800 + }, + { + "epoch": 0.18801, + "grad_norm": 1.0084511041641235, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 18801 + }, + { + "epoch": 0.18802, + "grad_norm": 0.9665206670761108, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 18802 + }, + { + "epoch": 0.18803, + "grad_norm": 0.9788021445274353, + "learning_rate": 0.003, + "loss": 4.014, + "step": 18803 + }, + { + "epoch": 0.18804, + "grad_norm": 0.9931485652923584, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 18804 + }, + { + "epoch": 0.18805, + "grad_norm": 1.2854509353637695, + "learning_rate": 0.003, + "loss": 4.047, + "step": 18805 + }, + { + "epoch": 0.18806, + "grad_norm": 0.9651660919189453, + "learning_rate": 0.003, + "loss": 4.021, + "step": 18806 + }, + { + "epoch": 0.18807, + "grad_norm": 1.035075068473816, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 18807 + }, + { + "epoch": 0.18808, + "grad_norm": 1.0494685173034668, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 18808 + }, + { + "epoch": 0.18809, + "grad_norm": 0.9226809144020081, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 18809 + }, + { + "epoch": 0.1881, + "grad_norm": 0.9085474014282227, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 18810 + }, + { + "epoch": 0.18811, + "grad_norm": 0.9232672452926636, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 18811 + }, + { + "epoch": 0.18812, + "grad_norm": 0.8821086883544922, + "learning_rate": 0.003, + "loss": 3.981, + "step": 18812 + }, + { + "epoch": 0.18813, + "grad_norm": 0.783420205116272, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 18813 + }, + { + "epoch": 0.18814, + "grad_norm": 0.7733287215232849, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 18814 + }, + { + "epoch": 0.18815, + "grad_norm": 0.6677088737487793, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 18815 + }, + { + "epoch": 0.18816, + "grad_norm": 0.6723862886428833, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 18816 + }, + { + "epoch": 0.18817, + "grad_norm": 0.9003461003303528, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 18817 + }, + { + "epoch": 0.18818, + "grad_norm": 1.1270079612731934, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 18818 + }, + { + "epoch": 0.18819, + "grad_norm": 0.8880910873413086, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 18819 + }, + { + "epoch": 0.1882, + "grad_norm": 0.7390077710151672, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 18820 + }, + { + "epoch": 0.18821, + "grad_norm": 0.5664216876029968, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 18821 + }, + { + "epoch": 0.18822, + "grad_norm": 0.5701670050621033, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 18822 + }, + { + "epoch": 0.18823, + "grad_norm": 0.5538231134414673, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 18823 + }, + { + "epoch": 0.18824, + "grad_norm": 0.5930655598640442, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 18824 + }, + { + "epoch": 0.18825, + "grad_norm": 0.6896167397499084, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 18825 + }, + { + "epoch": 0.18826, + "grad_norm": 0.7935672998428345, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 18826 + }, + { + "epoch": 0.18827, + "grad_norm": 0.7783662676811218, + "learning_rate": 0.003, + "loss": 4.009, + "step": 18827 + }, + { + "epoch": 0.18828, + "grad_norm": 0.838836133480072, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 18828 + }, + { + "epoch": 0.18829, + "grad_norm": 1.024140477180481, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 18829 + }, + { + "epoch": 0.1883, + "grad_norm": 1.1629550457000732, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 18830 + }, + { + "epoch": 0.18831, + "grad_norm": 0.7892631888389587, + "learning_rate": 0.003, + "loss": 3.975, + "step": 18831 + }, + { + "epoch": 0.18832, + "grad_norm": 0.7690302729606628, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 18832 + }, + { + "epoch": 0.18833, + "grad_norm": 0.8231732845306396, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 18833 + }, + { + "epoch": 0.18834, + "grad_norm": 0.766808271408081, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 18834 + }, + { + "epoch": 0.18835, + "grad_norm": 0.8518886566162109, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 18835 + }, + { + "epoch": 0.18836, + "grad_norm": 0.8174446821212769, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 18836 + }, + { + "epoch": 0.18837, + "grad_norm": 0.7299855351448059, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 18837 + }, + { + "epoch": 0.18838, + "grad_norm": 0.8137608766555786, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 18838 + }, + { + "epoch": 0.18839, + "grad_norm": 0.8550666570663452, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 18839 + }, + { + "epoch": 0.1884, + "grad_norm": 0.9712225198745728, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 18840 + }, + { + "epoch": 0.18841, + "grad_norm": 1.0288853645324707, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 18841 + }, + { + "epoch": 0.18842, + "grad_norm": 1.2778677940368652, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 18842 + }, + { + "epoch": 0.18843, + "grad_norm": 0.6840931177139282, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 18843 + }, + { + "epoch": 0.18844, + "grad_norm": 0.6637877225875854, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 18844 + }, + { + "epoch": 0.18845, + "grad_norm": 0.6504637002944946, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 18845 + }, + { + "epoch": 0.18846, + "grad_norm": 0.726256787776947, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 18846 + }, + { + "epoch": 0.18847, + "grad_norm": 0.9666273593902588, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 18847 + }, + { + "epoch": 0.18848, + "grad_norm": 0.9877365827560425, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 18848 + }, + { + "epoch": 0.18849, + "grad_norm": 0.9723947048187256, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 18849 + }, + { + "epoch": 0.1885, + "grad_norm": 1.0214213132858276, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 18850 + }, + { + "epoch": 0.18851, + "grad_norm": 0.7957757711410522, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 18851 + }, + { + "epoch": 0.18852, + "grad_norm": 0.7545115947723389, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 18852 + }, + { + "epoch": 0.18853, + "grad_norm": 0.743475615978241, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 18853 + }, + { + "epoch": 0.18854, + "grad_norm": 0.7947479486465454, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 18854 + }, + { + "epoch": 0.18855, + "grad_norm": 0.7675407528877258, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 18855 + }, + { + "epoch": 0.18856, + "grad_norm": 0.8497804999351501, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 18856 + }, + { + "epoch": 0.18857, + "grad_norm": 1.2066291570663452, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 18857 + }, + { + "epoch": 0.18858, + "grad_norm": 1.2024381160736084, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 18858 + }, + { + "epoch": 0.18859, + "grad_norm": 0.6818134188652039, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 18859 + }, + { + "epoch": 0.1886, + "grad_norm": 0.6120265126228333, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 18860 + }, + { + "epoch": 0.18861, + "grad_norm": 0.7907565236091614, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 18861 + }, + { + "epoch": 0.18862, + "grad_norm": 1.1322532892227173, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 18862 + }, + { + "epoch": 0.18863, + "grad_norm": 1.09321129322052, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 18863 + }, + { + "epoch": 0.18864, + "grad_norm": 0.7416861057281494, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 18864 + }, + { + "epoch": 0.18865, + "grad_norm": 0.6728402376174927, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 18865 + }, + { + "epoch": 0.18866, + "grad_norm": 0.7465229630470276, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 18866 + }, + { + "epoch": 0.18867, + "grad_norm": 0.8051891922950745, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 18867 + }, + { + "epoch": 0.18868, + "grad_norm": 0.8132879137992859, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 18868 + }, + { + "epoch": 0.18869, + "grad_norm": 0.7474648356437683, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 18869 + }, + { + "epoch": 0.1887, + "grad_norm": 0.7271384000778198, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 18870 + }, + { + "epoch": 0.18871, + "grad_norm": 0.7931719422340393, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 18871 + }, + { + "epoch": 0.18872, + "grad_norm": 0.7534982562065125, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 18872 + }, + { + "epoch": 0.18873, + "grad_norm": 0.7345370054244995, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 18873 + }, + { + "epoch": 0.18874, + "grad_norm": 0.7610151767730713, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 18874 + }, + { + "epoch": 0.18875, + "grad_norm": 0.7372151613235474, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 18875 + }, + { + "epoch": 0.18876, + "grad_norm": 0.8092154264450073, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 18876 + }, + { + "epoch": 0.18877, + "grad_norm": 1.111495852470398, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 18877 + }, + { + "epoch": 0.18878, + "grad_norm": 1.1198842525482178, + "learning_rate": 0.003, + "loss": 3.983, + "step": 18878 + }, + { + "epoch": 0.18879, + "grad_norm": 0.9886401295661926, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 18879 + }, + { + "epoch": 0.1888, + "grad_norm": 1.0660293102264404, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 18880 + }, + { + "epoch": 0.18881, + "grad_norm": 0.9938003420829773, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 18881 + }, + { + "epoch": 0.18882, + "grad_norm": 0.8295101523399353, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 18882 + }, + { + "epoch": 0.18883, + "grad_norm": 0.7040286064147949, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 18883 + }, + { + "epoch": 0.18884, + "grad_norm": 0.7084823250770569, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 18884 + }, + { + "epoch": 0.18885, + "grad_norm": 0.7420821189880371, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 18885 + }, + { + "epoch": 0.18886, + "grad_norm": 0.7304326891899109, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 18886 + }, + { + "epoch": 0.18887, + "grad_norm": 0.8151340484619141, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 18887 + }, + { + "epoch": 0.18888, + "grad_norm": 1.025665283203125, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 18888 + }, + { + "epoch": 0.18889, + "grad_norm": 1.026984691619873, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 18889 + }, + { + "epoch": 0.1889, + "grad_norm": 0.9665703177452087, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 18890 + }, + { + "epoch": 0.18891, + "grad_norm": 0.8499642014503479, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 18891 + }, + { + "epoch": 0.18892, + "grad_norm": 0.7022868990898132, + "learning_rate": 0.003, + "loss": 4.005, + "step": 18892 + }, + { + "epoch": 0.18893, + "grad_norm": 0.7038338780403137, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 18893 + }, + { + "epoch": 0.18894, + "grad_norm": 0.7314153909683228, + "learning_rate": 0.003, + "loss": 3.994, + "step": 18894 + }, + { + "epoch": 0.18895, + "grad_norm": 0.7689028382301331, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 18895 + }, + { + "epoch": 0.18896, + "grad_norm": 0.6190382838249207, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 18896 + }, + { + "epoch": 0.18897, + "grad_norm": 0.6298418045043945, + "learning_rate": 0.003, + "loss": 3.977, + "step": 18897 + }, + { + "epoch": 0.18898, + "grad_norm": 0.6815434098243713, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 18898 + }, + { + "epoch": 0.18899, + "grad_norm": 0.7141243815422058, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 18899 + }, + { + "epoch": 0.189, + "grad_norm": 0.9865851402282715, + "learning_rate": 0.003, + "loss": 3.978, + "step": 18900 + }, + { + "epoch": 0.18901, + "grad_norm": 1.3485716581344604, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 18901 + }, + { + "epoch": 0.18902, + "grad_norm": 0.6510710120201111, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 18902 + }, + { + "epoch": 0.18903, + "grad_norm": 0.6483829021453857, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 18903 + }, + { + "epoch": 0.18904, + "grad_norm": 0.7830336689949036, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 18904 + }, + { + "epoch": 0.18905, + "grad_norm": 1.0328387022018433, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 18905 + }, + { + "epoch": 0.18906, + "grad_norm": 1.0227880477905273, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 18906 + }, + { + "epoch": 0.18907, + "grad_norm": 0.9766803979873657, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 18907 + }, + { + "epoch": 0.18908, + "grad_norm": 0.9505747556686401, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 18908 + }, + { + "epoch": 0.18909, + "grad_norm": 0.9789347052574158, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 18909 + }, + { + "epoch": 0.1891, + "grad_norm": 1.2579947710037231, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 18910 + }, + { + "epoch": 0.18911, + "grad_norm": 0.6997345685958862, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 18911 + }, + { + "epoch": 0.18912, + "grad_norm": 0.741587221622467, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 18912 + }, + { + "epoch": 0.18913, + "grad_norm": 0.8010073900222778, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 18913 + }, + { + "epoch": 0.18914, + "grad_norm": 0.8004756569862366, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 18914 + }, + { + "epoch": 0.18915, + "grad_norm": 0.8011345863342285, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 18915 + }, + { + "epoch": 0.18916, + "grad_norm": 0.7441811561584473, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 18916 + }, + { + "epoch": 0.18917, + "grad_norm": 0.6722574234008789, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 18917 + }, + { + "epoch": 0.18918, + "grad_norm": 0.702165961265564, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 18918 + }, + { + "epoch": 0.18919, + "grad_norm": 0.7118338346481323, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 18919 + }, + { + "epoch": 0.1892, + "grad_norm": 0.6990228295326233, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 18920 + }, + { + "epoch": 0.18921, + "grad_norm": 0.8419510722160339, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 18921 + }, + { + "epoch": 0.18922, + "grad_norm": 1.246588110923767, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 18922 + }, + { + "epoch": 0.18923, + "grad_norm": 0.9596294164657593, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 18923 + }, + { + "epoch": 0.18924, + "grad_norm": 0.8327478766441345, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 18924 + }, + { + "epoch": 0.18925, + "grad_norm": 0.856834352016449, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 18925 + }, + { + "epoch": 0.18926, + "grad_norm": 0.8613009452819824, + "learning_rate": 0.003, + "loss": 4.038, + "step": 18926 + }, + { + "epoch": 0.18927, + "grad_norm": 0.861362874507904, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 18927 + }, + { + "epoch": 0.18928, + "grad_norm": 0.8827495574951172, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 18928 + }, + { + "epoch": 0.18929, + "grad_norm": 0.9191708564758301, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 18929 + }, + { + "epoch": 0.1893, + "grad_norm": 1.1141074895858765, + "learning_rate": 0.003, + "loss": 3.991, + "step": 18930 + }, + { + "epoch": 0.18931, + "grad_norm": 0.9931434392929077, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 18931 + }, + { + "epoch": 0.18932, + "grad_norm": 0.8666828274726868, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 18932 + }, + { + "epoch": 0.18933, + "grad_norm": 0.7398908138275146, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 18933 + }, + { + "epoch": 0.18934, + "grad_norm": 0.6053960919380188, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 18934 + }, + { + "epoch": 0.18935, + "grad_norm": 0.5848657488822937, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 18935 + }, + { + "epoch": 0.18936, + "grad_norm": 0.6164701581001282, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 18936 + }, + { + "epoch": 0.18937, + "grad_norm": 0.6691602468490601, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 18937 + }, + { + "epoch": 0.18938, + "grad_norm": 0.7680910229682922, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 18938 + }, + { + "epoch": 0.18939, + "grad_norm": 0.906510055065155, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 18939 + }, + { + "epoch": 0.1894, + "grad_norm": 1.1088699102401733, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 18940 + }, + { + "epoch": 0.18941, + "grad_norm": 0.809749960899353, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 18941 + }, + { + "epoch": 0.18942, + "grad_norm": 0.7769731283187866, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 18942 + }, + { + "epoch": 0.18943, + "grad_norm": 0.7073752284049988, + "learning_rate": 0.003, + "loss": 3.987, + "step": 18943 + }, + { + "epoch": 0.18944, + "grad_norm": 0.6503266096115112, + "learning_rate": 0.003, + "loss": 3.987, + "step": 18944 + }, + { + "epoch": 0.18945, + "grad_norm": 0.6349859237670898, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 18945 + }, + { + "epoch": 0.18946, + "grad_norm": 0.6565268039703369, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 18946 + }, + { + "epoch": 0.18947, + "grad_norm": 0.689110517501831, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 18947 + }, + { + "epoch": 0.18948, + "grad_norm": 0.6516901850700378, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 18948 + }, + { + "epoch": 0.18949, + "grad_norm": 0.6973980665206909, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 18949 + }, + { + "epoch": 0.1895, + "grad_norm": 0.8721265196800232, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 18950 + }, + { + "epoch": 0.18951, + "grad_norm": 1.3417309522628784, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 18951 + }, + { + "epoch": 0.18952, + "grad_norm": 0.8508768677711487, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 18952 + }, + { + "epoch": 0.18953, + "grad_norm": 0.8588517308235168, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 18953 + }, + { + "epoch": 0.18954, + "grad_norm": 0.9401569962501526, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 18954 + }, + { + "epoch": 0.18955, + "grad_norm": 0.9512640833854675, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 18955 + }, + { + "epoch": 0.18956, + "grad_norm": 1.1748946905136108, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 18956 + }, + { + "epoch": 0.18957, + "grad_norm": 0.8970866203308105, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 18957 + }, + { + "epoch": 0.18958, + "grad_norm": 0.970586359500885, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 18958 + }, + { + "epoch": 0.18959, + "grad_norm": 1.0547508001327515, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 18959 + }, + { + "epoch": 0.1896, + "grad_norm": 1.1379255056381226, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 18960 + }, + { + "epoch": 0.18961, + "grad_norm": 1.0756850242614746, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 18961 + }, + { + "epoch": 0.18962, + "grad_norm": 0.9282955527305603, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 18962 + }, + { + "epoch": 0.18963, + "grad_norm": 0.9438974261283875, + "learning_rate": 0.003, + "loss": 4.018, + "step": 18963 + }, + { + "epoch": 0.18964, + "grad_norm": 0.9445635080337524, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 18964 + }, + { + "epoch": 0.18965, + "grad_norm": 1.079493761062622, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 18965 + }, + { + "epoch": 0.18966, + "grad_norm": 1.120069146156311, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 18966 + }, + { + "epoch": 0.18967, + "grad_norm": 0.7256782054901123, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 18967 + }, + { + "epoch": 0.18968, + "grad_norm": 0.6131082773208618, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 18968 + }, + { + "epoch": 0.18969, + "grad_norm": 0.6501275897026062, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 18969 + }, + { + "epoch": 0.1897, + "grad_norm": 0.6911957859992981, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 18970 + }, + { + "epoch": 0.18971, + "grad_norm": 0.653016984462738, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 18971 + }, + { + "epoch": 0.18972, + "grad_norm": 0.5951827168464661, + "learning_rate": 0.003, + "loss": 4.001, + "step": 18972 + }, + { + "epoch": 0.18973, + "grad_norm": 0.5641265511512756, + "learning_rate": 0.003, + "loss": 3.968, + "step": 18973 + }, + { + "epoch": 0.18974, + "grad_norm": 0.615556538105011, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 18974 + }, + { + "epoch": 0.18975, + "grad_norm": 0.5846765637397766, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 18975 + }, + { + "epoch": 0.18976, + "grad_norm": 0.642020046710968, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 18976 + }, + { + "epoch": 0.18977, + "grad_norm": 0.6776500344276428, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 18977 + }, + { + "epoch": 0.18978, + "grad_norm": 0.8168145418167114, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 18978 + }, + { + "epoch": 0.18979, + "grad_norm": 1.0492795705795288, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 18979 + }, + { + "epoch": 0.1898, + "grad_norm": 1.2692487239837646, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 18980 + }, + { + "epoch": 0.18981, + "grad_norm": 0.7477732300758362, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 18981 + }, + { + "epoch": 0.18982, + "grad_norm": 0.7049261331558228, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 18982 + }, + { + "epoch": 0.18983, + "grad_norm": 0.8238551020622253, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 18983 + }, + { + "epoch": 0.18984, + "grad_norm": 0.8532552719116211, + "learning_rate": 0.003, + "loss": 4.005, + "step": 18984 + }, + { + "epoch": 0.18985, + "grad_norm": 0.854641854763031, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 18985 + }, + { + "epoch": 0.18986, + "grad_norm": 0.9322021007537842, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 18986 + }, + { + "epoch": 0.18987, + "grad_norm": 1.0125163793563843, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 18987 + }, + { + "epoch": 0.18988, + "grad_norm": 1.0155051946640015, + "learning_rate": 0.003, + "loss": 4.007, + "step": 18988 + }, + { + "epoch": 0.18989, + "grad_norm": 0.958838939666748, + "learning_rate": 0.003, + "loss": 4.003, + "step": 18989 + }, + { + "epoch": 0.1899, + "grad_norm": 1.0734632015228271, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 18990 + }, + { + "epoch": 0.18991, + "grad_norm": 0.9874166250228882, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 18991 + }, + { + "epoch": 0.18992, + "grad_norm": 1.1062909364700317, + "learning_rate": 0.003, + "loss": 3.997, + "step": 18992 + }, + { + "epoch": 0.18993, + "grad_norm": 0.9920216798782349, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 18993 + }, + { + "epoch": 0.18994, + "grad_norm": 0.9129654765129089, + "learning_rate": 0.003, + "loss": 4.007, + "step": 18994 + }, + { + "epoch": 0.18995, + "grad_norm": 0.9695618748664856, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 18995 + }, + { + "epoch": 0.18996, + "grad_norm": 1.1071839332580566, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 18996 + }, + { + "epoch": 0.18997, + "grad_norm": 0.8101666569709778, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 18997 + }, + { + "epoch": 0.18998, + "grad_norm": 0.7454321980476379, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 18998 + }, + { + "epoch": 0.18999, + "grad_norm": 0.76347416639328, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 18999 + }, + { + "epoch": 0.19, + "grad_norm": 0.676956057548523, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 19000 + }, + { + "epoch": 0.19001, + "grad_norm": 0.5809410214424133, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 19001 + }, + { + "epoch": 0.19002, + "grad_norm": 0.6385517120361328, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 19002 + }, + { + "epoch": 0.19003, + "grad_norm": 0.72422856092453, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 19003 + }, + { + "epoch": 0.19004, + "grad_norm": 0.8340731263160706, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 19004 + }, + { + "epoch": 0.19005, + "grad_norm": 0.9573755860328674, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 19005 + }, + { + "epoch": 0.19006, + "grad_norm": 1.1783071756362915, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 19006 + }, + { + "epoch": 0.19007, + "grad_norm": 0.8298691511154175, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 19007 + }, + { + "epoch": 0.19008, + "grad_norm": 0.6903021931648254, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 19008 + }, + { + "epoch": 0.19009, + "grad_norm": 0.6378114223480225, + "learning_rate": 0.003, + "loss": 4.013, + "step": 19009 + }, + { + "epoch": 0.1901, + "grad_norm": 0.6948537230491638, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 19010 + }, + { + "epoch": 0.19011, + "grad_norm": 0.8318427801132202, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 19011 + }, + { + "epoch": 0.19012, + "grad_norm": 1.0148303508758545, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 19012 + }, + { + "epoch": 0.19013, + "grad_norm": 1.0587249994277954, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 19013 + }, + { + "epoch": 0.19014, + "grad_norm": 0.9003660082817078, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 19014 + }, + { + "epoch": 0.19015, + "grad_norm": 0.869741678237915, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 19015 + }, + { + "epoch": 0.19016, + "grad_norm": 0.8709609508514404, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 19016 + }, + { + "epoch": 0.19017, + "grad_norm": 0.8974391222000122, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 19017 + }, + { + "epoch": 0.19018, + "grad_norm": 0.9310207962989807, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 19018 + }, + { + "epoch": 0.19019, + "grad_norm": 0.9279577136039734, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 19019 + }, + { + "epoch": 0.1902, + "grad_norm": 1.0117379426956177, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 19020 + }, + { + "epoch": 0.19021, + "grad_norm": 1.0991427898406982, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 19021 + }, + { + "epoch": 0.19022, + "grad_norm": 0.9016971588134766, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 19022 + }, + { + "epoch": 0.19023, + "grad_norm": 0.9247835278511047, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 19023 + }, + { + "epoch": 0.19024, + "grad_norm": 1.007381558418274, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 19024 + }, + { + "epoch": 0.19025, + "grad_norm": 1.206356167793274, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 19025 + }, + { + "epoch": 0.19026, + "grad_norm": 0.7278400659561157, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 19026 + }, + { + "epoch": 0.19027, + "grad_norm": 0.6956689953804016, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 19027 + }, + { + "epoch": 0.19028, + "grad_norm": 0.6170609593391418, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 19028 + }, + { + "epoch": 0.19029, + "grad_norm": 0.5841855406761169, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 19029 + }, + { + "epoch": 0.1903, + "grad_norm": 0.615669846534729, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 19030 + }, + { + "epoch": 0.19031, + "grad_norm": 0.7128541469573975, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 19031 + }, + { + "epoch": 0.19032, + "grad_norm": 0.9266338348388672, + "learning_rate": 0.003, + "loss": 4.0, + "step": 19032 + }, + { + "epoch": 0.19033, + "grad_norm": 1.13957679271698, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 19033 + }, + { + "epoch": 0.19034, + "grad_norm": 0.836190938949585, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 19034 + }, + { + "epoch": 0.19035, + "grad_norm": 0.6466774344444275, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 19035 + }, + { + "epoch": 0.19036, + "grad_norm": 0.5852211713790894, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 19036 + }, + { + "epoch": 0.19037, + "grad_norm": 0.7002070546150208, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 19037 + }, + { + "epoch": 0.19038, + "grad_norm": 0.7181115746498108, + "learning_rate": 0.003, + "loss": 4.013, + "step": 19038 + }, + { + "epoch": 0.19039, + "grad_norm": 0.696847140789032, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 19039 + }, + { + "epoch": 0.1904, + "grad_norm": 0.7739638090133667, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 19040 + }, + { + "epoch": 0.19041, + "grad_norm": 0.8265454173088074, + "learning_rate": 0.003, + "loss": 3.992, + "step": 19041 + }, + { + "epoch": 0.19042, + "grad_norm": 0.9292768836021423, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 19042 + }, + { + "epoch": 0.19043, + "grad_norm": 1.05363130569458, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 19043 + }, + { + "epoch": 0.19044, + "grad_norm": 0.8629478216171265, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 19044 + }, + { + "epoch": 0.19045, + "grad_norm": 0.7886390686035156, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 19045 + }, + { + "epoch": 0.19046, + "grad_norm": 0.8244084715843201, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 19046 + }, + { + "epoch": 0.19047, + "grad_norm": 0.8120138049125671, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 19047 + }, + { + "epoch": 0.19048, + "grad_norm": 0.8467174768447876, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 19048 + }, + { + "epoch": 0.19049, + "grad_norm": 0.9316357374191284, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 19049 + }, + { + "epoch": 0.1905, + "grad_norm": 0.9798911809921265, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 19050 + }, + { + "epoch": 0.19051, + "grad_norm": 0.8504247665405273, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 19051 + }, + { + "epoch": 0.19052, + "grad_norm": 0.748024046421051, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 19052 + }, + { + "epoch": 0.19053, + "grad_norm": 0.7716191411018372, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 19053 + }, + { + "epoch": 0.19054, + "grad_norm": 0.7982362508773804, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 19054 + }, + { + "epoch": 0.19055, + "grad_norm": 0.8361937403678894, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 19055 + }, + { + "epoch": 0.19056, + "grad_norm": 1.080262541770935, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 19056 + }, + { + "epoch": 0.19057, + "grad_norm": 1.1439944505691528, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 19057 + }, + { + "epoch": 0.19058, + "grad_norm": 0.9121918678283691, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 19058 + }, + { + "epoch": 0.19059, + "grad_norm": 0.9394298791885376, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 19059 + }, + { + "epoch": 0.1906, + "grad_norm": 0.9266657829284668, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 19060 + }, + { + "epoch": 0.19061, + "grad_norm": 0.881291389465332, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 19061 + }, + { + "epoch": 0.19062, + "grad_norm": 0.8577966690063477, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 19062 + }, + { + "epoch": 0.19063, + "grad_norm": 1.003316879272461, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 19063 + }, + { + "epoch": 0.19064, + "grad_norm": 1.2838621139526367, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 19064 + }, + { + "epoch": 0.19065, + "grad_norm": 0.8092989921569824, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 19065 + }, + { + "epoch": 0.19066, + "grad_norm": 0.6843506097793579, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 19066 + }, + { + "epoch": 0.19067, + "grad_norm": 0.5812391638755798, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 19067 + }, + { + "epoch": 0.19068, + "grad_norm": 0.5177496075630188, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 19068 + }, + { + "epoch": 0.19069, + "grad_norm": 0.49588555097579956, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 19069 + }, + { + "epoch": 0.1907, + "grad_norm": 0.580472469329834, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 19070 + }, + { + "epoch": 0.19071, + "grad_norm": 0.5863806009292603, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 19071 + }, + { + "epoch": 0.19072, + "grad_norm": 0.6271872520446777, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 19072 + }, + { + "epoch": 0.19073, + "grad_norm": 0.7241467237472534, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 19073 + }, + { + "epoch": 0.19074, + "grad_norm": 0.9060819149017334, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 19074 + }, + { + "epoch": 0.19075, + "grad_norm": 1.0471937656402588, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 19075 + }, + { + "epoch": 0.19076, + "grad_norm": 1.0599033832550049, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 19076 + }, + { + "epoch": 0.19077, + "grad_norm": 0.9345073699951172, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 19077 + }, + { + "epoch": 0.19078, + "grad_norm": 0.8938175439834595, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 19078 + }, + { + "epoch": 0.19079, + "grad_norm": 0.9085829257965088, + "learning_rate": 0.003, + "loss": 4.019, + "step": 19079 + }, + { + "epoch": 0.1908, + "grad_norm": 0.8755341172218323, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 19080 + }, + { + "epoch": 0.19081, + "grad_norm": 0.9034981727600098, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 19081 + }, + { + "epoch": 0.19082, + "grad_norm": 1.0624445676803589, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 19082 + }, + { + "epoch": 0.19083, + "grad_norm": 1.2125186920166016, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 19083 + }, + { + "epoch": 0.19084, + "grad_norm": 0.8801583647727966, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 19084 + }, + { + "epoch": 0.19085, + "grad_norm": 0.8632391691207886, + "learning_rate": 0.003, + "loss": 4.012, + "step": 19085 + }, + { + "epoch": 0.19086, + "grad_norm": 0.7941170930862427, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 19086 + }, + { + "epoch": 0.19087, + "grad_norm": 0.7477037906646729, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 19087 + }, + { + "epoch": 0.19088, + "grad_norm": 0.6981642246246338, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 19088 + }, + { + "epoch": 0.19089, + "grad_norm": 0.7463194131851196, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 19089 + }, + { + "epoch": 0.1909, + "grad_norm": 0.904951810836792, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 19090 + }, + { + "epoch": 0.19091, + "grad_norm": 0.9113186001777649, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 19091 + }, + { + "epoch": 0.19092, + "grad_norm": 0.7952855825424194, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 19092 + }, + { + "epoch": 0.19093, + "grad_norm": 0.8860151171684265, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 19093 + }, + { + "epoch": 0.19094, + "grad_norm": 0.8939390778541565, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 19094 + }, + { + "epoch": 0.19095, + "grad_norm": 0.8042697906494141, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 19095 + }, + { + "epoch": 0.19096, + "grad_norm": 0.8192028999328613, + "learning_rate": 0.003, + "loss": 4.017, + "step": 19096 + }, + { + "epoch": 0.19097, + "grad_norm": 0.9775270223617554, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 19097 + }, + { + "epoch": 0.19098, + "grad_norm": 1.206665277481079, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 19098 + }, + { + "epoch": 0.19099, + "grad_norm": 0.8887525200843811, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 19099 + }, + { + "epoch": 0.191, + "grad_norm": 0.9043484330177307, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 19100 + }, + { + "epoch": 0.19101, + "grad_norm": 0.9640260338783264, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 19101 + }, + { + "epoch": 0.19102, + "grad_norm": 0.8492767810821533, + "learning_rate": 0.003, + "loss": 3.988, + "step": 19102 + }, + { + "epoch": 0.19103, + "grad_norm": 0.8072956204414368, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 19103 + }, + { + "epoch": 0.19104, + "grad_norm": 0.8654136657714844, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 19104 + }, + { + "epoch": 0.19105, + "grad_norm": 0.8672546148300171, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 19105 + }, + { + "epoch": 0.19106, + "grad_norm": 0.7615546584129333, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 19106 + }, + { + "epoch": 0.19107, + "grad_norm": 0.8596553802490234, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 19107 + }, + { + "epoch": 0.19108, + "grad_norm": 0.8763645887374878, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 19108 + }, + { + "epoch": 0.19109, + "grad_norm": 0.880420446395874, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 19109 + }, + { + "epoch": 0.1911, + "grad_norm": 0.8416774272918701, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 19110 + }, + { + "epoch": 0.19111, + "grad_norm": 0.7089952826499939, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 19111 + }, + { + "epoch": 0.19112, + "grad_norm": 0.6287502646446228, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 19112 + }, + { + "epoch": 0.19113, + "grad_norm": 0.6182113289833069, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 19113 + }, + { + "epoch": 0.19114, + "grad_norm": 0.7214407324790955, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 19114 + }, + { + "epoch": 0.19115, + "grad_norm": 0.9095062613487244, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 19115 + }, + { + "epoch": 0.19116, + "grad_norm": 1.3402451276779175, + "learning_rate": 0.003, + "loss": 4.029, + "step": 19116 + }, + { + "epoch": 0.19117, + "grad_norm": 0.6122109293937683, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 19117 + }, + { + "epoch": 0.19118, + "grad_norm": 0.6436371803283691, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 19118 + }, + { + "epoch": 0.19119, + "grad_norm": 0.8043946623802185, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 19119 + }, + { + "epoch": 0.1912, + "grad_norm": 0.8193957209587097, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 19120 + }, + { + "epoch": 0.19121, + "grad_norm": 0.778981626033783, + "learning_rate": 0.003, + "loss": 4.013, + "step": 19121 + }, + { + "epoch": 0.19122, + "grad_norm": 0.7985784411430359, + "learning_rate": 0.003, + "loss": 4.013, + "step": 19122 + }, + { + "epoch": 0.19123, + "grad_norm": 0.7375852465629578, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 19123 + }, + { + "epoch": 0.19124, + "grad_norm": 0.8971546292304993, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 19124 + }, + { + "epoch": 0.19125, + "grad_norm": 0.9998382329940796, + "learning_rate": 0.003, + "loss": 3.998, + "step": 19125 + }, + { + "epoch": 0.19126, + "grad_norm": 0.9625898003578186, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 19126 + }, + { + "epoch": 0.19127, + "grad_norm": 1.0204427242279053, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 19127 + }, + { + "epoch": 0.19128, + "grad_norm": 1.2511576414108276, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 19128 + }, + { + "epoch": 0.19129, + "grad_norm": 0.7735320925712585, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 19129 + }, + { + "epoch": 0.1913, + "grad_norm": 0.6742091178894043, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 19130 + }, + { + "epoch": 0.19131, + "grad_norm": 0.6724696755409241, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 19131 + }, + { + "epoch": 0.19132, + "grad_norm": 0.723955512046814, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 19132 + }, + { + "epoch": 0.19133, + "grad_norm": 0.8885080218315125, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 19133 + }, + { + "epoch": 0.19134, + "grad_norm": 0.9969792366027832, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 19134 + }, + { + "epoch": 0.19135, + "grad_norm": 1.0846734046936035, + "learning_rate": 0.003, + "loss": 4.021, + "step": 19135 + }, + { + "epoch": 0.19136, + "grad_norm": 0.7409761548042297, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 19136 + }, + { + "epoch": 0.19137, + "grad_norm": 0.5579649209976196, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 19137 + }, + { + "epoch": 0.19138, + "grad_norm": 0.7213162183761597, + "learning_rate": 0.003, + "loss": 4.005, + "step": 19138 + }, + { + "epoch": 0.19139, + "grad_norm": 0.8387914896011353, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 19139 + }, + { + "epoch": 0.1914, + "grad_norm": 0.996585488319397, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 19140 + }, + { + "epoch": 0.19141, + "grad_norm": 1.0790042877197266, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 19141 + }, + { + "epoch": 0.19142, + "grad_norm": 0.6569657325744629, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 19142 + }, + { + "epoch": 0.19143, + "grad_norm": 0.6758229732513428, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 19143 + }, + { + "epoch": 0.19144, + "grad_norm": 0.7549024224281311, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 19144 + }, + { + "epoch": 0.19145, + "grad_norm": 0.8701123595237732, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 19145 + }, + { + "epoch": 0.19146, + "grad_norm": 0.9431536793708801, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 19146 + }, + { + "epoch": 0.19147, + "grad_norm": 0.9425971508026123, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 19147 + }, + { + "epoch": 0.19148, + "grad_norm": 0.9396876692771912, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 19148 + }, + { + "epoch": 0.19149, + "grad_norm": 0.7899640798568726, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 19149 + }, + { + "epoch": 0.1915, + "grad_norm": 0.8626085519790649, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 19150 + }, + { + "epoch": 0.19151, + "grad_norm": 0.9163888096809387, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 19151 + }, + { + "epoch": 0.19152, + "grad_norm": 0.7615463137626648, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 19152 + }, + { + "epoch": 0.19153, + "grad_norm": 0.8965135216712952, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 19153 + }, + { + "epoch": 0.19154, + "grad_norm": 1.055113673210144, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 19154 + }, + { + "epoch": 0.19155, + "grad_norm": 0.943031907081604, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 19155 + }, + { + "epoch": 0.19156, + "grad_norm": 0.8581675887107849, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 19156 + }, + { + "epoch": 0.19157, + "grad_norm": 0.8582165837287903, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 19157 + }, + { + "epoch": 0.19158, + "grad_norm": 0.977011501789093, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 19158 + }, + { + "epoch": 0.19159, + "grad_norm": 1.2999720573425293, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 19159 + }, + { + "epoch": 0.1916, + "grad_norm": 0.5874431133270264, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 19160 + }, + { + "epoch": 0.19161, + "grad_norm": 0.6543835997581482, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 19161 + }, + { + "epoch": 0.19162, + "grad_norm": 0.6022617816925049, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 19162 + }, + { + "epoch": 0.19163, + "grad_norm": 0.6174903512001038, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 19163 + }, + { + "epoch": 0.19164, + "grad_norm": 0.6451676487922668, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 19164 + }, + { + "epoch": 0.19165, + "grad_norm": 0.7233970761299133, + "learning_rate": 0.003, + "loss": 3.987, + "step": 19165 + }, + { + "epoch": 0.19166, + "grad_norm": 0.8002468943595886, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 19166 + }, + { + "epoch": 0.19167, + "grad_norm": 0.8846295475959778, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 19167 + }, + { + "epoch": 0.19168, + "grad_norm": 0.8885087966918945, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 19168 + }, + { + "epoch": 0.19169, + "grad_norm": 0.8156308531761169, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 19169 + }, + { + "epoch": 0.1917, + "grad_norm": 0.753954291343689, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 19170 + }, + { + "epoch": 0.19171, + "grad_norm": 0.7089570164680481, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 19171 + }, + { + "epoch": 0.19172, + "grad_norm": 0.6770859360694885, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 19172 + }, + { + "epoch": 0.19173, + "grad_norm": 0.7039762735366821, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 19173 + }, + { + "epoch": 0.19174, + "grad_norm": 0.6847444176673889, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 19174 + }, + { + "epoch": 0.19175, + "grad_norm": 0.7717546820640564, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 19175 + }, + { + "epoch": 0.19176, + "grad_norm": 1.1402064561843872, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 19176 + }, + { + "epoch": 0.19177, + "grad_norm": 1.2233058214187622, + "learning_rate": 0.003, + "loss": 4.005, + "step": 19177 + }, + { + "epoch": 0.19178, + "grad_norm": 0.7143800854682922, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 19178 + }, + { + "epoch": 0.19179, + "grad_norm": 0.6129800081253052, + "learning_rate": 0.003, + "loss": 4.021, + "step": 19179 + }, + { + "epoch": 0.1918, + "grad_norm": 0.596025824546814, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 19180 + }, + { + "epoch": 0.19181, + "grad_norm": 0.637728214263916, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 19181 + }, + { + "epoch": 0.19182, + "grad_norm": 0.54209965467453, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 19182 + }, + { + "epoch": 0.19183, + "grad_norm": 0.5044422745704651, + "learning_rate": 0.003, + "loss": 4.001, + "step": 19183 + }, + { + "epoch": 0.19184, + "grad_norm": 0.5068732500076294, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 19184 + }, + { + "epoch": 0.19185, + "grad_norm": 0.6069113612174988, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 19185 + }, + { + "epoch": 0.19186, + "grad_norm": 0.797422468662262, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 19186 + }, + { + "epoch": 0.19187, + "grad_norm": 1.0681114196777344, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 19187 + }, + { + "epoch": 0.19188, + "grad_norm": 1.1297365427017212, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 19188 + }, + { + "epoch": 0.19189, + "grad_norm": 0.9015651941299438, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 19189 + }, + { + "epoch": 0.1919, + "grad_norm": 0.8453728556632996, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 19190 + }, + { + "epoch": 0.19191, + "grad_norm": 0.8107426762580872, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 19191 + }, + { + "epoch": 0.19192, + "grad_norm": 0.7438492774963379, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 19192 + }, + { + "epoch": 0.19193, + "grad_norm": 0.7621150016784668, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 19193 + }, + { + "epoch": 0.19194, + "grad_norm": 0.8221701979637146, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 19194 + }, + { + "epoch": 0.19195, + "grad_norm": 0.9234983921051025, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 19195 + }, + { + "epoch": 0.19196, + "grad_norm": 0.8986861705780029, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 19196 + }, + { + "epoch": 0.19197, + "grad_norm": 0.9112302660942078, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 19197 + }, + { + "epoch": 0.19198, + "grad_norm": 0.8387850522994995, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 19198 + }, + { + "epoch": 0.19199, + "grad_norm": 0.9891698956489563, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 19199 + }, + { + "epoch": 0.192, + "grad_norm": 1.352467656135559, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 19200 + }, + { + "epoch": 0.19201, + "grad_norm": 1.191960096359253, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 19201 + }, + { + "epoch": 0.19202, + "grad_norm": 0.7610149383544922, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 19202 + }, + { + "epoch": 0.19203, + "grad_norm": 0.7151528000831604, + "learning_rate": 0.003, + "loss": 3.982, + "step": 19203 + }, + { + "epoch": 0.19204, + "grad_norm": 0.8194527626037598, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 19204 + }, + { + "epoch": 0.19205, + "grad_norm": 0.8837549686431885, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 19205 + }, + { + "epoch": 0.19206, + "grad_norm": 0.9980619549751282, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 19206 + }, + { + "epoch": 0.19207, + "grad_norm": 1.1445870399475098, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 19207 + }, + { + "epoch": 0.19208, + "grad_norm": 0.8322120904922485, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 19208 + }, + { + "epoch": 0.19209, + "grad_norm": 0.8153591752052307, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 19209 + }, + { + "epoch": 0.1921, + "grad_norm": 0.9522150158882141, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 19210 + }, + { + "epoch": 0.19211, + "grad_norm": 1.006571888923645, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 19211 + }, + { + "epoch": 0.19212, + "grad_norm": 0.9249372482299805, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 19212 + }, + { + "epoch": 0.19213, + "grad_norm": 0.764064610004425, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 19213 + }, + { + "epoch": 0.19214, + "grad_norm": 0.9267817139625549, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 19214 + }, + { + "epoch": 0.19215, + "grad_norm": 0.907848596572876, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 19215 + }, + { + "epoch": 0.19216, + "grad_norm": 0.9351398348808289, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 19216 + }, + { + "epoch": 0.19217, + "grad_norm": 0.9645265936851501, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 19217 + }, + { + "epoch": 0.19218, + "grad_norm": 0.9986564517021179, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 19218 + }, + { + "epoch": 0.19219, + "grad_norm": 1.132468819618225, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 19219 + }, + { + "epoch": 0.1922, + "grad_norm": 1.048362374305725, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 19220 + }, + { + "epoch": 0.19221, + "grad_norm": 1.063513994216919, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 19221 + }, + { + "epoch": 0.19222, + "grad_norm": 0.8725884556770325, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 19222 + }, + { + "epoch": 0.19223, + "grad_norm": 0.8729296922683716, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 19223 + }, + { + "epoch": 0.19224, + "grad_norm": 1.0028876066207886, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 19224 + }, + { + "epoch": 0.19225, + "grad_norm": 1.1254569292068481, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 19225 + }, + { + "epoch": 0.19226, + "grad_norm": 0.6930163502693176, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 19226 + }, + { + "epoch": 0.19227, + "grad_norm": 0.7227094769477844, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 19227 + }, + { + "epoch": 0.19228, + "grad_norm": 0.7676095366477966, + "learning_rate": 0.003, + "loss": 4.001, + "step": 19228 + }, + { + "epoch": 0.19229, + "grad_norm": 0.9406114816665649, + "learning_rate": 0.003, + "loss": 3.995, + "step": 19229 + }, + { + "epoch": 0.1923, + "grad_norm": 1.07334566116333, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 19230 + }, + { + "epoch": 0.19231, + "grad_norm": 0.9809719324111938, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 19231 + }, + { + "epoch": 0.19232, + "grad_norm": 1.104724407196045, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 19232 + }, + { + "epoch": 0.19233, + "grad_norm": 0.8253532648086548, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 19233 + }, + { + "epoch": 0.19234, + "grad_norm": 0.6342723965644836, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 19234 + }, + { + "epoch": 0.19235, + "grad_norm": 0.7233535051345825, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 19235 + }, + { + "epoch": 0.19236, + "grad_norm": 0.8769366145133972, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 19236 + }, + { + "epoch": 0.19237, + "grad_norm": 1.0040518045425415, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 19237 + }, + { + "epoch": 0.19238, + "grad_norm": 1.0712950229644775, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 19238 + }, + { + "epoch": 0.19239, + "grad_norm": 0.8149884939193726, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 19239 + }, + { + "epoch": 0.1924, + "grad_norm": 0.7744317054748535, + "learning_rate": 0.003, + "loss": 3.974, + "step": 19240 + }, + { + "epoch": 0.19241, + "grad_norm": 0.7898107171058655, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 19241 + }, + { + "epoch": 0.19242, + "grad_norm": 0.8145323991775513, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 19242 + }, + { + "epoch": 0.19243, + "grad_norm": 0.8011947870254517, + "learning_rate": 0.003, + "loss": 4.024, + "step": 19243 + }, + { + "epoch": 0.19244, + "grad_norm": 0.6696740984916687, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 19244 + }, + { + "epoch": 0.19245, + "grad_norm": 0.7286134958267212, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 19245 + }, + { + "epoch": 0.19246, + "grad_norm": 0.809220016002655, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 19246 + }, + { + "epoch": 0.19247, + "grad_norm": 0.717673122882843, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 19247 + }, + { + "epoch": 0.19248, + "grad_norm": 0.6953349709510803, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 19248 + }, + { + "epoch": 0.19249, + "grad_norm": 0.5940714478492737, + "learning_rate": 0.003, + "loss": 3.97, + "step": 19249 + }, + { + "epoch": 0.1925, + "grad_norm": 0.6480825543403625, + "learning_rate": 0.003, + "loss": 4.011, + "step": 19250 + }, + { + "epoch": 0.19251, + "grad_norm": 0.75681072473526, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 19251 + }, + { + "epoch": 0.19252, + "grad_norm": 0.9951421022415161, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 19252 + }, + { + "epoch": 0.19253, + "grad_norm": 1.3604280948638916, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 19253 + }, + { + "epoch": 0.19254, + "grad_norm": 0.6712243556976318, + "learning_rate": 0.003, + "loss": 4.04, + "step": 19254 + }, + { + "epoch": 0.19255, + "grad_norm": 0.6681257486343384, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 19255 + }, + { + "epoch": 0.19256, + "grad_norm": 0.9547545313835144, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 19256 + }, + { + "epoch": 0.19257, + "grad_norm": 0.8940248489379883, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 19257 + }, + { + "epoch": 0.19258, + "grad_norm": 0.8410124778747559, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 19258 + }, + { + "epoch": 0.19259, + "grad_norm": 0.8101106286048889, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 19259 + }, + { + "epoch": 0.1926, + "grad_norm": 0.6973461508750916, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 19260 + }, + { + "epoch": 0.19261, + "grad_norm": 0.6412228345870972, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 19261 + }, + { + "epoch": 0.19262, + "grad_norm": 0.7030969262123108, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 19262 + }, + { + "epoch": 0.19263, + "grad_norm": 0.6626996397972107, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 19263 + }, + { + "epoch": 0.19264, + "grad_norm": 0.7193623185157776, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 19264 + }, + { + "epoch": 0.19265, + "grad_norm": 0.7803868651390076, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 19265 + }, + { + "epoch": 0.19266, + "grad_norm": 0.8606928586959839, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 19266 + }, + { + "epoch": 0.19267, + "grad_norm": 1.137913703918457, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 19267 + }, + { + "epoch": 0.19268, + "grad_norm": 1.2063173055648804, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 19268 + }, + { + "epoch": 0.19269, + "grad_norm": 0.8055838346481323, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 19269 + }, + { + "epoch": 0.1927, + "grad_norm": 0.8761873245239258, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 19270 + }, + { + "epoch": 0.19271, + "grad_norm": 0.8632670044898987, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 19271 + }, + { + "epoch": 0.19272, + "grad_norm": 0.8842520117759705, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 19272 + }, + { + "epoch": 0.19273, + "grad_norm": 0.9441511034965515, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 19273 + }, + { + "epoch": 0.19274, + "grad_norm": 1.003538727760315, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 19274 + }, + { + "epoch": 0.19275, + "grad_norm": 1.0713435411453247, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 19275 + }, + { + "epoch": 0.19276, + "grad_norm": 0.8656263947486877, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 19276 + }, + { + "epoch": 0.19277, + "grad_norm": 0.8232696056365967, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 19277 + }, + { + "epoch": 0.19278, + "grad_norm": 0.9967733025550842, + "learning_rate": 0.003, + "loss": 4.02, + "step": 19278 + }, + { + "epoch": 0.19279, + "grad_norm": 1.1772435903549194, + "learning_rate": 0.003, + "loss": 3.987, + "step": 19279 + }, + { + "epoch": 0.1928, + "grad_norm": 0.718956470489502, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 19280 + }, + { + "epoch": 0.19281, + "grad_norm": 0.6918644309043884, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 19281 + }, + { + "epoch": 0.19282, + "grad_norm": 0.6497114896774292, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 19282 + }, + { + "epoch": 0.19283, + "grad_norm": 0.7545351386070251, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 19283 + }, + { + "epoch": 0.19284, + "grad_norm": 0.7960440516471863, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 19284 + }, + { + "epoch": 0.19285, + "grad_norm": 0.820967972278595, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 19285 + }, + { + "epoch": 0.19286, + "grad_norm": 0.9108230471611023, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 19286 + }, + { + "epoch": 0.19287, + "grad_norm": 1.113358974456787, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 19287 + }, + { + "epoch": 0.19288, + "grad_norm": 1.094118356704712, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 19288 + }, + { + "epoch": 0.19289, + "grad_norm": 0.9445165991783142, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 19289 + }, + { + "epoch": 0.1929, + "grad_norm": 0.8943971395492554, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 19290 + }, + { + "epoch": 0.19291, + "grad_norm": 0.8869792819023132, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 19291 + }, + { + "epoch": 0.19292, + "grad_norm": 0.8029800057411194, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 19292 + }, + { + "epoch": 0.19293, + "grad_norm": 0.63421630859375, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 19293 + }, + { + "epoch": 0.19294, + "grad_norm": 0.6482874155044556, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 19294 + }, + { + "epoch": 0.19295, + "grad_norm": 0.6990712285041809, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 19295 + }, + { + "epoch": 0.19296, + "grad_norm": 0.6776507496833801, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 19296 + }, + { + "epoch": 0.19297, + "grad_norm": 0.690116286277771, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 19297 + }, + { + "epoch": 0.19298, + "grad_norm": 0.7109023928642273, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 19298 + }, + { + "epoch": 0.19299, + "grad_norm": 0.7003673911094666, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 19299 + }, + { + "epoch": 0.193, + "grad_norm": 0.7295224070549011, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 19300 + }, + { + "epoch": 0.19301, + "grad_norm": 0.8747197985649109, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 19301 + }, + { + "epoch": 0.19302, + "grad_norm": 0.9157381653785706, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 19302 + }, + { + "epoch": 0.19303, + "grad_norm": 0.9902534484863281, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 19303 + }, + { + "epoch": 0.19304, + "grad_norm": 1.1121066808700562, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 19304 + }, + { + "epoch": 0.19305, + "grad_norm": 0.913833737373352, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 19305 + }, + { + "epoch": 0.19306, + "grad_norm": 0.9779497981071472, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 19306 + }, + { + "epoch": 0.19307, + "grad_norm": 0.9935235977172852, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 19307 + }, + { + "epoch": 0.19308, + "grad_norm": 1.01974356174469, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 19308 + }, + { + "epoch": 0.19309, + "grad_norm": 0.8969842791557312, + "learning_rate": 0.003, + "loss": 4.016, + "step": 19309 + }, + { + "epoch": 0.1931, + "grad_norm": 0.9678723812103271, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 19310 + }, + { + "epoch": 0.19311, + "grad_norm": 0.9434349536895752, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 19311 + }, + { + "epoch": 0.19312, + "grad_norm": 0.8398012518882751, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 19312 + }, + { + "epoch": 0.19313, + "grad_norm": 0.9102538824081421, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 19313 + }, + { + "epoch": 0.19314, + "grad_norm": 1.021769404411316, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 19314 + }, + { + "epoch": 0.19315, + "grad_norm": 1.102475881576538, + "learning_rate": 0.003, + "loss": 3.982, + "step": 19315 + }, + { + "epoch": 0.19316, + "grad_norm": 0.7890679240226746, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 19316 + }, + { + "epoch": 0.19317, + "grad_norm": 0.6756047606468201, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 19317 + }, + { + "epoch": 0.19318, + "grad_norm": 0.5816415548324585, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 19318 + }, + { + "epoch": 0.19319, + "grad_norm": 0.5623725652694702, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 19319 + }, + { + "epoch": 0.1932, + "grad_norm": 0.6121018528938293, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 19320 + }, + { + "epoch": 0.19321, + "grad_norm": 0.590885579586029, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 19321 + }, + { + "epoch": 0.19322, + "grad_norm": 0.6558840870857239, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 19322 + }, + { + "epoch": 0.19323, + "grad_norm": 0.7619689702987671, + "learning_rate": 0.003, + "loss": 3.987, + "step": 19323 + }, + { + "epoch": 0.19324, + "grad_norm": 1.0216729640960693, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 19324 + }, + { + "epoch": 0.19325, + "grad_norm": 1.1044150590896606, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 19325 + }, + { + "epoch": 0.19326, + "grad_norm": 0.6996297836303711, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 19326 + }, + { + "epoch": 0.19327, + "grad_norm": 0.6888504028320312, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 19327 + }, + { + "epoch": 0.19328, + "grad_norm": 0.7001692652702332, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 19328 + }, + { + "epoch": 0.19329, + "grad_norm": 0.6823729276657104, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 19329 + }, + { + "epoch": 0.1933, + "grad_norm": 0.6538233757019043, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 19330 + }, + { + "epoch": 0.19331, + "grad_norm": 0.7303865551948547, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 19331 + }, + { + "epoch": 0.19332, + "grad_norm": 0.7956865429878235, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 19332 + }, + { + "epoch": 0.19333, + "grad_norm": 0.8402124047279358, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 19333 + }, + { + "epoch": 0.19334, + "grad_norm": 0.978168785572052, + "learning_rate": 0.003, + "loss": 3.963, + "step": 19334 + }, + { + "epoch": 0.19335, + "grad_norm": 1.1001943349838257, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 19335 + }, + { + "epoch": 0.19336, + "grad_norm": 0.8973298072814941, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 19336 + }, + { + "epoch": 0.19337, + "grad_norm": 0.8150404095649719, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 19337 + }, + { + "epoch": 0.19338, + "grad_norm": 0.8321810960769653, + "learning_rate": 0.003, + "loss": 4.035, + "step": 19338 + }, + { + "epoch": 0.19339, + "grad_norm": 0.8140332102775574, + "learning_rate": 0.003, + "loss": 4.002, + "step": 19339 + }, + { + "epoch": 0.1934, + "grad_norm": 0.8432743549346924, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 19340 + }, + { + "epoch": 0.19341, + "grad_norm": 1.1796387434005737, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 19341 + }, + { + "epoch": 0.19342, + "grad_norm": 0.8989206552505493, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 19342 + }, + { + "epoch": 0.19343, + "grad_norm": 0.770767331123352, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 19343 + }, + { + "epoch": 0.19344, + "grad_norm": 0.847355306148529, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 19344 + }, + { + "epoch": 0.19345, + "grad_norm": 1.0080876350402832, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 19345 + }, + { + "epoch": 0.19346, + "grad_norm": 1.0260003805160522, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 19346 + }, + { + "epoch": 0.19347, + "grad_norm": 0.8898026943206787, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 19347 + }, + { + "epoch": 0.19348, + "grad_norm": 1.0255355834960938, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 19348 + }, + { + "epoch": 0.19349, + "grad_norm": 0.9337112307548523, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 19349 + }, + { + "epoch": 0.1935, + "grad_norm": 0.9396280646324158, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 19350 + }, + { + "epoch": 0.19351, + "grad_norm": 1.0090826749801636, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 19351 + }, + { + "epoch": 0.19352, + "grad_norm": 0.9399703145027161, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 19352 + }, + { + "epoch": 0.19353, + "grad_norm": 0.900275468826294, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 19353 + }, + { + "epoch": 0.19354, + "grad_norm": 0.8167095184326172, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 19354 + }, + { + "epoch": 0.19355, + "grad_norm": 0.8206589221954346, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 19355 + }, + { + "epoch": 0.19356, + "grad_norm": 0.9308629631996155, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 19356 + }, + { + "epoch": 0.19357, + "grad_norm": 0.9931376576423645, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 19357 + }, + { + "epoch": 0.19358, + "grad_norm": 0.9372820854187012, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 19358 + }, + { + "epoch": 0.19359, + "grad_norm": 0.7388105392456055, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 19359 + }, + { + "epoch": 0.1936, + "grad_norm": 0.6335276961326599, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 19360 + }, + { + "epoch": 0.19361, + "grad_norm": 0.7078648209571838, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 19361 + }, + { + "epoch": 0.19362, + "grad_norm": 0.677641749382019, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 19362 + }, + { + "epoch": 0.19363, + "grad_norm": 0.6329396963119507, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 19363 + }, + { + "epoch": 0.19364, + "grad_norm": 0.6954342126846313, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 19364 + }, + { + "epoch": 0.19365, + "grad_norm": 0.6739953756332397, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 19365 + }, + { + "epoch": 0.19366, + "grad_norm": 0.6043066382408142, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 19366 + }, + { + "epoch": 0.19367, + "grad_norm": 0.7020654082298279, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 19367 + }, + { + "epoch": 0.19368, + "grad_norm": 0.7260381579399109, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 19368 + }, + { + "epoch": 0.19369, + "grad_norm": 0.8367782831192017, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 19369 + }, + { + "epoch": 0.1937, + "grad_norm": 1.1033228635787964, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 19370 + }, + { + "epoch": 0.19371, + "grad_norm": 1.3062193393707275, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 19371 + }, + { + "epoch": 0.19372, + "grad_norm": 0.5842729806900024, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 19372 + }, + { + "epoch": 0.19373, + "grad_norm": 0.659385621547699, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 19373 + }, + { + "epoch": 0.19374, + "grad_norm": 0.7550154328346252, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 19374 + }, + { + "epoch": 0.19375, + "grad_norm": 0.7876573801040649, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 19375 + }, + { + "epoch": 0.19376, + "grad_norm": 0.830193817615509, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 19376 + }, + { + "epoch": 0.19377, + "grad_norm": 0.8212099671363831, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 19377 + }, + { + "epoch": 0.19378, + "grad_norm": 0.6696993112564087, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 19378 + }, + { + "epoch": 0.19379, + "grad_norm": 0.679273247718811, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 19379 + }, + { + "epoch": 0.1938, + "grad_norm": 0.7368372082710266, + "learning_rate": 0.003, + "loss": 4.005, + "step": 19380 + }, + { + "epoch": 0.19381, + "grad_norm": 0.6955564022064209, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 19381 + }, + { + "epoch": 0.19382, + "grad_norm": 0.7486737370491028, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 19382 + }, + { + "epoch": 0.19383, + "grad_norm": 0.8705873489379883, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 19383 + }, + { + "epoch": 0.19384, + "grad_norm": 1.004651665687561, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 19384 + }, + { + "epoch": 0.19385, + "grad_norm": 0.9881071448326111, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 19385 + }, + { + "epoch": 0.19386, + "grad_norm": 1.0152630805969238, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 19386 + }, + { + "epoch": 0.19387, + "grad_norm": 1.1704293489456177, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 19387 + }, + { + "epoch": 0.19388, + "grad_norm": 0.6372362375259399, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 19388 + }, + { + "epoch": 0.19389, + "grad_norm": 0.8370605111122131, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 19389 + }, + { + "epoch": 0.1939, + "grad_norm": 1.0577499866485596, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 19390 + }, + { + "epoch": 0.19391, + "grad_norm": 1.250914216041565, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 19391 + }, + { + "epoch": 0.19392, + "grad_norm": 1.0160754919052124, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 19392 + }, + { + "epoch": 0.19393, + "grad_norm": 1.0917208194732666, + "learning_rate": 0.003, + "loss": 3.994, + "step": 19393 + }, + { + "epoch": 0.19394, + "grad_norm": 0.85653156042099, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 19394 + }, + { + "epoch": 0.19395, + "grad_norm": 0.8118630051612854, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 19395 + }, + { + "epoch": 0.19396, + "grad_norm": 0.6265334486961365, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 19396 + }, + { + "epoch": 0.19397, + "grad_norm": 0.6211968660354614, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 19397 + }, + { + "epoch": 0.19398, + "grad_norm": 0.5773246884346008, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 19398 + }, + { + "epoch": 0.19399, + "grad_norm": 0.6495667695999146, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 19399 + }, + { + "epoch": 0.194, + "grad_norm": 0.7733378410339355, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 19400 + }, + { + "epoch": 0.19401, + "grad_norm": 0.988730251789093, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 19401 + }, + { + "epoch": 0.19402, + "grad_norm": 1.137104868888855, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 19402 + }, + { + "epoch": 0.19403, + "grad_norm": 0.720878005027771, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 19403 + }, + { + "epoch": 0.19404, + "grad_norm": 0.5655572414398193, + "learning_rate": 0.003, + "loss": 3.997, + "step": 19404 + }, + { + "epoch": 0.19405, + "grad_norm": 0.6101178526878357, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 19405 + }, + { + "epoch": 0.19406, + "grad_norm": 0.6160560846328735, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 19406 + }, + { + "epoch": 0.19407, + "grad_norm": 0.6651063561439514, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 19407 + }, + { + "epoch": 0.19408, + "grad_norm": 0.7105965614318848, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 19408 + }, + { + "epoch": 0.19409, + "grad_norm": 0.7422831654548645, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 19409 + }, + { + "epoch": 0.1941, + "grad_norm": 0.6924363970756531, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 19410 + }, + { + "epoch": 0.19411, + "grad_norm": 0.7285566329956055, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 19411 + }, + { + "epoch": 0.19412, + "grad_norm": 0.5908989906311035, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 19412 + }, + { + "epoch": 0.19413, + "grad_norm": 0.5184929370880127, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 19413 + }, + { + "epoch": 0.19414, + "grad_norm": 0.5630707740783691, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 19414 + }, + { + "epoch": 0.19415, + "grad_norm": 0.6402706503868103, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 19415 + }, + { + "epoch": 0.19416, + "grad_norm": 0.8083771467208862, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 19416 + }, + { + "epoch": 0.19417, + "grad_norm": 1.0702193975448608, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 19417 + }, + { + "epoch": 0.19418, + "grad_norm": 1.3197736740112305, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 19418 + }, + { + "epoch": 0.19419, + "grad_norm": 0.9700169563293457, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 19419 + }, + { + "epoch": 0.1942, + "grad_norm": 0.9081416130065918, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 19420 + }, + { + "epoch": 0.19421, + "grad_norm": 0.9122806191444397, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 19421 + }, + { + "epoch": 0.19422, + "grad_norm": 1.0107626914978027, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 19422 + }, + { + "epoch": 0.19423, + "grad_norm": 1.0141414403915405, + "learning_rate": 0.003, + "loss": 4.012, + "step": 19423 + }, + { + "epoch": 0.19424, + "grad_norm": 0.9829725623130798, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 19424 + }, + { + "epoch": 0.19425, + "grad_norm": 0.8267939686775208, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 19425 + }, + { + "epoch": 0.19426, + "grad_norm": 0.8093584775924683, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 19426 + }, + { + "epoch": 0.19427, + "grad_norm": 1.0037287473678589, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 19427 + }, + { + "epoch": 0.19428, + "grad_norm": 1.0828919410705566, + "learning_rate": 0.003, + "loss": 3.992, + "step": 19428 + }, + { + "epoch": 0.19429, + "grad_norm": 0.894026517868042, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 19429 + }, + { + "epoch": 0.1943, + "grad_norm": 0.9650946855545044, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 19430 + }, + { + "epoch": 0.19431, + "grad_norm": 1.0324418544769287, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 19431 + }, + { + "epoch": 0.19432, + "grad_norm": 1.05412757396698, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 19432 + }, + { + "epoch": 0.19433, + "grad_norm": 0.862998366355896, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 19433 + }, + { + "epoch": 0.19434, + "grad_norm": 1.014135479927063, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 19434 + }, + { + "epoch": 0.19435, + "grad_norm": 1.3630157709121704, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 19435 + }, + { + "epoch": 0.19436, + "grad_norm": 0.762340247631073, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 19436 + }, + { + "epoch": 0.19437, + "grad_norm": 0.8234502673149109, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 19437 + }, + { + "epoch": 0.19438, + "grad_norm": 0.8225938677787781, + "learning_rate": 0.003, + "loss": 4.027, + "step": 19438 + }, + { + "epoch": 0.19439, + "grad_norm": 0.9233136177062988, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 19439 + }, + { + "epoch": 0.1944, + "grad_norm": 0.7754296660423279, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 19440 + }, + { + "epoch": 0.19441, + "grad_norm": 0.7464416027069092, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 19441 + }, + { + "epoch": 0.19442, + "grad_norm": 0.7560956478118896, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 19442 + }, + { + "epoch": 0.19443, + "grad_norm": 0.765295147895813, + "learning_rate": 0.003, + "loss": 4.031, + "step": 19443 + }, + { + "epoch": 0.19444, + "grad_norm": 0.694036066532135, + "learning_rate": 0.003, + "loss": 3.99, + "step": 19444 + }, + { + "epoch": 0.19445, + "grad_norm": 0.7585387825965881, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 19445 + }, + { + "epoch": 0.19446, + "grad_norm": 1.08774733543396, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 19446 + }, + { + "epoch": 0.19447, + "grad_norm": 1.089281678199768, + "learning_rate": 0.003, + "loss": 3.991, + "step": 19447 + }, + { + "epoch": 0.19448, + "grad_norm": 1.04354727268219, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 19448 + }, + { + "epoch": 0.19449, + "grad_norm": 1.0154160261154175, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 19449 + }, + { + "epoch": 0.1945, + "grad_norm": 0.9952034950256348, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 19450 + }, + { + "epoch": 0.19451, + "grad_norm": 0.9826450347900391, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 19451 + }, + { + "epoch": 0.19452, + "grad_norm": 0.8546826839447021, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 19452 + }, + { + "epoch": 0.19453, + "grad_norm": 0.9039189219474792, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 19453 + }, + { + "epoch": 0.19454, + "grad_norm": 0.8569737076759338, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 19454 + }, + { + "epoch": 0.19455, + "grad_norm": 0.8439573645591736, + "learning_rate": 0.003, + "loss": 4.025, + "step": 19455 + }, + { + "epoch": 0.19456, + "grad_norm": 0.8585202693939209, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 19456 + }, + { + "epoch": 0.19457, + "grad_norm": 0.7985356450080872, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 19457 + }, + { + "epoch": 0.19458, + "grad_norm": 0.8447664380073547, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 19458 + }, + { + "epoch": 0.19459, + "grad_norm": 0.948081374168396, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 19459 + }, + { + "epoch": 0.1946, + "grad_norm": 1.1912204027175903, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 19460 + }, + { + "epoch": 0.19461, + "grad_norm": 1.2770426273345947, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 19461 + }, + { + "epoch": 0.19462, + "grad_norm": 0.6283977627754211, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 19462 + }, + { + "epoch": 0.19463, + "grad_norm": 0.6860957741737366, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 19463 + }, + { + "epoch": 0.19464, + "grad_norm": 0.8585407137870789, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 19464 + }, + { + "epoch": 0.19465, + "grad_norm": 1.158394455909729, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 19465 + }, + { + "epoch": 0.19466, + "grad_norm": 1.021871566772461, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 19466 + }, + { + "epoch": 0.19467, + "grad_norm": 0.7340610027313232, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 19467 + }, + { + "epoch": 0.19468, + "grad_norm": 0.6287662982940674, + "learning_rate": 0.003, + "loss": 3.984, + "step": 19468 + }, + { + "epoch": 0.19469, + "grad_norm": 0.7421482801437378, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 19469 + }, + { + "epoch": 0.1947, + "grad_norm": 0.8674723505973816, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 19470 + }, + { + "epoch": 0.19471, + "grad_norm": 0.93992018699646, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 19471 + }, + { + "epoch": 0.19472, + "grad_norm": 0.7212894558906555, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 19472 + }, + { + "epoch": 0.19473, + "grad_norm": 0.6208124756813049, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 19473 + }, + { + "epoch": 0.19474, + "grad_norm": 0.7051081657409668, + "learning_rate": 0.003, + "loss": 4.007, + "step": 19474 + }, + { + "epoch": 0.19475, + "grad_norm": 0.7773585319519043, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 19475 + }, + { + "epoch": 0.19476, + "grad_norm": 0.9117393493652344, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 19476 + }, + { + "epoch": 0.19477, + "grad_norm": 0.8664668798446655, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 19477 + }, + { + "epoch": 0.19478, + "grad_norm": 0.7039976716041565, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 19478 + }, + { + "epoch": 0.19479, + "grad_norm": 0.6012693047523499, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 19479 + }, + { + "epoch": 0.1948, + "grad_norm": 0.6584872007369995, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 19480 + }, + { + "epoch": 0.19481, + "grad_norm": 0.7361558675765991, + "learning_rate": 0.003, + "loss": 4.014, + "step": 19481 + }, + { + "epoch": 0.19482, + "grad_norm": 0.7228919267654419, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 19482 + }, + { + "epoch": 0.19483, + "grad_norm": 0.6640205383300781, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 19483 + }, + { + "epoch": 0.19484, + "grad_norm": 0.6353579759597778, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 19484 + }, + { + "epoch": 0.19485, + "grad_norm": 0.5491459965705872, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 19485 + }, + { + "epoch": 0.19486, + "grad_norm": 0.5839203596115112, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 19486 + }, + { + "epoch": 0.19487, + "grad_norm": 0.7307979464530945, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 19487 + }, + { + "epoch": 0.19488, + "grad_norm": 0.8992578983306885, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 19488 + }, + { + "epoch": 0.19489, + "grad_norm": 1.2707570791244507, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 19489 + }, + { + "epoch": 0.1949, + "grad_norm": 0.8503918051719666, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 19490 + }, + { + "epoch": 0.19491, + "grad_norm": 0.6091768741607666, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 19491 + }, + { + "epoch": 0.19492, + "grad_norm": 0.5840878486633301, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 19492 + }, + { + "epoch": 0.19493, + "grad_norm": 0.6856347918510437, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 19493 + }, + { + "epoch": 0.19494, + "grad_norm": 0.9025638699531555, + "learning_rate": 0.003, + "loss": 4.01, + "step": 19494 + }, + { + "epoch": 0.19495, + "grad_norm": 1.0733050107955933, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 19495 + }, + { + "epoch": 0.19496, + "grad_norm": 1.009167194366455, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 19496 + }, + { + "epoch": 0.19497, + "grad_norm": 1.1027225255966187, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 19497 + }, + { + "epoch": 0.19498, + "grad_norm": 0.8099426031112671, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 19498 + }, + { + "epoch": 0.19499, + "grad_norm": 0.7486651539802551, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 19499 + }, + { + "epoch": 0.195, + "grad_norm": 0.7175552845001221, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 19500 + }, + { + "epoch": 0.19501, + "grad_norm": 0.592884361743927, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 19501 + }, + { + "epoch": 0.19502, + "grad_norm": 0.5978587865829468, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 19502 + }, + { + "epoch": 0.19503, + "grad_norm": 0.7668148875236511, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 19503 + }, + { + "epoch": 0.19504, + "grad_norm": 0.8724342584609985, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 19504 + }, + { + "epoch": 0.19505, + "grad_norm": 1.0184301137924194, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 19505 + }, + { + "epoch": 0.19506, + "grad_norm": 1.0616868734359741, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 19506 + }, + { + "epoch": 0.19507, + "grad_norm": 0.9919269680976868, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 19507 + }, + { + "epoch": 0.19508, + "grad_norm": 0.9941419363021851, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 19508 + }, + { + "epoch": 0.19509, + "grad_norm": 1.1667042970657349, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 19509 + }, + { + "epoch": 0.1951, + "grad_norm": 0.8843926787376404, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 19510 + }, + { + "epoch": 0.19511, + "grad_norm": 0.8106421828269958, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 19511 + }, + { + "epoch": 0.19512, + "grad_norm": 0.7645822167396545, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 19512 + }, + { + "epoch": 0.19513, + "grad_norm": 0.808374285697937, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 19513 + }, + { + "epoch": 0.19514, + "grad_norm": 0.8262529373168945, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 19514 + }, + { + "epoch": 0.19515, + "grad_norm": 0.9069766998291016, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 19515 + }, + { + "epoch": 0.19516, + "grad_norm": 0.9706578254699707, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 19516 + }, + { + "epoch": 0.19517, + "grad_norm": 0.9760490655899048, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 19517 + }, + { + "epoch": 0.19518, + "grad_norm": 0.8847740888595581, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 19518 + }, + { + "epoch": 0.19519, + "grad_norm": 1.038620948791504, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 19519 + }, + { + "epoch": 0.1952, + "grad_norm": 0.942561149597168, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 19520 + }, + { + "epoch": 0.19521, + "grad_norm": 0.871234655380249, + "learning_rate": 0.003, + "loss": 3.995, + "step": 19521 + }, + { + "epoch": 0.19522, + "grad_norm": 0.8343075513839722, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 19522 + }, + { + "epoch": 0.19523, + "grad_norm": 0.8660777807235718, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 19523 + }, + { + "epoch": 0.19524, + "grad_norm": 0.9518958926200867, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 19524 + }, + { + "epoch": 0.19525, + "grad_norm": 0.9712399840354919, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 19525 + }, + { + "epoch": 0.19526, + "grad_norm": 1.0904133319854736, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 19526 + }, + { + "epoch": 0.19527, + "grad_norm": 0.7620598077774048, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 19527 + }, + { + "epoch": 0.19528, + "grad_norm": 0.7478907704353333, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 19528 + }, + { + "epoch": 0.19529, + "grad_norm": 0.8900734186172485, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 19529 + }, + { + "epoch": 0.1953, + "grad_norm": 1.03106689453125, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 19530 + }, + { + "epoch": 0.19531, + "grad_norm": 1.0545998811721802, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 19531 + }, + { + "epoch": 0.19532, + "grad_norm": 0.9456791877746582, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 19532 + }, + { + "epoch": 0.19533, + "grad_norm": 1.0635806322097778, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 19533 + }, + { + "epoch": 0.19534, + "grad_norm": 1.0025690793991089, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 19534 + }, + { + "epoch": 0.19535, + "grad_norm": 0.9176450371742249, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 19535 + }, + { + "epoch": 0.19536, + "grad_norm": 0.7869552969932556, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 19536 + }, + { + "epoch": 0.19537, + "grad_norm": 0.7948800325393677, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 19537 + }, + { + "epoch": 0.19538, + "grad_norm": 0.7708257436752319, + "learning_rate": 0.003, + "loss": 4.02, + "step": 19538 + }, + { + "epoch": 0.19539, + "grad_norm": 0.8740930557250977, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 19539 + }, + { + "epoch": 0.1954, + "grad_norm": 0.950183093547821, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 19540 + }, + { + "epoch": 0.19541, + "grad_norm": 0.9351487159729004, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 19541 + }, + { + "epoch": 0.19542, + "grad_norm": 1.0730854272842407, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 19542 + }, + { + "epoch": 0.19543, + "grad_norm": 0.9327528476715088, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 19543 + }, + { + "epoch": 0.19544, + "grad_norm": 0.771660566329956, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 19544 + }, + { + "epoch": 0.19545, + "grad_norm": 0.7307870388031006, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 19545 + }, + { + "epoch": 0.19546, + "grad_norm": 0.8321375846862793, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 19546 + }, + { + "epoch": 0.19547, + "grad_norm": 0.8160489201545715, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 19547 + }, + { + "epoch": 0.19548, + "grad_norm": 0.9429007172584534, + "learning_rate": 0.003, + "loss": 4.014, + "step": 19548 + }, + { + "epoch": 0.19549, + "grad_norm": 0.9813238382339478, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 19549 + }, + { + "epoch": 0.1955, + "grad_norm": 0.9885159730911255, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 19550 + }, + { + "epoch": 0.19551, + "grad_norm": 0.8191841244697571, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 19551 + }, + { + "epoch": 0.19552, + "grad_norm": 0.7239416241645813, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 19552 + }, + { + "epoch": 0.19553, + "grad_norm": 0.7224628329277039, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 19553 + }, + { + "epoch": 0.19554, + "grad_norm": 0.7405498623847961, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 19554 + }, + { + "epoch": 0.19555, + "grad_norm": 0.6783916354179382, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 19555 + }, + { + "epoch": 0.19556, + "grad_norm": 0.7338733673095703, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 19556 + }, + { + "epoch": 0.19557, + "grad_norm": 0.690729558467865, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 19557 + }, + { + "epoch": 0.19558, + "grad_norm": 0.6429121494293213, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 19558 + }, + { + "epoch": 0.19559, + "grad_norm": 0.706646203994751, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 19559 + }, + { + "epoch": 0.1956, + "grad_norm": 0.8352551460266113, + "learning_rate": 0.003, + "loss": 4.025, + "step": 19560 + }, + { + "epoch": 0.19561, + "grad_norm": 0.8661195039749146, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 19561 + }, + { + "epoch": 0.19562, + "grad_norm": 0.8343972563743591, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 19562 + }, + { + "epoch": 0.19563, + "grad_norm": 0.7886388897895813, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 19563 + }, + { + "epoch": 0.19564, + "grad_norm": 0.7899730205535889, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 19564 + }, + { + "epoch": 0.19565, + "grad_norm": 0.7374923229217529, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 19565 + }, + { + "epoch": 0.19566, + "grad_norm": 0.6665852069854736, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 19566 + }, + { + "epoch": 0.19567, + "grad_norm": 0.6971161365509033, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 19567 + }, + { + "epoch": 0.19568, + "grad_norm": 1.001599907875061, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 19568 + }, + { + "epoch": 0.19569, + "grad_norm": 1.2593739032745361, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 19569 + }, + { + "epoch": 0.1957, + "grad_norm": 0.7797518968582153, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 19570 + }, + { + "epoch": 0.19571, + "grad_norm": 0.6890308260917664, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 19571 + }, + { + "epoch": 0.19572, + "grad_norm": 0.6744365692138672, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 19572 + }, + { + "epoch": 0.19573, + "grad_norm": 0.6337965130805969, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 19573 + }, + { + "epoch": 0.19574, + "grad_norm": 0.6815075278282166, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 19574 + }, + { + "epoch": 0.19575, + "grad_norm": 0.7398589849472046, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 19575 + }, + { + "epoch": 0.19576, + "grad_norm": 0.808785080909729, + "learning_rate": 0.003, + "loss": 3.984, + "step": 19576 + }, + { + "epoch": 0.19577, + "grad_norm": 0.9979984760284424, + "learning_rate": 0.003, + "loss": 4.017, + "step": 19577 + }, + { + "epoch": 0.19578, + "grad_norm": 1.3587226867675781, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 19578 + }, + { + "epoch": 0.19579, + "grad_norm": 0.7701467871665955, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 19579 + }, + { + "epoch": 0.1958, + "grad_norm": 0.8579768538475037, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 19580 + }, + { + "epoch": 0.19581, + "grad_norm": 0.8206691145896912, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 19581 + }, + { + "epoch": 0.19582, + "grad_norm": 0.825986921787262, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 19582 + }, + { + "epoch": 0.19583, + "grad_norm": 1.0285770893096924, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 19583 + }, + { + "epoch": 0.19584, + "grad_norm": 0.9314975738525391, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 19584 + }, + { + "epoch": 0.19585, + "grad_norm": 0.8427761197090149, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 19585 + }, + { + "epoch": 0.19586, + "grad_norm": 0.8255916237831116, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 19586 + }, + { + "epoch": 0.19587, + "grad_norm": 0.9761923551559448, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 19587 + }, + { + "epoch": 0.19588, + "grad_norm": 0.9684008955955505, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 19588 + }, + { + "epoch": 0.19589, + "grad_norm": 0.8309191465377808, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 19589 + }, + { + "epoch": 0.1959, + "grad_norm": 0.9165033102035522, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 19590 + }, + { + "epoch": 0.19591, + "grad_norm": 1.048823595046997, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 19591 + }, + { + "epoch": 0.19592, + "grad_norm": 0.855726957321167, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 19592 + }, + { + "epoch": 0.19593, + "grad_norm": 0.7062085866928101, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 19593 + }, + { + "epoch": 0.19594, + "grad_norm": 0.6082330942153931, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 19594 + }, + { + "epoch": 0.19595, + "grad_norm": 0.6740702986717224, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 19595 + }, + { + "epoch": 0.19596, + "grad_norm": 0.700225830078125, + "learning_rate": 0.003, + "loss": 3.991, + "step": 19596 + }, + { + "epoch": 0.19597, + "grad_norm": 0.7771080732345581, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 19597 + }, + { + "epoch": 0.19598, + "grad_norm": 0.8797142505645752, + "learning_rate": 0.003, + "loss": 3.985, + "step": 19598 + }, + { + "epoch": 0.19599, + "grad_norm": 0.9389474987983704, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 19599 + }, + { + "epoch": 0.196, + "grad_norm": 1.2271173000335693, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 19600 + }, + { + "epoch": 0.19601, + "grad_norm": 0.9719414710998535, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 19601 + }, + { + "epoch": 0.19602, + "grad_norm": 0.7353074550628662, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 19602 + }, + { + "epoch": 0.19603, + "grad_norm": 0.7225217819213867, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 19603 + }, + { + "epoch": 0.19604, + "grad_norm": 0.8830018043518066, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 19604 + }, + { + "epoch": 0.19605, + "grad_norm": 0.9781860113143921, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 19605 + }, + { + "epoch": 0.19606, + "grad_norm": 1.1129882335662842, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 19606 + }, + { + "epoch": 0.19607, + "grad_norm": 0.8997833728790283, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 19607 + }, + { + "epoch": 0.19608, + "grad_norm": 0.8203573822975159, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 19608 + }, + { + "epoch": 0.19609, + "grad_norm": 0.8649213314056396, + "learning_rate": 0.003, + "loss": 4.014, + "step": 19609 + }, + { + "epoch": 0.1961, + "grad_norm": 0.8213027715682983, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 19610 + }, + { + "epoch": 0.19611, + "grad_norm": 0.7620607614517212, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 19611 + }, + { + "epoch": 0.19612, + "grad_norm": 0.650044322013855, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 19612 + }, + { + "epoch": 0.19613, + "grad_norm": 0.7654820680618286, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 19613 + }, + { + "epoch": 0.19614, + "grad_norm": 0.8626854419708252, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 19614 + }, + { + "epoch": 0.19615, + "grad_norm": 0.9784452319145203, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 19615 + }, + { + "epoch": 0.19616, + "grad_norm": 1.17198646068573, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 19616 + }, + { + "epoch": 0.19617, + "grad_norm": 1.013183832168579, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 19617 + }, + { + "epoch": 0.19618, + "grad_norm": 1.0921857357025146, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 19618 + }, + { + "epoch": 0.19619, + "grad_norm": 0.7567834258079529, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 19619 + }, + { + "epoch": 0.1962, + "grad_norm": 0.6237739324569702, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 19620 + }, + { + "epoch": 0.19621, + "grad_norm": 0.6849679350852966, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 19621 + }, + { + "epoch": 0.19622, + "grad_norm": 0.9135372638702393, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 19622 + }, + { + "epoch": 0.19623, + "grad_norm": 1.1048378944396973, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 19623 + }, + { + "epoch": 0.19624, + "grad_norm": 0.7356999516487122, + "learning_rate": 0.003, + "loss": 4.001, + "step": 19624 + }, + { + "epoch": 0.19625, + "grad_norm": 0.5960215330123901, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 19625 + }, + { + "epoch": 0.19626, + "grad_norm": 0.6349772214889526, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 19626 + }, + { + "epoch": 0.19627, + "grad_norm": 0.6433651447296143, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 19627 + }, + { + "epoch": 0.19628, + "grad_norm": 0.5735981464385986, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 19628 + }, + { + "epoch": 0.19629, + "grad_norm": 0.5863465666770935, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 19629 + }, + { + "epoch": 0.1963, + "grad_norm": 0.6383113861083984, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 19630 + }, + { + "epoch": 0.19631, + "grad_norm": 0.7619834542274475, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 19631 + }, + { + "epoch": 0.19632, + "grad_norm": 0.9621742963790894, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 19632 + }, + { + "epoch": 0.19633, + "grad_norm": 1.2753996849060059, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 19633 + }, + { + "epoch": 0.19634, + "grad_norm": 0.7475082278251648, + "learning_rate": 0.003, + "loss": 4.018, + "step": 19634 + }, + { + "epoch": 0.19635, + "grad_norm": 0.6645981669425964, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 19635 + }, + { + "epoch": 0.19636, + "grad_norm": 0.6509314179420471, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 19636 + }, + { + "epoch": 0.19637, + "grad_norm": 0.6968465447425842, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 19637 + }, + { + "epoch": 0.19638, + "grad_norm": 0.695035457611084, + "learning_rate": 0.003, + "loss": 3.9423, + "step": 19638 + }, + { + "epoch": 0.19639, + "grad_norm": 0.6628406047821045, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 19639 + }, + { + "epoch": 0.1964, + "grad_norm": 0.7157939076423645, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 19640 + }, + { + "epoch": 0.19641, + "grad_norm": 0.707119345664978, + "learning_rate": 0.003, + "loss": 4.013, + "step": 19641 + }, + { + "epoch": 0.19642, + "grad_norm": 0.7395365834236145, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 19642 + }, + { + "epoch": 0.19643, + "grad_norm": 0.8493033647537231, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 19643 + }, + { + "epoch": 0.19644, + "grad_norm": 1.1865185499191284, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 19644 + }, + { + "epoch": 0.19645, + "grad_norm": 0.9355396032333374, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 19645 + }, + { + "epoch": 0.19646, + "grad_norm": 0.8840484023094177, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 19646 + }, + { + "epoch": 0.19647, + "grad_norm": 1.0170649290084839, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 19647 + }, + { + "epoch": 0.19648, + "grad_norm": 1.1426259279251099, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 19648 + }, + { + "epoch": 0.19649, + "grad_norm": 0.8664590120315552, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 19649 + }, + { + "epoch": 0.1965, + "grad_norm": 0.8678274154663086, + "learning_rate": 0.003, + "loss": 4.01, + "step": 19650 + }, + { + "epoch": 0.19651, + "grad_norm": 0.9149540662765503, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 19651 + }, + { + "epoch": 0.19652, + "grad_norm": 1.1072866916656494, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 19652 + }, + { + "epoch": 0.19653, + "grad_norm": 1.0925770998001099, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 19653 + }, + { + "epoch": 0.19654, + "grad_norm": 1.0328413248062134, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 19654 + }, + { + "epoch": 0.19655, + "grad_norm": 0.7894022464752197, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 19655 + }, + { + "epoch": 0.19656, + "grad_norm": 0.765109658241272, + "learning_rate": 0.003, + "loss": 3.989, + "step": 19656 + }, + { + "epoch": 0.19657, + "grad_norm": 0.7951400876045227, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 19657 + }, + { + "epoch": 0.19658, + "grad_norm": 0.7783117890357971, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 19658 + }, + { + "epoch": 0.19659, + "grad_norm": 0.7544420957565308, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 19659 + }, + { + "epoch": 0.1966, + "grad_norm": 0.7228208184242249, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 19660 + }, + { + "epoch": 0.19661, + "grad_norm": 0.8361641764640808, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 19661 + }, + { + "epoch": 0.19662, + "grad_norm": 0.9005764722824097, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 19662 + }, + { + "epoch": 0.19663, + "grad_norm": 0.8779737949371338, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 19663 + }, + { + "epoch": 0.19664, + "grad_norm": 0.9157434105873108, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 19664 + }, + { + "epoch": 0.19665, + "grad_norm": 0.8780549168586731, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 19665 + }, + { + "epoch": 0.19666, + "grad_norm": 0.7676054835319519, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 19666 + }, + { + "epoch": 0.19667, + "grad_norm": 0.8035011887550354, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 19667 + }, + { + "epoch": 0.19668, + "grad_norm": 1.096777319908142, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 19668 + }, + { + "epoch": 0.19669, + "grad_norm": 1.2337048053741455, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 19669 + }, + { + "epoch": 0.1967, + "grad_norm": 0.8524037599563599, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 19670 + }, + { + "epoch": 0.19671, + "grad_norm": 0.8955839276313782, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 19671 + }, + { + "epoch": 0.19672, + "grad_norm": 0.8179341554641724, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 19672 + }, + { + "epoch": 0.19673, + "grad_norm": 0.805479884147644, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 19673 + }, + { + "epoch": 0.19674, + "grad_norm": 0.9087765216827393, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 19674 + }, + { + "epoch": 0.19675, + "grad_norm": 0.9388098120689392, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 19675 + }, + { + "epoch": 0.19676, + "grad_norm": 0.7843668460845947, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 19676 + }, + { + "epoch": 0.19677, + "grad_norm": 0.8374168872833252, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 19677 + }, + { + "epoch": 0.19678, + "grad_norm": 0.8288348913192749, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 19678 + }, + { + "epoch": 0.19679, + "grad_norm": 0.9719083905220032, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 19679 + }, + { + "epoch": 0.1968, + "grad_norm": 1.1349914073944092, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 19680 + }, + { + "epoch": 0.19681, + "grad_norm": 0.9622350931167603, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 19681 + }, + { + "epoch": 0.19682, + "grad_norm": 0.9577870965003967, + "learning_rate": 0.003, + "loss": 4.004, + "step": 19682 + }, + { + "epoch": 0.19683, + "grad_norm": 0.9903905391693115, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 19683 + }, + { + "epoch": 0.19684, + "grad_norm": 0.9436802268028259, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 19684 + }, + { + "epoch": 0.19685, + "grad_norm": 0.8398043513298035, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 19685 + }, + { + "epoch": 0.19686, + "grad_norm": 0.7675806283950806, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 19686 + }, + { + "epoch": 0.19687, + "grad_norm": 0.7498754262924194, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 19687 + }, + { + "epoch": 0.19688, + "grad_norm": 0.7731102108955383, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 19688 + }, + { + "epoch": 0.19689, + "grad_norm": 0.8190988302230835, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 19689 + }, + { + "epoch": 0.1969, + "grad_norm": 0.9088810682296753, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 19690 + }, + { + "epoch": 0.19691, + "grad_norm": 0.9039297699928284, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 19691 + }, + { + "epoch": 0.19692, + "grad_norm": 0.9410934448242188, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 19692 + }, + { + "epoch": 0.19693, + "grad_norm": 0.8614537715911865, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 19693 + }, + { + "epoch": 0.19694, + "grad_norm": 0.8556104302406311, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 19694 + }, + { + "epoch": 0.19695, + "grad_norm": 1.0040862560272217, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 19695 + }, + { + "epoch": 0.19696, + "grad_norm": 1.0933338403701782, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 19696 + }, + { + "epoch": 0.19697, + "grad_norm": 1.2151308059692383, + "learning_rate": 0.003, + "loss": 4.002, + "step": 19697 + }, + { + "epoch": 0.19698, + "grad_norm": 0.6764607429504395, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 19698 + }, + { + "epoch": 0.19699, + "grad_norm": 0.6174522638320923, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 19699 + }, + { + "epoch": 0.197, + "grad_norm": 0.6862599849700928, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 19700 + }, + { + "epoch": 0.19701, + "grad_norm": 0.7461639642715454, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 19701 + }, + { + "epoch": 0.19702, + "grad_norm": 0.8803563714027405, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 19702 + }, + { + "epoch": 0.19703, + "grad_norm": 0.9222602248191833, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 19703 + }, + { + "epoch": 0.19704, + "grad_norm": 0.8738635778427124, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 19704 + }, + { + "epoch": 0.19705, + "grad_norm": 0.8284536004066467, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 19705 + }, + { + "epoch": 0.19706, + "grad_norm": 0.9120162725448608, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 19706 + }, + { + "epoch": 0.19707, + "grad_norm": 0.8574915528297424, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 19707 + }, + { + "epoch": 0.19708, + "grad_norm": 1.0177977085113525, + "learning_rate": 0.003, + "loss": 4.021, + "step": 19708 + }, + { + "epoch": 0.19709, + "grad_norm": 1.028807520866394, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 19709 + }, + { + "epoch": 0.1971, + "grad_norm": 1.077039361000061, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 19710 + }, + { + "epoch": 0.19711, + "grad_norm": 0.7538310885429382, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 19711 + }, + { + "epoch": 0.19712, + "grad_norm": 0.7804000973701477, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 19712 + }, + { + "epoch": 0.19713, + "grad_norm": 0.7949774265289307, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 19713 + }, + { + "epoch": 0.19714, + "grad_norm": 0.7602480053901672, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 19714 + }, + { + "epoch": 0.19715, + "grad_norm": 0.7396253943443298, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 19715 + }, + { + "epoch": 0.19716, + "grad_norm": 0.789199948310852, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 19716 + }, + { + "epoch": 0.19717, + "grad_norm": 0.9171184301376343, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 19717 + }, + { + "epoch": 0.19718, + "grad_norm": 1.1561167240142822, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 19718 + }, + { + "epoch": 0.19719, + "grad_norm": 0.8584945201873779, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 19719 + }, + { + "epoch": 0.1972, + "grad_norm": 0.7851689457893372, + "learning_rate": 0.003, + "loss": 3.967, + "step": 19720 + }, + { + "epoch": 0.19721, + "grad_norm": 0.8622748851776123, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 19721 + }, + { + "epoch": 0.19722, + "grad_norm": 0.9780141711235046, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 19722 + }, + { + "epoch": 0.19723, + "grad_norm": 1.0162153244018555, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 19723 + }, + { + "epoch": 0.19724, + "grad_norm": 0.8792099356651306, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 19724 + }, + { + "epoch": 0.19725, + "grad_norm": 0.7323285341262817, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 19725 + }, + { + "epoch": 0.19726, + "grad_norm": 0.5505766868591309, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 19726 + }, + { + "epoch": 0.19727, + "grad_norm": 0.5450384020805359, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 19727 + }, + { + "epoch": 0.19728, + "grad_norm": 0.6860783696174622, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 19728 + }, + { + "epoch": 0.19729, + "grad_norm": 0.8005902171134949, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 19729 + }, + { + "epoch": 0.1973, + "grad_norm": 0.8072187304496765, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 19730 + }, + { + "epoch": 0.19731, + "grad_norm": 0.7468485236167908, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 19731 + }, + { + "epoch": 0.19732, + "grad_norm": 0.6146760582923889, + "learning_rate": 0.003, + "loss": 3.995, + "step": 19732 + }, + { + "epoch": 0.19733, + "grad_norm": 0.6144158840179443, + "learning_rate": 0.003, + "loss": 3.988, + "step": 19733 + }, + { + "epoch": 0.19734, + "grad_norm": 0.6798016428947449, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 19734 + }, + { + "epoch": 0.19735, + "grad_norm": 0.6656534671783447, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 19735 + }, + { + "epoch": 0.19736, + "grad_norm": 0.7358046174049377, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 19736 + }, + { + "epoch": 0.19737, + "grad_norm": 0.8249279856681824, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 19737 + }, + { + "epoch": 0.19738, + "grad_norm": 0.8836650848388672, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 19738 + }, + { + "epoch": 0.19739, + "grad_norm": 0.847974419593811, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 19739 + }, + { + "epoch": 0.1974, + "grad_norm": 0.8489623069763184, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 19740 + }, + { + "epoch": 0.19741, + "grad_norm": 0.9032725691795349, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 19741 + }, + { + "epoch": 0.19742, + "grad_norm": 0.8884512186050415, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 19742 + }, + { + "epoch": 0.19743, + "grad_norm": 0.8677976727485657, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 19743 + }, + { + "epoch": 0.19744, + "grad_norm": 0.897802472114563, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 19744 + }, + { + "epoch": 0.19745, + "grad_norm": 0.899142324924469, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 19745 + }, + { + "epoch": 0.19746, + "grad_norm": 0.9474034905433655, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 19746 + }, + { + "epoch": 0.19747, + "grad_norm": 0.9039626121520996, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 19747 + }, + { + "epoch": 0.19748, + "grad_norm": 0.9217604398727417, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 19748 + }, + { + "epoch": 0.19749, + "grad_norm": 1.0116978883743286, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 19749 + }, + { + "epoch": 0.1975, + "grad_norm": 0.9110078811645508, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 19750 + }, + { + "epoch": 0.19751, + "grad_norm": 1.0594103336334229, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 19751 + }, + { + "epoch": 0.19752, + "grad_norm": 1.0353708267211914, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 19752 + }, + { + "epoch": 0.19753, + "grad_norm": 0.9697558879852295, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 19753 + }, + { + "epoch": 0.19754, + "grad_norm": 0.8715497255325317, + "learning_rate": 0.003, + "loss": 4.006, + "step": 19754 + }, + { + "epoch": 0.19755, + "grad_norm": 0.9651036262512207, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 19755 + }, + { + "epoch": 0.19756, + "grad_norm": 1.427270531654358, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 19756 + }, + { + "epoch": 0.19757, + "grad_norm": 0.6546532511711121, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 19757 + }, + { + "epoch": 0.19758, + "grad_norm": 0.7043947577476501, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 19758 + }, + { + "epoch": 0.19759, + "grad_norm": 0.8468017578125, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 19759 + }, + { + "epoch": 0.1976, + "grad_norm": 0.7900123596191406, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 19760 + }, + { + "epoch": 0.19761, + "grad_norm": 0.6871782541275024, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 19761 + }, + { + "epoch": 0.19762, + "grad_norm": 0.7135297656059265, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 19762 + }, + { + "epoch": 0.19763, + "grad_norm": 0.8189486265182495, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 19763 + }, + { + "epoch": 0.19764, + "grad_norm": 0.9873164892196655, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 19764 + }, + { + "epoch": 0.19765, + "grad_norm": 1.213226079940796, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 19765 + }, + { + "epoch": 0.19766, + "grad_norm": 0.634507417678833, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 19766 + }, + { + "epoch": 0.19767, + "grad_norm": 0.5030357837677002, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 19767 + }, + { + "epoch": 0.19768, + "grad_norm": 0.5763629674911499, + "learning_rate": 0.003, + "loss": 4.01, + "step": 19768 + }, + { + "epoch": 0.19769, + "grad_norm": 0.6826533079147339, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 19769 + }, + { + "epoch": 0.1977, + "grad_norm": 0.7585141062736511, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 19770 + }, + { + "epoch": 0.19771, + "grad_norm": 0.7298708558082581, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 19771 + }, + { + "epoch": 0.19772, + "grad_norm": 0.7219603657722473, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 19772 + }, + { + "epoch": 0.19773, + "grad_norm": 0.7906659245491028, + "learning_rate": 0.003, + "loss": 3.968, + "step": 19773 + }, + { + "epoch": 0.19774, + "grad_norm": 0.9170296788215637, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 19774 + }, + { + "epoch": 0.19775, + "grad_norm": 0.9445464015007019, + "learning_rate": 0.003, + "loss": 3.986, + "step": 19775 + }, + { + "epoch": 0.19776, + "grad_norm": 0.8738341927528381, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 19776 + }, + { + "epoch": 0.19777, + "grad_norm": 0.8931269645690918, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 19777 + }, + { + "epoch": 0.19778, + "grad_norm": 1.0102183818817139, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 19778 + }, + { + "epoch": 0.19779, + "grad_norm": 1.0455142259597778, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 19779 + }, + { + "epoch": 0.1978, + "grad_norm": 1.0050582885742188, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 19780 + }, + { + "epoch": 0.19781, + "grad_norm": 0.9604713320732117, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 19781 + }, + { + "epoch": 0.19782, + "grad_norm": 0.8043559789657593, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 19782 + }, + { + "epoch": 0.19783, + "grad_norm": 0.7254209518432617, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 19783 + }, + { + "epoch": 0.19784, + "grad_norm": 0.8698140978813171, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 19784 + }, + { + "epoch": 0.19785, + "grad_norm": 1.16221284866333, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 19785 + }, + { + "epoch": 0.19786, + "grad_norm": 0.8535551428794861, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 19786 + }, + { + "epoch": 0.19787, + "grad_norm": 0.6688913702964783, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 19787 + }, + { + "epoch": 0.19788, + "grad_norm": 0.6375434398651123, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 19788 + }, + { + "epoch": 0.19789, + "grad_norm": 0.6849889755249023, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 19789 + }, + { + "epoch": 0.1979, + "grad_norm": 0.8710851669311523, + "learning_rate": 0.003, + "loss": 3.968, + "step": 19790 + }, + { + "epoch": 0.19791, + "grad_norm": 1.0272157192230225, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 19791 + }, + { + "epoch": 0.19792, + "grad_norm": 0.9329808354377747, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 19792 + }, + { + "epoch": 0.19793, + "grad_norm": 0.9324360489845276, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 19793 + }, + { + "epoch": 0.19794, + "grad_norm": 1.0681499242782593, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 19794 + }, + { + "epoch": 0.19795, + "grad_norm": 1.0618821382522583, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 19795 + }, + { + "epoch": 0.19796, + "grad_norm": 0.8624000549316406, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 19796 + }, + { + "epoch": 0.19797, + "grad_norm": 0.7481711506843567, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 19797 + }, + { + "epoch": 0.19798, + "grad_norm": 0.7045367360115051, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 19798 + }, + { + "epoch": 0.19799, + "grad_norm": 0.7832326889038086, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 19799 + }, + { + "epoch": 0.198, + "grad_norm": 0.9828478693962097, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 19800 + }, + { + "epoch": 0.19801, + "grad_norm": 1.1521586179733276, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 19801 + }, + { + "epoch": 0.19802, + "grad_norm": 0.6974064111709595, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 19802 + }, + { + "epoch": 0.19803, + "grad_norm": 0.6272710561752319, + "learning_rate": 0.003, + "loss": 3.996, + "step": 19803 + }, + { + "epoch": 0.19804, + "grad_norm": 0.7898540496826172, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 19804 + }, + { + "epoch": 0.19805, + "grad_norm": 0.8566932678222656, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 19805 + }, + { + "epoch": 0.19806, + "grad_norm": 0.9286311268806458, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 19806 + }, + { + "epoch": 0.19807, + "grad_norm": 0.8343588709831238, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 19807 + }, + { + "epoch": 0.19808, + "grad_norm": 0.787165641784668, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 19808 + }, + { + "epoch": 0.19809, + "grad_norm": 0.7548251152038574, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 19809 + }, + { + "epoch": 0.1981, + "grad_norm": 0.7300617098808289, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 19810 + }, + { + "epoch": 0.19811, + "grad_norm": 0.7945849299430847, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 19811 + }, + { + "epoch": 0.19812, + "grad_norm": 0.8813292384147644, + "learning_rate": 0.003, + "loss": 3.981, + "step": 19812 + }, + { + "epoch": 0.19813, + "grad_norm": 0.781726062297821, + "learning_rate": 0.003, + "loss": 3.991, + "step": 19813 + }, + { + "epoch": 0.19814, + "grad_norm": 0.6853597164154053, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 19814 + }, + { + "epoch": 0.19815, + "grad_norm": 0.7190750241279602, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 19815 + }, + { + "epoch": 0.19816, + "grad_norm": 0.7978911399841309, + "learning_rate": 0.003, + "loss": 3.99, + "step": 19816 + }, + { + "epoch": 0.19817, + "grad_norm": 0.9732509851455688, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 19817 + }, + { + "epoch": 0.19818, + "grad_norm": 1.2398778200149536, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 19818 + }, + { + "epoch": 0.19819, + "grad_norm": 0.8849130272865295, + "learning_rate": 0.003, + "loss": 3.987, + "step": 19819 + }, + { + "epoch": 0.1982, + "grad_norm": 0.979040265083313, + "learning_rate": 0.003, + "loss": 4.003, + "step": 19820 + }, + { + "epoch": 0.19821, + "grad_norm": 1.0033127069473267, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 19821 + }, + { + "epoch": 0.19822, + "grad_norm": 0.8031641840934753, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 19822 + }, + { + "epoch": 0.19823, + "grad_norm": 0.700145959854126, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 19823 + }, + { + "epoch": 0.19824, + "grad_norm": 0.7709698677062988, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 19824 + }, + { + "epoch": 0.19825, + "grad_norm": 0.9196158051490784, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 19825 + }, + { + "epoch": 0.19826, + "grad_norm": 1.0031273365020752, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 19826 + }, + { + "epoch": 0.19827, + "grad_norm": 1.04696524143219, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 19827 + }, + { + "epoch": 0.19828, + "grad_norm": 0.8323779106140137, + "learning_rate": 0.003, + "loss": 3.993, + "step": 19828 + }, + { + "epoch": 0.19829, + "grad_norm": 0.6003943681716919, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 19829 + }, + { + "epoch": 0.1983, + "grad_norm": 0.587047815322876, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 19830 + }, + { + "epoch": 0.19831, + "grad_norm": 0.5091107487678528, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 19831 + }, + { + "epoch": 0.19832, + "grad_norm": 0.550421953201294, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 19832 + }, + { + "epoch": 0.19833, + "grad_norm": 0.633222758769989, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 19833 + }, + { + "epoch": 0.19834, + "grad_norm": 0.6866979598999023, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 19834 + }, + { + "epoch": 0.19835, + "grad_norm": 0.8001576066017151, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 19835 + }, + { + "epoch": 0.19836, + "grad_norm": 0.9836876392364502, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 19836 + }, + { + "epoch": 0.19837, + "grad_norm": 0.9438751339912415, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 19837 + }, + { + "epoch": 0.19838, + "grad_norm": 0.6752098798751831, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 19838 + }, + { + "epoch": 0.19839, + "grad_norm": 0.7575652003288269, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 19839 + }, + { + "epoch": 0.1984, + "grad_norm": 0.8691379427909851, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 19840 + }, + { + "epoch": 0.19841, + "grad_norm": 1.0416834354400635, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 19841 + }, + { + "epoch": 0.19842, + "grad_norm": 1.0642367601394653, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 19842 + }, + { + "epoch": 0.19843, + "grad_norm": 1.018958330154419, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 19843 + }, + { + "epoch": 0.19844, + "grad_norm": 0.9024566411972046, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 19844 + }, + { + "epoch": 0.19845, + "grad_norm": 0.8511251211166382, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 19845 + }, + { + "epoch": 0.19846, + "grad_norm": 0.9189441204071045, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 19846 + }, + { + "epoch": 0.19847, + "grad_norm": 1.092054009437561, + "learning_rate": 0.003, + "loss": 4.001, + "step": 19847 + }, + { + "epoch": 0.19848, + "grad_norm": 0.8711313605308533, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 19848 + }, + { + "epoch": 0.19849, + "grad_norm": 0.8742891550064087, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 19849 + }, + { + "epoch": 0.1985, + "grad_norm": 0.9483180046081543, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 19850 + }, + { + "epoch": 0.19851, + "grad_norm": 1.1380302906036377, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 19851 + }, + { + "epoch": 0.19852, + "grad_norm": 0.8698943257331848, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 19852 + }, + { + "epoch": 0.19853, + "grad_norm": 0.8115482926368713, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 19853 + }, + { + "epoch": 0.19854, + "grad_norm": 0.7584234476089478, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 19854 + }, + { + "epoch": 0.19855, + "grad_norm": 0.8206432461738586, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 19855 + }, + { + "epoch": 0.19856, + "grad_norm": 0.8196293115615845, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 19856 + }, + { + "epoch": 0.19857, + "grad_norm": 0.8289163708686829, + "learning_rate": 0.003, + "loss": 3.991, + "step": 19857 + }, + { + "epoch": 0.19858, + "grad_norm": 0.8808925747871399, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 19858 + }, + { + "epoch": 0.19859, + "grad_norm": 1.131120204925537, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 19859 + }, + { + "epoch": 0.1986, + "grad_norm": 1.0879793167114258, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 19860 + }, + { + "epoch": 0.19861, + "grad_norm": 0.780910074710846, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 19861 + }, + { + "epoch": 0.19862, + "grad_norm": 0.6187844276428223, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 19862 + }, + { + "epoch": 0.19863, + "grad_norm": 0.665023922920227, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 19863 + }, + { + "epoch": 0.19864, + "grad_norm": 0.9973242878913879, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 19864 + }, + { + "epoch": 0.19865, + "grad_norm": 1.1035375595092773, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 19865 + }, + { + "epoch": 0.19866, + "grad_norm": 0.7924714684486389, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 19866 + }, + { + "epoch": 0.19867, + "grad_norm": 0.7919667363166809, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 19867 + }, + { + "epoch": 0.19868, + "grad_norm": 0.8689961433410645, + "learning_rate": 0.003, + "loss": 4.01, + "step": 19868 + }, + { + "epoch": 0.19869, + "grad_norm": 0.7589074373245239, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 19869 + }, + { + "epoch": 0.1987, + "grad_norm": 0.6872889995574951, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 19870 + }, + { + "epoch": 0.19871, + "grad_norm": 0.7166623473167419, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 19871 + }, + { + "epoch": 0.19872, + "grad_norm": 0.758948564529419, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 19872 + }, + { + "epoch": 0.19873, + "grad_norm": 0.8943865299224854, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 19873 + }, + { + "epoch": 0.19874, + "grad_norm": 1.0847712755203247, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 19874 + }, + { + "epoch": 0.19875, + "grad_norm": 1.0528215169906616, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 19875 + }, + { + "epoch": 0.19876, + "grad_norm": 0.9561887979507446, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 19876 + }, + { + "epoch": 0.19877, + "grad_norm": 0.9552778005599976, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 19877 + }, + { + "epoch": 0.19878, + "grad_norm": 0.7830400466918945, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 19878 + }, + { + "epoch": 0.19879, + "grad_norm": 0.7615962028503418, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 19879 + }, + { + "epoch": 0.1988, + "grad_norm": 0.7598347067832947, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 19880 + }, + { + "epoch": 0.19881, + "grad_norm": 0.7968227863311768, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 19881 + }, + { + "epoch": 0.19882, + "grad_norm": 0.7083049416542053, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 19882 + }, + { + "epoch": 0.19883, + "grad_norm": 0.7206639051437378, + "learning_rate": 0.003, + "loss": 3.999, + "step": 19883 + }, + { + "epoch": 0.19884, + "grad_norm": 0.678696870803833, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 19884 + }, + { + "epoch": 0.19885, + "grad_norm": 0.7773205637931824, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 19885 + }, + { + "epoch": 0.19886, + "grad_norm": 0.8615788221359253, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 19886 + }, + { + "epoch": 0.19887, + "grad_norm": 0.9369819760322571, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 19887 + }, + { + "epoch": 0.19888, + "grad_norm": 1.0702393054962158, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 19888 + }, + { + "epoch": 0.19889, + "grad_norm": 1.0090556144714355, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 19889 + }, + { + "epoch": 0.1989, + "grad_norm": 0.9635480046272278, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 19890 + }, + { + "epoch": 0.19891, + "grad_norm": 0.9786340594291687, + "learning_rate": 0.003, + "loss": 4.005, + "step": 19891 + }, + { + "epoch": 0.19892, + "grad_norm": 1.0961414575576782, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 19892 + }, + { + "epoch": 0.19893, + "grad_norm": 0.8336204290390015, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 19893 + }, + { + "epoch": 0.19894, + "grad_norm": 0.8045467138290405, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 19894 + }, + { + "epoch": 0.19895, + "grad_norm": 0.7639846205711365, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 19895 + }, + { + "epoch": 0.19896, + "grad_norm": 0.711005449295044, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 19896 + }, + { + "epoch": 0.19897, + "grad_norm": 0.7749508619308472, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 19897 + }, + { + "epoch": 0.19898, + "grad_norm": 0.8815631866455078, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 19898 + }, + { + "epoch": 0.19899, + "grad_norm": 1.0683717727661133, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 19899 + }, + { + "epoch": 0.199, + "grad_norm": 1.163893461227417, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 19900 + }, + { + "epoch": 0.19901, + "grad_norm": 0.9210971593856812, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 19901 + }, + { + "epoch": 0.19902, + "grad_norm": 0.881843090057373, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 19902 + }, + { + "epoch": 0.19903, + "grad_norm": 0.7368955016136169, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 19903 + }, + { + "epoch": 0.19904, + "grad_norm": 0.710391640663147, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 19904 + }, + { + "epoch": 0.19905, + "grad_norm": 0.7400331497192383, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 19905 + }, + { + "epoch": 0.19906, + "grad_norm": 0.7363407015800476, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 19906 + }, + { + "epoch": 0.19907, + "grad_norm": 0.7485768795013428, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 19907 + }, + { + "epoch": 0.19908, + "grad_norm": 0.8457195162773132, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 19908 + }, + { + "epoch": 0.19909, + "grad_norm": 0.8688400387763977, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 19909 + }, + { + "epoch": 0.1991, + "grad_norm": 0.8740235567092896, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 19910 + }, + { + "epoch": 0.19911, + "grad_norm": 0.7607549428939819, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 19911 + }, + { + "epoch": 0.19912, + "grad_norm": 0.8285241723060608, + "learning_rate": 0.003, + "loss": 4.011, + "step": 19912 + }, + { + "epoch": 0.19913, + "grad_norm": 0.8437578678131104, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 19913 + }, + { + "epoch": 0.19914, + "grad_norm": 1.0094083547592163, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 19914 + }, + { + "epoch": 0.19915, + "grad_norm": 1.1903234720230103, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 19915 + }, + { + "epoch": 0.19916, + "grad_norm": 0.8025692701339722, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 19916 + }, + { + "epoch": 0.19917, + "grad_norm": 0.6294260621070862, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 19917 + }, + { + "epoch": 0.19918, + "grad_norm": 0.7292620539665222, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 19918 + }, + { + "epoch": 0.19919, + "grad_norm": 0.7832669615745544, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 19919 + }, + { + "epoch": 0.1992, + "grad_norm": 0.7429418563842773, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 19920 + }, + { + "epoch": 0.19921, + "grad_norm": 0.7528523206710815, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 19921 + }, + { + "epoch": 0.19922, + "grad_norm": 0.9220634698867798, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 19922 + }, + { + "epoch": 0.19923, + "grad_norm": 1.0650054216384888, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 19923 + }, + { + "epoch": 0.19924, + "grad_norm": 0.9716998934745789, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 19924 + }, + { + "epoch": 0.19925, + "grad_norm": 1.0343821048736572, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 19925 + }, + { + "epoch": 0.19926, + "grad_norm": 1.0323013067245483, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 19926 + }, + { + "epoch": 0.19927, + "grad_norm": 1.1438926458358765, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 19927 + }, + { + "epoch": 0.19928, + "grad_norm": 0.9947730898857117, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 19928 + }, + { + "epoch": 0.19929, + "grad_norm": 0.9139689803123474, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 19929 + }, + { + "epoch": 0.1993, + "grad_norm": 0.7781880497932434, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 19930 + }, + { + "epoch": 0.19931, + "grad_norm": 0.7079624533653259, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 19931 + }, + { + "epoch": 0.19932, + "grad_norm": 0.6574318408966064, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 19932 + }, + { + "epoch": 0.19933, + "grad_norm": 0.5009096264839172, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 19933 + }, + { + "epoch": 0.19934, + "grad_norm": 0.570099949836731, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 19934 + }, + { + "epoch": 0.19935, + "grad_norm": 0.740693211555481, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 19935 + }, + { + "epoch": 0.19936, + "grad_norm": 0.9218637943267822, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 19936 + }, + { + "epoch": 0.19937, + "grad_norm": 1.1842974424362183, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 19937 + }, + { + "epoch": 0.19938, + "grad_norm": 0.8443614840507507, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 19938 + }, + { + "epoch": 0.19939, + "grad_norm": 0.6035322546958923, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 19939 + }, + { + "epoch": 0.1994, + "grad_norm": 0.6982221007347107, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 19940 + }, + { + "epoch": 0.19941, + "grad_norm": 0.9117821455001831, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 19941 + }, + { + "epoch": 0.19942, + "grad_norm": 1.0192880630493164, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 19942 + }, + { + "epoch": 0.19943, + "grad_norm": 1.0437871217727661, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 19943 + }, + { + "epoch": 0.19944, + "grad_norm": 0.7494208216667175, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 19944 + }, + { + "epoch": 0.19945, + "grad_norm": 0.6562110781669617, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 19945 + }, + { + "epoch": 0.19946, + "grad_norm": 0.9326704144477844, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 19946 + }, + { + "epoch": 0.19947, + "grad_norm": 0.9937777519226074, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 19947 + }, + { + "epoch": 0.19948, + "grad_norm": 1.0566006898880005, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 19948 + }, + { + "epoch": 0.19949, + "grad_norm": 0.9444541931152344, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 19949 + }, + { + "epoch": 0.1995, + "grad_norm": 0.7949107885360718, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 19950 + }, + { + "epoch": 0.19951, + "grad_norm": 0.8902096748352051, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 19951 + }, + { + "epoch": 0.19952, + "grad_norm": 0.859947919845581, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 19952 + }, + { + "epoch": 0.19953, + "grad_norm": 0.6775383353233337, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 19953 + }, + { + "epoch": 0.19954, + "grad_norm": 0.6701100468635559, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 19954 + }, + { + "epoch": 0.19955, + "grad_norm": 0.6915163993835449, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 19955 + }, + { + "epoch": 0.19956, + "grad_norm": 0.7247843742370605, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 19956 + }, + { + "epoch": 0.19957, + "grad_norm": 0.7546032667160034, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 19957 + }, + { + "epoch": 0.19958, + "grad_norm": 0.8902223706245422, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 19958 + }, + { + "epoch": 0.19959, + "grad_norm": 1.1033283472061157, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 19959 + }, + { + "epoch": 0.1996, + "grad_norm": 1.0015894174575806, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 19960 + }, + { + "epoch": 0.19961, + "grad_norm": 0.8465616106987, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 19961 + }, + { + "epoch": 0.19962, + "grad_norm": 0.771872878074646, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 19962 + }, + { + "epoch": 0.19963, + "grad_norm": 0.8222938776016235, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 19963 + }, + { + "epoch": 0.19964, + "grad_norm": 0.7800220251083374, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 19964 + }, + { + "epoch": 0.19965, + "grad_norm": 0.8346208333969116, + "learning_rate": 0.003, + "loss": 4.011, + "step": 19965 + }, + { + "epoch": 0.19966, + "grad_norm": 0.8404595851898193, + "learning_rate": 0.003, + "loss": 4.001, + "step": 19966 + }, + { + "epoch": 0.19967, + "grad_norm": 0.7356424927711487, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 19967 + }, + { + "epoch": 0.19968, + "grad_norm": 0.7217704653739929, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 19968 + }, + { + "epoch": 0.19969, + "grad_norm": 0.7107825875282288, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 19969 + }, + { + "epoch": 0.1997, + "grad_norm": 0.773546040058136, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 19970 + }, + { + "epoch": 0.19971, + "grad_norm": 0.8497675061225891, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 19971 + }, + { + "epoch": 0.19972, + "grad_norm": 0.9392867088317871, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 19972 + }, + { + "epoch": 0.19973, + "grad_norm": 0.9622828364372253, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 19973 + }, + { + "epoch": 0.19974, + "grad_norm": 1.161346435546875, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 19974 + }, + { + "epoch": 0.19975, + "grad_norm": 0.9697285294532776, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 19975 + }, + { + "epoch": 0.19976, + "grad_norm": 0.9470873475074768, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 19976 + }, + { + "epoch": 0.19977, + "grad_norm": 0.9103334546089172, + "learning_rate": 0.003, + "loss": 3.968, + "step": 19977 + }, + { + "epoch": 0.19978, + "grad_norm": 1.0092731714248657, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 19978 + }, + { + "epoch": 0.19979, + "grad_norm": 1.1104042530059814, + "learning_rate": 0.003, + "loss": 4.018, + "step": 19979 + }, + { + "epoch": 0.1998, + "grad_norm": 0.9779863357543945, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 19980 + }, + { + "epoch": 0.19981, + "grad_norm": 1.0093814134597778, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 19981 + }, + { + "epoch": 0.19982, + "grad_norm": 1.0757005214691162, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 19982 + }, + { + "epoch": 0.19983, + "grad_norm": 0.9751392006874084, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 19983 + }, + { + "epoch": 0.19984, + "grad_norm": 0.9791036248207092, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 19984 + }, + { + "epoch": 0.19985, + "grad_norm": 0.9474849700927734, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 19985 + }, + { + "epoch": 0.19986, + "grad_norm": 1.0115426778793335, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 19986 + }, + { + "epoch": 0.19987, + "grad_norm": 0.925239086151123, + "learning_rate": 0.003, + "loss": 3.984, + "step": 19987 + }, + { + "epoch": 0.19988, + "grad_norm": 0.8519060611724854, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 19988 + }, + { + "epoch": 0.19989, + "grad_norm": 0.986173689365387, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 19989 + }, + { + "epoch": 0.1999, + "grad_norm": 1.378746509552002, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 19990 + }, + { + "epoch": 0.19991, + "grad_norm": 0.5946156978607178, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 19991 + }, + { + "epoch": 0.19992, + "grad_norm": 0.6783791780471802, + "learning_rate": 0.003, + "loss": 3.981, + "step": 19992 + }, + { + "epoch": 0.19993, + "grad_norm": 0.8321949243545532, + "learning_rate": 0.003, + "loss": 4.024, + "step": 19993 + }, + { + "epoch": 0.19994, + "grad_norm": 0.7344717383384705, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 19994 + }, + { + "epoch": 0.19995, + "grad_norm": 0.7066532969474792, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 19995 + }, + { + "epoch": 0.19996, + "grad_norm": 0.7640671133995056, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 19996 + }, + { + "epoch": 0.19997, + "grad_norm": 0.7876488566398621, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 19997 + }, + { + "epoch": 0.19998, + "grad_norm": 0.7732638716697693, + "learning_rate": 0.003, + "loss": 4.013, + "step": 19998 + }, + { + "epoch": 0.19999, + "grad_norm": 0.9254476428031921, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 19999 + }, + { + "epoch": 0.2, + "grad_norm": 1.0530250072479248, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 20000 + }, + { + "epoch": 0.20001, + "grad_norm": 0.9983343482017517, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 20001 + }, + { + "epoch": 0.20002, + "grad_norm": 0.9355091452598572, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 20002 + }, + { + "epoch": 0.20003, + "grad_norm": 0.7862496972084045, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 20003 + }, + { + "epoch": 0.20004, + "grad_norm": 0.7292976379394531, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 20004 + }, + { + "epoch": 0.20005, + "grad_norm": 0.6682198643684387, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 20005 + }, + { + "epoch": 0.20006, + "grad_norm": 0.5940819978713989, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 20006 + }, + { + "epoch": 0.20007, + "grad_norm": 0.65861576795578, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 20007 + }, + { + "epoch": 0.20008, + "grad_norm": 0.7443798184394836, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 20008 + }, + { + "epoch": 0.20009, + "grad_norm": 0.8729384541511536, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 20009 + }, + { + "epoch": 0.2001, + "grad_norm": 0.9064518809318542, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 20010 + }, + { + "epoch": 0.20011, + "grad_norm": 0.7785030007362366, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 20011 + }, + { + "epoch": 0.20012, + "grad_norm": 0.6019753217697144, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 20012 + }, + { + "epoch": 0.20013, + "grad_norm": 0.6211569309234619, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 20013 + }, + { + "epoch": 0.20014, + "grad_norm": 0.7714516520500183, + "learning_rate": 0.003, + "loss": 3.997, + "step": 20014 + }, + { + "epoch": 0.20015, + "grad_norm": 0.9249976873397827, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 20015 + }, + { + "epoch": 0.20016, + "grad_norm": 0.8645272850990295, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 20016 + }, + { + "epoch": 0.20017, + "grad_norm": 0.7886145114898682, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 20017 + }, + { + "epoch": 0.20018, + "grad_norm": 0.7523903250694275, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 20018 + }, + { + "epoch": 0.20019, + "grad_norm": 0.7617239356040955, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 20019 + }, + { + "epoch": 0.2002, + "grad_norm": 0.762170135974884, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 20020 + }, + { + "epoch": 0.20021, + "grad_norm": 0.7641904354095459, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 20021 + }, + { + "epoch": 0.20022, + "grad_norm": 0.7135266661643982, + "learning_rate": 0.003, + "loss": 3.995, + "step": 20022 + }, + { + "epoch": 0.20023, + "grad_norm": 0.6172149777412415, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 20023 + }, + { + "epoch": 0.20024, + "grad_norm": 0.620384693145752, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 20024 + }, + { + "epoch": 0.20025, + "grad_norm": 0.68131422996521, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 20025 + }, + { + "epoch": 0.20026, + "grad_norm": 0.8223885297775269, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 20026 + }, + { + "epoch": 0.20027, + "grad_norm": 0.9747222661972046, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 20027 + }, + { + "epoch": 0.20028, + "grad_norm": 1.0941600799560547, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 20028 + }, + { + "epoch": 0.20029, + "grad_norm": 0.9600298404693604, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 20029 + }, + { + "epoch": 0.2003, + "grad_norm": 0.9398216009140015, + "learning_rate": 0.003, + "loss": 4.001, + "step": 20030 + }, + { + "epoch": 0.20031, + "grad_norm": 0.8767874240875244, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 20031 + }, + { + "epoch": 0.20032, + "grad_norm": 0.8055996894836426, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 20032 + }, + { + "epoch": 0.20033, + "grad_norm": 0.9127088189125061, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 20033 + }, + { + "epoch": 0.20034, + "grad_norm": 1.1113648414611816, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 20034 + }, + { + "epoch": 0.20035, + "grad_norm": 1.2661253213882446, + "learning_rate": 0.003, + "loss": 4.007, + "step": 20035 + }, + { + "epoch": 0.20036, + "grad_norm": 0.8963114619255066, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 20036 + }, + { + "epoch": 0.20037, + "grad_norm": 0.907471776008606, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 20037 + }, + { + "epoch": 0.20038, + "grad_norm": 0.8981186151504517, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 20038 + }, + { + "epoch": 0.20039, + "grad_norm": 0.9879725575447083, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 20039 + }, + { + "epoch": 0.2004, + "grad_norm": 1.06181800365448, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 20040 + }, + { + "epoch": 0.20041, + "grad_norm": 0.9043117761611938, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 20041 + }, + { + "epoch": 0.20042, + "grad_norm": 0.9589005708694458, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 20042 + }, + { + "epoch": 0.20043, + "grad_norm": 1.015989065170288, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 20043 + }, + { + "epoch": 0.20044, + "grad_norm": 1.1127653121948242, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 20044 + }, + { + "epoch": 0.20045, + "grad_norm": 0.867679238319397, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 20045 + }, + { + "epoch": 0.20046, + "grad_norm": 0.6572129726409912, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 20046 + }, + { + "epoch": 0.20047, + "grad_norm": 0.7014610767364502, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 20047 + }, + { + "epoch": 0.20048, + "grad_norm": 0.5642576217651367, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 20048 + }, + { + "epoch": 0.20049, + "grad_norm": 0.5769412517547607, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 20049 + }, + { + "epoch": 0.2005, + "grad_norm": 0.5566190481185913, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 20050 + }, + { + "epoch": 0.20051, + "grad_norm": 0.538754940032959, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 20051 + }, + { + "epoch": 0.20052, + "grad_norm": 0.49352458119392395, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 20052 + }, + { + "epoch": 0.20053, + "grad_norm": 0.5705218315124512, + "learning_rate": 0.003, + "loss": 3.982, + "step": 20053 + }, + { + "epoch": 0.20054, + "grad_norm": 0.7290521860122681, + "learning_rate": 0.003, + "loss": 4.001, + "step": 20054 + }, + { + "epoch": 0.20055, + "grad_norm": 0.956404447555542, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 20055 + }, + { + "epoch": 0.20056, + "grad_norm": 1.1518845558166504, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 20056 + }, + { + "epoch": 0.20057, + "grad_norm": 0.8350433707237244, + "learning_rate": 0.003, + "loss": 3.981, + "step": 20057 + }, + { + "epoch": 0.20058, + "grad_norm": 0.6605261564254761, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 20058 + }, + { + "epoch": 0.20059, + "grad_norm": 0.5880551934242249, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 20059 + }, + { + "epoch": 0.2006, + "grad_norm": 0.5347115993499756, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 20060 + }, + { + "epoch": 0.20061, + "grad_norm": 0.5903977751731873, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 20061 + }, + { + "epoch": 0.20062, + "grad_norm": 0.6846911311149597, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 20062 + }, + { + "epoch": 0.20063, + "grad_norm": 0.7614838480949402, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 20063 + }, + { + "epoch": 0.20064, + "grad_norm": 0.9693281650543213, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 20064 + }, + { + "epoch": 0.20065, + "grad_norm": 1.1056090593338013, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 20065 + }, + { + "epoch": 0.20066, + "grad_norm": 0.9413695931434631, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 20066 + }, + { + "epoch": 0.20067, + "grad_norm": 1.2127419710159302, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 20067 + }, + { + "epoch": 0.20068, + "grad_norm": 0.9101815223693848, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 20068 + }, + { + "epoch": 0.20069, + "grad_norm": 0.911109447479248, + "learning_rate": 0.003, + "loss": 4.02, + "step": 20069 + }, + { + "epoch": 0.2007, + "grad_norm": 0.8823415637016296, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 20070 + }, + { + "epoch": 0.20071, + "grad_norm": 0.761504590511322, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 20071 + }, + { + "epoch": 0.20072, + "grad_norm": 0.6968681812286377, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 20072 + }, + { + "epoch": 0.20073, + "grad_norm": 0.7271966338157654, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 20073 + }, + { + "epoch": 0.20074, + "grad_norm": 0.8523584008216858, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 20074 + }, + { + "epoch": 0.20075, + "grad_norm": 1.0339120626449585, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 20075 + }, + { + "epoch": 0.20076, + "grad_norm": 1.1442428827285767, + "learning_rate": 0.003, + "loss": 4.011, + "step": 20076 + }, + { + "epoch": 0.20077, + "grad_norm": 0.8703557848930359, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 20077 + }, + { + "epoch": 0.20078, + "grad_norm": 0.8325510621070862, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 20078 + }, + { + "epoch": 0.20079, + "grad_norm": 0.9348880052566528, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 20079 + }, + { + "epoch": 0.2008, + "grad_norm": 1.0314216613769531, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 20080 + }, + { + "epoch": 0.20081, + "grad_norm": 1.0552459955215454, + "learning_rate": 0.003, + "loss": 3.996, + "step": 20081 + }, + { + "epoch": 0.20082, + "grad_norm": 0.9345569014549255, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 20082 + }, + { + "epoch": 0.20083, + "grad_norm": 0.8296226859092712, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 20083 + }, + { + "epoch": 0.20084, + "grad_norm": 0.7488117218017578, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 20084 + }, + { + "epoch": 0.20085, + "grad_norm": 0.9322945475578308, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 20085 + }, + { + "epoch": 0.20086, + "grad_norm": 1.038449764251709, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 20086 + }, + { + "epoch": 0.20087, + "grad_norm": 1.1149946451187134, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 20087 + }, + { + "epoch": 0.20088, + "grad_norm": 0.9755486249923706, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 20088 + }, + { + "epoch": 0.20089, + "grad_norm": 0.8299311995506287, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 20089 + }, + { + "epoch": 0.2009, + "grad_norm": 0.7166337370872498, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 20090 + }, + { + "epoch": 0.20091, + "grad_norm": 0.6959244012832642, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 20091 + }, + { + "epoch": 0.20092, + "grad_norm": 0.7742272615432739, + "learning_rate": 0.003, + "loss": 4.017, + "step": 20092 + }, + { + "epoch": 0.20093, + "grad_norm": 0.8766480684280396, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 20093 + }, + { + "epoch": 0.20094, + "grad_norm": 0.7357550859451294, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 20094 + }, + { + "epoch": 0.20095, + "grad_norm": 0.962944507598877, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 20095 + }, + { + "epoch": 0.20096, + "grad_norm": 1.0833114385604858, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 20096 + }, + { + "epoch": 0.20097, + "grad_norm": 1.0890945196151733, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 20097 + }, + { + "epoch": 0.20098, + "grad_norm": 0.8043980598449707, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 20098 + }, + { + "epoch": 0.20099, + "grad_norm": 0.6822202801704407, + "learning_rate": 0.003, + "loss": 4.01, + "step": 20099 + }, + { + "epoch": 0.201, + "grad_norm": 0.7577584385871887, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 20100 + }, + { + "epoch": 0.20101, + "grad_norm": 0.8958604335784912, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 20101 + }, + { + "epoch": 0.20102, + "grad_norm": 0.9075361490249634, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 20102 + }, + { + "epoch": 0.20103, + "grad_norm": 0.8133460879325867, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 20103 + }, + { + "epoch": 0.20104, + "grad_norm": 0.8210949897766113, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 20104 + }, + { + "epoch": 0.20105, + "grad_norm": 0.8700422048568726, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 20105 + }, + { + "epoch": 0.20106, + "grad_norm": 0.9867281317710876, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 20106 + }, + { + "epoch": 0.20107, + "grad_norm": 1.0752992630004883, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 20107 + }, + { + "epoch": 0.20108, + "grad_norm": 0.8385269045829773, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 20108 + }, + { + "epoch": 0.20109, + "grad_norm": 0.689854621887207, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 20109 + }, + { + "epoch": 0.2011, + "grad_norm": 0.7017920017242432, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 20110 + }, + { + "epoch": 0.20111, + "grad_norm": 0.7412782311439514, + "learning_rate": 0.003, + "loss": 3.999, + "step": 20111 + }, + { + "epoch": 0.20112, + "grad_norm": 0.703317403793335, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 20112 + }, + { + "epoch": 0.20113, + "grad_norm": 0.655783474445343, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 20113 + }, + { + "epoch": 0.20114, + "grad_norm": 0.6221177577972412, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 20114 + }, + { + "epoch": 0.20115, + "grad_norm": 0.6611085534095764, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 20115 + }, + { + "epoch": 0.20116, + "grad_norm": 0.7219901084899902, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 20116 + }, + { + "epoch": 0.20117, + "grad_norm": 1.033886432647705, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 20117 + }, + { + "epoch": 0.20118, + "grad_norm": 1.2467701435089111, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 20118 + }, + { + "epoch": 0.20119, + "grad_norm": 0.7234686613082886, + "learning_rate": 0.003, + "loss": 4.002, + "step": 20119 + }, + { + "epoch": 0.2012, + "grad_norm": 0.6081365942955017, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 20120 + }, + { + "epoch": 0.20121, + "grad_norm": 0.6261841058731079, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 20121 + }, + { + "epoch": 0.20122, + "grad_norm": 0.7835150361061096, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 20122 + }, + { + "epoch": 0.20123, + "grad_norm": 0.8453176021575928, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 20123 + }, + { + "epoch": 0.20124, + "grad_norm": 0.9139364957809448, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 20124 + }, + { + "epoch": 0.20125, + "grad_norm": 1.1292014122009277, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 20125 + }, + { + "epoch": 0.20126, + "grad_norm": 1.171099066734314, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 20126 + }, + { + "epoch": 0.20127, + "grad_norm": 0.7680856585502625, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 20127 + }, + { + "epoch": 0.20128, + "grad_norm": 0.6378629803657532, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 20128 + }, + { + "epoch": 0.20129, + "grad_norm": 0.7174150943756104, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 20129 + }, + { + "epoch": 0.2013, + "grad_norm": 0.91487056016922, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 20130 + }, + { + "epoch": 0.20131, + "grad_norm": 1.1376959085464478, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 20131 + }, + { + "epoch": 0.20132, + "grad_norm": 0.8521286845207214, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 20132 + }, + { + "epoch": 0.20133, + "grad_norm": 0.7289743423461914, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 20133 + }, + { + "epoch": 0.20134, + "grad_norm": 0.7509118318557739, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 20134 + }, + { + "epoch": 0.20135, + "grad_norm": 1.0695552825927734, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 20135 + }, + { + "epoch": 0.20136, + "grad_norm": 1.0982223749160767, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 20136 + }, + { + "epoch": 0.20137, + "grad_norm": 0.9279093146324158, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 20137 + }, + { + "epoch": 0.20138, + "grad_norm": 0.8908368349075317, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 20138 + }, + { + "epoch": 0.20139, + "grad_norm": 0.762957751750946, + "learning_rate": 0.003, + "loss": 3.995, + "step": 20139 + }, + { + "epoch": 0.2014, + "grad_norm": 0.6879292130470276, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 20140 + }, + { + "epoch": 0.20141, + "grad_norm": 0.7439958453178406, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 20141 + }, + { + "epoch": 0.20142, + "grad_norm": 0.8566961884498596, + "learning_rate": 0.003, + "loss": 3.981, + "step": 20142 + }, + { + "epoch": 0.20143, + "grad_norm": 0.8951508402824402, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 20143 + }, + { + "epoch": 0.20144, + "grad_norm": 1.1969650983810425, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 20144 + }, + { + "epoch": 0.20145, + "grad_norm": 0.8651864528656006, + "learning_rate": 0.003, + "loss": 4.005, + "step": 20145 + }, + { + "epoch": 0.20146, + "grad_norm": 0.7943025827407837, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 20146 + }, + { + "epoch": 0.20147, + "grad_norm": 0.8146642446517944, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 20147 + }, + { + "epoch": 0.20148, + "grad_norm": 1.0197103023529053, + "learning_rate": 0.003, + "loss": 4.011, + "step": 20148 + }, + { + "epoch": 0.20149, + "grad_norm": 1.257275104522705, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 20149 + }, + { + "epoch": 0.2015, + "grad_norm": 0.7094889283180237, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 20150 + }, + { + "epoch": 0.20151, + "grad_norm": 0.660546064376831, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 20151 + }, + { + "epoch": 0.20152, + "grad_norm": 0.6721454858779907, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 20152 + }, + { + "epoch": 0.20153, + "grad_norm": 0.7225807905197144, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 20153 + }, + { + "epoch": 0.20154, + "grad_norm": 0.9250208139419556, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 20154 + }, + { + "epoch": 0.20155, + "grad_norm": 0.9858287572860718, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 20155 + }, + { + "epoch": 0.20156, + "grad_norm": 0.8439344763755798, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 20156 + }, + { + "epoch": 0.20157, + "grad_norm": 0.7250839471817017, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 20157 + }, + { + "epoch": 0.20158, + "grad_norm": 0.7875615954399109, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 20158 + }, + { + "epoch": 0.20159, + "grad_norm": 0.9504771828651428, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 20159 + }, + { + "epoch": 0.2016, + "grad_norm": 1.1765533685684204, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 20160 + }, + { + "epoch": 0.20161, + "grad_norm": 1.008900761604309, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 20161 + }, + { + "epoch": 0.20162, + "grad_norm": 0.9553276300430298, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 20162 + }, + { + "epoch": 0.20163, + "grad_norm": 0.8626341819763184, + "learning_rate": 0.003, + "loss": 4.015, + "step": 20163 + }, + { + "epoch": 0.20164, + "grad_norm": 0.7288162708282471, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 20164 + }, + { + "epoch": 0.20165, + "grad_norm": 0.6690157651901245, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 20165 + }, + { + "epoch": 0.20166, + "grad_norm": 0.6751003265380859, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 20166 + }, + { + "epoch": 0.20167, + "grad_norm": 0.7453999519348145, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 20167 + }, + { + "epoch": 0.20168, + "grad_norm": 0.8011784553527832, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 20168 + }, + { + "epoch": 0.20169, + "grad_norm": 0.901778519153595, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 20169 + }, + { + "epoch": 0.2017, + "grad_norm": 0.9532800316810608, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 20170 + }, + { + "epoch": 0.20171, + "grad_norm": 1.0813367366790771, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 20171 + }, + { + "epoch": 0.20172, + "grad_norm": 0.961879312992096, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 20172 + }, + { + "epoch": 0.20173, + "grad_norm": 0.7712802290916443, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 20173 + }, + { + "epoch": 0.20174, + "grad_norm": 0.7062988877296448, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 20174 + }, + { + "epoch": 0.20175, + "grad_norm": 0.7877758741378784, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 20175 + }, + { + "epoch": 0.20176, + "grad_norm": 0.9011003375053406, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 20176 + }, + { + "epoch": 0.20177, + "grad_norm": 1.0270823240280151, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 20177 + }, + { + "epoch": 0.20178, + "grad_norm": 0.9703652262687683, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 20178 + }, + { + "epoch": 0.20179, + "grad_norm": 0.9204882979393005, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 20179 + }, + { + "epoch": 0.2018, + "grad_norm": 0.9217516779899597, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 20180 + }, + { + "epoch": 0.20181, + "grad_norm": 0.811480700969696, + "learning_rate": 0.003, + "loss": 3.984, + "step": 20181 + }, + { + "epoch": 0.20182, + "grad_norm": 0.953285813331604, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 20182 + }, + { + "epoch": 0.20183, + "grad_norm": 1.112351894378662, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 20183 + }, + { + "epoch": 0.20184, + "grad_norm": 1.1086965799331665, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 20184 + }, + { + "epoch": 0.20185, + "grad_norm": 0.8778813481330872, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 20185 + }, + { + "epoch": 0.20186, + "grad_norm": 0.8389912843704224, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 20186 + }, + { + "epoch": 0.20187, + "grad_norm": 0.9361748695373535, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 20187 + }, + { + "epoch": 0.20188, + "grad_norm": 0.8386715054512024, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 20188 + }, + { + "epoch": 0.20189, + "grad_norm": 0.7701196670532227, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 20189 + }, + { + "epoch": 0.2019, + "grad_norm": 0.6367450952529907, + "learning_rate": 0.003, + "loss": 4.004, + "step": 20190 + }, + { + "epoch": 0.20191, + "grad_norm": 0.5951288938522339, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 20191 + }, + { + "epoch": 0.20192, + "grad_norm": 0.7320143580436707, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 20192 + }, + { + "epoch": 0.20193, + "grad_norm": 0.7183142304420471, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 20193 + }, + { + "epoch": 0.20194, + "grad_norm": 0.7571724057197571, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 20194 + }, + { + "epoch": 0.20195, + "grad_norm": 0.6767758727073669, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 20195 + }, + { + "epoch": 0.20196, + "grad_norm": 0.7127629518508911, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 20196 + }, + { + "epoch": 0.20197, + "grad_norm": 0.7226316928863525, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 20197 + }, + { + "epoch": 0.20198, + "grad_norm": 0.9264624118804932, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 20198 + }, + { + "epoch": 0.20199, + "grad_norm": 1.3039424419403076, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 20199 + }, + { + "epoch": 0.202, + "grad_norm": 0.7310620546340942, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 20200 + }, + { + "epoch": 0.20201, + "grad_norm": 0.6267219185829163, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 20201 + }, + { + "epoch": 0.20202, + "grad_norm": 0.5859816074371338, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 20202 + }, + { + "epoch": 0.20203, + "grad_norm": 0.6759727001190186, + "learning_rate": 0.003, + "loss": 3.993, + "step": 20203 + }, + { + "epoch": 0.20204, + "grad_norm": 0.6308323740959167, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 20204 + }, + { + "epoch": 0.20205, + "grad_norm": 0.714514970779419, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 20205 + }, + { + "epoch": 0.20206, + "grad_norm": 0.9549436569213867, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 20206 + }, + { + "epoch": 0.20207, + "grad_norm": 1.2413829565048218, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 20207 + }, + { + "epoch": 0.20208, + "grad_norm": 0.9563367962837219, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 20208 + }, + { + "epoch": 0.20209, + "grad_norm": 0.8096930980682373, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 20209 + }, + { + "epoch": 0.2021, + "grad_norm": 0.6866564750671387, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 20210 + }, + { + "epoch": 0.20211, + "grad_norm": 0.7624746561050415, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 20211 + }, + { + "epoch": 0.20212, + "grad_norm": 0.8786949515342712, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 20212 + }, + { + "epoch": 0.20213, + "grad_norm": 0.8951416611671448, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 20213 + }, + { + "epoch": 0.20214, + "grad_norm": 0.9443157911300659, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 20214 + }, + { + "epoch": 0.20215, + "grad_norm": 0.933879554271698, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 20215 + }, + { + "epoch": 0.20216, + "grad_norm": 0.8607663512229919, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 20216 + }, + { + "epoch": 0.20217, + "grad_norm": 0.9158996939659119, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 20217 + }, + { + "epoch": 0.20218, + "grad_norm": 0.8870666027069092, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 20218 + }, + { + "epoch": 0.20219, + "grad_norm": 0.8453297019004822, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 20219 + }, + { + "epoch": 0.2022, + "grad_norm": 0.9020869731903076, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 20220 + }, + { + "epoch": 0.20221, + "grad_norm": 0.9760621190071106, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 20221 + }, + { + "epoch": 0.20222, + "grad_norm": 1.164143443107605, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 20222 + }, + { + "epoch": 0.20223, + "grad_norm": 0.7390578985214233, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 20223 + }, + { + "epoch": 0.20224, + "grad_norm": 0.7947418689727783, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 20224 + }, + { + "epoch": 0.20225, + "grad_norm": 0.769934892654419, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 20225 + }, + { + "epoch": 0.20226, + "grad_norm": 0.8830723166465759, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 20226 + }, + { + "epoch": 0.20227, + "grad_norm": 1.1716639995574951, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 20227 + }, + { + "epoch": 0.20228, + "grad_norm": 0.7477881908416748, + "learning_rate": 0.003, + "loss": 3.974, + "step": 20228 + }, + { + "epoch": 0.20229, + "grad_norm": 0.6853918433189392, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 20229 + }, + { + "epoch": 0.2023, + "grad_norm": 0.6665154099464417, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 20230 + }, + { + "epoch": 0.20231, + "grad_norm": 0.5964629054069519, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 20231 + }, + { + "epoch": 0.20232, + "grad_norm": 0.6783205270767212, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 20232 + }, + { + "epoch": 0.20233, + "grad_norm": 0.7677664756774902, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 20233 + }, + { + "epoch": 0.20234, + "grad_norm": 0.7461740374565125, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 20234 + }, + { + "epoch": 0.20235, + "grad_norm": 0.6478622555732727, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 20235 + }, + { + "epoch": 0.20236, + "grad_norm": 0.7441017627716064, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 20236 + }, + { + "epoch": 0.20237, + "grad_norm": 0.771283745765686, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 20237 + }, + { + "epoch": 0.20238, + "grad_norm": 0.8057474493980408, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 20238 + }, + { + "epoch": 0.20239, + "grad_norm": 0.7607077360153198, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 20239 + }, + { + "epoch": 0.2024, + "grad_norm": 0.7734676599502563, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 20240 + }, + { + "epoch": 0.20241, + "grad_norm": 1.048604965209961, + "learning_rate": 0.003, + "loss": 4.012, + "step": 20241 + }, + { + "epoch": 0.20242, + "grad_norm": 1.2987990379333496, + "learning_rate": 0.003, + "loss": 4.022, + "step": 20242 + }, + { + "epoch": 0.20243, + "grad_norm": 1.064478874206543, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 20243 + }, + { + "epoch": 0.20244, + "grad_norm": 1.1175570487976074, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 20244 + }, + { + "epoch": 0.20245, + "grad_norm": 0.850923478603363, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 20245 + }, + { + "epoch": 0.20246, + "grad_norm": 0.7859277725219727, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 20246 + }, + { + "epoch": 0.20247, + "grad_norm": 0.7892622351646423, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 20247 + }, + { + "epoch": 0.20248, + "grad_norm": 0.7633906006813049, + "learning_rate": 0.003, + "loss": 3.983, + "step": 20248 + }, + { + "epoch": 0.20249, + "grad_norm": 0.6468591690063477, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 20249 + }, + { + "epoch": 0.2025, + "grad_norm": 0.6362141370773315, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 20250 + }, + { + "epoch": 0.20251, + "grad_norm": 0.7041432857513428, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 20251 + }, + { + "epoch": 0.20252, + "grad_norm": 0.8176714777946472, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 20252 + }, + { + "epoch": 0.20253, + "grad_norm": 0.97947096824646, + "learning_rate": 0.003, + "loss": 3.985, + "step": 20253 + }, + { + "epoch": 0.20254, + "grad_norm": 1.2270678281784058, + "learning_rate": 0.003, + "loss": 4.045, + "step": 20254 + }, + { + "epoch": 0.20255, + "grad_norm": 0.7587008476257324, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 20255 + }, + { + "epoch": 0.20256, + "grad_norm": 0.7780997157096863, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 20256 + }, + { + "epoch": 0.20257, + "grad_norm": 0.8538438677787781, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 20257 + }, + { + "epoch": 0.20258, + "grad_norm": 0.9861294627189636, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 20258 + }, + { + "epoch": 0.20259, + "grad_norm": 0.9398753643035889, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 20259 + }, + { + "epoch": 0.2026, + "grad_norm": 1.0167715549468994, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 20260 + }, + { + "epoch": 0.20261, + "grad_norm": 1.0032497644424438, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 20261 + }, + { + "epoch": 0.20262, + "grad_norm": 0.9985441565513611, + "learning_rate": 0.003, + "loss": 4.007, + "step": 20262 + }, + { + "epoch": 0.20263, + "grad_norm": 1.136631727218628, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 20263 + }, + { + "epoch": 0.20264, + "grad_norm": 1.066340446472168, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 20264 + }, + { + "epoch": 0.20265, + "grad_norm": 0.9510641694068909, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 20265 + }, + { + "epoch": 0.20266, + "grad_norm": 0.969109833240509, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 20266 + }, + { + "epoch": 0.20267, + "grad_norm": 0.8656512498855591, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 20267 + }, + { + "epoch": 0.20268, + "grad_norm": 0.8168076276779175, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 20268 + }, + { + "epoch": 0.20269, + "grad_norm": 0.8075709939002991, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 20269 + }, + { + "epoch": 0.2027, + "grad_norm": 0.7317737936973572, + "learning_rate": 0.003, + "loss": 4.006, + "step": 20270 + }, + { + "epoch": 0.20271, + "grad_norm": 0.7332939505577087, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 20271 + }, + { + "epoch": 0.20272, + "grad_norm": 0.7304376363754272, + "learning_rate": 0.003, + "loss": 4.003, + "step": 20272 + }, + { + "epoch": 0.20273, + "grad_norm": 0.8709371089935303, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 20273 + }, + { + "epoch": 0.20274, + "grad_norm": 1.1287729740142822, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 20274 + }, + { + "epoch": 0.20275, + "grad_norm": 0.9655367136001587, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 20275 + }, + { + "epoch": 0.20276, + "grad_norm": 1.0948092937469482, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 20276 + }, + { + "epoch": 0.20277, + "grad_norm": 0.9808290600776672, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 20277 + }, + { + "epoch": 0.20278, + "grad_norm": 0.841752290725708, + "learning_rate": 0.003, + "loss": 3.974, + "step": 20278 + }, + { + "epoch": 0.20279, + "grad_norm": 0.7497544884681702, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 20279 + }, + { + "epoch": 0.2028, + "grad_norm": 0.6757055521011353, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 20280 + }, + { + "epoch": 0.20281, + "grad_norm": 0.6535053253173828, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 20281 + }, + { + "epoch": 0.20282, + "grad_norm": 0.7720223665237427, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 20282 + }, + { + "epoch": 0.20283, + "grad_norm": 1.0587536096572876, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 20283 + }, + { + "epoch": 0.20284, + "grad_norm": 1.1201404333114624, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 20284 + }, + { + "epoch": 0.20285, + "grad_norm": 0.8129140138626099, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 20285 + }, + { + "epoch": 0.20286, + "grad_norm": 0.6837013363838196, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 20286 + }, + { + "epoch": 0.20287, + "grad_norm": 0.6548781991004944, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 20287 + }, + { + "epoch": 0.20288, + "grad_norm": 0.8084627389907837, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 20288 + }, + { + "epoch": 0.20289, + "grad_norm": 1.0328177213668823, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 20289 + }, + { + "epoch": 0.2029, + "grad_norm": 1.1478947401046753, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 20290 + }, + { + "epoch": 0.20291, + "grad_norm": 0.7494494318962097, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 20291 + }, + { + "epoch": 0.20292, + "grad_norm": 0.679472029209137, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 20292 + }, + { + "epoch": 0.20293, + "grad_norm": 0.6625007390975952, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 20293 + }, + { + "epoch": 0.20294, + "grad_norm": 0.6561849117279053, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 20294 + }, + { + "epoch": 0.20295, + "grad_norm": 0.7200389504432678, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 20295 + }, + { + "epoch": 0.20296, + "grad_norm": 0.7838565111160278, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 20296 + }, + { + "epoch": 0.20297, + "grad_norm": 0.9034474492073059, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 20297 + }, + { + "epoch": 0.20298, + "grad_norm": 1.0330101251602173, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 20298 + }, + { + "epoch": 0.20299, + "grad_norm": 0.9066010117530823, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 20299 + }, + { + "epoch": 0.203, + "grad_norm": 0.9214524626731873, + "learning_rate": 0.003, + "loss": 4.017, + "step": 20300 + }, + { + "epoch": 0.20301, + "grad_norm": 1.0282068252563477, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 20301 + }, + { + "epoch": 0.20302, + "grad_norm": 1.0103950500488281, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 20302 + }, + { + "epoch": 0.20303, + "grad_norm": 0.9253962635993958, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 20303 + }, + { + "epoch": 0.20304, + "grad_norm": 1.0147264003753662, + "learning_rate": 0.003, + "loss": 3.998, + "step": 20304 + }, + { + "epoch": 0.20305, + "grad_norm": 0.978193461894989, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 20305 + }, + { + "epoch": 0.20306, + "grad_norm": 1.0849967002868652, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 20306 + }, + { + "epoch": 0.20307, + "grad_norm": 1.0655522346496582, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 20307 + }, + { + "epoch": 0.20308, + "grad_norm": 0.8298773765563965, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 20308 + }, + { + "epoch": 0.20309, + "grad_norm": 0.7420493960380554, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 20309 + }, + { + "epoch": 0.2031, + "grad_norm": 0.6811785101890564, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 20310 + }, + { + "epoch": 0.20311, + "grad_norm": 0.6202403903007507, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 20311 + }, + { + "epoch": 0.20312, + "grad_norm": 0.6893637180328369, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 20312 + }, + { + "epoch": 0.20313, + "grad_norm": 0.7760577201843262, + "learning_rate": 0.003, + "loss": 4.004, + "step": 20313 + }, + { + "epoch": 0.20314, + "grad_norm": 0.8740729689598083, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 20314 + }, + { + "epoch": 0.20315, + "grad_norm": 1.0910762548446655, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 20315 + }, + { + "epoch": 0.20316, + "grad_norm": 0.9553356766700745, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 20316 + }, + { + "epoch": 0.20317, + "grad_norm": 0.8944961428642273, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 20317 + }, + { + "epoch": 0.20318, + "grad_norm": 1.0081169605255127, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 20318 + }, + { + "epoch": 0.20319, + "grad_norm": 0.8911204934120178, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 20319 + }, + { + "epoch": 0.2032, + "grad_norm": 0.8690194487571716, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 20320 + }, + { + "epoch": 0.20321, + "grad_norm": 0.7861948013305664, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 20321 + }, + { + "epoch": 0.20322, + "grad_norm": 0.7284369468688965, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 20322 + }, + { + "epoch": 0.20323, + "grad_norm": 0.7329832315444946, + "learning_rate": 0.003, + "loss": 4.022, + "step": 20323 + }, + { + "epoch": 0.20324, + "grad_norm": 0.5863949060440063, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 20324 + }, + { + "epoch": 0.20325, + "grad_norm": 0.6903998851776123, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 20325 + }, + { + "epoch": 0.20326, + "grad_norm": 0.7285104990005493, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 20326 + }, + { + "epoch": 0.20327, + "grad_norm": 1.0294723510742188, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 20327 + }, + { + "epoch": 0.20328, + "grad_norm": 1.311887264251709, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 20328 + }, + { + "epoch": 0.20329, + "grad_norm": 0.5420759320259094, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 20329 + }, + { + "epoch": 0.2033, + "grad_norm": 0.7763385772705078, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 20330 + }, + { + "epoch": 0.20331, + "grad_norm": 0.9855422377586365, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 20331 + }, + { + "epoch": 0.20332, + "grad_norm": 1.1434701681137085, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 20332 + }, + { + "epoch": 0.20333, + "grad_norm": 0.7839611768722534, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 20333 + }, + { + "epoch": 0.20334, + "grad_norm": 0.6856549978256226, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 20334 + }, + { + "epoch": 0.20335, + "grad_norm": 0.7158058881759644, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 20335 + }, + { + "epoch": 0.20336, + "grad_norm": 0.7355321049690247, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 20336 + }, + { + "epoch": 0.20337, + "grad_norm": 0.8203396797180176, + "learning_rate": 0.003, + "loss": 3.994, + "step": 20337 + }, + { + "epoch": 0.20338, + "grad_norm": 0.8416463136672974, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 20338 + }, + { + "epoch": 0.20339, + "grad_norm": 0.8109787702560425, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 20339 + }, + { + "epoch": 0.2034, + "grad_norm": 0.7734821438789368, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 20340 + }, + { + "epoch": 0.20341, + "grad_norm": 0.8469385504722595, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 20341 + }, + { + "epoch": 0.20342, + "grad_norm": 1.1418490409851074, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 20342 + }, + { + "epoch": 0.20343, + "grad_norm": 0.8195468783378601, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 20343 + }, + { + "epoch": 0.20344, + "grad_norm": 0.7609240412712097, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 20344 + }, + { + "epoch": 0.20345, + "grad_norm": 0.8220378160476685, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 20345 + }, + { + "epoch": 0.20346, + "grad_norm": 0.9295966625213623, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 20346 + }, + { + "epoch": 0.20347, + "grad_norm": 0.8783484101295471, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 20347 + }, + { + "epoch": 0.20348, + "grad_norm": 0.8919439315795898, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 20348 + }, + { + "epoch": 0.20349, + "grad_norm": 0.7911661863327026, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 20349 + }, + { + "epoch": 0.2035, + "grad_norm": 0.7926628589630127, + "learning_rate": 0.003, + "loss": 4.008, + "step": 20350 + }, + { + "epoch": 0.20351, + "grad_norm": 0.7087711095809937, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 20351 + }, + { + "epoch": 0.20352, + "grad_norm": 0.7259397506713867, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 20352 + }, + { + "epoch": 0.20353, + "grad_norm": 0.6317782998085022, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 20353 + }, + { + "epoch": 0.20354, + "grad_norm": 0.7174623608589172, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 20354 + }, + { + "epoch": 0.20355, + "grad_norm": 0.6670953035354614, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 20355 + }, + { + "epoch": 0.20356, + "grad_norm": 0.6677290797233582, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 20356 + }, + { + "epoch": 0.20357, + "grad_norm": 0.6100268363952637, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 20357 + }, + { + "epoch": 0.20358, + "grad_norm": 0.7424579858779907, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 20358 + }, + { + "epoch": 0.20359, + "grad_norm": 0.956942081451416, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 20359 + }, + { + "epoch": 0.2036, + "grad_norm": 1.204334020614624, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 20360 + }, + { + "epoch": 0.20361, + "grad_norm": 0.9877221584320068, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 20361 + }, + { + "epoch": 0.20362, + "grad_norm": 1.0937947034835815, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 20362 + }, + { + "epoch": 0.20363, + "grad_norm": 0.8922629952430725, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 20363 + }, + { + "epoch": 0.20364, + "grad_norm": 0.8023450970649719, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 20364 + }, + { + "epoch": 0.20365, + "grad_norm": 0.7399235367774963, + "learning_rate": 0.003, + "loss": 3.994, + "step": 20365 + }, + { + "epoch": 0.20366, + "grad_norm": 0.9052486419677734, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 20366 + }, + { + "epoch": 0.20367, + "grad_norm": 1.0596771240234375, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 20367 + }, + { + "epoch": 0.20368, + "grad_norm": 0.8454614281654358, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 20368 + }, + { + "epoch": 0.20369, + "grad_norm": 0.7641255855560303, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 20369 + }, + { + "epoch": 0.2037, + "grad_norm": 0.7677737474441528, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 20370 + }, + { + "epoch": 0.20371, + "grad_norm": 0.7251455187797546, + "learning_rate": 0.003, + "loss": 3.998, + "step": 20371 + }, + { + "epoch": 0.20372, + "grad_norm": 0.6663448214530945, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 20372 + }, + { + "epoch": 0.20373, + "grad_norm": 0.6636404395103455, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 20373 + }, + { + "epoch": 0.20374, + "grad_norm": 0.6882753968238831, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 20374 + }, + { + "epoch": 0.20375, + "grad_norm": 0.7860545516014099, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 20375 + }, + { + "epoch": 0.20376, + "grad_norm": 1.0693397521972656, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 20376 + }, + { + "epoch": 0.20377, + "grad_norm": 1.2248581647872925, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 20377 + }, + { + "epoch": 0.20378, + "grad_norm": 0.8462511301040649, + "learning_rate": 0.003, + "loss": 3.994, + "step": 20378 + }, + { + "epoch": 0.20379, + "grad_norm": 0.8828621506690979, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 20379 + }, + { + "epoch": 0.2038, + "grad_norm": 0.9216285943984985, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 20380 + }, + { + "epoch": 0.20381, + "grad_norm": 0.9714720249176025, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 20381 + }, + { + "epoch": 0.20382, + "grad_norm": 0.9852808713912964, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 20382 + }, + { + "epoch": 0.20383, + "grad_norm": 0.8930702209472656, + "learning_rate": 0.003, + "loss": 4.013, + "step": 20383 + }, + { + "epoch": 0.20384, + "grad_norm": 1.0646729469299316, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 20384 + }, + { + "epoch": 0.20385, + "grad_norm": 1.072921633720398, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 20385 + }, + { + "epoch": 0.20386, + "grad_norm": 1.1347222328186035, + "learning_rate": 0.003, + "loss": 3.999, + "step": 20386 + }, + { + "epoch": 0.20387, + "grad_norm": 0.7293988466262817, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 20387 + }, + { + "epoch": 0.20388, + "grad_norm": 0.6437301635742188, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 20388 + }, + { + "epoch": 0.20389, + "grad_norm": 0.6251480579376221, + "learning_rate": 0.003, + "loss": 3.971, + "step": 20389 + }, + { + "epoch": 0.2039, + "grad_norm": 0.6301602721214294, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 20390 + }, + { + "epoch": 0.20391, + "grad_norm": 0.6458478569984436, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 20391 + }, + { + "epoch": 0.20392, + "grad_norm": 0.788723886013031, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 20392 + }, + { + "epoch": 0.20393, + "grad_norm": 0.9218687415122986, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 20393 + }, + { + "epoch": 0.20394, + "grad_norm": 1.1501364707946777, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 20394 + }, + { + "epoch": 0.20395, + "grad_norm": 1.05363130569458, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 20395 + }, + { + "epoch": 0.20396, + "grad_norm": 0.8909705877304077, + "learning_rate": 0.003, + "loss": 3.968, + "step": 20396 + }, + { + "epoch": 0.20397, + "grad_norm": 0.7462480664253235, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 20397 + }, + { + "epoch": 0.20398, + "grad_norm": 0.694731593132019, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 20398 + }, + { + "epoch": 0.20399, + "grad_norm": 0.667031466960907, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 20399 + }, + { + "epoch": 0.204, + "grad_norm": 0.6090689897537231, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 20400 + }, + { + "epoch": 0.20401, + "grad_norm": 0.627703070640564, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 20401 + }, + { + "epoch": 0.20402, + "grad_norm": 0.6416513323783875, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 20402 + }, + { + "epoch": 0.20403, + "grad_norm": 0.7873333692550659, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 20403 + }, + { + "epoch": 0.20404, + "grad_norm": 1.1909332275390625, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 20404 + }, + { + "epoch": 0.20405, + "grad_norm": 0.9008625745773315, + "learning_rate": 0.003, + "loss": 3.98, + "step": 20405 + }, + { + "epoch": 0.20406, + "grad_norm": 0.7145691514015198, + "learning_rate": 0.003, + "loss": 3.993, + "step": 20406 + }, + { + "epoch": 0.20407, + "grad_norm": 0.7303884029388428, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 20407 + }, + { + "epoch": 0.20408, + "grad_norm": 0.681465744972229, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 20408 + }, + { + "epoch": 0.20409, + "grad_norm": 0.7974075078964233, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 20409 + }, + { + "epoch": 0.2041, + "grad_norm": 0.9380106329917908, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 20410 + }, + { + "epoch": 0.20411, + "grad_norm": 1.085464358329773, + "learning_rate": 0.003, + "loss": 3.98, + "step": 20411 + }, + { + "epoch": 0.20412, + "grad_norm": 0.9400927424430847, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 20412 + }, + { + "epoch": 0.20413, + "grad_norm": 0.886445939540863, + "learning_rate": 0.003, + "loss": 3.996, + "step": 20413 + }, + { + "epoch": 0.20414, + "grad_norm": 0.9608079791069031, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 20414 + }, + { + "epoch": 0.20415, + "grad_norm": 1.0392273664474487, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 20415 + }, + { + "epoch": 0.20416, + "grad_norm": 0.870465099811554, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 20416 + }, + { + "epoch": 0.20417, + "grad_norm": 0.8516085147857666, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 20417 + }, + { + "epoch": 0.20418, + "grad_norm": 1.0546752214431763, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 20418 + }, + { + "epoch": 0.20419, + "grad_norm": 1.0786575078964233, + "learning_rate": 0.003, + "loss": 4.008, + "step": 20419 + }, + { + "epoch": 0.2042, + "grad_norm": 0.9115407466888428, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 20420 + }, + { + "epoch": 0.20421, + "grad_norm": 0.8256513476371765, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 20421 + }, + { + "epoch": 0.20422, + "grad_norm": 0.907086968421936, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 20422 + }, + { + "epoch": 0.20423, + "grad_norm": 1.1340298652648926, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 20423 + }, + { + "epoch": 0.20424, + "grad_norm": 0.8989710211753845, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 20424 + }, + { + "epoch": 0.20425, + "grad_norm": 0.8972444534301758, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 20425 + }, + { + "epoch": 0.20426, + "grad_norm": 1.0549273490905762, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 20426 + }, + { + "epoch": 0.20427, + "grad_norm": 1.1332916021347046, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 20427 + }, + { + "epoch": 0.20428, + "grad_norm": 0.774750828742981, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 20428 + }, + { + "epoch": 0.20429, + "grad_norm": 0.6685149073600769, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 20429 + }, + { + "epoch": 0.2043, + "grad_norm": 0.6455409526824951, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 20430 + }, + { + "epoch": 0.20431, + "grad_norm": 0.6255378723144531, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 20431 + }, + { + "epoch": 0.20432, + "grad_norm": 0.7297254204750061, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 20432 + }, + { + "epoch": 0.20433, + "grad_norm": 0.8061620593070984, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 20433 + }, + { + "epoch": 0.20434, + "grad_norm": 0.9574712514877319, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 20434 + }, + { + "epoch": 0.20435, + "grad_norm": 0.9712481498718262, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 20435 + }, + { + "epoch": 0.20436, + "grad_norm": 1.0066169500350952, + "learning_rate": 0.003, + "loss": 3.99, + "step": 20436 + }, + { + "epoch": 0.20437, + "grad_norm": 1.0666333436965942, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 20437 + }, + { + "epoch": 0.20438, + "grad_norm": 0.9452031254768372, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 20438 + }, + { + "epoch": 0.20439, + "grad_norm": 1.003495693206787, + "learning_rate": 0.003, + "loss": 4.012, + "step": 20439 + }, + { + "epoch": 0.2044, + "grad_norm": 0.910893440246582, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 20440 + }, + { + "epoch": 0.20441, + "grad_norm": 0.8472775816917419, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 20441 + }, + { + "epoch": 0.20442, + "grad_norm": 0.682340681552887, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 20442 + }, + { + "epoch": 0.20443, + "grad_norm": 0.7600615620613098, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 20443 + }, + { + "epoch": 0.20444, + "grad_norm": 0.9301922917366028, + "learning_rate": 0.003, + "loss": 4.01, + "step": 20444 + }, + { + "epoch": 0.20445, + "grad_norm": 1.0311607122421265, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 20445 + }, + { + "epoch": 0.20446, + "grad_norm": 0.96709144115448, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 20446 + }, + { + "epoch": 0.20447, + "grad_norm": 1.1222336292266846, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 20447 + }, + { + "epoch": 0.20448, + "grad_norm": 0.8827340006828308, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 20448 + }, + { + "epoch": 0.20449, + "grad_norm": 0.7806417346000671, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 20449 + }, + { + "epoch": 0.2045, + "grad_norm": 0.8268476724624634, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 20450 + }, + { + "epoch": 0.20451, + "grad_norm": 0.8784385919570923, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 20451 + }, + { + "epoch": 0.20452, + "grad_norm": 0.8739992380142212, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 20452 + }, + { + "epoch": 0.20453, + "grad_norm": 0.9219098687171936, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 20453 + }, + { + "epoch": 0.20454, + "grad_norm": 1.051000952720642, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 20454 + }, + { + "epoch": 0.20455, + "grad_norm": 1.1686969995498657, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 20455 + }, + { + "epoch": 0.20456, + "grad_norm": 0.7915661931037903, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 20456 + }, + { + "epoch": 0.20457, + "grad_norm": 0.7893812656402588, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 20457 + }, + { + "epoch": 0.20458, + "grad_norm": 0.9086239337921143, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 20458 + }, + { + "epoch": 0.20459, + "grad_norm": 1.0326051712036133, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 20459 + }, + { + "epoch": 0.2046, + "grad_norm": 1.0527185201644897, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 20460 + }, + { + "epoch": 0.20461, + "grad_norm": 0.8683894276618958, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 20461 + }, + { + "epoch": 0.20462, + "grad_norm": 0.636903703212738, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 20462 + }, + { + "epoch": 0.20463, + "grad_norm": 0.6364380717277527, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 20463 + }, + { + "epoch": 0.20464, + "grad_norm": 0.6431009769439697, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 20464 + }, + { + "epoch": 0.20465, + "grad_norm": 0.7139235734939575, + "learning_rate": 0.003, + "loss": 4.002, + "step": 20465 + }, + { + "epoch": 0.20466, + "grad_norm": 0.6889108419418335, + "learning_rate": 0.003, + "loss": 4.022, + "step": 20466 + }, + { + "epoch": 0.20467, + "grad_norm": 0.668367326259613, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 20467 + }, + { + "epoch": 0.20468, + "grad_norm": 0.6962161660194397, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 20468 + }, + { + "epoch": 0.20469, + "grad_norm": 0.7134402990341187, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 20469 + }, + { + "epoch": 0.2047, + "grad_norm": 0.7592452764511108, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 20470 + }, + { + "epoch": 0.20471, + "grad_norm": 0.8259060978889465, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 20471 + }, + { + "epoch": 0.20472, + "grad_norm": 0.7741798758506775, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 20472 + }, + { + "epoch": 0.20473, + "grad_norm": 0.7275305986404419, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 20473 + }, + { + "epoch": 0.20474, + "grad_norm": 0.8968299627304077, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 20474 + }, + { + "epoch": 0.20475, + "grad_norm": 1.1683197021484375, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 20475 + }, + { + "epoch": 0.20476, + "grad_norm": 1.089591383934021, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 20476 + }, + { + "epoch": 0.20477, + "grad_norm": 0.8169175982475281, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 20477 + }, + { + "epoch": 0.20478, + "grad_norm": 0.625155508518219, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 20478 + }, + { + "epoch": 0.20479, + "grad_norm": 0.6189466714859009, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 20479 + }, + { + "epoch": 0.2048, + "grad_norm": 0.741943359375, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 20480 + }, + { + "epoch": 0.20481, + "grad_norm": 0.8120701909065247, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 20481 + }, + { + "epoch": 0.20482, + "grad_norm": 0.8646007180213928, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 20482 + }, + { + "epoch": 0.20483, + "grad_norm": 0.9611397385597229, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 20483 + }, + { + "epoch": 0.20484, + "grad_norm": 0.9605675935745239, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 20484 + }, + { + "epoch": 0.20485, + "grad_norm": 0.7831942439079285, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 20485 + }, + { + "epoch": 0.20486, + "grad_norm": 0.6748376488685608, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 20486 + }, + { + "epoch": 0.20487, + "grad_norm": 0.7112252116203308, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 20487 + }, + { + "epoch": 0.20488, + "grad_norm": 0.8787015080451965, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 20488 + }, + { + "epoch": 0.20489, + "grad_norm": 1.1302354335784912, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 20489 + }, + { + "epoch": 0.2049, + "grad_norm": 1.0216718912124634, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 20490 + }, + { + "epoch": 0.20491, + "grad_norm": 0.9042589664459229, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 20491 + }, + { + "epoch": 0.20492, + "grad_norm": 0.8357610702514648, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 20492 + }, + { + "epoch": 0.20493, + "grad_norm": 0.8341541290283203, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 20493 + }, + { + "epoch": 0.20494, + "grad_norm": 0.9072319865226746, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 20494 + }, + { + "epoch": 0.20495, + "grad_norm": 0.8294616341590881, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 20495 + }, + { + "epoch": 0.20496, + "grad_norm": 0.8708383440971375, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 20496 + }, + { + "epoch": 0.20497, + "grad_norm": 0.918099582195282, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 20497 + }, + { + "epoch": 0.20498, + "grad_norm": 0.979347288608551, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 20498 + }, + { + "epoch": 0.20499, + "grad_norm": 0.9320781230926514, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 20499 + }, + { + "epoch": 0.205, + "grad_norm": 0.7862676978111267, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 20500 + }, + { + "epoch": 0.20501, + "grad_norm": 0.8444691300392151, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 20501 + }, + { + "epoch": 0.20502, + "grad_norm": 1.0602794885635376, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 20502 + }, + { + "epoch": 0.20503, + "grad_norm": 0.9569676518440247, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 20503 + }, + { + "epoch": 0.20504, + "grad_norm": 1.1059049367904663, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 20504 + }, + { + "epoch": 0.20505, + "grad_norm": 1.0884469747543335, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 20505 + }, + { + "epoch": 0.20506, + "grad_norm": 0.8607928156852722, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 20506 + }, + { + "epoch": 0.20507, + "grad_norm": 0.8740077614784241, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 20507 + }, + { + "epoch": 0.20508, + "grad_norm": 0.8393558859825134, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 20508 + }, + { + "epoch": 0.20509, + "grad_norm": 0.8504719734191895, + "learning_rate": 0.003, + "loss": 4.026, + "step": 20509 + }, + { + "epoch": 0.2051, + "grad_norm": 0.8674973249435425, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 20510 + }, + { + "epoch": 0.20511, + "grad_norm": 0.8348132967948914, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 20511 + }, + { + "epoch": 0.20512, + "grad_norm": 0.7985794544219971, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 20512 + }, + { + "epoch": 0.20513, + "grad_norm": 0.771939218044281, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 20513 + }, + { + "epoch": 0.20514, + "grad_norm": 0.7467824220657349, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 20514 + }, + { + "epoch": 0.20515, + "grad_norm": 0.7036437392234802, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 20515 + }, + { + "epoch": 0.20516, + "grad_norm": 0.688878059387207, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 20516 + }, + { + "epoch": 0.20517, + "grad_norm": 0.6202580332756042, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 20517 + }, + { + "epoch": 0.20518, + "grad_norm": 0.6291422843933105, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 20518 + }, + { + "epoch": 0.20519, + "grad_norm": 0.6769140362739563, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 20519 + }, + { + "epoch": 0.2052, + "grad_norm": 0.7055646777153015, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 20520 + }, + { + "epoch": 0.20521, + "grad_norm": 0.7506475448608398, + "learning_rate": 0.003, + "loss": 3.969, + "step": 20521 + }, + { + "epoch": 0.20522, + "grad_norm": 0.7492371797561646, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 20522 + }, + { + "epoch": 0.20523, + "grad_norm": 0.7688263654708862, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 20523 + }, + { + "epoch": 0.20524, + "grad_norm": 0.7825839519500732, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 20524 + }, + { + "epoch": 0.20525, + "grad_norm": 0.9201285243034363, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 20525 + }, + { + "epoch": 0.20526, + "grad_norm": 1.0699529647827148, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 20526 + }, + { + "epoch": 0.20527, + "grad_norm": 0.8841052651405334, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 20527 + }, + { + "epoch": 0.20528, + "grad_norm": 0.7189367413520813, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 20528 + }, + { + "epoch": 0.20529, + "grad_norm": 0.689725935459137, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 20529 + }, + { + "epoch": 0.2053, + "grad_norm": 0.7628370523452759, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 20530 + }, + { + "epoch": 0.20531, + "grad_norm": 0.8612552285194397, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 20531 + }, + { + "epoch": 0.20532, + "grad_norm": 1.033647894859314, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 20532 + }, + { + "epoch": 0.20533, + "grad_norm": 1.1336464881896973, + "learning_rate": 0.003, + "loss": 4.004, + "step": 20533 + }, + { + "epoch": 0.20534, + "grad_norm": 0.7695611715316772, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 20534 + }, + { + "epoch": 0.20535, + "grad_norm": 0.6660147309303284, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 20535 + }, + { + "epoch": 0.20536, + "grad_norm": 0.6618660688400269, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 20536 + }, + { + "epoch": 0.20537, + "grad_norm": 0.5880175828933716, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 20537 + }, + { + "epoch": 0.20538, + "grad_norm": 0.7051225900650024, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 20538 + }, + { + "epoch": 0.20539, + "grad_norm": 0.7596197724342346, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 20539 + }, + { + "epoch": 0.2054, + "grad_norm": 0.8910565972328186, + "learning_rate": 0.003, + "loss": 4.007, + "step": 20540 + }, + { + "epoch": 0.20541, + "grad_norm": 1.1327186822891235, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 20541 + }, + { + "epoch": 0.20542, + "grad_norm": 1.0431007146835327, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 20542 + }, + { + "epoch": 0.20543, + "grad_norm": 1.0606658458709717, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 20543 + }, + { + "epoch": 0.20544, + "grad_norm": 1.0663015842437744, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 20544 + }, + { + "epoch": 0.20545, + "grad_norm": 0.9315351247787476, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 20545 + }, + { + "epoch": 0.20546, + "grad_norm": 0.8413115739822388, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 20546 + }, + { + "epoch": 0.20547, + "grad_norm": 0.8238143920898438, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 20547 + }, + { + "epoch": 0.20548, + "grad_norm": 0.9945659041404724, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 20548 + }, + { + "epoch": 0.20549, + "grad_norm": 1.103825569152832, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 20549 + }, + { + "epoch": 0.2055, + "grad_norm": 0.9757719039916992, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 20550 + }, + { + "epoch": 0.20551, + "grad_norm": 1.0624744892120361, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 20551 + }, + { + "epoch": 0.20552, + "grad_norm": 0.9679355621337891, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 20552 + }, + { + "epoch": 0.20553, + "grad_norm": 1.1138170957565308, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 20553 + }, + { + "epoch": 0.20554, + "grad_norm": 1.2719899415969849, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 20554 + }, + { + "epoch": 0.20555, + "grad_norm": 0.926088809967041, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 20555 + }, + { + "epoch": 0.20556, + "grad_norm": 0.9676206111907959, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 20556 + }, + { + "epoch": 0.20557, + "grad_norm": 0.8924266695976257, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 20557 + }, + { + "epoch": 0.20558, + "grad_norm": 0.6859793066978455, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 20558 + }, + { + "epoch": 0.20559, + "grad_norm": 0.7555159330368042, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 20559 + }, + { + "epoch": 0.2056, + "grad_norm": 0.7331491112709045, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 20560 + }, + { + "epoch": 0.20561, + "grad_norm": 0.7226747870445251, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 20561 + }, + { + "epoch": 0.20562, + "grad_norm": 0.8749589323997498, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 20562 + }, + { + "epoch": 0.20563, + "grad_norm": 1.0173287391662598, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 20563 + }, + { + "epoch": 0.20564, + "grad_norm": 0.9920432567596436, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 20564 + }, + { + "epoch": 0.20565, + "grad_norm": 0.8987650871276855, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 20565 + }, + { + "epoch": 0.20566, + "grad_norm": 0.9268227219581604, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 20566 + }, + { + "epoch": 0.20567, + "grad_norm": 0.8564572930335999, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 20567 + }, + { + "epoch": 0.20568, + "grad_norm": 0.6740842461585999, + "learning_rate": 0.003, + "loss": 3.971, + "step": 20568 + }, + { + "epoch": 0.20569, + "grad_norm": 0.5516361594200134, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 20569 + }, + { + "epoch": 0.2057, + "grad_norm": 0.5278864502906799, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 20570 + }, + { + "epoch": 0.20571, + "grad_norm": 0.5913801193237305, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 20571 + }, + { + "epoch": 0.20572, + "grad_norm": 0.6298001408576965, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 20572 + }, + { + "epoch": 0.20573, + "grad_norm": 0.756517767906189, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 20573 + }, + { + "epoch": 0.20574, + "grad_norm": 0.9460106492042542, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 20574 + }, + { + "epoch": 0.20575, + "grad_norm": 1.0529441833496094, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 20575 + }, + { + "epoch": 0.20576, + "grad_norm": 0.7504254579544067, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 20576 + }, + { + "epoch": 0.20577, + "grad_norm": 0.7312063574790955, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 20577 + }, + { + "epoch": 0.20578, + "grad_norm": 0.8165370225906372, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 20578 + }, + { + "epoch": 0.20579, + "grad_norm": 0.8489643931388855, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 20579 + }, + { + "epoch": 0.2058, + "grad_norm": 0.749298095703125, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 20580 + }, + { + "epoch": 0.20581, + "grad_norm": 0.8343328833580017, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 20581 + }, + { + "epoch": 0.20582, + "grad_norm": 0.9363728165626526, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 20582 + }, + { + "epoch": 0.20583, + "grad_norm": 0.9963992238044739, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 20583 + }, + { + "epoch": 0.20584, + "grad_norm": 0.9953159689903259, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 20584 + }, + { + "epoch": 0.20585, + "grad_norm": 1.0232398509979248, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 20585 + }, + { + "epoch": 0.20586, + "grad_norm": 0.9123949408531189, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 20586 + }, + { + "epoch": 0.20587, + "grad_norm": 0.9217385649681091, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 20587 + }, + { + "epoch": 0.20588, + "grad_norm": 1.0196894407272339, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 20588 + }, + { + "epoch": 0.20589, + "grad_norm": 1.2130417823791504, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 20589 + }, + { + "epoch": 0.2059, + "grad_norm": 0.8121554255485535, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 20590 + }, + { + "epoch": 0.20591, + "grad_norm": 0.7075498104095459, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 20591 + }, + { + "epoch": 0.20592, + "grad_norm": 0.6768858432769775, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 20592 + }, + { + "epoch": 0.20593, + "grad_norm": 0.8094514608383179, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 20593 + }, + { + "epoch": 0.20594, + "grad_norm": 1.0636597871780396, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 20594 + }, + { + "epoch": 0.20595, + "grad_norm": 1.216055154800415, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 20595 + }, + { + "epoch": 0.20596, + "grad_norm": 0.8186025619506836, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 20596 + }, + { + "epoch": 0.20597, + "grad_norm": 0.7135723233222961, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 20597 + }, + { + "epoch": 0.20598, + "grad_norm": 0.7067899107933044, + "learning_rate": 0.003, + "loss": 4.009, + "step": 20598 + }, + { + "epoch": 0.20599, + "grad_norm": 0.6274234056472778, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 20599 + }, + { + "epoch": 0.206, + "grad_norm": 0.7237199544906616, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 20600 + }, + { + "epoch": 0.20601, + "grad_norm": 0.7106822729110718, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 20601 + }, + { + "epoch": 0.20602, + "grad_norm": 0.8345945477485657, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 20602 + }, + { + "epoch": 0.20603, + "grad_norm": 1.030349612236023, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 20603 + }, + { + "epoch": 0.20604, + "grad_norm": 1.2641339302062988, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 20604 + }, + { + "epoch": 0.20605, + "grad_norm": 0.5765160918235779, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 20605 + }, + { + "epoch": 0.20606, + "grad_norm": 0.7353900074958801, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 20606 + }, + { + "epoch": 0.20607, + "grad_norm": 0.771910548210144, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 20607 + }, + { + "epoch": 0.20608, + "grad_norm": 0.9372097253799438, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 20608 + }, + { + "epoch": 0.20609, + "grad_norm": 1.223936676979065, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 20609 + }, + { + "epoch": 0.2061, + "grad_norm": 0.6746963858604431, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 20610 + }, + { + "epoch": 0.20611, + "grad_norm": 0.6881647109985352, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 20611 + }, + { + "epoch": 0.20612, + "grad_norm": 0.9571933150291443, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 20612 + }, + { + "epoch": 0.20613, + "grad_norm": 1.0425173044204712, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 20613 + }, + { + "epoch": 0.20614, + "grad_norm": 1.009878158569336, + "learning_rate": 0.003, + "loss": 4.024, + "step": 20614 + }, + { + "epoch": 0.20615, + "grad_norm": 0.8970416784286499, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 20615 + }, + { + "epoch": 0.20616, + "grad_norm": 0.800406277179718, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 20616 + }, + { + "epoch": 0.20617, + "grad_norm": 0.9475623965263367, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 20617 + }, + { + "epoch": 0.20618, + "grad_norm": 1.1195120811462402, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 20618 + }, + { + "epoch": 0.20619, + "grad_norm": 0.9776735901832581, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 20619 + }, + { + "epoch": 0.2062, + "grad_norm": 0.8268507719039917, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 20620 + }, + { + "epoch": 0.20621, + "grad_norm": 0.6747649312019348, + "learning_rate": 0.003, + "loss": 4.012, + "step": 20621 + }, + { + "epoch": 0.20622, + "grad_norm": 0.6576265096664429, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 20622 + }, + { + "epoch": 0.20623, + "grad_norm": 0.72371506690979, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 20623 + }, + { + "epoch": 0.20624, + "grad_norm": 0.7435429096221924, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 20624 + }, + { + "epoch": 0.20625, + "grad_norm": 0.6706408262252808, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 20625 + }, + { + "epoch": 0.20626, + "grad_norm": 0.6452346444129944, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 20626 + }, + { + "epoch": 0.20627, + "grad_norm": 0.6868218183517456, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 20627 + }, + { + "epoch": 0.20628, + "grad_norm": 0.6958888173103333, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 20628 + }, + { + "epoch": 0.20629, + "grad_norm": 0.9264516830444336, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 20629 + }, + { + "epoch": 0.2063, + "grad_norm": 1.2161481380462646, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 20630 + }, + { + "epoch": 0.20631, + "grad_norm": 0.6730556488037109, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 20631 + }, + { + "epoch": 0.20632, + "grad_norm": 0.6242616176605225, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 20632 + }, + { + "epoch": 0.20633, + "grad_norm": 0.7033523321151733, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 20633 + }, + { + "epoch": 0.20634, + "grad_norm": 0.7647010684013367, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 20634 + }, + { + "epoch": 0.20635, + "grad_norm": 0.9941380023956299, + "learning_rate": 0.003, + "loss": 3.98, + "step": 20635 + }, + { + "epoch": 0.20636, + "grad_norm": 1.1974639892578125, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 20636 + }, + { + "epoch": 0.20637, + "grad_norm": 0.6021768450737, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 20637 + }, + { + "epoch": 0.20638, + "grad_norm": 0.7827711701393127, + "learning_rate": 0.003, + "loss": 3.981, + "step": 20638 + }, + { + "epoch": 0.20639, + "grad_norm": 0.9160796403884888, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 20639 + }, + { + "epoch": 0.2064, + "grad_norm": 1.1267950534820557, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 20640 + }, + { + "epoch": 0.20641, + "grad_norm": 0.9441350698471069, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 20641 + }, + { + "epoch": 0.20642, + "grad_norm": 0.6999184489250183, + "learning_rate": 0.003, + "loss": 3.999, + "step": 20642 + }, + { + "epoch": 0.20643, + "grad_norm": 0.6596429347991943, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 20643 + }, + { + "epoch": 0.20644, + "grad_norm": 0.6299886703491211, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 20644 + }, + { + "epoch": 0.20645, + "grad_norm": 0.6572009921073914, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 20645 + }, + { + "epoch": 0.20646, + "grad_norm": 0.7711091637611389, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 20646 + }, + { + "epoch": 0.20647, + "grad_norm": 0.8173057436943054, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 20647 + }, + { + "epoch": 0.20648, + "grad_norm": 1.027649998664856, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 20648 + }, + { + "epoch": 0.20649, + "grad_norm": 1.249096155166626, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 20649 + }, + { + "epoch": 0.2065, + "grad_norm": 0.6178358793258667, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 20650 + }, + { + "epoch": 0.20651, + "grad_norm": 0.6693252921104431, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 20651 + }, + { + "epoch": 0.20652, + "grad_norm": 0.7295918464660645, + "learning_rate": 0.003, + "loss": 3.9404, + "step": 20652 + }, + { + "epoch": 0.20653, + "grad_norm": 0.789218544960022, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 20653 + }, + { + "epoch": 0.20654, + "grad_norm": 0.8371591567993164, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 20654 + }, + { + "epoch": 0.20655, + "grad_norm": 0.7333207726478577, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 20655 + }, + { + "epoch": 0.20656, + "grad_norm": 0.6638398170471191, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 20656 + }, + { + "epoch": 0.20657, + "grad_norm": 0.7411385178565979, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 20657 + }, + { + "epoch": 0.20658, + "grad_norm": 0.8418099284172058, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 20658 + }, + { + "epoch": 0.20659, + "grad_norm": 0.8741323351860046, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 20659 + }, + { + "epoch": 0.2066, + "grad_norm": 0.7878667712211609, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 20660 + }, + { + "epoch": 0.20661, + "grad_norm": 0.6896508932113647, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 20661 + }, + { + "epoch": 0.20662, + "grad_norm": 0.7355780601501465, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 20662 + }, + { + "epoch": 0.20663, + "grad_norm": 0.8107819557189941, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 20663 + }, + { + "epoch": 0.20664, + "grad_norm": 1.0781714916229248, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 20664 + }, + { + "epoch": 0.20665, + "grad_norm": 1.2095935344696045, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 20665 + }, + { + "epoch": 0.20666, + "grad_norm": 0.5785348415374756, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 20666 + }, + { + "epoch": 0.20667, + "grad_norm": 0.6400589942932129, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 20667 + }, + { + "epoch": 0.20668, + "grad_norm": 0.7440767288208008, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 20668 + }, + { + "epoch": 0.20669, + "grad_norm": 0.782042384147644, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 20669 + }, + { + "epoch": 0.2067, + "grad_norm": 0.8038446307182312, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 20670 + }, + { + "epoch": 0.20671, + "grad_norm": 0.9218879342079163, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 20671 + }, + { + "epoch": 0.20672, + "grad_norm": 1.0722054243087769, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 20672 + }, + { + "epoch": 0.20673, + "grad_norm": 1.2632372379302979, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 20673 + }, + { + "epoch": 0.20674, + "grad_norm": 0.712864100933075, + "learning_rate": 0.003, + "loss": 3.975, + "step": 20674 + }, + { + "epoch": 0.20675, + "grad_norm": 0.7353287935256958, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 20675 + }, + { + "epoch": 0.20676, + "grad_norm": 0.7543138861656189, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 20676 + }, + { + "epoch": 0.20677, + "grad_norm": 0.9637554287910461, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 20677 + }, + { + "epoch": 0.20678, + "grad_norm": 0.9501532912254333, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 20678 + }, + { + "epoch": 0.20679, + "grad_norm": 1.0105867385864258, + "learning_rate": 0.003, + "loss": 4.041, + "step": 20679 + }, + { + "epoch": 0.2068, + "grad_norm": 1.3445831537246704, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 20680 + }, + { + "epoch": 0.20681, + "grad_norm": 0.7950366139411926, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 20681 + }, + { + "epoch": 0.20682, + "grad_norm": 0.7824738025665283, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 20682 + }, + { + "epoch": 0.20683, + "grad_norm": 0.757445752620697, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 20683 + }, + { + "epoch": 0.20684, + "grad_norm": 0.7416343688964844, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 20684 + }, + { + "epoch": 0.20685, + "grad_norm": 0.8872725367546082, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 20685 + }, + { + "epoch": 0.20686, + "grad_norm": 1.1165724992752075, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 20686 + }, + { + "epoch": 0.20687, + "grad_norm": 0.9379168152809143, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 20687 + }, + { + "epoch": 0.20688, + "grad_norm": 1.0233606100082397, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 20688 + }, + { + "epoch": 0.20689, + "grad_norm": 0.9732743501663208, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 20689 + }, + { + "epoch": 0.2069, + "grad_norm": 0.8878822326660156, + "learning_rate": 0.003, + "loss": 4.009, + "step": 20690 + }, + { + "epoch": 0.20691, + "grad_norm": 0.9220402836799622, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 20691 + }, + { + "epoch": 0.20692, + "grad_norm": 0.9080644845962524, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 20692 + }, + { + "epoch": 0.20693, + "grad_norm": 0.8790196180343628, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 20693 + }, + { + "epoch": 0.20694, + "grad_norm": 0.7577972412109375, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 20694 + }, + { + "epoch": 0.20695, + "grad_norm": 0.658717930316925, + "learning_rate": 0.003, + "loss": 3.988, + "step": 20695 + }, + { + "epoch": 0.20696, + "grad_norm": 0.6440026760101318, + "learning_rate": 0.003, + "loss": 3.999, + "step": 20696 + }, + { + "epoch": 0.20697, + "grad_norm": 0.7962105870246887, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 20697 + }, + { + "epoch": 0.20698, + "grad_norm": 0.8110194802284241, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 20698 + }, + { + "epoch": 0.20699, + "grad_norm": 0.8429470658302307, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 20699 + }, + { + "epoch": 0.207, + "grad_norm": 0.925254762172699, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 20700 + }, + { + "epoch": 0.20701, + "grad_norm": 0.9512263536453247, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 20701 + }, + { + "epoch": 0.20702, + "grad_norm": 0.9852175116539001, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 20702 + }, + { + "epoch": 0.20703, + "grad_norm": 1.0767631530761719, + "learning_rate": 0.003, + "loss": 3.991, + "step": 20703 + }, + { + "epoch": 0.20704, + "grad_norm": 1.0503569841384888, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 20704 + }, + { + "epoch": 0.20705, + "grad_norm": 0.9670467972755432, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 20705 + }, + { + "epoch": 0.20706, + "grad_norm": 0.7200356721878052, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 20706 + }, + { + "epoch": 0.20707, + "grad_norm": 0.6849318742752075, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 20707 + }, + { + "epoch": 0.20708, + "grad_norm": 0.7746059894561768, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 20708 + }, + { + "epoch": 0.20709, + "grad_norm": 0.9228622913360596, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 20709 + }, + { + "epoch": 0.2071, + "grad_norm": 1.082072377204895, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 20710 + }, + { + "epoch": 0.20711, + "grad_norm": 0.9330927133560181, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 20711 + }, + { + "epoch": 0.20712, + "grad_norm": 0.7052283883094788, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 20712 + }, + { + "epoch": 0.20713, + "grad_norm": 0.6981417536735535, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 20713 + }, + { + "epoch": 0.20714, + "grad_norm": 0.8098515272140503, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 20714 + }, + { + "epoch": 0.20715, + "grad_norm": 0.654692530632019, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 20715 + }, + { + "epoch": 0.20716, + "grad_norm": 0.6670729517936707, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 20716 + }, + { + "epoch": 0.20717, + "grad_norm": 0.6163148880004883, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 20717 + }, + { + "epoch": 0.20718, + "grad_norm": 0.6592658162117004, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 20718 + }, + { + "epoch": 0.20719, + "grad_norm": 0.6538748741149902, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 20719 + }, + { + "epoch": 0.2072, + "grad_norm": 0.7887983918190002, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 20720 + }, + { + "epoch": 0.20721, + "grad_norm": 0.9480772614479065, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 20721 + }, + { + "epoch": 0.20722, + "grad_norm": 1.082365870475769, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 20722 + }, + { + "epoch": 0.20723, + "grad_norm": 1.0134780406951904, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 20723 + }, + { + "epoch": 0.20724, + "grad_norm": 0.8613335490226746, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 20724 + }, + { + "epoch": 0.20725, + "grad_norm": 0.7501160502433777, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 20725 + }, + { + "epoch": 0.20726, + "grad_norm": 0.7889297604560852, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 20726 + }, + { + "epoch": 0.20727, + "grad_norm": 0.8222788572311401, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 20727 + }, + { + "epoch": 0.20728, + "grad_norm": 0.8500961661338806, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 20728 + }, + { + "epoch": 0.20729, + "grad_norm": 1.0210720300674438, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 20729 + }, + { + "epoch": 0.2073, + "grad_norm": 1.1979697942733765, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 20730 + }, + { + "epoch": 0.20731, + "grad_norm": 0.8383584022521973, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 20731 + }, + { + "epoch": 0.20732, + "grad_norm": 0.7667802572250366, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 20732 + }, + { + "epoch": 0.20733, + "grad_norm": 0.6863619089126587, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 20733 + }, + { + "epoch": 0.20734, + "grad_norm": 0.7429079413414001, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 20734 + }, + { + "epoch": 0.20735, + "grad_norm": 0.7810400724411011, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 20735 + }, + { + "epoch": 0.20736, + "grad_norm": 0.7995291352272034, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 20736 + }, + { + "epoch": 0.20737, + "grad_norm": 1.1064059734344482, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 20737 + }, + { + "epoch": 0.20738, + "grad_norm": 1.1041861772537231, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 20738 + }, + { + "epoch": 0.20739, + "grad_norm": 0.9637973308563232, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 20739 + }, + { + "epoch": 0.2074, + "grad_norm": 0.9645422697067261, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 20740 + }, + { + "epoch": 0.20741, + "grad_norm": 1.0687763690948486, + "learning_rate": 0.003, + "loss": 4.01, + "step": 20741 + }, + { + "epoch": 0.20742, + "grad_norm": 0.9210143089294434, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 20742 + }, + { + "epoch": 0.20743, + "grad_norm": 0.753804087638855, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 20743 + }, + { + "epoch": 0.20744, + "grad_norm": 0.8265671133995056, + "learning_rate": 0.003, + "loss": 4.032, + "step": 20744 + }, + { + "epoch": 0.20745, + "grad_norm": 0.8992443680763245, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 20745 + }, + { + "epoch": 0.20746, + "grad_norm": 0.948488175868988, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 20746 + }, + { + "epoch": 0.20747, + "grad_norm": 0.8744862675666809, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 20747 + }, + { + "epoch": 0.20748, + "grad_norm": 0.8626583218574524, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 20748 + }, + { + "epoch": 0.20749, + "grad_norm": 1.053406000137329, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 20749 + }, + { + "epoch": 0.2075, + "grad_norm": 0.8808895349502563, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 20750 + }, + { + "epoch": 0.20751, + "grad_norm": 0.7860477566719055, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 20751 + }, + { + "epoch": 0.20752, + "grad_norm": 0.887262225151062, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 20752 + }, + { + "epoch": 0.20753, + "grad_norm": 0.9119638204574585, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 20753 + }, + { + "epoch": 0.20754, + "grad_norm": 0.8986718654632568, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 20754 + }, + { + "epoch": 0.20755, + "grad_norm": 0.8674590587615967, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 20755 + }, + { + "epoch": 0.20756, + "grad_norm": 1.0189762115478516, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 20756 + }, + { + "epoch": 0.20757, + "grad_norm": 1.087554693222046, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 20757 + }, + { + "epoch": 0.20758, + "grad_norm": 0.9033676981925964, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 20758 + }, + { + "epoch": 0.20759, + "grad_norm": 0.8090853095054626, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 20759 + }, + { + "epoch": 0.2076, + "grad_norm": 0.7151592969894409, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 20760 + }, + { + "epoch": 0.20761, + "grad_norm": 0.6113814115524292, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 20761 + }, + { + "epoch": 0.20762, + "grad_norm": 0.5708845257759094, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 20762 + }, + { + "epoch": 0.20763, + "grad_norm": 0.6272398829460144, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 20763 + }, + { + "epoch": 0.20764, + "grad_norm": 0.6918250918388367, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 20764 + }, + { + "epoch": 0.20765, + "grad_norm": 0.6999816298484802, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 20765 + }, + { + "epoch": 0.20766, + "grad_norm": 0.6208664774894714, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 20766 + }, + { + "epoch": 0.20767, + "grad_norm": 0.7076413035392761, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 20767 + }, + { + "epoch": 0.20768, + "grad_norm": 0.8253567814826965, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 20768 + }, + { + "epoch": 0.20769, + "grad_norm": 0.9532641768455505, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 20769 + }, + { + "epoch": 0.2077, + "grad_norm": 1.3245623111724854, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 20770 + }, + { + "epoch": 0.20771, + "grad_norm": 0.6137048006057739, + "learning_rate": 0.003, + "loss": 3.9498, + "step": 20771 + }, + { + "epoch": 0.20772, + "grad_norm": 0.5572570562362671, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 20772 + }, + { + "epoch": 0.20773, + "grad_norm": 0.5751827955245972, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 20773 + }, + { + "epoch": 0.20774, + "grad_norm": 0.624350905418396, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 20774 + }, + { + "epoch": 0.20775, + "grad_norm": 0.6695300340652466, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 20775 + }, + { + "epoch": 0.20776, + "grad_norm": 0.7807313203811646, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 20776 + }, + { + "epoch": 0.20777, + "grad_norm": 0.8226802349090576, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 20777 + }, + { + "epoch": 0.20778, + "grad_norm": 0.7035885453224182, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 20778 + }, + { + "epoch": 0.20779, + "grad_norm": 0.5964787602424622, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 20779 + }, + { + "epoch": 0.2078, + "grad_norm": 0.6474304795265198, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 20780 + }, + { + "epoch": 0.20781, + "grad_norm": 0.7350100874900818, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 20781 + }, + { + "epoch": 0.20782, + "grad_norm": 0.8760716319084167, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 20782 + }, + { + "epoch": 0.20783, + "grad_norm": 1.0762218236923218, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 20783 + }, + { + "epoch": 0.20784, + "grad_norm": 1.0604761838912964, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 20784 + }, + { + "epoch": 0.20785, + "grad_norm": 1.0759546756744385, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 20785 + }, + { + "epoch": 0.20786, + "grad_norm": 1.051599144935608, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 20786 + }, + { + "epoch": 0.20787, + "grad_norm": 1.355031967163086, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 20787 + }, + { + "epoch": 0.20788, + "grad_norm": 0.9172400236129761, + "learning_rate": 0.003, + "loss": 3.976, + "step": 20788 + }, + { + "epoch": 0.20789, + "grad_norm": 0.9965242147445679, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 20789 + }, + { + "epoch": 0.2079, + "grad_norm": 1.13826584815979, + "learning_rate": 0.003, + "loss": 4.025, + "step": 20790 + }, + { + "epoch": 0.20791, + "grad_norm": 0.9952272772789001, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 20791 + }, + { + "epoch": 0.20792, + "grad_norm": 0.8852728009223938, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 20792 + }, + { + "epoch": 0.20793, + "grad_norm": 0.950574517250061, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 20793 + }, + { + "epoch": 0.20794, + "grad_norm": 1.0464208126068115, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 20794 + }, + { + "epoch": 0.20795, + "grad_norm": 0.9873263239860535, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 20795 + }, + { + "epoch": 0.20796, + "grad_norm": 1.309618592262268, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 20796 + }, + { + "epoch": 0.20797, + "grad_norm": 0.74017333984375, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 20797 + }, + { + "epoch": 0.20798, + "grad_norm": 0.6632053256034851, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 20798 + }, + { + "epoch": 0.20799, + "grad_norm": 0.6369159817695618, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 20799 + }, + { + "epoch": 0.208, + "grad_norm": 0.6135809421539307, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 20800 + }, + { + "epoch": 0.20801, + "grad_norm": 0.5961670875549316, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 20801 + }, + { + "epoch": 0.20802, + "grad_norm": 0.6939438581466675, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 20802 + }, + { + "epoch": 0.20803, + "grad_norm": 0.8525596857070923, + "learning_rate": 0.003, + "loss": 3.989, + "step": 20803 + }, + { + "epoch": 0.20804, + "grad_norm": 0.898422360420227, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 20804 + }, + { + "epoch": 0.20805, + "grad_norm": 0.8727675080299377, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 20805 + }, + { + "epoch": 0.20806, + "grad_norm": 0.8255506753921509, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 20806 + }, + { + "epoch": 0.20807, + "grad_norm": 0.8468108773231506, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 20807 + }, + { + "epoch": 0.20808, + "grad_norm": 0.9557967782020569, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 20808 + }, + { + "epoch": 0.20809, + "grad_norm": 1.1428208351135254, + "learning_rate": 0.003, + "loss": 4.016, + "step": 20809 + }, + { + "epoch": 0.2081, + "grad_norm": 0.9441230893135071, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 20810 + }, + { + "epoch": 0.20811, + "grad_norm": 1.04192316532135, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 20811 + }, + { + "epoch": 0.20812, + "grad_norm": 0.8832850456237793, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 20812 + }, + { + "epoch": 0.20813, + "grad_norm": 0.7046552300453186, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 20813 + }, + { + "epoch": 0.20814, + "grad_norm": 0.7719017267227173, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 20814 + }, + { + "epoch": 0.20815, + "grad_norm": 0.8758888840675354, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 20815 + }, + { + "epoch": 0.20816, + "grad_norm": 1.175139307975769, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 20816 + }, + { + "epoch": 0.20817, + "grad_norm": 0.7121004462242126, + "learning_rate": 0.003, + "loss": 4.002, + "step": 20817 + }, + { + "epoch": 0.20818, + "grad_norm": 0.6497937440872192, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 20818 + }, + { + "epoch": 0.20819, + "grad_norm": 0.8046772480010986, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 20819 + }, + { + "epoch": 0.2082, + "grad_norm": 0.9324015378952026, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 20820 + }, + { + "epoch": 0.20821, + "grad_norm": 1.0890274047851562, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 20821 + }, + { + "epoch": 0.20822, + "grad_norm": 0.8312063813209534, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 20822 + }, + { + "epoch": 0.20823, + "grad_norm": 0.7680968046188354, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 20823 + }, + { + "epoch": 0.20824, + "grad_norm": 0.792281448841095, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 20824 + }, + { + "epoch": 0.20825, + "grad_norm": 0.9612289667129517, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 20825 + }, + { + "epoch": 0.20826, + "grad_norm": 1.093168020248413, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 20826 + }, + { + "epoch": 0.20827, + "grad_norm": 0.8961457014083862, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 20827 + }, + { + "epoch": 0.20828, + "grad_norm": 0.94273841381073, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 20828 + }, + { + "epoch": 0.20829, + "grad_norm": 1.0375275611877441, + "learning_rate": 0.003, + "loss": 4.01, + "step": 20829 + }, + { + "epoch": 0.2083, + "grad_norm": 1.0365378856658936, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 20830 + }, + { + "epoch": 0.20831, + "grad_norm": 1.000985026359558, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 20831 + }, + { + "epoch": 0.20832, + "grad_norm": 0.8756856918334961, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 20832 + }, + { + "epoch": 0.20833, + "grad_norm": 0.8284798264503479, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 20833 + }, + { + "epoch": 0.20834, + "grad_norm": 0.7230366468429565, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 20834 + }, + { + "epoch": 0.20835, + "grad_norm": 0.6954699158668518, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 20835 + }, + { + "epoch": 0.20836, + "grad_norm": 0.7048358917236328, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 20836 + }, + { + "epoch": 0.20837, + "grad_norm": 0.8838732838630676, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 20837 + }, + { + "epoch": 0.20838, + "grad_norm": 1.0553494691848755, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 20838 + }, + { + "epoch": 0.20839, + "grad_norm": 0.996741533279419, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 20839 + }, + { + "epoch": 0.2084, + "grad_norm": 0.8984896540641785, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 20840 + }, + { + "epoch": 0.20841, + "grad_norm": 0.7694340944290161, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 20841 + }, + { + "epoch": 0.20842, + "grad_norm": 0.8103636503219604, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 20842 + }, + { + "epoch": 0.20843, + "grad_norm": 0.8145402669906616, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 20843 + }, + { + "epoch": 0.20844, + "grad_norm": 0.6706199049949646, + "learning_rate": 0.003, + "loss": 3.987, + "step": 20844 + }, + { + "epoch": 0.20845, + "grad_norm": 0.6789239645004272, + "learning_rate": 0.003, + "loss": 3.998, + "step": 20845 + }, + { + "epoch": 0.20846, + "grad_norm": 0.659748375415802, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 20846 + }, + { + "epoch": 0.20847, + "grad_norm": 0.7569118738174438, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 20847 + }, + { + "epoch": 0.20848, + "grad_norm": 0.7948017120361328, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 20848 + }, + { + "epoch": 0.20849, + "grad_norm": 0.7857460975646973, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 20849 + }, + { + "epoch": 0.2085, + "grad_norm": 0.8948972821235657, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 20850 + }, + { + "epoch": 0.20851, + "grad_norm": 1.0538241863250732, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 20851 + }, + { + "epoch": 0.20852, + "grad_norm": 0.944847047328949, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 20852 + }, + { + "epoch": 0.20853, + "grad_norm": 0.7649935483932495, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 20853 + }, + { + "epoch": 0.20854, + "grad_norm": 0.7309855222702026, + "learning_rate": 0.003, + "loss": 4.015, + "step": 20854 + }, + { + "epoch": 0.20855, + "grad_norm": 0.839871883392334, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 20855 + }, + { + "epoch": 0.20856, + "grad_norm": 1.0781958103179932, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 20856 + }, + { + "epoch": 0.20857, + "grad_norm": 1.0436609983444214, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 20857 + }, + { + "epoch": 0.20858, + "grad_norm": 1.0899194478988647, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 20858 + }, + { + "epoch": 0.20859, + "grad_norm": 0.8802366256713867, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 20859 + }, + { + "epoch": 0.2086, + "grad_norm": 0.8422168493270874, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 20860 + }, + { + "epoch": 0.20861, + "grad_norm": 0.8001992106437683, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 20861 + }, + { + "epoch": 0.20862, + "grad_norm": 0.7634277939796448, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 20862 + }, + { + "epoch": 0.20863, + "grad_norm": 0.7780807018280029, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 20863 + }, + { + "epoch": 0.20864, + "grad_norm": 0.7725064754486084, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 20864 + }, + { + "epoch": 0.20865, + "grad_norm": 0.8087694644927979, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 20865 + }, + { + "epoch": 0.20866, + "grad_norm": 0.8541216850280762, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 20866 + }, + { + "epoch": 0.20867, + "grad_norm": 1.0422252416610718, + "learning_rate": 0.003, + "loss": 3.999, + "step": 20867 + }, + { + "epoch": 0.20868, + "grad_norm": 0.872579038143158, + "learning_rate": 0.003, + "loss": 3.992, + "step": 20868 + }, + { + "epoch": 0.20869, + "grad_norm": 0.9293691515922546, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 20869 + }, + { + "epoch": 0.2087, + "grad_norm": 0.9818077087402344, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 20870 + }, + { + "epoch": 0.20871, + "grad_norm": 0.9938639402389526, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 20871 + }, + { + "epoch": 0.20872, + "grad_norm": 1.0270440578460693, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 20872 + }, + { + "epoch": 0.20873, + "grad_norm": 1.033719539642334, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 20873 + }, + { + "epoch": 0.20874, + "grad_norm": 0.9151277542114258, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 20874 + }, + { + "epoch": 0.20875, + "grad_norm": 0.809447705745697, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 20875 + }, + { + "epoch": 0.20876, + "grad_norm": 0.7976981997489929, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 20876 + }, + { + "epoch": 0.20877, + "grad_norm": 0.7487211227416992, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 20877 + }, + { + "epoch": 0.20878, + "grad_norm": 0.7258038520812988, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 20878 + }, + { + "epoch": 0.20879, + "grad_norm": 0.656012237071991, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 20879 + }, + { + "epoch": 0.2088, + "grad_norm": 0.5595047473907471, + "learning_rate": 0.003, + "loss": 4.001, + "step": 20880 + }, + { + "epoch": 0.20881, + "grad_norm": 0.6370016932487488, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 20881 + }, + { + "epoch": 0.20882, + "grad_norm": 0.6755586266517639, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 20882 + }, + { + "epoch": 0.20883, + "grad_norm": 0.6820712089538574, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 20883 + }, + { + "epoch": 0.20884, + "grad_norm": 0.6668835878372192, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 20884 + }, + { + "epoch": 0.20885, + "grad_norm": 0.7880050539970398, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 20885 + }, + { + "epoch": 0.20886, + "grad_norm": 0.7801070213317871, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 20886 + }, + { + "epoch": 0.20887, + "grad_norm": 0.7368282079696655, + "learning_rate": 0.003, + "loss": 4.037, + "step": 20887 + }, + { + "epoch": 0.20888, + "grad_norm": 0.6302228569984436, + "learning_rate": 0.003, + "loss": 4.017, + "step": 20888 + }, + { + "epoch": 0.20889, + "grad_norm": 0.6666967272758484, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 20889 + }, + { + "epoch": 0.2089, + "grad_norm": 0.7467374205589294, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 20890 + }, + { + "epoch": 0.20891, + "grad_norm": 0.8886362910270691, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 20891 + }, + { + "epoch": 0.20892, + "grad_norm": 1.063424825668335, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 20892 + }, + { + "epoch": 0.20893, + "grad_norm": 1.1250430345535278, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 20893 + }, + { + "epoch": 0.20894, + "grad_norm": 0.7578840255737305, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 20894 + }, + { + "epoch": 0.20895, + "grad_norm": 0.6238217353820801, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 20895 + }, + { + "epoch": 0.20896, + "grad_norm": 0.6875156164169312, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 20896 + }, + { + "epoch": 0.20897, + "grad_norm": 0.731509268283844, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 20897 + }, + { + "epoch": 0.20898, + "grad_norm": 0.6633671522140503, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 20898 + }, + { + "epoch": 0.20899, + "grad_norm": 0.6270951628684998, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 20899 + }, + { + "epoch": 0.209, + "grad_norm": 0.7258926033973694, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 20900 + }, + { + "epoch": 0.20901, + "grad_norm": 0.6589219570159912, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 20901 + }, + { + "epoch": 0.20902, + "grad_norm": 0.7099064588546753, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 20902 + }, + { + "epoch": 0.20903, + "grad_norm": 0.9100574254989624, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 20903 + }, + { + "epoch": 0.20904, + "grad_norm": 1.1960123777389526, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 20904 + }, + { + "epoch": 0.20905, + "grad_norm": 0.8406502604484558, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 20905 + }, + { + "epoch": 0.20906, + "grad_norm": 0.8279140591621399, + "learning_rate": 0.003, + "loss": 4.008, + "step": 20906 + }, + { + "epoch": 0.20907, + "grad_norm": 0.8723680377006531, + "learning_rate": 0.003, + "loss": 4.008, + "step": 20907 + }, + { + "epoch": 0.20908, + "grad_norm": 0.9509908556938171, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 20908 + }, + { + "epoch": 0.20909, + "grad_norm": 1.0323761701583862, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 20909 + }, + { + "epoch": 0.2091, + "grad_norm": 0.9811918139457703, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 20910 + }, + { + "epoch": 0.20911, + "grad_norm": 0.8328064680099487, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 20911 + }, + { + "epoch": 0.20912, + "grad_norm": 0.8476513028144836, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 20912 + }, + { + "epoch": 0.20913, + "grad_norm": 0.8520492315292358, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 20913 + }, + { + "epoch": 0.20914, + "grad_norm": 0.9105722904205322, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 20914 + }, + { + "epoch": 0.20915, + "grad_norm": 0.8135495185852051, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 20915 + }, + { + "epoch": 0.20916, + "grad_norm": 0.8004677295684814, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 20916 + }, + { + "epoch": 0.20917, + "grad_norm": 0.8024815917015076, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 20917 + }, + { + "epoch": 0.20918, + "grad_norm": 0.8938299417495728, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 20918 + }, + { + "epoch": 0.20919, + "grad_norm": 1.1373519897460938, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 20919 + }, + { + "epoch": 0.2092, + "grad_norm": 1.09964919090271, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 20920 + }, + { + "epoch": 0.20921, + "grad_norm": 0.9594411849975586, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 20921 + }, + { + "epoch": 0.20922, + "grad_norm": 1.1219273805618286, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 20922 + }, + { + "epoch": 0.20923, + "grad_norm": 1.0501848459243774, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 20923 + }, + { + "epoch": 0.20924, + "grad_norm": 0.8713542222976685, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 20924 + }, + { + "epoch": 0.20925, + "grad_norm": 0.8268400430679321, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 20925 + }, + { + "epoch": 0.20926, + "grad_norm": 0.9429669976234436, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 20926 + }, + { + "epoch": 0.20927, + "grad_norm": 1.1184543371200562, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 20927 + }, + { + "epoch": 0.20928, + "grad_norm": 0.8756439089775085, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 20928 + }, + { + "epoch": 0.20929, + "grad_norm": 0.7752197980880737, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 20929 + }, + { + "epoch": 0.2093, + "grad_norm": 0.7061675190925598, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 20930 + }, + { + "epoch": 0.20931, + "grad_norm": 0.7310569286346436, + "learning_rate": 0.003, + "loss": 4.013, + "step": 20931 + }, + { + "epoch": 0.20932, + "grad_norm": 0.8134058713912964, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 20932 + }, + { + "epoch": 0.20933, + "grad_norm": 0.8310827016830444, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 20933 + }, + { + "epoch": 0.20934, + "grad_norm": 0.8122318387031555, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 20934 + }, + { + "epoch": 0.20935, + "grad_norm": 0.718316376209259, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 20935 + }, + { + "epoch": 0.20936, + "grad_norm": 0.6134288311004639, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 20936 + }, + { + "epoch": 0.20937, + "grad_norm": 0.564063549041748, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 20937 + }, + { + "epoch": 0.20938, + "grad_norm": 0.6001213192939758, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 20938 + }, + { + "epoch": 0.20939, + "grad_norm": 0.7452312707901001, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 20939 + }, + { + "epoch": 0.2094, + "grad_norm": 0.7726470232009888, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 20940 + }, + { + "epoch": 0.20941, + "grad_norm": 0.9299268126487732, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 20941 + }, + { + "epoch": 0.20942, + "grad_norm": 1.3489012718200684, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 20942 + }, + { + "epoch": 0.20943, + "grad_norm": 0.9157458543777466, + "learning_rate": 0.003, + "loss": 4.001, + "step": 20943 + }, + { + "epoch": 0.20944, + "grad_norm": 0.850161612033844, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 20944 + }, + { + "epoch": 0.20945, + "grad_norm": 0.9274651408195496, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 20945 + }, + { + "epoch": 0.20946, + "grad_norm": 0.9364386796951294, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 20946 + }, + { + "epoch": 0.20947, + "grad_norm": 0.9944769144058228, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 20947 + }, + { + "epoch": 0.20948, + "grad_norm": 0.9468872547149658, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 20948 + }, + { + "epoch": 0.20949, + "grad_norm": 0.8152547478675842, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 20949 + }, + { + "epoch": 0.2095, + "grad_norm": 0.8274593353271484, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 20950 + }, + { + "epoch": 0.20951, + "grad_norm": 0.7041265368461609, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 20951 + }, + { + "epoch": 0.20952, + "grad_norm": 0.6787787079811096, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 20952 + }, + { + "epoch": 0.20953, + "grad_norm": 0.7629432678222656, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 20953 + }, + { + "epoch": 0.20954, + "grad_norm": 0.751412034034729, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 20954 + }, + { + "epoch": 0.20955, + "grad_norm": 0.7168709635734558, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 20955 + }, + { + "epoch": 0.20956, + "grad_norm": 0.7899840474128723, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 20956 + }, + { + "epoch": 0.20957, + "grad_norm": 0.8560346364974976, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 20957 + }, + { + "epoch": 0.20958, + "grad_norm": 0.945913553237915, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 20958 + }, + { + "epoch": 0.20959, + "grad_norm": 1.2707791328430176, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 20959 + }, + { + "epoch": 0.2096, + "grad_norm": 0.9422053694725037, + "learning_rate": 0.003, + "loss": 3.99, + "step": 20960 + }, + { + "epoch": 0.20961, + "grad_norm": 0.9197450876235962, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 20961 + }, + { + "epoch": 0.20962, + "grad_norm": 1.1179391145706177, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 20962 + }, + { + "epoch": 0.20963, + "grad_norm": 1.0478678941726685, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 20963 + }, + { + "epoch": 0.20964, + "grad_norm": 0.9024834632873535, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 20964 + }, + { + "epoch": 0.20965, + "grad_norm": 0.8023054599761963, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 20965 + }, + { + "epoch": 0.20966, + "grad_norm": 0.8307892680168152, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 20966 + }, + { + "epoch": 0.20967, + "grad_norm": 0.9591643214225769, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 20967 + }, + { + "epoch": 0.20968, + "grad_norm": 1.2103545665740967, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 20968 + }, + { + "epoch": 0.20969, + "grad_norm": 0.9418818354606628, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 20969 + }, + { + "epoch": 0.2097, + "grad_norm": 0.6856959462165833, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 20970 + }, + { + "epoch": 0.20971, + "grad_norm": 0.5946695804595947, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 20971 + }, + { + "epoch": 0.20972, + "grad_norm": 0.7504004240036011, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 20972 + }, + { + "epoch": 0.20973, + "grad_norm": 0.9070709347724915, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 20973 + }, + { + "epoch": 0.20974, + "grad_norm": 1.0570318698883057, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 20974 + }, + { + "epoch": 0.20975, + "grad_norm": 0.8538148403167725, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 20975 + }, + { + "epoch": 0.20976, + "grad_norm": 0.6629658341407776, + "learning_rate": 0.003, + "loss": 3.979, + "step": 20976 + }, + { + "epoch": 0.20977, + "grad_norm": 0.6608455777168274, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 20977 + }, + { + "epoch": 0.20978, + "grad_norm": 0.749142050743103, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 20978 + }, + { + "epoch": 0.20979, + "grad_norm": 0.8641400337219238, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 20979 + }, + { + "epoch": 0.2098, + "grad_norm": 0.871356725692749, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 20980 + }, + { + "epoch": 0.20981, + "grad_norm": 0.7936810255050659, + "learning_rate": 0.003, + "loss": 3.97, + "step": 20981 + }, + { + "epoch": 0.20982, + "grad_norm": 0.7631054520606995, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 20982 + }, + { + "epoch": 0.20983, + "grad_norm": 0.6764727830886841, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 20983 + }, + { + "epoch": 0.20984, + "grad_norm": 0.7330240607261658, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 20984 + }, + { + "epoch": 0.20985, + "grad_norm": 0.7792537212371826, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 20985 + }, + { + "epoch": 0.20986, + "grad_norm": 0.8118487596511841, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 20986 + }, + { + "epoch": 0.20987, + "grad_norm": 0.679811418056488, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 20987 + }, + { + "epoch": 0.20988, + "grad_norm": 0.6592569947242737, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 20988 + }, + { + "epoch": 0.20989, + "grad_norm": 0.6194579005241394, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 20989 + }, + { + "epoch": 0.2099, + "grad_norm": 0.653477668762207, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 20990 + }, + { + "epoch": 0.20991, + "grad_norm": 0.8853954672813416, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 20991 + }, + { + "epoch": 0.20992, + "grad_norm": 1.1030426025390625, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 20992 + }, + { + "epoch": 0.20993, + "grad_norm": 0.9298046827316284, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 20993 + }, + { + "epoch": 0.20994, + "grad_norm": 0.9900102615356445, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 20994 + }, + { + "epoch": 0.20995, + "grad_norm": 1.241927146911621, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 20995 + }, + { + "epoch": 0.20996, + "grad_norm": 0.8189589381217957, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 20996 + }, + { + "epoch": 0.20997, + "grad_norm": 0.6784688830375671, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 20997 + }, + { + "epoch": 0.20998, + "grad_norm": 0.8254179954528809, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 20998 + }, + { + "epoch": 0.20999, + "grad_norm": 1.0192513465881348, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 20999 + }, + { + "epoch": 0.21, + "grad_norm": 1.1719720363616943, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 21000 + }, + { + "epoch": 0.21001, + "grad_norm": 0.9457380771636963, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 21001 + }, + { + "epoch": 0.21002, + "grad_norm": 1.0332539081573486, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 21002 + }, + { + "epoch": 0.21003, + "grad_norm": 0.8509064316749573, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 21003 + }, + { + "epoch": 0.21004, + "grad_norm": 0.8835716843605042, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 21004 + }, + { + "epoch": 0.21005, + "grad_norm": 0.7943909168243408, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 21005 + }, + { + "epoch": 0.21006, + "grad_norm": 0.7949368953704834, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 21006 + }, + { + "epoch": 0.21007, + "grad_norm": 0.8772362470626831, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 21007 + }, + { + "epoch": 0.21008, + "grad_norm": 0.9262873530387878, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 21008 + }, + { + "epoch": 0.21009, + "grad_norm": 1.020415186882019, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 21009 + }, + { + "epoch": 0.2101, + "grad_norm": 1.1163513660430908, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 21010 + }, + { + "epoch": 0.21011, + "grad_norm": 0.9820419549942017, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 21011 + }, + { + "epoch": 0.21012, + "grad_norm": 1.000394582748413, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 21012 + }, + { + "epoch": 0.21013, + "grad_norm": 0.9880903959274292, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 21013 + }, + { + "epoch": 0.21014, + "grad_norm": 0.9880756139755249, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 21014 + }, + { + "epoch": 0.21015, + "grad_norm": 0.8214560747146606, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 21015 + }, + { + "epoch": 0.21016, + "grad_norm": 0.8155760765075684, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 21016 + }, + { + "epoch": 0.21017, + "grad_norm": 0.7967983484268188, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 21017 + }, + { + "epoch": 0.21018, + "grad_norm": 0.7826992869377136, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 21018 + }, + { + "epoch": 0.21019, + "grad_norm": 0.7505625486373901, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 21019 + }, + { + "epoch": 0.2102, + "grad_norm": 0.7374244332313538, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 21020 + }, + { + "epoch": 0.21021, + "grad_norm": 0.8816823363304138, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 21021 + }, + { + "epoch": 0.21022, + "grad_norm": 1.1566390991210938, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 21022 + }, + { + "epoch": 0.21023, + "grad_norm": 0.8181625008583069, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 21023 + }, + { + "epoch": 0.21024, + "grad_norm": 0.7529526948928833, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 21024 + }, + { + "epoch": 0.21025, + "grad_norm": 0.8288283348083496, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 21025 + }, + { + "epoch": 0.21026, + "grad_norm": 0.8934691548347473, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 21026 + }, + { + "epoch": 0.21027, + "grad_norm": 0.892758846282959, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 21027 + }, + { + "epoch": 0.21028, + "grad_norm": 0.7953301668167114, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 21028 + }, + { + "epoch": 0.21029, + "grad_norm": 0.8886691331863403, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 21029 + }, + { + "epoch": 0.2103, + "grad_norm": 1.1598305702209473, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 21030 + }, + { + "epoch": 0.21031, + "grad_norm": 1.253692626953125, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 21031 + }, + { + "epoch": 0.21032, + "grad_norm": 0.7313674092292786, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 21032 + }, + { + "epoch": 0.21033, + "grad_norm": 0.6906070709228516, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 21033 + }, + { + "epoch": 0.21034, + "grad_norm": 0.6842135787010193, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 21034 + }, + { + "epoch": 0.21035, + "grad_norm": 0.7087405920028687, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 21035 + }, + { + "epoch": 0.21036, + "grad_norm": 0.7700204253196716, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 21036 + }, + { + "epoch": 0.21037, + "grad_norm": 0.7962941527366638, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 21037 + }, + { + "epoch": 0.21038, + "grad_norm": 0.8052358627319336, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 21038 + }, + { + "epoch": 0.21039, + "grad_norm": 0.7158791422843933, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 21039 + }, + { + "epoch": 0.2104, + "grad_norm": 0.7184638381004333, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 21040 + }, + { + "epoch": 0.21041, + "grad_norm": 0.7577104568481445, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 21041 + }, + { + "epoch": 0.21042, + "grad_norm": 0.7389146089553833, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 21042 + }, + { + "epoch": 0.21043, + "grad_norm": 0.8012667298316956, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 21043 + }, + { + "epoch": 0.21044, + "grad_norm": 0.9212856888771057, + "learning_rate": 0.003, + "loss": 3.983, + "step": 21044 + }, + { + "epoch": 0.21045, + "grad_norm": 0.9559372663497925, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 21045 + }, + { + "epoch": 0.21046, + "grad_norm": 0.8581951260566711, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 21046 + }, + { + "epoch": 0.21047, + "grad_norm": 1.0302932262420654, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 21047 + }, + { + "epoch": 0.21048, + "grad_norm": 0.9309453964233398, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 21048 + }, + { + "epoch": 0.21049, + "grad_norm": 0.7509773373603821, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 21049 + }, + { + "epoch": 0.2105, + "grad_norm": 0.7801438570022583, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 21050 + }, + { + "epoch": 0.21051, + "grad_norm": 0.787688136100769, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 21051 + }, + { + "epoch": 0.21052, + "grad_norm": 0.961395263671875, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 21052 + }, + { + "epoch": 0.21053, + "grad_norm": 1.1159189939498901, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 21053 + }, + { + "epoch": 0.21054, + "grad_norm": 0.883037269115448, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 21054 + }, + { + "epoch": 0.21055, + "grad_norm": 0.7064389586448669, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 21055 + }, + { + "epoch": 0.21056, + "grad_norm": 0.6360693573951721, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 21056 + }, + { + "epoch": 0.21057, + "grad_norm": 0.7037835717201233, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 21057 + }, + { + "epoch": 0.21058, + "grad_norm": 0.8194627165794373, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 21058 + }, + { + "epoch": 0.21059, + "grad_norm": 0.8633446097373962, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 21059 + }, + { + "epoch": 0.2106, + "grad_norm": 1.052125096321106, + "learning_rate": 0.003, + "loss": 4.011, + "step": 21060 + }, + { + "epoch": 0.21061, + "grad_norm": 1.1290466785430908, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 21061 + }, + { + "epoch": 0.21062, + "grad_norm": 0.7299408912658691, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 21062 + }, + { + "epoch": 0.21063, + "grad_norm": 0.7068364024162292, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 21063 + }, + { + "epoch": 0.21064, + "grad_norm": 0.7383988499641418, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 21064 + }, + { + "epoch": 0.21065, + "grad_norm": 0.7136686444282532, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 21065 + }, + { + "epoch": 0.21066, + "grad_norm": 0.8387410640716553, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 21066 + }, + { + "epoch": 0.21067, + "grad_norm": 0.9602887630462646, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 21067 + }, + { + "epoch": 0.21068, + "grad_norm": 0.8212478756904602, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 21068 + }, + { + "epoch": 0.21069, + "grad_norm": 0.6330386996269226, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 21069 + }, + { + "epoch": 0.2107, + "grad_norm": 0.8101123571395874, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 21070 + }, + { + "epoch": 0.21071, + "grad_norm": 1.0983551740646362, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 21071 + }, + { + "epoch": 0.21072, + "grad_norm": 0.9737641215324402, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 21072 + }, + { + "epoch": 0.21073, + "grad_norm": 0.839832067489624, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 21073 + }, + { + "epoch": 0.21074, + "grad_norm": 0.858851432800293, + "learning_rate": 0.003, + "loss": 4.0, + "step": 21074 + }, + { + "epoch": 0.21075, + "grad_norm": 0.9766066670417786, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 21075 + }, + { + "epoch": 0.21076, + "grad_norm": 0.980224072933197, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 21076 + }, + { + "epoch": 0.21077, + "grad_norm": 0.9520453810691833, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 21077 + }, + { + "epoch": 0.21078, + "grad_norm": 1.0487115383148193, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 21078 + }, + { + "epoch": 0.21079, + "grad_norm": 1.1287165880203247, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 21079 + }, + { + "epoch": 0.2108, + "grad_norm": 0.998401403427124, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 21080 + }, + { + "epoch": 0.21081, + "grad_norm": 0.9656020998954773, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 21081 + }, + { + "epoch": 0.21082, + "grad_norm": 1.0451844930648804, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 21082 + }, + { + "epoch": 0.21083, + "grad_norm": 1.0403201580047607, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 21083 + }, + { + "epoch": 0.21084, + "grad_norm": 0.8141436576843262, + "learning_rate": 0.003, + "loss": 3.972, + "step": 21084 + }, + { + "epoch": 0.21085, + "grad_norm": 0.9922561645507812, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 21085 + }, + { + "epoch": 0.21086, + "grad_norm": 1.0949798822402954, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 21086 + }, + { + "epoch": 0.21087, + "grad_norm": 0.7363989353179932, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 21087 + }, + { + "epoch": 0.21088, + "grad_norm": 0.6821371912956238, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 21088 + }, + { + "epoch": 0.21089, + "grad_norm": 0.6545329093933105, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 21089 + }, + { + "epoch": 0.2109, + "grad_norm": 0.7614679932594299, + "learning_rate": 0.003, + "loss": 4.006, + "step": 21090 + }, + { + "epoch": 0.21091, + "grad_norm": 0.8033214807510376, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 21091 + }, + { + "epoch": 0.21092, + "grad_norm": 0.8122650980949402, + "learning_rate": 0.003, + "loss": 4.021, + "step": 21092 + }, + { + "epoch": 0.21093, + "grad_norm": 0.8499550223350525, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 21093 + }, + { + "epoch": 0.21094, + "grad_norm": 0.7690388560295105, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 21094 + }, + { + "epoch": 0.21095, + "grad_norm": 0.7322508692741394, + "learning_rate": 0.003, + "loss": 4.012, + "step": 21095 + }, + { + "epoch": 0.21096, + "grad_norm": 0.7588611841201782, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 21096 + }, + { + "epoch": 0.21097, + "grad_norm": 0.7608283162117004, + "learning_rate": 0.003, + "loss": 3.994, + "step": 21097 + }, + { + "epoch": 0.21098, + "grad_norm": 0.8392338156700134, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 21098 + }, + { + "epoch": 0.21099, + "grad_norm": 0.9615392088890076, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 21099 + }, + { + "epoch": 0.211, + "grad_norm": 0.9732010364532471, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 21100 + }, + { + "epoch": 0.21101, + "grad_norm": 0.9136336445808411, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 21101 + }, + { + "epoch": 0.21102, + "grad_norm": 1.1355119943618774, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 21102 + }, + { + "epoch": 0.21103, + "grad_norm": 0.918555736541748, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 21103 + }, + { + "epoch": 0.21104, + "grad_norm": 0.9605987668037415, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 21104 + }, + { + "epoch": 0.21105, + "grad_norm": 0.9908633828163147, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 21105 + }, + { + "epoch": 0.21106, + "grad_norm": 1.0930907726287842, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 21106 + }, + { + "epoch": 0.21107, + "grad_norm": 0.8681524991989136, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 21107 + }, + { + "epoch": 0.21108, + "grad_norm": 0.7648935317993164, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 21108 + }, + { + "epoch": 0.21109, + "grad_norm": 0.7801657915115356, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 21109 + }, + { + "epoch": 0.2111, + "grad_norm": 0.6971256136894226, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 21110 + }, + { + "epoch": 0.21111, + "grad_norm": 0.7200019359588623, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 21111 + }, + { + "epoch": 0.21112, + "grad_norm": 0.6669362783432007, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 21112 + }, + { + "epoch": 0.21113, + "grad_norm": 0.6524654030799866, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 21113 + }, + { + "epoch": 0.21114, + "grad_norm": 0.6392310261726379, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 21114 + }, + { + "epoch": 0.21115, + "grad_norm": 0.8195772767066956, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 21115 + }, + { + "epoch": 0.21116, + "grad_norm": 1.2931321859359741, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 21116 + }, + { + "epoch": 0.21117, + "grad_norm": 1.125009536743164, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 21117 + }, + { + "epoch": 0.21118, + "grad_norm": 0.6511727571487427, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 21118 + }, + { + "epoch": 0.21119, + "grad_norm": 0.6132333874702454, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 21119 + }, + { + "epoch": 0.2112, + "grad_norm": 0.8177005052566528, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 21120 + }, + { + "epoch": 0.21121, + "grad_norm": 0.8262137770652771, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 21121 + }, + { + "epoch": 0.21122, + "grad_norm": 0.9091663956642151, + "learning_rate": 0.003, + "loss": 3.977, + "step": 21122 + }, + { + "epoch": 0.21123, + "grad_norm": 1.029575228691101, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 21123 + }, + { + "epoch": 0.21124, + "grad_norm": 0.9782790541648865, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 21124 + }, + { + "epoch": 0.21125, + "grad_norm": 0.9406898617744446, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 21125 + }, + { + "epoch": 0.21126, + "grad_norm": 0.7773774266242981, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 21126 + }, + { + "epoch": 0.21127, + "grad_norm": 0.7181642055511475, + "learning_rate": 0.003, + "loss": 4.012, + "step": 21127 + }, + { + "epoch": 0.21128, + "grad_norm": 0.6219280362129211, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 21128 + }, + { + "epoch": 0.21129, + "grad_norm": 0.6637054085731506, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 21129 + }, + { + "epoch": 0.2113, + "grad_norm": 0.6986822485923767, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 21130 + }, + { + "epoch": 0.21131, + "grad_norm": 0.7793785333633423, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 21131 + }, + { + "epoch": 0.21132, + "grad_norm": 0.8745966553688049, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 21132 + }, + { + "epoch": 0.21133, + "grad_norm": 1.0361571311950684, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 21133 + }, + { + "epoch": 0.21134, + "grad_norm": 0.960414707660675, + "learning_rate": 0.003, + "loss": 3.998, + "step": 21134 + }, + { + "epoch": 0.21135, + "grad_norm": 0.8092941045761108, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 21135 + }, + { + "epoch": 0.21136, + "grad_norm": 0.7796742916107178, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 21136 + }, + { + "epoch": 0.21137, + "grad_norm": 0.8545747399330139, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 21137 + }, + { + "epoch": 0.21138, + "grad_norm": 0.8861588835716248, + "learning_rate": 0.003, + "loss": 3.972, + "step": 21138 + }, + { + "epoch": 0.21139, + "grad_norm": 0.9746605753898621, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 21139 + }, + { + "epoch": 0.2114, + "grad_norm": 1.0485796928405762, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 21140 + }, + { + "epoch": 0.21141, + "grad_norm": 0.9233496785163879, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 21141 + }, + { + "epoch": 0.21142, + "grad_norm": 1.0705775022506714, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 21142 + }, + { + "epoch": 0.21143, + "grad_norm": 1.004017949104309, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 21143 + }, + { + "epoch": 0.21144, + "grad_norm": 1.067195177078247, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 21144 + }, + { + "epoch": 0.21145, + "grad_norm": 0.95079505443573, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 21145 + }, + { + "epoch": 0.21146, + "grad_norm": 0.8161368370056152, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 21146 + }, + { + "epoch": 0.21147, + "grad_norm": 0.7655653357505798, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 21147 + }, + { + "epoch": 0.21148, + "grad_norm": 0.7744003534317017, + "learning_rate": 0.003, + "loss": 3.993, + "step": 21148 + }, + { + "epoch": 0.21149, + "grad_norm": 0.8398821353912354, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 21149 + }, + { + "epoch": 0.2115, + "grad_norm": 0.8834649324417114, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 21150 + }, + { + "epoch": 0.21151, + "grad_norm": 0.8834509253501892, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 21151 + }, + { + "epoch": 0.21152, + "grad_norm": 0.8296899795532227, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 21152 + }, + { + "epoch": 0.21153, + "grad_norm": 0.8983633518218994, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 21153 + }, + { + "epoch": 0.21154, + "grad_norm": 0.9647314548492432, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 21154 + }, + { + "epoch": 0.21155, + "grad_norm": 1.0629037618637085, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 21155 + }, + { + "epoch": 0.21156, + "grad_norm": 1.043571949005127, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 21156 + }, + { + "epoch": 0.21157, + "grad_norm": 0.969160258769989, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 21157 + }, + { + "epoch": 0.21158, + "grad_norm": 0.938548743724823, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 21158 + }, + { + "epoch": 0.21159, + "grad_norm": 0.7767465710639954, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 21159 + }, + { + "epoch": 0.2116, + "grad_norm": 0.5851399898529053, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 21160 + }, + { + "epoch": 0.21161, + "grad_norm": 0.6357923746109009, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 21161 + }, + { + "epoch": 0.21162, + "grad_norm": 0.5589550733566284, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 21162 + }, + { + "epoch": 0.21163, + "grad_norm": 0.5388723611831665, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 21163 + }, + { + "epoch": 0.21164, + "grad_norm": 0.5704944729804993, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 21164 + }, + { + "epoch": 0.21165, + "grad_norm": 0.6750659346580505, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 21165 + }, + { + "epoch": 0.21166, + "grad_norm": 0.8935434818267822, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 21166 + }, + { + "epoch": 0.21167, + "grad_norm": 1.1401575803756714, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 21167 + }, + { + "epoch": 0.21168, + "grad_norm": 1.0440500974655151, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 21168 + }, + { + "epoch": 0.21169, + "grad_norm": 0.7498615980148315, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 21169 + }, + { + "epoch": 0.2117, + "grad_norm": 0.6301253437995911, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 21170 + }, + { + "epoch": 0.21171, + "grad_norm": 0.8668792843818665, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 21171 + }, + { + "epoch": 0.21172, + "grad_norm": 1.1318798065185547, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 21172 + }, + { + "epoch": 0.21173, + "grad_norm": 1.0271556377410889, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 21173 + }, + { + "epoch": 0.21174, + "grad_norm": 0.8347497582435608, + "learning_rate": 0.003, + "loss": 4.001, + "step": 21174 + }, + { + "epoch": 0.21175, + "grad_norm": 0.7902694940567017, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 21175 + }, + { + "epoch": 0.21176, + "grad_norm": 0.6845634579658508, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 21176 + }, + { + "epoch": 0.21177, + "grad_norm": 0.7081102728843689, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 21177 + }, + { + "epoch": 0.21178, + "grad_norm": 0.8384816646575928, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 21178 + }, + { + "epoch": 0.21179, + "grad_norm": 0.9143149256706238, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 21179 + }, + { + "epoch": 0.2118, + "grad_norm": 1.0327494144439697, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 21180 + }, + { + "epoch": 0.21181, + "grad_norm": 1.0821131467819214, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 21181 + }, + { + "epoch": 0.21182, + "grad_norm": 0.9439168572425842, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 21182 + }, + { + "epoch": 0.21183, + "grad_norm": 0.7780779004096985, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 21183 + }, + { + "epoch": 0.21184, + "grad_norm": 0.7189350128173828, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 21184 + }, + { + "epoch": 0.21185, + "grad_norm": 0.7029850482940674, + "learning_rate": 0.003, + "loss": 3.999, + "step": 21185 + }, + { + "epoch": 0.21186, + "grad_norm": 0.7703216075897217, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 21186 + }, + { + "epoch": 0.21187, + "grad_norm": 0.5767636895179749, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 21187 + }, + { + "epoch": 0.21188, + "grad_norm": 0.7411084771156311, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 21188 + }, + { + "epoch": 0.21189, + "grad_norm": 0.8170750141143799, + "learning_rate": 0.003, + "loss": 3.969, + "step": 21189 + }, + { + "epoch": 0.2119, + "grad_norm": 0.8514722585678101, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 21190 + }, + { + "epoch": 0.21191, + "grad_norm": 0.9461209177970886, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 21191 + }, + { + "epoch": 0.21192, + "grad_norm": 0.9616516828536987, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 21192 + }, + { + "epoch": 0.21193, + "grad_norm": 0.9413039088249207, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 21193 + }, + { + "epoch": 0.21194, + "grad_norm": 0.8806334733963013, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 21194 + }, + { + "epoch": 0.21195, + "grad_norm": 0.842782199382782, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 21195 + }, + { + "epoch": 0.21196, + "grad_norm": 0.7331745624542236, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 21196 + }, + { + "epoch": 0.21197, + "grad_norm": 0.7475547194480896, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 21197 + }, + { + "epoch": 0.21198, + "grad_norm": 0.8262837529182434, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 21198 + }, + { + "epoch": 0.21199, + "grad_norm": 0.8500202298164368, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 21199 + }, + { + "epoch": 0.212, + "grad_norm": 0.9070703983306885, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 21200 + }, + { + "epoch": 0.21201, + "grad_norm": 0.8915760517120361, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 21201 + }, + { + "epoch": 0.21202, + "grad_norm": 0.8790895342826843, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 21202 + }, + { + "epoch": 0.21203, + "grad_norm": 0.8754155039787292, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 21203 + }, + { + "epoch": 0.21204, + "grad_norm": 1.0038352012634277, + "learning_rate": 0.003, + "loss": 4.005, + "step": 21204 + }, + { + "epoch": 0.21205, + "grad_norm": 1.1724965572357178, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 21205 + }, + { + "epoch": 0.21206, + "grad_norm": 0.7596562504768372, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 21206 + }, + { + "epoch": 0.21207, + "grad_norm": 0.6416271328926086, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 21207 + }, + { + "epoch": 0.21208, + "grad_norm": 0.8359499573707581, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 21208 + }, + { + "epoch": 0.21209, + "grad_norm": 0.8451243042945862, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 21209 + }, + { + "epoch": 0.2121, + "grad_norm": 0.8369916677474976, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 21210 + }, + { + "epoch": 0.21211, + "grad_norm": 0.980510950088501, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 21211 + }, + { + "epoch": 0.21212, + "grad_norm": 1.2077192068099976, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 21212 + }, + { + "epoch": 0.21213, + "grad_norm": 0.8718260526657104, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 21213 + }, + { + "epoch": 0.21214, + "grad_norm": 0.8959618210792542, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 21214 + }, + { + "epoch": 0.21215, + "grad_norm": 0.8953097462654114, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 21215 + }, + { + "epoch": 0.21216, + "grad_norm": 0.9954930543899536, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 21216 + }, + { + "epoch": 0.21217, + "grad_norm": 0.8618950843811035, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 21217 + }, + { + "epoch": 0.21218, + "grad_norm": 0.9115675687789917, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 21218 + }, + { + "epoch": 0.21219, + "grad_norm": 1.0541709661483765, + "learning_rate": 0.003, + "loss": 4.012, + "step": 21219 + }, + { + "epoch": 0.2122, + "grad_norm": 1.0102969408035278, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 21220 + }, + { + "epoch": 0.21221, + "grad_norm": 0.9446771144866943, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 21221 + }, + { + "epoch": 0.21222, + "grad_norm": 0.8503367304801941, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 21222 + }, + { + "epoch": 0.21223, + "grad_norm": 0.8016173839569092, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 21223 + }, + { + "epoch": 0.21224, + "grad_norm": 0.7813971042633057, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 21224 + }, + { + "epoch": 0.21225, + "grad_norm": 0.7670163512229919, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 21225 + }, + { + "epoch": 0.21226, + "grad_norm": 0.7613299489021301, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 21226 + }, + { + "epoch": 0.21227, + "grad_norm": 0.8657068014144897, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 21227 + }, + { + "epoch": 0.21228, + "grad_norm": 0.8493665456771851, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 21228 + }, + { + "epoch": 0.21229, + "grad_norm": 0.7315042018890381, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 21229 + }, + { + "epoch": 0.2123, + "grad_norm": 0.8328627347946167, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 21230 + }, + { + "epoch": 0.21231, + "grad_norm": 0.7406572103500366, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 21231 + }, + { + "epoch": 0.21232, + "grad_norm": 0.645925760269165, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 21232 + }, + { + "epoch": 0.21233, + "grad_norm": 0.7460606694221497, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 21233 + }, + { + "epoch": 0.21234, + "grad_norm": 0.7875913977622986, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 21234 + }, + { + "epoch": 0.21235, + "grad_norm": 0.7267428636550903, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 21235 + }, + { + "epoch": 0.21236, + "grad_norm": 0.7423426508903503, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 21236 + }, + { + "epoch": 0.21237, + "grad_norm": 1.041614055633545, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 21237 + }, + { + "epoch": 0.21238, + "grad_norm": 0.9236260652542114, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 21238 + }, + { + "epoch": 0.21239, + "grad_norm": 0.9289224743843079, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 21239 + }, + { + "epoch": 0.2124, + "grad_norm": 1.0022330284118652, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 21240 + }, + { + "epoch": 0.21241, + "grad_norm": 1.310619592666626, + "learning_rate": 0.003, + "loss": 4.027, + "step": 21241 + }, + { + "epoch": 0.21242, + "grad_norm": 0.7588840126991272, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 21242 + }, + { + "epoch": 0.21243, + "grad_norm": 0.708224356174469, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 21243 + }, + { + "epoch": 0.21244, + "grad_norm": 0.6868268251419067, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 21244 + }, + { + "epoch": 0.21245, + "grad_norm": 0.6447989344596863, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 21245 + }, + { + "epoch": 0.21246, + "grad_norm": 0.6886347532272339, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 21246 + }, + { + "epoch": 0.21247, + "grad_norm": 1.069603443145752, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 21247 + }, + { + "epoch": 0.21248, + "grad_norm": 1.2370485067367554, + "learning_rate": 0.003, + "loss": 3.995, + "step": 21248 + }, + { + "epoch": 0.21249, + "grad_norm": 0.889805018901825, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 21249 + }, + { + "epoch": 0.2125, + "grad_norm": 0.8083428144454956, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 21250 + }, + { + "epoch": 0.21251, + "grad_norm": 0.8232893347740173, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 21251 + }, + { + "epoch": 0.21252, + "grad_norm": 0.7635586261749268, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 21252 + }, + { + "epoch": 0.21253, + "grad_norm": 0.635540783405304, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 21253 + }, + { + "epoch": 0.21254, + "grad_norm": 0.5873146057128906, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 21254 + }, + { + "epoch": 0.21255, + "grad_norm": 0.6101182699203491, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 21255 + }, + { + "epoch": 0.21256, + "grad_norm": 0.64992755651474, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 21256 + }, + { + "epoch": 0.21257, + "grad_norm": 0.738416314125061, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 21257 + }, + { + "epoch": 0.21258, + "grad_norm": 0.7467118501663208, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 21258 + }, + { + "epoch": 0.21259, + "grad_norm": 0.7452531456947327, + "learning_rate": 0.003, + "loss": 3.989, + "step": 21259 + }, + { + "epoch": 0.2126, + "grad_norm": 0.8863193988800049, + "learning_rate": 0.003, + "loss": 4.001, + "step": 21260 + }, + { + "epoch": 0.21261, + "grad_norm": 1.1656006574630737, + "learning_rate": 0.003, + "loss": 4.005, + "step": 21261 + }, + { + "epoch": 0.21262, + "grad_norm": 0.8003297448158264, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 21262 + }, + { + "epoch": 0.21263, + "grad_norm": 0.6419789791107178, + "learning_rate": 0.003, + "loss": 3.9467, + "step": 21263 + }, + { + "epoch": 0.21264, + "grad_norm": 0.6226710081100464, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 21264 + }, + { + "epoch": 0.21265, + "grad_norm": 0.7710648775100708, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 21265 + }, + { + "epoch": 0.21266, + "grad_norm": 0.9740917086601257, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 21266 + }, + { + "epoch": 0.21267, + "grad_norm": 1.2377678155899048, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 21267 + }, + { + "epoch": 0.21268, + "grad_norm": 0.6565141081809998, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 21268 + }, + { + "epoch": 0.21269, + "grad_norm": 0.7450768947601318, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 21269 + }, + { + "epoch": 0.2127, + "grad_norm": 0.9690152406692505, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 21270 + }, + { + "epoch": 0.21271, + "grad_norm": 0.9610605835914612, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 21271 + }, + { + "epoch": 0.21272, + "grad_norm": 0.9422609806060791, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 21272 + }, + { + "epoch": 0.21273, + "grad_norm": 0.8732948899269104, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 21273 + }, + { + "epoch": 0.21274, + "grad_norm": 0.7308747172355652, + "learning_rate": 0.003, + "loss": 4.022, + "step": 21274 + }, + { + "epoch": 0.21275, + "grad_norm": 0.7389112710952759, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 21275 + }, + { + "epoch": 0.21276, + "grad_norm": 0.708928644657135, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 21276 + }, + { + "epoch": 0.21277, + "grad_norm": 0.7040172815322876, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 21277 + }, + { + "epoch": 0.21278, + "grad_norm": 0.838904619216919, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 21278 + }, + { + "epoch": 0.21279, + "grad_norm": 0.9159785509109497, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 21279 + }, + { + "epoch": 0.2128, + "grad_norm": 0.8771882653236389, + "learning_rate": 0.003, + "loss": 3.982, + "step": 21280 + }, + { + "epoch": 0.21281, + "grad_norm": 1.0078885555267334, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 21281 + }, + { + "epoch": 0.21282, + "grad_norm": 1.3111568689346313, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 21282 + }, + { + "epoch": 0.21283, + "grad_norm": 0.7286000847816467, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 21283 + }, + { + "epoch": 0.21284, + "grad_norm": 0.7487033605575562, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 21284 + }, + { + "epoch": 0.21285, + "grad_norm": 0.8190486431121826, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 21285 + }, + { + "epoch": 0.21286, + "grad_norm": 0.7764313817024231, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 21286 + }, + { + "epoch": 0.21287, + "grad_norm": 0.8552969694137573, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 21287 + }, + { + "epoch": 0.21288, + "grad_norm": 1.0460505485534668, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 21288 + }, + { + "epoch": 0.21289, + "grad_norm": 1.0178496837615967, + "learning_rate": 0.003, + "loss": 4.0, + "step": 21289 + }, + { + "epoch": 0.2129, + "grad_norm": 1.0043630599975586, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 21290 + }, + { + "epoch": 0.21291, + "grad_norm": 1.036853313446045, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 21291 + }, + { + "epoch": 0.21292, + "grad_norm": 1.2530320882797241, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 21292 + }, + { + "epoch": 0.21293, + "grad_norm": 0.7688485383987427, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 21293 + }, + { + "epoch": 0.21294, + "grad_norm": 0.6272863745689392, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 21294 + }, + { + "epoch": 0.21295, + "grad_norm": 0.7582641243934631, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 21295 + }, + { + "epoch": 0.21296, + "grad_norm": 0.8384027481079102, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 21296 + }, + { + "epoch": 0.21297, + "grad_norm": 0.8716685175895691, + "learning_rate": 0.003, + "loss": 4.006, + "step": 21297 + }, + { + "epoch": 0.21298, + "grad_norm": 0.8398637175559998, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 21298 + }, + { + "epoch": 0.21299, + "grad_norm": 0.9078550338745117, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 21299 + }, + { + "epoch": 0.213, + "grad_norm": 0.977097749710083, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 21300 + }, + { + "epoch": 0.21301, + "grad_norm": 0.8427779674530029, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 21301 + }, + { + "epoch": 0.21302, + "grad_norm": 0.8695405125617981, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 21302 + }, + { + "epoch": 0.21303, + "grad_norm": 0.8224928975105286, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 21303 + }, + { + "epoch": 0.21304, + "grad_norm": 0.7738889455795288, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 21304 + }, + { + "epoch": 0.21305, + "grad_norm": 0.7187053561210632, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 21305 + }, + { + "epoch": 0.21306, + "grad_norm": 0.792472779750824, + "learning_rate": 0.003, + "loss": 3.997, + "step": 21306 + }, + { + "epoch": 0.21307, + "grad_norm": 0.9581801891326904, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 21307 + }, + { + "epoch": 0.21308, + "grad_norm": 1.2714388370513916, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 21308 + }, + { + "epoch": 0.21309, + "grad_norm": 0.9173094034194946, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 21309 + }, + { + "epoch": 0.2131, + "grad_norm": 0.7521931529045105, + "learning_rate": 0.003, + "loss": 3.973, + "step": 21310 + }, + { + "epoch": 0.21311, + "grad_norm": 0.7684593200683594, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 21311 + }, + { + "epoch": 0.21312, + "grad_norm": 0.8680782318115234, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 21312 + }, + { + "epoch": 0.21313, + "grad_norm": 0.9322457313537598, + "learning_rate": 0.003, + "loss": 4.004, + "step": 21313 + }, + { + "epoch": 0.21314, + "grad_norm": 0.9743700623512268, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 21314 + }, + { + "epoch": 0.21315, + "grad_norm": 1.0613746643066406, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 21315 + }, + { + "epoch": 0.21316, + "grad_norm": 0.7336246371269226, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 21316 + }, + { + "epoch": 0.21317, + "grad_norm": 0.8099008798599243, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 21317 + }, + { + "epoch": 0.21318, + "grad_norm": 0.7765265703201294, + "learning_rate": 0.003, + "loss": 4.012, + "step": 21318 + }, + { + "epoch": 0.21319, + "grad_norm": 0.9255688786506653, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 21319 + }, + { + "epoch": 0.2132, + "grad_norm": 1.1583796739578247, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 21320 + }, + { + "epoch": 0.21321, + "grad_norm": 1.160841941833496, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 21321 + }, + { + "epoch": 0.21322, + "grad_norm": 0.9834440350532532, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 21322 + }, + { + "epoch": 0.21323, + "grad_norm": 0.8796864748001099, + "learning_rate": 0.003, + "loss": 3.991, + "step": 21323 + }, + { + "epoch": 0.21324, + "grad_norm": 0.670418381690979, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 21324 + }, + { + "epoch": 0.21325, + "grad_norm": 0.6408448815345764, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 21325 + }, + { + "epoch": 0.21326, + "grad_norm": 0.6484048962593079, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 21326 + }, + { + "epoch": 0.21327, + "grad_norm": 0.6597508192062378, + "learning_rate": 0.003, + "loss": 3.977, + "step": 21327 + }, + { + "epoch": 0.21328, + "grad_norm": 0.6739736199378967, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 21328 + }, + { + "epoch": 0.21329, + "grad_norm": 0.6913688778877258, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 21329 + }, + { + "epoch": 0.2133, + "grad_norm": 0.7158641219139099, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 21330 + }, + { + "epoch": 0.21331, + "grad_norm": 0.8251495957374573, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 21331 + }, + { + "epoch": 0.21332, + "grad_norm": 0.9861782193183899, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 21332 + }, + { + "epoch": 0.21333, + "grad_norm": 1.1715059280395508, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 21333 + }, + { + "epoch": 0.21334, + "grad_norm": 0.8657821416854858, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 21334 + }, + { + "epoch": 0.21335, + "grad_norm": 0.7464960217475891, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 21335 + }, + { + "epoch": 0.21336, + "grad_norm": 0.6485677361488342, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 21336 + }, + { + "epoch": 0.21337, + "grad_norm": 0.7432356476783752, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 21337 + }, + { + "epoch": 0.21338, + "grad_norm": 0.7690950036048889, + "learning_rate": 0.003, + "loss": 4.009, + "step": 21338 + }, + { + "epoch": 0.21339, + "grad_norm": 0.7918177843093872, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 21339 + }, + { + "epoch": 0.2134, + "grad_norm": 0.9233434200286865, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 21340 + }, + { + "epoch": 0.21341, + "grad_norm": 0.9883318543434143, + "learning_rate": 0.003, + "loss": 4.011, + "step": 21341 + }, + { + "epoch": 0.21342, + "grad_norm": 0.9573469758033752, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 21342 + }, + { + "epoch": 0.21343, + "grad_norm": 1.0672484636306763, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 21343 + }, + { + "epoch": 0.21344, + "grad_norm": 0.8774200081825256, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 21344 + }, + { + "epoch": 0.21345, + "grad_norm": 0.6955781579017639, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 21345 + }, + { + "epoch": 0.21346, + "grad_norm": 0.7631359696388245, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 21346 + }, + { + "epoch": 0.21347, + "grad_norm": 0.8767004013061523, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 21347 + }, + { + "epoch": 0.21348, + "grad_norm": 0.8137421011924744, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 21348 + }, + { + "epoch": 0.21349, + "grad_norm": 0.8356567621231079, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 21349 + }, + { + "epoch": 0.2135, + "grad_norm": 0.729219377040863, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 21350 + }, + { + "epoch": 0.21351, + "grad_norm": 0.7466780543327332, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 21351 + }, + { + "epoch": 0.21352, + "grad_norm": 0.8320849537849426, + "learning_rate": 0.003, + "loss": 4.002, + "step": 21352 + }, + { + "epoch": 0.21353, + "grad_norm": 0.9109935164451599, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 21353 + }, + { + "epoch": 0.21354, + "grad_norm": 0.7894145846366882, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 21354 + }, + { + "epoch": 0.21355, + "grad_norm": 0.9256765246391296, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 21355 + }, + { + "epoch": 0.21356, + "grad_norm": 1.1972886323928833, + "learning_rate": 0.003, + "loss": 4.008, + "step": 21356 + }, + { + "epoch": 0.21357, + "grad_norm": 0.8073228001594543, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 21357 + }, + { + "epoch": 0.21358, + "grad_norm": 0.645750880241394, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 21358 + }, + { + "epoch": 0.21359, + "grad_norm": 0.5913841724395752, + "learning_rate": 0.003, + "loss": 3.986, + "step": 21359 + }, + { + "epoch": 0.2136, + "grad_norm": 0.6498050689697266, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 21360 + }, + { + "epoch": 0.21361, + "grad_norm": 0.7870457768440247, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 21361 + }, + { + "epoch": 0.21362, + "grad_norm": 0.9344431757926941, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 21362 + }, + { + "epoch": 0.21363, + "grad_norm": 1.3138923645019531, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 21363 + }, + { + "epoch": 0.21364, + "grad_norm": 0.9131081700325012, + "learning_rate": 0.003, + "loss": 3.997, + "step": 21364 + }, + { + "epoch": 0.21365, + "grad_norm": 0.6897475123405457, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 21365 + }, + { + "epoch": 0.21366, + "grad_norm": 0.6339974403381348, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 21366 + }, + { + "epoch": 0.21367, + "grad_norm": 0.6924554109573364, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 21367 + }, + { + "epoch": 0.21368, + "grad_norm": 0.7020323872566223, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 21368 + }, + { + "epoch": 0.21369, + "grad_norm": 0.7593525648117065, + "learning_rate": 0.003, + "loss": 3.97, + "step": 21369 + }, + { + "epoch": 0.2137, + "grad_norm": 0.9868069887161255, + "learning_rate": 0.003, + "loss": 4.037, + "step": 21370 + }, + { + "epoch": 0.21371, + "grad_norm": 1.468988299369812, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 21371 + }, + { + "epoch": 0.21372, + "grad_norm": 0.6237825751304626, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 21372 + }, + { + "epoch": 0.21373, + "grad_norm": 0.8894487619400024, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 21373 + }, + { + "epoch": 0.21374, + "grad_norm": 1.1508939266204834, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 21374 + }, + { + "epoch": 0.21375, + "grad_norm": 0.8118572235107422, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 21375 + }, + { + "epoch": 0.21376, + "grad_norm": 0.8020175099372864, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 21376 + }, + { + "epoch": 0.21377, + "grad_norm": 0.8234835267066956, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 21377 + }, + { + "epoch": 0.21378, + "grad_norm": 0.7949830293655396, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 21378 + }, + { + "epoch": 0.21379, + "grad_norm": 0.7271310091018677, + "learning_rate": 0.003, + "loss": 3.998, + "step": 21379 + }, + { + "epoch": 0.2138, + "grad_norm": 0.7402547597885132, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 21380 + }, + { + "epoch": 0.21381, + "grad_norm": 0.8389953374862671, + "learning_rate": 0.003, + "loss": 3.987, + "step": 21381 + }, + { + "epoch": 0.21382, + "grad_norm": 0.9322521090507507, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 21382 + }, + { + "epoch": 0.21383, + "grad_norm": 0.8542722463607788, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 21383 + }, + { + "epoch": 0.21384, + "grad_norm": 0.739112913608551, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 21384 + }, + { + "epoch": 0.21385, + "grad_norm": 0.6925649642944336, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 21385 + }, + { + "epoch": 0.21386, + "grad_norm": 0.7998360991477966, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 21386 + }, + { + "epoch": 0.21387, + "grad_norm": 0.9737545847892761, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 21387 + }, + { + "epoch": 0.21388, + "grad_norm": 1.1811580657958984, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 21388 + }, + { + "epoch": 0.21389, + "grad_norm": 0.8689839839935303, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 21389 + }, + { + "epoch": 0.2139, + "grad_norm": 0.8986697196960449, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 21390 + }, + { + "epoch": 0.21391, + "grad_norm": 0.8463107347488403, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 21391 + }, + { + "epoch": 0.21392, + "grad_norm": 0.8827471733093262, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 21392 + }, + { + "epoch": 0.21393, + "grad_norm": 1.0404102802276611, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 21393 + }, + { + "epoch": 0.21394, + "grad_norm": 1.0666475296020508, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 21394 + }, + { + "epoch": 0.21395, + "grad_norm": 0.7487891316413879, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 21395 + }, + { + "epoch": 0.21396, + "grad_norm": 0.6302778124809265, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 21396 + }, + { + "epoch": 0.21397, + "grad_norm": 0.6236978769302368, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 21397 + }, + { + "epoch": 0.21398, + "grad_norm": 0.6740934252738953, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 21398 + }, + { + "epoch": 0.21399, + "grad_norm": 0.7326478958129883, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 21399 + }, + { + "epoch": 0.214, + "grad_norm": 0.7422345280647278, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 21400 + }, + { + "epoch": 0.21401, + "grad_norm": 0.8307872414588928, + "learning_rate": 0.003, + "loss": 3.997, + "step": 21401 + }, + { + "epoch": 0.21402, + "grad_norm": 0.996636688709259, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 21402 + }, + { + "epoch": 0.21403, + "grad_norm": 1.3118737936019897, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 21403 + }, + { + "epoch": 0.21404, + "grad_norm": 0.5969485640525818, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 21404 + }, + { + "epoch": 0.21405, + "grad_norm": 0.6585992574691772, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 21405 + }, + { + "epoch": 0.21406, + "grad_norm": 0.8573005795478821, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 21406 + }, + { + "epoch": 0.21407, + "grad_norm": 0.9111630916595459, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 21407 + }, + { + "epoch": 0.21408, + "grad_norm": 0.8855857253074646, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 21408 + }, + { + "epoch": 0.21409, + "grad_norm": 0.8993113040924072, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 21409 + }, + { + "epoch": 0.2141, + "grad_norm": 1.2753902673721313, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 21410 + }, + { + "epoch": 0.21411, + "grad_norm": 0.7424344420433044, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 21411 + }, + { + "epoch": 0.21412, + "grad_norm": 0.6525055766105652, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 21412 + }, + { + "epoch": 0.21413, + "grad_norm": 0.626751184463501, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 21413 + }, + { + "epoch": 0.21414, + "grad_norm": 0.665280818939209, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 21414 + }, + { + "epoch": 0.21415, + "grad_norm": 0.700272262096405, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 21415 + }, + { + "epoch": 0.21416, + "grad_norm": 0.8077136874198914, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 21416 + }, + { + "epoch": 0.21417, + "grad_norm": 0.878995954990387, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 21417 + }, + { + "epoch": 0.21418, + "grad_norm": 0.912872314453125, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 21418 + }, + { + "epoch": 0.21419, + "grad_norm": 1.1034051179885864, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 21419 + }, + { + "epoch": 0.2142, + "grad_norm": 0.8531668782234192, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 21420 + }, + { + "epoch": 0.21421, + "grad_norm": 0.7344503402709961, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 21421 + }, + { + "epoch": 0.21422, + "grad_norm": 0.777967631816864, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 21422 + }, + { + "epoch": 0.21423, + "grad_norm": 0.8980423212051392, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 21423 + }, + { + "epoch": 0.21424, + "grad_norm": 1.0109288692474365, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 21424 + }, + { + "epoch": 0.21425, + "grad_norm": 1.206906795501709, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 21425 + }, + { + "epoch": 0.21426, + "grad_norm": 0.9823929667472839, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 21426 + }, + { + "epoch": 0.21427, + "grad_norm": 1.0044265985488892, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 21427 + }, + { + "epoch": 0.21428, + "grad_norm": 1.1148549318313599, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 21428 + }, + { + "epoch": 0.21429, + "grad_norm": 1.2754542827606201, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 21429 + }, + { + "epoch": 0.2143, + "grad_norm": 1.0673002004623413, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 21430 + }, + { + "epoch": 0.21431, + "grad_norm": 0.9503219723701477, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 21431 + }, + { + "epoch": 0.21432, + "grad_norm": 0.8529365062713623, + "learning_rate": 0.003, + "loss": 4.025, + "step": 21432 + }, + { + "epoch": 0.21433, + "grad_norm": 0.8356447219848633, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 21433 + }, + { + "epoch": 0.21434, + "grad_norm": 0.8862406611442566, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 21434 + }, + { + "epoch": 0.21435, + "grad_norm": 0.9108374714851379, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 21435 + }, + { + "epoch": 0.21436, + "grad_norm": 1.031944990158081, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 21436 + }, + { + "epoch": 0.21437, + "grad_norm": 0.8075357675552368, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 21437 + }, + { + "epoch": 0.21438, + "grad_norm": 0.6826090216636658, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 21438 + }, + { + "epoch": 0.21439, + "grad_norm": 0.6600490212440491, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 21439 + }, + { + "epoch": 0.2144, + "grad_norm": 0.8710610866546631, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 21440 + }, + { + "epoch": 0.21441, + "grad_norm": 0.9636276364326477, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 21441 + }, + { + "epoch": 0.21442, + "grad_norm": 1.1399955749511719, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 21442 + }, + { + "epoch": 0.21443, + "grad_norm": 0.8459378480911255, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 21443 + }, + { + "epoch": 0.21444, + "grad_norm": 0.7782046794891357, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 21444 + }, + { + "epoch": 0.21445, + "grad_norm": 0.7522889971733093, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 21445 + }, + { + "epoch": 0.21446, + "grad_norm": 0.7701814770698547, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 21446 + }, + { + "epoch": 0.21447, + "grad_norm": 0.669678807258606, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 21447 + }, + { + "epoch": 0.21448, + "grad_norm": 0.7141163349151611, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 21448 + }, + { + "epoch": 0.21449, + "grad_norm": 0.7170905470848083, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 21449 + }, + { + "epoch": 0.2145, + "grad_norm": 0.8248327970504761, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 21450 + }, + { + "epoch": 0.21451, + "grad_norm": 0.8136173486709595, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 21451 + }, + { + "epoch": 0.21452, + "grad_norm": 0.9080814719200134, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 21452 + }, + { + "epoch": 0.21453, + "grad_norm": 0.9823253154754639, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 21453 + }, + { + "epoch": 0.21454, + "grad_norm": 1.1566261053085327, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 21454 + }, + { + "epoch": 0.21455, + "grad_norm": 0.9019991755485535, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 21455 + }, + { + "epoch": 0.21456, + "grad_norm": 0.9257096648216248, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 21456 + }, + { + "epoch": 0.21457, + "grad_norm": 0.8937629461288452, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 21457 + }, + { + "epoch": 0.21458, + "grad_norm": 0.7520314455032349, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 21458 + }, + { + "epoch": 0.21459, + "grad_norm": 0.6758750081062317, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 21459 + }, + { + "epoch": 0.2146, + "grad_norm": 0.7649513483047485, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 21460 + }, + { + "epoch": 0.21461, + "grad_norm": 1.1376173496246338, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 21461 + }, + { + "epoch": 0.21462, + "grad_norm": 0.9588015675544739, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 21462 + }, + { + "epoch": 0.21463, + "grad_norm": 0.8070213198661804, + "learning_rate": 0.003, + "loss": 4.022, + "step": 21463 + }, + { + "epoch": 0.21464, + "grad_norm": 0.6913764476776123, + "learning_rate": 0.003, + "loss": 4.013, + "step": 21464 + }, + { + "epoch": 0.21465, + "grad_norm": 0.7754470109939575, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 21465 + }, + { + "epoch": 0.21466, + "grad_norm": 0.8009421825408936, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 21466 + }, + { + "epoch": 0.21467, + "grad_norm": 0.739096999168396, + "learning_rate": 0.003, + "loss": 4.017, + "step": 21467 + }, + { + "epoch": 0.21468, + "grad_norm": 0.7636803984642029, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 21468 + }, + { + "epoch": 0.21469, + "grad_norm": 0.7612506747245789, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 21469 + }, + { + "epoch": 0.2147, + "grad_norm": 0.7444126009941101, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 21470 + }, + { + "epoch": 0.21471, + "grad_norm": 0.7803741693496704, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 21471 + }, + { + "epoch": 0.21472, + "grad_norm": 0.7790638208389282, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 21472 + }, + { + "epoch": 0.21473, + "grad_norm": 0.8299775123596191, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 21473 + }, + { + "epoch": 0.21474, + "grad_norm": 0.9028633236885071, + "learning_rate": 0.003, + "loss": 3.999, + "step": 21474 + }, + { + "epoch": 0.21475, + "grad_norm": 1.0119889974594116, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 21475 + }, + { + "epoch": 0.21476, + "grad_norm": 0.9485741853713989, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 21476 + }, + { + "epoch": 0.21477, + "grad_norm": 0.9185463190078735, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 21477 + }, + { + "epoch": 0.21478, + "grad_norm": 0.9477009773254395, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 21478 + }, + { + "epoch": 0.21479, + "grad_norm": 1.134232521057129, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 21479 + }, + { + "epoch": 0.2148, + "grad_norm": 1.0365551710128784, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 21480 + }, + { + "epoch": 0.21481, + "grad_norm": 0.8751606941223145, + "learning_rate": 0.003, + "loss": 3.983, + "step": 21481 + }, + { + "epoch": 0.21482, + "grad_norm": 0.8953807950019836, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 21482 + }, + { + "epoch": 0.21483, + "grad_norm": 1.025256872177124, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 21483 + }, + { + "epoch": 0.21484, + "grad_norm": 1.1377090215682983, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 21484 + }, + { + "epoch": 0.21485, + "grad_norm": 0.7435174584388733, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 21485 + }, + { + "epoch": 0.21486, + "grad_norm": 0.5592331290245056, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 21486 + }, + { + "epoch": 0.21487, + "grad_norm": 0.5993446111679077, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 21487 + }, + { + "epoch": 0.21488, + "grad_norm": 0.7046895027160645, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 21488 + }, + { + "epoch": 0.21489, + "grad_norm": 0.7890210747718811, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 21489 + }, + { + "epoch": 0.2149, + "grad_norm": 0.8567792177200317, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 21490 + }, + { + "epoch": 0.21491, + "grad_norm": 1.0247083902359009, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 21491 + }, + { + "epoch": 0.21492, + "grad_norm": 0.9478097558021545, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 21492 + }, + { + "epoch": 0.21493, + "grad_norm": 0.759121835231781, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 21493 + }, + { + "epoch": 0.21494, + "grad_norm": 0.718255877494812, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 21494 + }, + { + "epoch": 0.21495, + "grad_norm": 0.7934027314186096, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 21495 + }, + { + "epoch": 0.21496, + "grad_norm": 0.8434153199195862, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 21496 + }, + { + "epoch": 0.21497, + "grad_norm": 0.9614878296852112, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 21497 + }, + { + "epoch": 0.21498, + "grad_norm": 1.1912420988082886, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 21498 + }, + { + "epoch": 0.21499, + "grad_norm": 0.8249834179878235, + "learning_rate": 0.003, + "loss": 3.98, + "step": 21499 + }, + { + "epoch": 0.215, + "grad_norm": 0.7677482962608337, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 21500 + }, + { + "epoch": 0.21501, + "grad_norm": 0.7687137126922607, + "learning_rate": 0.003, + "loss": 4.021, + "step": 21501 + }, + { + "epoch": 0.21502, + "grad_norm": 0.8445277810096741, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 21502 + }, + { + "epoch": 0.21503, + "grad_norm": 0.8190683722496033, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 21503 + }, + { + "epoch": 0.21504, + "grad_norm": 0.8961876034736633, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 21504 + }, + { + "epoch": 0.21505, + "grad_norm": 0.8899973630905151, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 21505 + }, + { + "epoch": 0.21506, + "grad_norm": 0.7235985994338989, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 21506 + }, + { + "epoch": 0.21507, + "grad_norm": 0.6671749353408813, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 21507 + }, + { + "epoch": 0.21508, + "grad_norm": 0.7030459642410278, + "learning_rate": 0.003, + "loss": 3.996, + "step": 21508 + }, + { + "epoch": 0.21509, + "grad_norm": 0.7492251396179199, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 21509 + }, + { + "epoch": 0.2151, + "grad_norm": 0.7980300784111023, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 21510 + }, + { + "epoch": 0.21511, + "grad_norm": 0.8603885173797607, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 21511 + }, + { + "epoch": 0.21512, + "grad_norm": 1.0846902132034302, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 21512 + }, + { + "epoch": 0.21513, + "grad_norm": 1.1388031244277954, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 21513 + }, + { + "epoch": 0.21514, + "grad_norm": 1.119376540184021, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 21514 + }, + { + "epoch": 0.21515, + "grad_norm": 0.9386163353919983, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 21515 + }, + { + "epoch": 0.21516, + "grad_norm": 0.6738733649253845, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 21516 + }, + { + "epoch": 0.21517, + "grad_norm": 0.7374707460403442, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 21517 + }, + { + "epoch": 0.21518, + "grad_norm": 0.9816496968269348, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 21518 + }, + { + "epoch": 0.21519, + "grad_norm": 1.312325119972229, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 21519 + }, + { + "epoch": 0.2152, + "grad_norm": 0.6848953366279602, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 21520 + }, + { + "epoch": 0.21521, + "grad_norm": 0.7131217122077942, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 21521 + }, + { + "epoch": 0.21522, + "grad_norm": 0.7658547759056091, + "learning_rate": 0.003, + "loss": 3.997, + "step": 21522 + }, + { + "epoch": 0.21523, + "grad_norm": 0.7071949243545532, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 21523 + }, + { + "epoch": 0.21524, + "grad_norm": 0.8120801448822021, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 21524 + }, + { + "epoch": 0.21525, + "grad_norm": 0.910185694694519, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 21525 + }, + { + "epoch": 0.21526, + "grad_norm": 0.9224514365196228, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 21526 + }, + { + "epoch": 0.21527, + "grad_norm": 0.9379438161849976, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 21527 + }, + { + "epoch": 0.21528, + "grad_norm": 0.8752361536026001, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 21528 + }, + { + "epoch": 0.21529, + "grad_norm": 0.8849471211433411, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 21529 + }, + { + "epoch": 0.2153, + "grad_norm": 0.8519395589828491, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 21530 + }, + { + "epoch": 0.21531, + "grad_norm": 0.7978440523147583, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 21531 + }, + { + "epoch": 0.21532, + "grad_norm": 0.8714047074317932, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 21532 + }, + { + "epoch": 0.21533, + "grad_norm": 1.1668659448623657, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 21533 + }, + { + "epoch": 0.21534, + "grad_norm": 1.114955186843872, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 21534 + }, + { + "epoch": 0.21535, + "grad_norm": 0.8509944081306458, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 21535 + }, + { + "epoch": 0.21536, + "grad_norm": 0.8822514414787292, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 21536 + }, + { + "epoch": 0.21537, + "grad_norm": 0.9310247302055359, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 21537 + }, + { + "epoch": 0.21538, + "grad_norm": 0.9640525579452515, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 21538 + }, + { + "epoch": 0.21539, + "grad_norm": 0.7971515655517578, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 21539 + }, + { + "epoch": 0.2154, + "grad_norm": 0.6938146948814392, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 21540 + }, + { + "epoch": 0.21541, + "grad_norm": 0.7298148274421692, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 21541 + }, + { + "epoch": 0.21542, + "grad_norm": 0.8091928958892822, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 21542 + }, + { + "epoch": 0.21543, + "grad_norm": 0.9020199775695801, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 21543 + }, + { + "epoch": 0.21544, + "grad_norm": 1.0660803318023682, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 21544 + }, + { + "epoch": 0.21545, + "grad_norm": 0.8591522574424744, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 21545 + }, + { + "epoch": 0.21546, + "grad_norm": 0.7764420509338379, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 21546 + }, + { + "epoch": 0.21547, + "grad_norm": 0.8070088028907776, + "learning_rate": 0.003, + "loss": 4.0, + "step": 21547 + }, + { + "epoch": 0.21548, + "grad_norm": 0.9091497659683228, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 21548 + }, + { + "epoch": 0.21549, + "grad_norm": 0.9033742547035217, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 21549 + }, + { + "epoch": 0.2155, + "grad_norm": 0.9027680158615112, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 21550 + }, + { + "epoch": 0.21551, + "grad_norm": 0.8090609908103943, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 21551 + }, + { + "epoch": 0.21552, + "grad_norm": 1.0181396007537842, + "learning_rate": 0.003, + "loss": 3.99, + "step": 21552 + }, + { + "epoch": 0.21553, + "grad_norm": 1.0318619012832642, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 21553 + }, + { + "epoch": 0.21554, + "grad_norm": 1.1289538145065308, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 21554 + }, + { + "epoch": 0.21555, + "grad_norm": 0.810670018196106, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 21555 + }, + { + "epoch": 0.21556, + "grad_norm": 0.6473629474639893, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 21556 + }, + { + "epoch": 0.21557, + "grad_norm": 0.6656478643417358, + "learning_rate": 0.003, + "loss": 3.994, + "step": 21557 + }, + { + "epoch": 0.21558, + "grad_norm": 0.7705038785934448, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 21558 + }, + { + "epoch": 0.21559, + "grad_norm": 0.7846300005912781, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 21559 + }, + { + "epoch": 0.2156, + "grad_norm": 0.6978834867477417, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 21560 + }, + { + "epoch": 0.21561, + "grad_norm": 0.7100029587745667, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 21561 + }, + { + "epoch": 0.21562, + "grad_norm": 0.7772417068481445, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 21562 + }, + { + "epoch": 0.21563, + "grad_norm": 0.849115252494812, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 21563 + }, + { + "epoch": 0.21564, + "grad_norm": 0.782450795173645, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 21564 + }, + { + "epoch": 0.21565, + "grad_norm": 0.7190684080123901, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 21565 + }, + { + "epoch": 0.21566, + "grad_norm": 0.7838330268859863, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 21566 + }, + { + "epoch": 0.21567, + "grad_norm": 0.770158588886261, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 21567 + }, + { + "epoch": 0.21568, + "grad_norm": 0.7984477877616882, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 21568 + }, + { + "epoch": 0.21569, + "grad_norm": 0.8937097191810608, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 21569 + }, + { + "epoch": 0.2157, + "grad_norm": 0.968723714351654, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 21570 + }, + { + "epoch": 0.21571, + "grad_norm": 0.9115933179855347, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 21571 + }, + { + "epoch": 0.21572, + "grad_norm": 1.1553531885147095, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 21572 + }, + { + "epoch": 0.21573, + "grad_norm": 1.3083298206329346, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 21573 + }, + { + "epoch": 0.21574, + "grad_norm": 1.010390281677246, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 21574 + }, + { + "epoch": 0.21575, + "grad_norm": 0.9186457991600037, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 21575 + }, + { + "epoch": 0.21576, + "grad_norm": 0.8118718266487122, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 21576 + }, + { + "epoch": 0.21577, + "grad_norm": 0.802436113357544, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 21577 + }, + { + "epoch": 0.21578, + "grad_norm": 0.7805765271186829, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 21578 + }, + { + "epoch": 0.21579, + "grad_norm": 0.7462113499641418, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 21579 + }, + { + "epoch": 0.2158, + "grad_norm": 0.7370455265045166, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 21580 + }, + { + "epoch": 0.21581, + "grad_norm": 0.6179274916648865, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 21581 + }, + { + "epoch": 0.21582, + "grad_norm": 0.6713892221450806, + "learning_rate": 0.003, + "loss": 3.992, + "step": 21582 + }, + { + "epoch": 0.21583, + "grad_norm": 0.702362596988678, + "learning_rate": 0.003, + "loss": 3.993, + "step": 21583 + }, + { + "epoch": 0.21584, + "grad_norm": 0.8408122658729553, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 21584 + }, + { + "epoch": 0.21585, + "grad_norm": 1.3136085271835327, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 21585 + }, + { + "epoch": 0.21586, + "grad_norm": 0.8202847242355347, + "learning_rate": 0.003, + "loss": 3.976, + "step": 21586 + }, + { + "epoch": 0.21587, + "grad_norm": 0.7213858366012573, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 21587 + }, + { + "epoch": 0.21588, + "grad_norm": 0.7149998545646667, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 21588 + }, + { + "epoch": 0.21589, + "grad_norm": 0.6210320591926575, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 21589 + }, + { + "epoch": 0.2159, + "grad_norm": 0.5500255227088928, + "learning_rate": 0.003, + "loss": 3.978, + "step": 21590 + }, + { + "epoch": 0.21591, + "grad_norm": 0.5912858843803406, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 21591 + }, + { + "epoch": 0.21592, + "grad_norm": 0.6400618553161621, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 21592 + }, + { + "epoch": 0.21593, + "grad_norm": 0.7033151984214783, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 21593 + }, + { + "epoch": 0.21594, + "grad_norm": 0.7388778328895569, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 21594 + }, + { + "epoch": 0.21595, + "grad_norm": 1.0031992197036743, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 21595 + }, + { + "epoch": 0.21596, + "grad_norm": 1.3574848175048828, + "learning_rate": 0.003, + "loss": 3.986, + "step": 21596 + }, + { + "epoch": 0.21597, + "grad_norm": 0.6363314986228943, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 21597 + }, + { + "epoch": 0.21598, + "grad_norm": 0.5633575916290283, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 21598 + }, + { + "epoch": 0.21599, + "grad_norm": 0.5635490417480469, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 21599 + }, + { + "epoch": 0.216, + "grad_norm": 0.5104173421859741, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 21600 + }, + { + "epoch": 0.21601, + "grad_norm": 0.5622468590736389, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 21601 + }, + { + "epoch": 0.21602, + "grad_norm": 0.5432456135749817, + "learning_rate": 0.003, + "loss": 3.9406, + "step": 21602 + }, + { + "epoch": 0.21603, + "grad_norm": 0.6619833111763, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 21603 + }, + { + "epoch": 0.21604, + "grad_norm": 0.7667783498764038, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 21604 + }, + { + "epoch": 0.21605, + "grad_norm": 0.8937801122665405, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 21605 + }, + { + "epoch": 0.21606, + "grad_norm": 1.0920990705490112, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 21606 + }, + { + "epoch": 0.21607, + "grad_norm": 1.1117022037506104, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 21607 + }, + { + "epoch": 0.21608, + "grad_norm": 1.0428662300109863, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 21608 + }, + { + "epoch": 0.21609, + "grad_norm": 1.1177937984466553, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 21609 + }, + { + "epoch": 0.2161, + "grad_norm": 0.9072465300559998, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 21610 + }, + { + "epoch": 0.21611, + "grad_norm": 0.7594946622848511, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 21611 + }, + { + "epoch": 0.21612, + "grad_norm": 0.7143943309783936, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 21612 + }, + { + "epoch": 0.21613, + "grad_norm": 0.7479103207588196, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 21613 + }, + { + "epoch": 0.21614, + "grad_norm": 0.7915438413619995, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 21614 + }, + { + "epoch": 0.21615, + "grad_norm": 0.765329122543335, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 21615 + }, + { + "epoch": 0.21616, + "grad_norm": 0.9077237248420715, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 21616 + }, + { + "epoch": 0.21617, + "grad_norm": 0.9856047630310059, + "learning_rate": 0.003, + "loss": 4.004, + "step": 21617 + }, + { + "epoch": 0.21618, + "grad_norm": 1.047212839126587, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 21618 + }, + { + "epoch": 0.21619, + "grad_norm": 1.0377479791641235, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 21619 + }, + { + "epoch": 0.2162, + "grad_norm": 0.9774327278137207, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 21620 + }, + { + "epoch": 0.21621, + "grad_norm": 1.0613621473312378, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 21621 + }, + { + "epoch": 0.21622, + "grad_norm": 1.2862685918807983, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 21622 + }, + { + "epoch": 0.21623, + "grad_norm": 0.9352540373802185, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 21623 + }, + { + "epoch": 0.21624, + "grad_norm": 0.9812989234924316, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 21624 + }, + { + "epoch": 0.21625, + "grad_norm": 1.1307133436203003, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 21625 + }, + { + "epoch": 0.21626, + "grad_norm": 1.0251976251602173, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 21626 + }, + { + "epoch": 0.21627, + "grad_norm": 0.9355548620223999, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 21627 + }, + { + "epoch": 0.21628, + "grad_norm": 0.8433915972709656, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 21628 + }, + { + "epoch": 0.21629, + "grad_norm": 0.8060188293457031, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 21629 + }, + { + "epoch": 0.2163, + "grad_norm": 0.8156616687774658, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 21630 + }, + { + "epoch": 0.21631, + "grad_norm": 0.7303346991539001, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 21631 + }, + { + "epoch": 0.21632, + "grad_norm": 0.7202138304710388, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 21632 + }, + { + "epoch": 0.21633, + "grad_norm": 0.6567436456680298, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 21633 + }, + { + "epoch": 0.21634, + "grad_norm": 0.6734424829483032, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 21634 + }, + { + "epoch": 0.21635, + "grad_norm": 0.7806814908981323, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 21635 + }, + { + "epoch": 0.21636, + "grad_norm": 0.8717731237411499, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 21636 + }, + { + "epoch": 0.21637, + "grad_norm": 0.9170213341712952, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 21637 + }, + { + "epoch": 0.21638, + "grad_norm": 0.9883865118026733, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 21638 + }, + { + "epoch": 0.21639, + "grad_norm": 1.0601403713226318, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 21639 + }, + { + "epoch": 0.2164, + "grad_norm": 1.1392728090286255, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 21640 + }, + { + "epoch": 0.21641, + "grad_norm": 1.009885311126709, + "learning_rate": 0.003, + "loss": 4.019, + "step": 21641 + }, + { + "epoch": 0.21642, + "grad_norm": 0.8549948930740356, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 21642 + }, + { + "epoch": 0.21643, + "grad_norm": 0.6906864643096924, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 21643 + }, + { + "epoch": 0.21644, + "grad_norm": 0.7085233926773071, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 21644 + }, + { + "epoch": 0.21645, + "grad_norm": 0.7718926072120667, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 21645 + }, + { + "epoch": 0.21646, + "grad_norm": 0.9482201337814331, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 21646 + }, + { + "epoch": 0.21647, + "grad_norm": 1.035446286201477, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 21647 + }, + { + "epoch": 0.21648, + "grad_norm": 0.9593836069107056, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 21648 + }, + { + "epoch": 0.21649, + "grad_norm": 1.111215591430664, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 21649 + }, + { + "epoch": 0.2165, + "grad_norm": 0.7875701189041138, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 21650 + }, + { + "epoch": 0.21651, + "grad_norm": 0.8115482926368713, + "learning_rate": 0.003, + "loss": 4.009, + "step": 21651 + }, + { + "epoch": 0.21652, + "grad_norm": 0.9459853768348694, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 21652 + }, + { + "epoch": 0.21653, + "grad_norm": 1.3868050575256348, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 21653 + }, + { + "epoch": 0.21654, + "grad_norm": 0.7253365516662598, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 21654 + }, + { + "epoch": 0.21655, + "grad_norm": 0.5790639519691467, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 21655 + }, + { + "epoch": 0.21656, + "grad_norm": 0.7314210534095764, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 21656 + }, + { + "epoch": 0.21657, + "grad_norm": 0.7323318719863892, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 21657 + }, + { + "epoch": 0.21658, + "grad_norm": 0.7297476530075073, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 21658 + }, + { + "epoch": 0.21659, + "grad_norm": 0.7540402412414551, + "learning_rate": 0.003, + "loss": 3.986, + "step": 21659 + }, + { + "epoch": 0.2166, + "grad_norm": 0.9123125076293945, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 21660 + }, + { + "epoch": 0.21661, + "grad_norm": 1.0856797695159912, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 21661 + }, + { + "epoch": 0.21662, + "grad_norm": 0.8925968408584595, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 21662 + }, + { + "epoch": 0.21663, + "grad_norm": 0.804034411907196, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 21663 + }, + { + "epoch": 0.21664, + "grad_norm": 0.7565365433692932, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 21664 + }, + { + "epoch": 0.21665, + "grad_norm": 0.7280241250991821, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 21665 + }, + { + "epoch": 0.21666, + "grad_norm": 0.7137926816940308, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 21666 + }, + { + "epoch": 0.21667, + "grad_norm": 0.6572439074516296, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 21667 + }, + { + "epoch": 0.21668, + "grad_norm": 0.6711937785148621, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 21668 + }, + { + "epoch": 0.21669, + "grad_norm": 0.9364628195762634, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 21669 + }, + { + "epoch": 0.2167, + "grad_norm": 1.2037913799285889, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 21670 + }, + { + "epoch": 0.21671, + "grad_norm": 0.7352555394172668, + "learning_rate": 0.003, + "loss": 4.013, + "step": 21671 + }, + { + "epoch": 0.21672, + "grad_norm": 0.8104667067527771, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 21672 + }, + { + "epoch": 0.21673, + "grad_norm": 0.9006807208061218, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 21673 + }, + { + "epoch": 0.21674, + "grad_norm": 0.822273850440979, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 21674 + }, + { + "epoch": 0.21675, + "grad_norm": 0.7177990674972534, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 21675 + }, + { + "epoch": 0.21676, + "grad_norm": 0.7374477982521057, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 21676 + }, + { + "epoch": 0.21677, + "grad_norm": 0.8151225447654724, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 21677 + }, + { + "epoch": 0.21678, + "grad_norm": 0.7467964291572571, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 21678 + }, + { + "epoch": 0.21679, + "grad_norm": 0.8282430768013, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 21679 + }, + { + "epoch": 0.2168, + "grad_norm": 0.8612915873527527, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 21680 + }, + { + "epoch": 0.21681, + "grad_norm": 0.8159078359603882, + "learning_rate": 0.003, + "loss": 3.994, + "step": 21681 + }, + { + "epoch": 0.21682, + "grad_norm": 0.8112987875938416, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 21682 + }, + { + "epoch": 0.21683, + "grad_norm": 0.9414745569229126, + "learning_rate": 0.003, + "loss": 4.031, + "step": 21683 + }, + { + "epoch": 0.21684, + "grad_norm": 1.0101068019866943, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 21684 + }, + { + "epoch": 0.21685, + "grad_norm": 1.2375761270523071, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 21685 + }, + { + "epoch": 0.21686, + "grad_norm": 0.940755307674408, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 21686 + }, + { + "epoch": 0.21687, + "grad_norm": 0.8457043170928955, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 21687 + }, + { + "epoch": 0.21688, + "grad_norm": 0.7019209861755371, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 21688 + }, + { + "epoch": 0.21689, + "grad_norm": 0.6815224289894104, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 21689 + }, + { + "epoch": 0.2169, + "grad_norm": 0.6714164614677429, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 21690 + }, + { + "epoch": 0.21691, + "grad_norm": 0.6608923673629761, + "learning_rate": 0.003, + "loss": 3.97, + "step": 21691 + }, + { + "epoch": 0.21692, + "grad_norm": 0.6295070052146912, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 21692 + }, + { + "epoch": 0.21693, + "grad_norm": 0.6505048274993896, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 21693 + }, + { + "epoch": 0.21694, + "grad_norm": 0.7221375703811646, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 21694 + }, + { + "epoch": 0.21695, + "grad_norm": 0.808851420879364, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 21695 + }, + { + "epoch": 0.21696, + "grad_norm": 0.866611123085022, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 21696 + }, + { + "epoch": 0.21697, + "grad_norm": 1.0013079643249512, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 21697 + }, + { + "epoch": 0.21698, + "grad_norm": 1.250712513923645, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 21698 + }, + { + "epoch": 0.21699, + "grad_norm": 0.6898664832115173, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 21699 + }, + { + "epoch": 0.217, + "grad_norm": 0.6790230870246887, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 21700 + }, + { + "epoch": 0.21701, + "grad_norm": 0.7735278010368347, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 21701 + }, + { + "epoch": 0.21702, + "grad_norm": 0.8030800819396973, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 21702 + }, + { + "epoch": 0.21703, + "grad_norm": 0.7727534174919128, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 21703 + }, + { + "epoch": 0.21704, + "grad_norm": 0.7794595956802368, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 21704 + }, + { + "epoch": 0.21705, + "grad_norm": 0.7921155095100403, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 21705 + }, + { + "epoch": 0.21706, + "grad_norm": 0.8238611817359924, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 21706 + }, + { + "epoch": 0.21707, + "grad_norm": 1.0892044305801392, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 21707 + }, + { + "epoch": 0.21708, + "grad_norm": 1.425960659980774, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 21708 + }, + { + "epoch": 0.21709, + "grad_norm": 0.5476034283638, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 21709 + }, + { + "epoch": 0.2171, + "grad_norm": 0.8816379308700562, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 21710 + }, + { + "epoch": 0.21711, + "grad_norm": 1.0819042921066284, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 21711 + }, + { + "epoch": 0.21712, + "grad_norm": 0.8465421199798584, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 21712 + }, + { + "epoch": 0.21713, + "grad_norm": 0.7796959280967712, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 21713 + }, + { + "epoch": 0.21714, + "grad_norm": 0.7378363013267517, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 21714 + }, + { + "epoch": 0.21715, + "grad_norm": 0.6723987460136414, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 21715 + }, + { + "epoch": 0.21716, + "grad_norm": 0.7222424745559692, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 21716 + }, + { + "epoch": 0.21717, + "grad_norm": 0.7406854629516602, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 21717 + }, + { + "epoch": 0.21718, + "grad_norm": 0.7715611457824707, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 21718 + }, + { + "epoch": 0.21719, + "grad_norm": 1.0326521396636963, + "learning_rate": 0.003, + "loss": 3.989, + "step": 21719 + }, + { + "epoch": 0.2172, + "grad_norm": 1.1519089937210083, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 21720 + }, + { + "epoch": 0.21721, + "grad_norm": 0.6841143369674683, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 21721 + }, + { + "epoch": 0.21722, + "grad_norm": 0.6327571868896484, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 21722 + }, + { + "epoch": 0.21723, + "grad_norm": 0.6629207134246826, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 21723 + }, + { + "epoch": 0.21724, + "grad_norm": 0.7101562023162842, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 21724 + }, + { + "epoch": 0.21725, + "grad_norm": 0.894132673740387, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 21725 + }, + { + "epoch": 0.21726, + "grad_norm": 0.9501135945320129, + "learning_rate": 0.003, + "loss": 3.979, + "step": 21726 + }, + { + "epoch": 0.21727, + "grad_norm": 1.0270566940307617, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 21727 + }, + { + "epoch": 0.21728, + "grad_norm": 0.9478116035461426, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 21728 + }, + { + "epoch": 0.21729, + "grad_norm": 0.9073694348335266, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 21729 + }, + { + "epoch": 0.2173, + "grad_norm": 1.0418245792388916, + "learning_rate": 0.003, + "loss": 3.989, + "step": 21730 + }, + { + "epoch": 0.21731, + "grad_norm": 1.1172386407852173, + "learning_rate": 0.003, + "loss": 4.048, + "step": 21731 + }, + { + "epoch": 0.21732, + "grad_norm": 1.1806389093399048, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 21732 + }, + { + "epoch": 0.21733, + "grad_norm": 0.8895754218101501, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 21733 + }, + { + "epoch": 0.21734, + "grad_norm": 0.9886199831962585, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 21734 + }, + { + "epoch": 0.21735, + "grad_norm": 1.1685677766799927, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 21735 + }, + { + "epoch": 0.21736, + "grad_norm": 0.9962295293807983, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 21736 + }, + { + "epoch": 0.21737, + "grad_norm": 0.9883165955543518, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 21737 + }, + { + "epoch": 0.21738, + "grad_norm": 1.1411752700805664, + "learning_rate": 0.003, + "loss": 4.028, + "step": 21738 + }, + { + "epoch": 0.21739, + "grad_norm": 0.7186645865440369, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 21739 + }, + { + "epoch": 0.2174, + "grad_norm": 0.8156120181083679, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 21740 + }, + { + "epoch": 0.21741, + "grad_norm": 0.9368207454681396, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 21741 + }, + { + "epoch": 0.21742, + "grad_norm": 0.9107393026351929, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 21742 + }, + { + "epoch": 0.21743, + "grad_norm": 0.9836253523826599, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 21743 + }, + { + "epoch": 0.21744, + "grad_norm": 0.9611278772354126, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 21744 + }, + { + "epoch": 0.21745, + "grad_norm": 0.9981785416603088, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 21745 + }, + { + "epoch": 0.21746, + "grad_norm": 0.930952250957489, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 21746 + }, + { + "epoch": 0.21747, + "grad_norm": 0.8511864542961121, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 21747 + }, + { + "epoch": 0.21748, + "grad_norm": 0.6964396834373474, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 21748 + }, + { + "epoch": 0.21749, + "grad_norm": 0.7083607912063599, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 21749 + }, + { + "epoch": 0.2175, + "grad_norm": 0.6942681670188904, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 21750 + }, + { + "epoch": 0.21751, + "grad_norm": 0.691506028175354, + "learning_rate": 0.003, + "loss": 3.967, + "step": 21751 + }, + { + "epoch": 0.21752, + "grad_norm": 0.6675229072570801, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 21752 + }, + { + "epoch": 0.21753, + "grad_norm": 0.6643424034118652, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 21753 + }, + { + "epoch": 0.21754, + "grad_norm": 0.7258358001708984, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 21754 + }, + { + "epoch": 0.21755, + "grad_norm": 0.7555429339408875, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 21755 + }, + { + "epoch": 0.21756, + "grad_norm": 0.9071589112281799, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 21756 + }, + { + "epoch": 0.21757, + "grad_norm": 1.1698025465011597, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 21757 + }, + { + "epoch": 0.21758, + "grad_norm": 0.8683314919471741, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 21758 + }, + { + "epoch": 0.21759, + "grad_norm": 0.7756586074829102, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 21759 + }, + { + "epoch": 0.2176, + "grad_norm": 0.7192184925079346, + "learning_rate": 0.003, + "loss": 3.972, + "step": 21760 + }, + { + "epoch": 0.21761, + "grad_norm": 0.637648344039917, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 21761 + }, + { + "epoch": 0.21762, + "grad_norm": 0.6303030848503113, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 21762 + }, + { + "epoch": 0.21763, + "grad_norm": 0.614669680595398, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 21763 + }, + { + "epoch": 0.21764, + "grad_norm": 0.5687729716300964, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 21764 + }, + { + "epoch": 0.21765, + "grad_norm": 0.5657177567481995, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 21765 + }, + { + "epoch": 0.21766, + "grad_norm": 0.6852759122848511, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 21766 + }, + { + "epoch": 0.21767, + "grad_norm": 0.8339049220085144, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 21767 + }, + { + "epoch": 0.21768, + "grad_norm": 1.0641227960586548, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 21768 + }, + { + "epoch": 0.21769, + "grad_norm": 1.1641342639923096, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 21769 + }, + { + "epoch": 0.2177, + "grad_norm": 0.7870425581932068, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 21770 + }, + { + "epoch": 0.21771, + "grad_norm": 0.8347501158714294, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 21771 + }, + { + "epoch": 0.21772, + "grad_norm": 1.1029120683670044, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 21772 + }, + { + "epoch": 0.21773, + "grad_norm": 1.1193777322769165, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 21773 + }, + { + "epoch": 0.21774, + "grad_norm": 1.2279517650604248, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 21774 + }, + { + "epoch": 0.21775, + "grad_norm": 0.7667878866195679, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 21775 + }, + { + "epoch": 0.21776, + "grad_norm": 0.7371446490287781, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 21776 + }, + { + "epoch": 0.21777, + "grad_norm": 0.8275866508483887, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 21777 + }, + { + "epoch": 0.21778, + "grad_norm": 0.7437096834182739, + "learning_rate": 0.003, + "loss": 3.966, + "step": 21778 + }, + { + "epoch": 0.21779, + "grad_norm": 0.7907671332359314, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 21779 + }, + { + "epoch": 0.2178, + "grad_norm": 0.7602505683898926, + "learning_rate": 0.003, + "loss": 4.008, + "step": 21780 + }, + { + "epoch": 0.21781, + "grad_norm": 0.8467785716056824, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 21781 + }, + { + "epoch": 0.21782, + "grad_norm": 0.8706849217414856, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 21782 + }, + { + "epoch": 0.21783, + "grad_norm": 0.7256757616996765, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 21783 + }, + { + "epoch": 0.21784, + "grad_norm": 0.8991482853889465, + "learning_rate": 0.003, + "loss": 3.992, + "step": 21784 + }, + { + "epoch": 0.21785, + "grad_norm": 1.0459821224212646, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 21785 + }, + { + "epoch": 0.21786, + "grad_norm": 1.0255707502365112, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 21786 + }, + { + "epoch": 0.21787, + "grad_norm": 1.1089587211608887, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 21787 + }, + { + "epoch": 0.21788, + "grad_norm": 1.123881220817566, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 21788 + }, + { + "epoch": 0.21789, + "grad_norm": 0.780867338180542, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 21789 + }, + { + "epoch": 0.2179, + "grad_norm": 0.660801112651825, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 21790 + }, + { + "epoch": 0.21791, + "grad_norm": 0.693256676197052, + "learning_rate": 0.003, + "loss": 4.0, + "step": 21791 + }, + { + "epoch": 0.21792, + "grad_norm": 0.8057222962379456, + "learning_rate": 0.003, + "loss": 3.973, + "step": 21792 + }, + { + "epoch": 0.21793, + "grad_norm": 1.0628511905670166, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 21793 + }, + { + "epoch": 0.21794, + "grad_norm": 0.9622427821159363, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 21794 + }, + { + "epoch": 0.21795, + "grad_norm": 0.9382039904594421, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 21795 + }, + { + "epoch": 0.21796, + "grad_norm": 0.8528504967689514, + "learning_rate": 0.003, + "loss": 3.952, + "step": 21796 + }, + { + "epoch": 0.21797, + "grad_norm": 0.7184682488441467, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 21797 + }, + { + "epoch": 0.21798, + "grad_norm": 0.8047919869422913, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 21798 + }, + { + "epoch": 0.21799, + "grad_norm": 0.8219313621520996, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 21799 + }, + { + "epoch": 0.218, + "grad_norm": 0.7478654980659485, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 21800 + }, + { + "epoch": 0.21801, + "grad_norm": 0.8288010954856873, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 21801 + }, + { + "epoch": 0.21802, + "grad_norm": 1.1118663549423218, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 21802 + }, + { + "epoch": 0.21803, + "grad_norm": 0.9051061272621155, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 21803 + }, + { + "epoch": 0.21804, + "grad_norm": 0.9293070435523987, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 21804 + }, + { + "epoch": 0.21805, + "grad_norm": 1.1804019212722778, + "learning_rate": 0.003, + "loss": 4.013, + "step": 21805 + }, + { + "epoch": 0.21806, + "grad_norm": 0.8746418356895447, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 21806 + }, + { + "epoch": 0.21807, + "grad_norm": 0.6622724533081055, + "learning_rate": 0.003, + "loss": 3.996, + "step": 21807 + }, + { + "epoch": 0.21808, + "grad_norm": 0.8090115189552307, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 21808 + }, + { + "epoch": 0.21809, + "grad_norm": 0.9673647880554199, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 21809 + }, + { + "epoch": 0.2181, + "grad_norm": 1.0920014381408691, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 21810 + }, + { + "epoch": 0.21811, + "grad_norm": 0.9011911749839783, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 21811 + }, + { + "epoch": 0.21812, + "grad_norm": 0.7135774493217468, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 21812 + }, + { + "epoch": 0.21813, + "grad_norm": 0.7775082588195801, + "learning_rate": 0.003, + "loss": 4.001, + "step": 21813 + }, + { + "epoch": 0.21814, + "grad_norm": 0.8009023666381836, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 21814 + }, + { + "epoch": 0.21815, + "grad_norm": 0.9355428814888, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 21815 + }, + { + "epoch": 0.21816, + "grad_norm": 1.066105604171753, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 21816 + }, + { + "epoch": 0.21817, + "grad_norm": 0.8733352422714233, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 21817 + }, + { + "epoch": 0.21818, + "grad_norm": 0.7580428719520569, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 21818 + }, + { + "epoch": 0.21819, + "grad_norm": 0.7471960186958313, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 21819 + }, + { + "epoch": 0.2182, + "grad_norm": 0.7892679572105408, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 21820 + }, + { + "epoch": 0.21821, + "grad_norm": 0.9344003200531006, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 21821 + }, + { + "epoch": 0.21822, + "grad_norm": 0.9431074261665344, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 21822 + }, + { + "epoch": 0.21823, + "grad_norm": 0.7924651503562927, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 21823 + }, + { + "epoch": 0.21824, + "grad_norm": 0.6665099263191223, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 21824 + }, + { + "epoch": 0.21825, + "grad_norm": 0.6552866101264954, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 21825 + }, + { + "epoch": 0.21826, + "grad_norm": 0.7465530633926392, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 21826 + }, + { + "epoch": 0.21827, + "grad_norm": 0.9469466209411621, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 21827 + }, + { + "epoch": 0.21828, + "grad_norm": 1.1896499395370483, + "learning_rate": 0.003, + "loss": 4.017, + "step": 21828 + }, + { + "epoch": 0.21829, + "grad_norm": 0.741629958152771, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 21829 + }, + { + "epoch": 0.2183, + "grad_norm": 0.7058972120285034, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 21830 + }, + { + "epoch": 0.21831, + "grad_norm": 0.7743293046951294, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 21831 + }, + { + "epoch": 0.21832, + "grad_norm": 0.8948047757148743, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 21832 + }, + { + "epoch": 0.21833, + "grad_norm": 1.0396822690963745, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 21833 + }, + { + "epoch": 0.21834, + "grad_norm": 1.0337562561035156, + "learning_rate": 0.003, + "loss": 4.013, + "step": 21834 + }, + { + "epoch": 0.21835, + "grad_norm": 1.0034815073013306, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 21835 + }, + { + "epoch": 0.21836, + "grad_norm": 1.1195166110992432, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 21836 + }, + { + "epoch": 0.21837, + "grad_norm": 0.9589192867279053, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 21837 + }, + { + "epoch": 0.21838, + "grad_norm": 0.9210767149925232, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 21838 + }, + { + "epoch": 0.21839, + "grad_norm": 0.8899030685424805, + "learning_rate": 0.003, + "loss": 3.99, + "step": 21839 + }, + { + "epoch": 0.2184, + "grad_norm": 0.8589850664138794, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 21840 + }, + { + "epoch": 0.21841, + "grad_norm": 0.8063408136367798, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 21841 + }, + { + "epoch": 0.21842, + "grad_norm": 0.7477491497993469, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 21842 + }, + { + "epoch": 0.21843, + "grad_norm": 0.767784059047699, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 21843 + }, + { + "epoch": 0.21844, + "grad_norm": 0.8069847226142883, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 21844 + }, + { + "epoch": 0.21845, + "grad_norm": 0.9541254639625549, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 21845 + }, + { + "epoch": 0.21846, + "grad_norm": 1.0547032356262207, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 21846 + }, + { + "epoch": 0.21847, + "grad_norm": 1.011212706565857, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 21847 + }, + { + "epoch": 0.21848, + "grad_norm": 0.9120833277702332, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 21848 + }, + { + "epoch": 0.21849, + "grad_norm": 0.8835930228233337, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 21849 + }, + { + "epoch": 0.2185, + "grad_norm": 0.8800077438354492, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 21850 + }, + { + "epoch": 0.21851, + "grad_norm": 0.8210106492042542, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 21851 + }, + { + "epoch": 0.21852, + "grad_norm": 0.7035391330718994, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 21852 + }, + { + "epoch": 0.21853, + "grad_norm": 0.6045515537261963, + "learning_rate": 0.003, + "loss": 3.998, + "step": 21853 + }, + { + "epoch": 0.21854, + "grad_norm": 0.6367928385734558, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 21854 + }, + { + "epoch": 0.21855, + "grad_norm": 0.81891930103302, + "learning_rate": 0.003, + "loss": 4.003, + "step": 21855 + }, + { + "epoch": 0.21856, + "grad_norm": 1.0347527265548706, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 21856 + }, + { + "epoch": 0.21857, + "grad_norm": 0.8652042746543884, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 21857 + }, + { + "epoch": 0.21858, + "grad_norm": 0.7179366946220398, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 21858 + }, + { + "epoch": 0.21859, + "grad_norm": 0.6417927742004395, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 21859 + }, + { + "epoch": 0.2186, + "grad_norm": 0.6499086618423462, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 21860 + }, + { + "epoch": 0.21861, + "grad_norm": 0.848762571811676, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 21861 + }, + { + "epoch": 0.21862, + "grad_norm": 1.1485133171081543, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 21862 + }, + { + "epoch": 0.21863, + "grad_norm": 0.7762075662612915, + "learning_rate": 0.003, + "loss": 3.978, + "step": 21863 + }, + { + "epoch": 0.21864, + "grad_norm": 0.6452844142913818, + "learning_rate": 0.003, + "loss": 3.949, + "step": 21864 + }, + { + "epoch": 0.21865, + "grad_norm": 0.6497505307197571, + "learning_rate": 0.003, + "loss": 4.015, + "step": 21865 + }, + { + "epoch": 0.21866, + "grad_norm": 0.6114002466201782, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 21866 + }, + { + "epoch": 0.21867, + "grad_norm": 0.6360263228416443, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 21867 + }, + { + "epoch": 0.21868, + "grad_norm": 0.6664139628410339, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 21868 + }, + { + "epoch": 0.21869, + "grad_norm": 0.7450810074806213, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 21869 + }, + { + "epoch": 0.2187, + "grad_norm": 0.7692160606384277, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 21870 + }, + { + "epoch": 0.21871, + "grad_norm": 0.7508399486541748, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 21871 + }, + { + "epoch": 0.21872, + "grad_norm": 0.7897133231163025, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 21872 + }, + { + "epoch": 0.21873, + "grad_norm": 1.0052375793457031, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 21873 + }, + { + "epoch": 0.21874, + "grad_norm": 1.2442702054977417, + "learning_rate": 0.003, + "loss": 4.008, + "step": 21874 + }, + { + "epoch": 0.21875, + "grad_norm": 0.7757917642593384, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 21875 + }, + { + "epoch": 0.21876, + "grad_norm": 0.6992387175559998, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 21876 + }, + { + "epoch": 0.21877, + "grad_norm": 0.9269700646400452, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 21877 + }, + { + "epoch": 0.21878, + "grad_norm": 0.9496790170669556, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 21878 + }, + { + "epoch": 0.21879, + "grad_norm": 1.0214134454727173, + "learning_rate": 0.003, + "loss": 4.006, + "step": 21879 + }, + { + "epoch": 0.2188, + "grad_norm": 1.0221530199050903, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 21880 + }, + { + "epoch": 0.21881, + "grad_norm": 0.9862629771232605, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 21881 + }, + { + "epoch": 0.21882, + "grad_norm": 0.9332275390625, + "learning_rate": 0.003, + "loss": 3.986, + "step": 21882 + }, + { + "epoch": 0.21883, + "grad_norm": 0.9583297967910767, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 21883 + }, + { + "epoch": 0.21884, + "grad_norm": 0.8877280950546265, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 21884 + }, + { + "epoch": 0.21885, + "grad_norm": 0.7536664009094238, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 21885 + }, + { + "epoch": 0.21886, + "grad_norm": 0.8727545142173767, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 21886 + }, + { + "epoch": 0.21887, + "grad_norm": 1.2611030340194702, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 21887 + }, + { + "epoch": 0.21888, + "grad_norm": 0.9947317838668823, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 21888 + }, + { + "epoch": 0.21889, + "grad_norm": 1.0494213104248047, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 21889 + }, + { + "epoch": 0.2189, + "grad_norm": 1.1365302801132202, + "learning_rate": 0.003, + "loss": 3.992, + "step": 21890 + }, + { + "epoch": 0.21891, + "grad_norm": 0.7793552279472351, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 21891 + }, + { + "epoch": 0.21892, + "grad_norm": 0.5573461651802063, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 21892 + }, + { + "epoch": 0.21893, + "grad_norm": 0.5648072361946106, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 21893 + }, + { + "epoch": 0.21894, + "grad_norm": 0.7403696179389954, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 21894 + }, + { + "epoch": 0.21895, + "grad_norm": 0.8550360798835754, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 21895 + }, + { + "epoch": 0.21896, + "grad_norm": 0.8431591391563416, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 21896 + }, + { + "epoch": 0.21897, + "grad_norm": 0.6262039542198181, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 21897 + }, + { + "epoch": 0.21898, + "grad_norm": 0.5236126184463501, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 21898 + }, + { + "epoch": 0.21899, + "grad_norm": 0.6156030893325806, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 21899 + }, + { + "epoch": 0.219, + "grad_norm": 0.7020008563995361, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 21900 + }, + { + "epoch": 0.21901, + "grad_norm": 0.8077492713928223, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 21901 + }, + { + "epoch": 0.21902, + "grad_norm": 0.8745711445808411, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 21902 + }, + { + "epoch": 0.21903, + "grad_norm": 0.8810880184173584, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 21903 + }, + { + "epoch": 0.21904, + "grad_norm": 0.8140197992324829, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 21904 + }, + { + "epoch": 0.21905, + "grad_norm": 0.7950108647346497, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 21905 + }, + { + "epoch": 0.21906, + "grad_norm": 0.972731351852417, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 21906 + }, + { + "epoch": 0.21907, + "grad_norm": 1.0046824216842651, + "learning_rate": 0.003, + "loss": 3.985, + "step": 21907 + }, + { + "epoch": 0.21908, + "grad_norm": 1.0718578100204468, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 21908 + }, + { + "epoch": 0.21909, + "grad_norm": 1.0573465824127197, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 21909 + }, + { + "epoch": 0.2191, + "grad_norm": 0.955188512802124, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 21910 + }, + { + "epoch": 0.21911, + "grad_norm": 0.8179667592048645, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 21911 + }, + { + "epoch": 0.21912, + "grad_norm": 0.6760953068733215, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 21912 + }, + { + "epoch": 0.21913, + "grad_norm": 0.7233975529670715, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 21913 + }, + { + "epoch": 0.21914, + "grad_norm": 0.9044110178947449, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 21914 + }, + { + "epoch": 0.21915, + "grad_norm": 1.2637940645217896, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 21915 + }, + { + "epoch": 0.21916, + "grad_norm": 0.8100100755691528, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 21916 + }, + { + "epoch": 0.21917, + "grad_norm": 0.9130622744560242, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 21917 + }, + { + "epoch": 0.21918, + "grad_norm": 1.0892486572265625, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 21918 + }, + { + "epoch": 0.21919, + "grad_norm": 0.861940324306488, + "learning_rate": 0.003, + "loss": 4.03, + "step": 21919 + }, + { + "epoch": 0.2192, + "grad_norm": 0.8659204840660095, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 21920 + }, + { + "epoch": 0.21921, + "grad_norm": 0.9533199071884155, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 21921 + }, + { + "epoch": 0.21922, + "grad_norm": 1.0614013671875, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 21922 + }, + { + "epoch": 0.21923, + "grad_norm": 1.188806414604187, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 21923 + }, + { + "epoch": 0.21924, + "grad_norm": 0.733652651309967, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 21924 + }, + { + "epoch": 0.21925, + "grad_norm": 0.8300747275352478, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 21925 + }, + { + "epoch": 0.21926, + "grad_norm": 0.801517903804779, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 21926 + }, + { + "epoch": 0.21927, + "grad_norm": 0.8466152548789978, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 21927 + }, + { + "epoch": 0.21928, + "grad_norm": 0.7429540157318115, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 21928 + }, + { + "epoch": 0.21929, + "grad_norm": 0.6662804484367371, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 21929 + }, + { + "epoch": 0.2193, + "grad_norm": 0.7149642109870911, + "learning_rate": 0.003, + "loss": 3.992, + "step": 21930 + }, + { + "epoch": 0.21931, + "grad_norm": 0.7264842391014099, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 21931 + }, + { + "epoch": 0.21932, + "grad_norm": 0.7160820960998535, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 21932 + }, + { + "epoch": 0.21933, + "grad_norm": 0.6801262497901917, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 21933 + }, + { + "epoch": 0.21934, + "grad_norm": 0.8299878835678101, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 21934 + }, + { + "epoch": 0.21935, + "grad_norm": 1.001130223274231, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 21935 + }, + { + "epoch": 0.21936, + "grad_norm": 1.205831527709961, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 21936 + }, + { + "epoch": 0.21937, + "grad_norm": 0.667312741279602, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 21937 + }, + { + "epoch": 0.21938, + "grad_norm": 0.7152754664421082, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 21938 + }, + { + "epoch": 0.21939, + "grad_norm": 0.8823091983795166, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 21939 + }, + { + "epoch": 0.2194, + "grad_norm": 0.9329628944396973, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 21940 + }, + { + "epoch": 0.21941, + "grad_norm": 0.8481011986732483, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 21941 + }, + { + "epoch": 0.21942, + "grad_norm": 0.8525370955467224, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 21942 + }, + { + "epoch": 0.21943, + "grad_norm": 0.7130926251411438, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 21943 + }, + { + "epoch": 0.21944, + "grad_norm": 0.6061363220214844, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 21944 + }, + { + "epoch": 0.21945, + "grad_norm": 0.7238110899925232, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 21945 + }, + { + "epoch": 0.21946, + "grad_norm": 0.6811885833740234, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 21946 + }, + { + "epoch": 0.21947, + "grad_norm": 0.6604867577552795, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 21947 + }, + { + "epoch": 0.21948, + "grad_norm": 0.6574642658233643, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 21948 + }, + { + "epoch": 0.21949, + "grad_norm": 0.7350143194198608, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 21949 + }, + { + "epoch": 0.2195, + "grad_norm": 0.8567330241203308, + "learning_rate": 0.003, + "loss": 3.971, + "step": 21950 + }, + { + "epoch": 0.21951, + "grad_norm": 0.8811405301094055, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 21951 + }, + { + "epoch": 0.21952, + "grad_norm": 0.8789968490600586, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 21952 + }, + { + "epoch": 0.21953, + "grad_norm": 1.1621136665344238, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 21953 + }, + { + "epoch": 0.21954, + "grad_norm": 1.098154902458191, + "learning_rate": 0.003, + "loss": 4.008, + "step": 21954 + }, + { + "epoch": 0.21955, + "grad_norm": 0.9401182532310486, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 21955 + }, + { + "epoch": 0.21956, + "grad_norm": 0.9764015674591064, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 21956 + }, + { + "epoch": 0.21957, + "grad_norm": 1.0041645765304565, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 21957 + }, + { + "epoch": 0.21958, + "grad_norm": 0.9567504525184631, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 21958 + }, + { + "epoch": 0.21959, + "grad_norm": 1.0383940935134888, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 21959 + }, + { + "epoch": 0.2196, + "grad_norm": 0.8846251368522644, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 21960 + }, + { + "epoch": 0.21961, + "grad_norm": 0.9069587588310242, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 21961 + }, + { + "epoch": 0.21962, + "grad_norm": 0.9837573766708374, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 21962 + }, + { + "epoch": 0.21963, + "grad_norm": 1.1362138986587524, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 21963 + }, + { + "epoch": 0.21964, + "grad_norm": 0.7696027755737305, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 21964 + }, + { + "epoch": 0.21965, + "grad_norm": 0.7514989376068115, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 21965 + }, + { + "epoch": 0.21966, + "grad_norm": 0.8975114822387695, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 21966 + }, + { + "epoch": 0.21967, + "grad_norm": 1.233403205871582, + "learning_rate": 0.003, + "loss": 4.007, + "step": 21967 + }, + { + "epoch": 0.21968, + "grad_norm": 0.7966341376304626, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 21968 + }, + { + "epoch": 0.21969, + "grad_norm": 0.7406032085418701, + "learning_rate": 0.003, + "loss": 4.01, + "step": 21969 + }, + { + "epoch": 0.2197, + "grad_norm": 0.7861499190330505, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 21970 + }, + { + "epoch": 0.21971, + "grad_norm": 0.7460372447967529, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 21971 + }, + { + "epoch": 0.21972, + "grad_norm": 0.687766432762146, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 21972 + }, + { + "epoch": 0.21973, + "grad_norm": 0.6905523538589478, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 21973 + }, + { + "epoch": 0.21974, + "grad_norm": 0.8351119756698608, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 21974 + }, + { + "epoch": 0.21975, + "grad_norm": 1.0248225927352905, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 21975 + }, + { + "epoch": 0.21976, + "grad_norm": 1.071418285369873, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 21976 + }, + { + "epoch": 0.21977, + "grad_norm": 0.8034154772758484, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 21977 + }, + { + "epoch": 0.21978, + "grad_norm": 0.6141573190689087, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 21978 + }, + { + "epoch": 0.21979, + "grad_norm": 0.6174970865249634, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 21979 + }, + { + "epoch": 0.2198, + "grad_norm": 0.6216170787811279, + "learning_rate": 0.003, + "loss": 3.97, + "step": 21980 + }, + { + "epoch": 0.21981, + "grad_norm": 0.6464536190032959, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 21981 + }, + { + "epoch": 0.21982, + "grad_norm": 0.7951745390892029, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 21982 + }, + { + "epoch": 0.21983, + "grad_norm": 0.8922833800315857, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 21983 + }, + { + "epoch": 0.21984, + "grad_norm": 0.8015391230583191, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 21984 + }, + { + "epoch": 0.21985, + "grad_norm": 0.6743970513343811, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 21985 + }, + { + "epoch": 0.21986, + "grad_norm": 0.7552760243415833, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 21986 + }, + { + "epoch": 0.21987, + "grad_norm": 0.9202058911323547, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 21987 + }, + { + "epoch": 0.21988, + "grad_norm": 1.0355278253555298, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 21988 + }, + { + "epoch": 0.21989, + "grad_norm": 1.0612934827804565, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 21989 + }, + { + "epoch": 0.2199, + "grad_norm": 1.1252192258834839, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 21990 + }, + { + "epoch": 0.21991, + "grad_norm": 0.8886119723320007, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 21991 + }, + { + "epoch": 0.21992, + "grad_norm": 1.0190086364746094, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 21992 + }, + { + "epoch": 0.21993, + "grad_norm": 1.0807583332061768, + "learning_rate": 0.003, + "loss": 3.991, + "step": 21993 + }, + { + "epoch": 0.21994, + "grad_norm": 1.0260684490203857, + "learning_rate": 0.003, + "loss": 3.994, + "step": 21994 + }, + { + "epoch": 0.21995, + "grad_norm": 0.8468531370162964, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 21995 + }, + { + "epoch": 0.21996, + "grad_norm": 0.738538384437561, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 21996 + }, + { + "epoch": 0.21997, + "grad_norm": 0.7444962859153748, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 21997 + }, + { + "epoch": 0.21998, + "grad_norm": 0.8011465072631836, + "learning_rate": 0.003, + "loss": 3.996, + "step": 21998 + }, + { + "epoch": 0.21999, + "grad_norm": 0.9042376279830933, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 21999 + }, + { + "epoch": 0.22, + "grad_norm": 1.055024266242981, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 22000 + }, + { + "epoch": 0.22001, + "grad_norm": 1.0458046197891235, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 22001 + }, + { + "epoch": 0.22002, + "grad_norm": 0.9207172989845276, + "learning_rate": 0.003, + "loss": 4.017, + "step": 22002 + }, + { + "epoch": 0.22003, + "grad_norm": 0.8619856834411621, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 22003 + }, + { + "epoch": 0.22004, + "grad_norm": 0.8821252584457397, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 22004 + }, + { + "epoch": 0.22005, + "grad_norm": 0.8186073303222656, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 22005 + }, + { + "epoch": 0.22006, + "grad_norm": 0.6712355613708496, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 22006 + }, + { + "epoch": 0.22007, + "grad_norm": 0.7234464883804321, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 22007 + }, + { + "epoch": 0.22008, + "grad_norm": 0.8740297555923462, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 22008 + }, + { + "epoch": 0.22009, + "grad_norm": 1.2004238367080688, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 22009 + }, + { + "epoch": 0.2201, + "grad_norm": 0.8302531838417053, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 22010 + }, + { + "epoch": 0.22011, + "grad_norm": 0.8586046695709229, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 22011 + }, + { + "epoch": 0.22012, + "grad_norm": 0.9329288005828857, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 22012 + }, + { + "epoch": 0.22013, + "grad_norm": 1.199567437171936, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 22013 + }, + { + "epoch": 0.22014, + "grad_norm": 0.8196219205856323, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 22014 + }, + { + "epoch": 0.22015, + "grad_norm": 0.8068655729293823, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 22015 + }, + { + "epoch": 0.22016, + "grad_norm": 0.843953549861908, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 22016 + }, + { + "epoch": 0.22017, + "grad_norm": 0.9907034635543823, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 22017 + }, + { + "epoch": 0.22018, + "grad_norm": 1.2731077671051025, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 22018 + }, + { + "epoch": 0.22019, + "grad_norm": 0.7847492694854736, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 22019 + }, + { + "epoch": 0.2202, + "grad_norm": 0.80507493019104, + "learning_rate": 0.003, + "loss": 4.007, + "step": 22020 + }, + { + "epoch": 0.22021, + "grad_norm": 0.914775550365448, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 22021 + }, + { + "epoch": 0.22022, + "grad_norm": 1.069144368171692, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 22022 + }, + { + "epoch": 0.22023, + "grad_norm": 0.9461555480957031, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 22023 + }, + { + "epoch": 0.22024, + "grad_norm": 0.8630251884460449, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 22024 + }, + { + "epoch": 0.22025, + "grad_norm": 0.7855283617973328, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 22025 + }, + { + "epoch": 0.22026, + "grad_norm": 0.7197895646095276, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 22026 + }, + { + "epoch": 0.22027, + "grad_norm": 0.6618131399154663, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 22027 + }, + { + "epoch": 0.22028, + "grad_norm": 0.6987737417221069, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 22028 + }, + { + "epoch": 0.22029, + "grad_norm": 0.699775755405426, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 22029 + }, + { + "epoch": 0.2203, + "grad_norm": 0.7354749441146851, + "learning_rate": 0.003, + "loss": 4.01, + "step": 22030 + }, + { + "epoch": 0.22031, + "grad_norm": 0.7739908695220947, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 22031 + }, + { + "epoch": 0.22032, + "grad_norm": 0.6414990425109863, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 22032 + }, + { + "epoch": 0.22033, + "grad_norm": 0.5843051671981812, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 22033 + }, + { + "epoch": 0.22034, + "grad_norm": 0.7568508386611938, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 22034 + }, + { + "epoch": 0.22035, + "grad_norm": 0.995360255241394, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 22035 + }, + { + "epoch": 0.22036, + "grad_norm": 1.3896152973175049, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 22036 + }, + { + "epoch": 0.22037, + "grad_norm": 0.58205646276474, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 22037 + }, + { + "epoch": 0.22038, + "grad_norm": 0.7497769594192505, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 22038 + }, + { + "epoch": 0.22039, + "grad_norm": 0.8326087594032288, + "learning_rate": 0.003, + "loss": 4.028, + "step": 22039 + }, + { + "epoch": 0.2204, + "grad_norm": 0.7921262979507446, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 22040 + }, + { + "epoch": 0.22041, + "grad_norm": 0.9724779725074768, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 22041 + }, + { + "epoch": 0.22042, + "grad_norm": 0.9791088700294495, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 22042 + }, + { + "epoch": 0.22043, + "grad_norm": 0.8840323686599731, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 22043 + }, + { + "epoch": 0.22044, + "grad_norm": 0.9133219718933105, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 22044 + }, + { + "epoch": 0.22045, + "grad_norm": 0.903985321521759, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 22045 + }, + { + "epoch": 0.22046, + "grad_norm": 0.8336799144744873, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 22046 + }, + { + "epoch": 0.22047, + "grad_norm": 0.8146893978118896, + "learning_rate": 0.003, + "loss": 4.006, + "step": 22047 + }, + { + "epoch": 0.22048, + "grad_norm": 0.8821060061454773, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 22048 + }, + { + "epoch": 0.22049, + "grad_norm": 0.7554482817649841, + "learning_rate": 0.003, + "loss": 3.98, + "step": 22049 + }, + { + "epoch": 0.2205, + "grad_norm": 0.6880313158035278, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 22050 + }, + { + "epoch": 0.22051, + "grad_norm": 0.768394410610199, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 22051 + }, + { + "epoch": 0.22052, + "grad_norm": 0.8815709948539734, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 22052 + }, + { + "epoch": 0.22053, + "grad_norm": 1.0536171197891235, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 22053 + }, + { + "epoch": 0.22054, + "grad_norm": 1.0781521797180176, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 22054 + }, + { + "epoch": 0.22055, + "grad_norm": 0.9600433111190796, + "learning_rate": 0.003, + "loss": 3.966, + "step": 22055 + }, + { + "epoch": 0.22056, + "grad_norm": 1.2109897136688232, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 22056 + }, + { + "epoch": 0.22057, + "grad_norm": 0.9964637756347656, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 22057 + }, + { + "epoch": 0.22058, + "grad_norm": 1.011827826499939, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 22058 + }, + { + "epoch": 0.22059, + "grad_norm": 1.0942292213439941, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 22059 + }, + { + "epoch": 0.2206, + "grad_norm": 0.8283538222312927, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 22060 + }, + { + "epoch": 0.22061, + "grad_norm": 0.7590895295143127, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 22061 + }, + { + "epoch": 0.22062, + "grad_norm": 0.7482670545578003, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 22062 + }, + { + "epoch": 0.22063, + "grad_norm": 0.8592257499694824, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 22063 + }, + { + "epoch": 0.22064, + "grad_norm": 1.224790334701538, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 22064 + }, + { + "epoch": 0.22065, + "grad_norm": 0.7667205333709717, + "learning_rate": 0.003, + "loss": 3.958, + "step": 22065 + }, + { + "epoch": 0.22066, + "grad_norm": 0.6214425563812256, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 22066 + }, + { + "epoch": 0.22067, + "grad_norm": 0.6794809103012085, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 22067 + }, + { + "epoch": 0.22068, + "grad_norm": 0.6893031597137451, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 22068 + }, + { + "epoch": 0.22069, + "grad_norm": 0.731228232383728, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 22069 + }, + { + "epoch": 0.2207, + "grad_norm": 0.6863489747047424, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 22070 + }, + { + "epoch": 0.22071, + "grad_norm": 0.6219610571861267, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 22071 + }, + { + "epoch": 0.22072, + "grad_norm": 0.6319617629051208, + "learning_rate": 0.003, + "loss": 4.006, + "step": 22072 + }, + { + "epoch": 0.22073, + "grad_norm": 0.5473001003265381, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 22073 + }, + { + "epoch": 0.22074, + "grad_norm": 0.45465850830078125, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 22074 + }, + { + "epoch": 0.22075, + "grad_norm": 0.4719337224960327, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 22075 + }, + { + "epoch": 0.22076, + "grad_norm": 0.5699535012245178, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 22076 + }, + { + "epoch": 0.22077, + "grad_norm": 0.6660944819450378, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 22077 + }, + { + "epoch": 0.22078, + "grad_norm": 0.7694702744483948, + "learning_rate": 0.003, + "loss": 3.986, + "step": 22078 + }, + { + "epoch": 0.22079, + "grad_norm": 0.8310136795043945, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 22079 + }, + { + "epoch": 0.2208, + "grad_norm": 0.8229924440383911, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 22080 + }, + { + "epoch": 0.22081, + "grad_norm": 1.0219606161117554, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 22081 + }, + { + "epoch": 0.22082, + "grad_norm": 1.138608694076538, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 22082 + }, + { + "epoch": 0.22083, + "grad_norm": 1.0048757791519165, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 22083 + }, + { + "epoch": 0.22084, + "grad_norm": 0.9989445209503174, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 22084 + }, + { + "epoch": 0.22085, + "grad_norm": 0.9029433131217957, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 22085 + }, + { + "epoch": 0.22086, + "grad_norm": 0.8330425024032593, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 22086 + }, + { + "epoch": 0.22087, + "grad_norm": 0.8408195376396179, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 22087 + }, + { + "epoch": 0.22088, + "grad_norm": 0.9590644836425781, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 22088 + }, + { + "epoch": 0.22089, + "grad_norm": 1.1778767108917236, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 22089 + }, + { + "epoch": 0.2209, + "grad_norm": 0.9013878703117371, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 22090 + }, + { + "epoch": 0.22091, + "grad_norm": 1.0042451620101929, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 22091 + }, + { + "epoch": 0.22092, + "grad_norm": 1.206027626991272, + "learning_rate": 0.003, + "loss": 4.012, + "step": 22092 + }, + { + "epoch": 0.22093, + "grad_norm": 0.8067894577980042, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 22093 + }, + { + "epoch": 0.22094, + "grad_norm": 0.8525177836418152, + "learning_rate": 0.003, + "loss": 3.992, + "step": 22094 + }, + { + "epoch": 0.22095, + "grad_norm": 0.8194411396980286, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 22095 + }, + { + "epoch": 0.22096, + "grad_norm": 0.7573952674865723, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 22096 + }, + { + "epoch": 0.22097, + "grad_norm": 0.8444618582725525, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 22097 + }, + { + "epoch": 0.22098, + "grad_norm": 0.8737071752548218, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 22098 + }, + { + "epoch": 0.22099, + "grad_norm": 0.859525203704834, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 22099 + }, + { + "epoch": 0.221, + "grad_norm": 1.0038325786590576, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 22100 + }, + { + "epoch": 0.22101, + "grad_norm": 0.8679660558700562, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 22101 + }, + { + "epoch": 0.22102, + "grad_norm": 0.8583188056945801, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 22102 + }, + { + "epoch": 0.22103, + "grad_norm": 0.9698607325553894, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 22103 + }, + { + "epoch": 0.22104, + "grad_norm": 1.1323961019515991, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 22104 + }, + { + "epoch": 0.22105, + "grad_norm": 0.9282779097557068, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 22105 + }, + { + "epoch": 0.22106, + "grad_norm": 1.0819494724273682, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 22106 + }, + { + "epoch": 0.22107, + "grad_norm": 1.019161343574524, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 22107 + }, + { + "epoch": 0.22108, + "grad_norm": 0.6851383447647095, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 22108 + }, + { + "epoch": 0.22109, + "grad_norm": 0.6658185720443726, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 22109 + }, + { + "epoch": 0.2211, + "grad_norm": 0.7594912052154541, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 22110 + }, + { + "epoch": 0.22111, + "grad_norm": 0.717314600944519, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 22111 + }, + { + "epoch": 0.22112, + "grad_norm": 0.7401524782180786, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 22112 + }, + { + "epoch": 0.22113, + "grad_norm": 0.712432861328125, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 22113 + }, + { + "epoch": 0.22114, + "grad_norm": 0.6184938549995422, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 22114 + }, + { + "epoch": 0.22115, + "grad_norm": 0.7260004281997681, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 22115 + }, + { + "epoch": 0.22116, + "grad_norm": 0.9034877419471741, + "learning_rate": 0.003, + "loss": 3.987, + "step": 22116 + }, + { + "epoch": 0.22117, + "grad_norm": 0.8833739757537842, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 22117 + }, + { + "epoch": 0.22118, + "grad_norm": 0.8505811095237732, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 22118 + }, + { + "epoch": 0.22119, + "grad_norm": 0.9156543612480164, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 22119 + }, + { + "epoch": 0.2212, + "grad_norm": 0.9952896237373352, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 22120 + }, + { + "epoch": 0.22121, + "grad_norm": 1.087693691253662, + "learning_rate": 0.003, + "loss": 4.008, + "step": 22121 + }, + { + "epoch": 0.22122, + "grad_norm": 0.950884997844696, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 22122 + }, + { + "epoch": 0.22123, + "grad_norm": 0.8316933512687683, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 22123 + }, + { + "epoch": 0.22124, + "grad_norm": 1.000860571861267, + "learning_rate": 0.003, + "loss": 3.995, + "step": 22124 + }, + { + "epoch": 0.22125, + "grad_norm": 1.3743723630905151, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 22125 + }, + { + "epoch": 0.22126, + "grad_norm": 0.6166684031486511, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 22126 + }, + { + "epoch": 0.22127, + "grad_norm": 0.7940456867218018, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 22127 + }, + { + "epoch": 0.22128, + "grad_norm": 0.8608361482620239, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 22128 + }, + { + "epoch": 0.22129, + "grad_norm": 0.7950621247291565, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 22129 + }, + { + "epoch": 0.2213, + "grad_norm": 0.7726272940635681, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 22130 + }, + { + "epoch": 0.22131, + "grad_norm": 0.8329786658287048, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 22131 + }, + { + "epoch": 0.22132, + "grad_norm": 0.917313277721405, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 22132 + }, + { + "epoch": 0.22133, + "grad_norm": 1.0228408575057983, + "learning_rate": 0.003, + "loss": 3.99, + "step": 22133 + }, + { + "epoch": 0.22134, + "grad_norm": 1.1284728050231934, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 22134 + }, + { + "epoch": 0.22135, + "grad_norm": 0.8940958976745605, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 22135 + }, + { + "epoch": 0.22136, + "grad_norm": 0.9187617301940918, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 22136 + }, + { + "epoch": 0.22137, + "grad_norm": 0.8356421589851379, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 22137 + }, + { + "epoch": 0.22138, + "grad_norm": 0.9083290100097656, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 22138 + }, + { + "epoch": 0.22139, + "grad_norm": 0.8453940153121948, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 22139 + }, + { + "epoch": 0.2214, + "grad_norm": 0.8249866962432861, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 22140 + }, + { + "epoch": 0.22141, + "grad_norm": 0.8216676115989685, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 22141 + }, + { + "epoch": 0.22142, + "grad_norm": 0.7377619743347168, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 22142 + }, + { + "epoch": 0.22143, + "grad_norm": 0.6523450613021851, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 22143 + }, + { + "epoch": 0.22144, + "grad_norm": 0.5541719794273376, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 22144 + }, + { + "epoch": 0.22145, + "grad_norm": 0.6150355339050293, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 22145 + }, + { + "epoch": 0.22146, + "grad_norm": 0.6009129285812378, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 22146 + }, + { + "epoch": 0.22147, + "grad_norm": 0.6190749406814575, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 22147 + }, + { + "epoch": 0.22148, + "grad_norm": 0.817204475402832, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 22148 + }, + { + "epoch": 0.22149, + "grad_norm": 1.0945972204208374, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 22149 + }, + { + "epoch": 0.2215, + "grad_norm": 1.1067050695419312, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 22150 + }, + { + "epoch": 0.22151, + "grad_norm": 0.8544173240661621, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 22151 + }, + { + "epoch": 0.22152, + "grad_norm": 0.7296759486198425, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 22152 + }, + { + "epoch": 0.22153, + "grad_norm": 0.7495543956756592, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 22153 + }, + { + "epoch": 0.22154, + "grad_norm": 0.8824262619018555, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 22154 + }, + { + "epoch": 0.22155, + "grad_norm": 1.009552001953125, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 22155 + }, + { + "epoch": 0.22156, + "grad_norm": 1.0163651704788208, + "learning_rate": 0.003, + "loss": 4.016, + "step": 22156 + }, + { + "epoch": 0.22157, + "grad_norm": 1.0127520561218262, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 22157 + }, + { + "epoch": 0.22158, + "grad_norm": 1.123732328414917, + "learning_rate": 0.003, + "loss": 4.045, + "step": 22158 + }, + { + "epoch": 0.22159, + "grad_norm": 0.9984117746353149, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 22159 + }, + { + "epoch": 0.2216, + "grad_norm": 1.1209408044815063, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 22160 + }, + { + "epoch": 0.22161, + "grad_norm": 0.8375343084335327, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 22161 + }, + { + "epoch": 0.22162, + "grad_norm": 0.8259281516075134, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 22162 + }, + { + "epoch": 0.22163, + "grad_norm": 0.8779699206352234, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 22163 + }, + { + "epoch": 0.22164, + "grad_norm": 0.7958191633224487, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 22164 + }, + { + "epoch": 0.22165, + "grad_norm": 0.8405134677886963, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 22165 + }, + { + "epoch": 0.22166, + "grad_norm": 0.8765663504600525, + "learning_rate": 0.003, + "loss": 3.992, + "step": 22166 + }, + { + "epoch": 0.22167, + "grad_norm": 0.8923958539962769, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 22167 + }, + { + "epoch": 0.22168, + "grad_norm": 1.0607088804244995, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 22168 + }, + { + "epoch": 0.22169, + "grad_norm": 1.018362283706665, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 22169 + }, + { + "epoch": 0.2217, + "grad_norm": 0.9706133008003235, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 22170 + }, + { + "epoch": 0.22171, + "grad_norm": 0.857482373714447, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 22171 + }, + { + "epoch": 0.22172, + "grad_norm": 0.7722411751747131, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 22172 + }, + { + "epoch": 0.22173, + "grad_norm": 0.7258930802345276, + "learning_rate": 0.003, + "loss": 3.996, + "step": 22173 + }, + { + "epoch": 0.22174, + "grad_norm": 0.7837193012237549, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 22174 + }, + { + "epoch": 0.22175, + "grad_norm": 0.908833920955658, + "learning_rate": 0.003, + "loss": 3.987, + "step": 22175 + }, + { + "epoch": 0.22176, + "grad_norm": 1.2276118993759155, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 22176 + }, + { + "epoch": 0.22177, + "grad_norm": 0.8051838278770447, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 22177 + }, + { + "epoch": 0.22178, + "grad_norm": 0.7251731753349304, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 22178 + }, + { + "epoch": 0.22179, + "grad_norm": 0.7389411330223083, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 22179 + }, + { + "epoch": 0.2218, + "grad_norm": 0.78200763463974, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 22180 + }, + { + "epoch": 0.22181, + "grad_norm": 0.6772249341011047, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 22181 + }, + { + "epoch": 0.22182, + "grad_norm": 0.7646527886390686, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 22182 + }, + { + "epoch": 0.22183, + "grad_norm": 0.9512960910797119, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 22183 + }, + { + "epoch": 0.22184, + "grad_norm": 1.070918321609497, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 22184 + }, + { + "epoch": 0.22185, + "grad_norm": 0.8389847874641418, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 22185 + }, + { + "epoch": 0.22186, + "grad_norm": 0.7842741012573242, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 22186 + }, + { + "epoch": 0.22187, + "grad_norm": 0.6231105923652649, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 22187 + }, + { + "epoch": 0.22188, + "grad_norm": 0.5638734698295593, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 22188 + }, + { + "epoch": 0.22189, + "grad_norm": 0.5618468523025513, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 22189 + }, + { + "epoch": 0.2219, + "grad_norm": 0.5843892097473145, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 22190 + }, + { + "epoch": 0.22191, + "grad_norm": 0.638679027557373, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 22191 + }, + { + "epoch": 0.22192, + "grad_norm": 0.6256392598152161, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 22192 + }, + { + "epoch": 0.22193, + "grad_norm": 0.5351330041885376, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 22193 + }, + { + "epoch": 0.22194, + "grad_norm": 0.5749831199645996, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 22194 + }, + { + "epoch": 0.22195, + "grad_norm": 0.7873042821884155, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 22195 + }, + { + "epoch": 0.22196, + "grad_norm": 1.008075475692749, + "learning_rate": 0.003, + "loss": 4.014, + "step": 22196 + }, + { + "epoch": 0.22197, + "grad_norm": 1.3661165237426758, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 22197 + }, + { + "epoch": 0.22198, + "grad_norm": 0.6056413054466248, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 22198 + }, + { + "epoch": 0.22199, + "grad_norm": 0.6689925193786621, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 22199 + }, + { + "epoch": 0.222, + "grad_norm": 0.8478479385375977, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 22200 + }, + { + "epoch": 0.22201, + "grad_norm": 1.0504173040390015, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 22201 + }, + { + "epoch": 0.22202, + "grad_norm": 1.135113000869751, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 22202 + }, + { + "epoch": 0.22203, + "grad_norm": 1.0248078107833862, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 22203 + }, + { + "epoch": 0.22204, + "grad_norm": 1.0481826066970825, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 22204 + }, + { + "epoch": 0.22205, + "grad_norm": 1.1051968336105347, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 22205 + }, + { + "epoch": 0.22206, + "grad_norm": 0.7789502739906311, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 22206 + }, + { + "epoch": 0.22207, + "grad_norm": 0.9116061329841614, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 22207 + }, + { + "epoch": 0.22208, + "grad_norm": 1.0341863632202148, + "learning_rate": 0.003, + "loss": 4.029, + "step": 22208 + }, + { + "epoch": 0.22209, + "grad_norm": 1.0729049444198608, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 22209 + }, + { + "epoch": 0.2221, + "grad_norm": 0.88841712474823, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 22210 + }, + { + "epoch": 0.22211, + "grad_norm": 0.7122446298599243, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 22211 + }, + { + "epoch": 0.22212, + "grad_norm": 0.8000245094299316, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 22212 + }, + { + "epoch": 0.22213, + "grad_norm": 0.8583983778953552, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 22213 + }, + { + "epoch": 0.22214, + "grad_norm": 0.7325140833854675, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 22214 + }, + { + "epoch": 0.22215, + "grad_norm": 0.8608267903327942, + "learning_rate": 0.003, + "loss": 4.004, + "step": 22215 + }, + { + "epoch": 0.22216, + "grad_norm": 1.0634552240371704, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 22216 + }, + { + "epoch": 0.22217, + "grad_norm": 1.17488431930542, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 22217 + }, + { + "epoch": 0.22218, + "grad_norm": 0.8430327773094177, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 22218 + }, + { + "epoch": 0.22219, + "grad_norm": 0.8326776027679443, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 22219 + }, + { + "epoch": 0.2222, + "grad_norm": 1.0416381359100342, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 22220 + }, + { + "epoch": 0.22221, + "grad_norm": 1.185704231262207, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 22221 + }, + { + "epoch": 0.22222, + "grad_norm": 0.8133190274238586, + "learning_rate": 0.003, + "loss": 4.024, + "step": 22222 + }, + { + "epoch": 0.22223, + "grad_norm": 0.7522664666175842, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 22223 + }, + { + "epoch": 0.22224, + "grad_norm": 0.7580513954162598, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 22224 + }, + { + "epoch": 0.22225, + "grad_norm": 0.7679519057273865, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 22225 + }, + { + "epoch": 0.22226, + "grad_norm": 0.7913228273391724, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 22226 + }, + { + "epoch": 0.22227, + "grad_norm": 1.041473150253296, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 22227 + }, + { + "epoch": 0.22228, + "grad_norm": 1.074540615081787, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 22228 + }, + { + "epoch": 0.22229, + "grad_norm": 0.7825056910514832, + "learning_rate": 0.003, + "loss": 3.996, + "step": 22229 + }, + { + "epoch": 0.2223, + "grad_norm": 0.7808905243873596, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 22230 + }, + { + "epoch": 0.22231, + "grad_norm": 0.890830397605896, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 22231 + }, + { + "epoch": 0.22232, + "grad_norm": 1.039552927017212, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 22232 + }, + { + "epoch": 0.22233, + "grad_norm": 0.901796817779541, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 22233 + }, + { + "epoch": 0.22234, + "grad_norm": 0.6827748417854309, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 22234 + }, + { + "epoch": 0.22235, + "grad_norm": 0.6562280654907227, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 22235 + }, + { + "epoch": 0.22236, + "grad_norm": 0.7415642142295837, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 22236 + }, + { + "epoch": 0.22237, + "grad_norm": 0.8057253360748291, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 22237 + }, + { + "epoch": 0.22238, + "grad_norm": 0.7927719354629517, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 22238 + }, + { + "epoch": 0.22239, + "grad_norm": 0.8244659900665283, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 22239 + }, + { + "epoch": 0.2224, + "grad_norm": 0.882388710975647, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 22240 + }, + { + "epoch": 0.22241, + "grad_norm": 0.9439489245414734, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 22241 + }, + { + "epoch": 0.22242, + "grad_norm": 1.1597942113876343, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 22242 + }, + { + "epoch": 0.22243, + "grad_norm": 1.118087649345398, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 22243 + }, + { + "epoch": 0.22244, + "grad_norm": 0.7988182902336121, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 22244 + }, + { + "epoch": 0.22245, + "grad_norm": 0.7414251565933228, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 22245 + }, + { + "epoch": 0.22246, + "grad_norm": 0.7245816588401794, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 22246 + }, + { + "epoch": 0.22247, + "grad_norm": 0.6117534041404724, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 22247 + }, + { + "epoch": 0.22248, + "grad_norm": 0.5136464834213257, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 22248 + }, + { + "epoch": 0.22249, + "grad_norm": 0.5456826090812683, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 22249 + }, + { + "epoch": 0.2225, + "grad_norm": 0.5742612481117249, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 22250 + }, + { + "epoch": 0.22251, + "grad_norm": 0.549299418926239, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 22251 + }, + { + "epoch": 0.22252, + "grad_norm": 0.6208416223526001, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 22252 + }, + { + "epoch": 0.22253, + "grad_norm": 0.7547205090522766, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 22253 + }, + { + "epoch": 0.22254, + "grad_norm": 0.9424986839294434, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 22254 + }, + { + "epoch": 0.22255, + "grad_norm": 1.1643556356430054, + "learning_rate": 0.003, + "loss": 4.024, + "step": 22255 + }, + { + "epoch": 0.22256, + "grad_norm": 0.9346198439598083, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 22256 + }, + { + "epoch": 0.22257, + "grad_norm": 0.9100785851478577, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 22257 + }, + { + "epoch": 0.22258, + "grad_norm": 0.7921284437179565, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 22258 + }, + { + "epoch": 0.22259, + "grad_norm": 0.8944505453109741, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 22259 + }, + { + "epoch": 0.2226, + "grad_norm": 1.0874114036560059, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 22260 + }, + { + "epoch": 0.22261, + "grad_norm": 1.373734951019287, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 22261 + }, + { + "epoch": 0.22262, + "grad_norm": 0.695056676864624, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 22262 + }, + { + "epoch": 0.22263, + "grad_norm": 0.6417421102523804, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 22263 + }, + { + "epoch": 0.22264, + "grad_norm": 0.6516551971435547, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 22264 + }, + { + "epoch": 0.22265, + "grad_norm": 0.6863132119178772, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 22265 + }, + { + "epoch": 0.22266, + "grad_norm": 0.7354105114936829, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 22266 + }, + { + "epoch": 0.22267, + "grad_norm": 0.8992321491241455, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 22267 + }, + { + "epoch": 0.22268, + "grad_norm": 1.1727731227874756, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 22268 + }, + { + "epoch": 0.22269, + "grad_norm": 0.9738989472389221, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 22269 + }, + { + "epoch": 0.2227, + "grad_norm": 1.1270623207092285, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 22270 + }, + { + "epoch": 0.22271, + "grad_norm": 1.000229001045227, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 22271 + }, + { + "epoch": 0.22272, + "grad_norm": 0.965679943561554, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 22272 + }, + { + "epoch": 0.22273, + "grad_norm": 0.865630030632019, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 22273 + }, + { + "epoch": 0.22274, + "grad_norm": 0.8891963958740234, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 22274 + }, + { + "epoch": 0.22275, + "grad_norm": 0.9803498387336731, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 22275 + }, + { + "epoch": 0.22276, + "grad_norm": 1.0342193841934204, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 22276 + }, + { + "epoch": 0.22277, + "grad_norm": 0.9761995077133179, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 22277 + }, + { + "epoch": 0.22278, + "grad_norm": 1.0215781927108765, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 22278 + }, + { + "epoch": 0.22279, + "grad_norm": 0.9841533899307251, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 22279 + }, + { + "epoch": 0.2228, + "grad_norm": 1.119734525680542, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 22280 + }, + { + "epoch": 0.22281, + "grad_norm": 1.0128929615020752, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 22281 + }, + { + "epoch": 0.22282, + "grad_norm": 0.8592376112937927, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 22282 + }, + { + "epoch": 0.22283, + "grad_norm": 0.6808568239212036, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 22283 + }, + { + "epoch": 0.22284, + "grad_norm": 0.7002740502357483, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 22284 + }, + { + "epoch": 0.22285, + "grad_norm": 0.7724105715751648, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 22285 + }, + { + "epoch": 0.22286, + "grad_norm": 0.9008498191833496, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 22286 + }, + { + "epoch": 0.22287, + "grad_norm": 0.9009160399436951, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 22287 + }, + { + "epoch": 0.22288, + "grad_norm": 0.8375632762908936, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 22288 + }, + { + "epoch": 0.22289, + "grad_norm": 0.8899363875389099, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 22289 + }, + { + "epoch": 0.2229, + "grad_norm": 0.9187003374099731, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 22290 + }, + { + "epoch": 0.22291, + "grad_norm": 1.0064581632614136, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 22291 + }, + { + "epoch": 0.22292, + "grad_norm": 1.0947022438049316, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 22292 + }, + { + "epoch": 0.22293, + "grad_norm": 0.9522907733917236, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 22293 + }, + { + "epoch": 0.22294, + "grad_norm": 0.9779766201972961, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 22294 + }, + { + "epoch": 0.22295, + "grad_norm": 0.9188073873519897, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 22295 + }, + { + "epoch": 0.22296, + "grad_norm": 0.8763561844825745, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 22296 + }, + { + "epoch": 0.22297, + "grad_norm": 1.0179646015167236, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 22297 + }, + { + "epoch": 0.22298, + "grad_norm": 1.1319931745529175, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 22298 + }, + { + "epoch": 0.22299, + "grad_norm": 0.8164716362953186, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 22299 + }, + { + "epoch": 0.223, + "grad_norm": 0.661088764667511, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 22300 + }, + { + "epoch": 0.22301, + "grad_norm": 0.7075953483581543, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 22301 + }, + { + "epoch": 0.22302, + "grad_norm": 0.8688150644302368, + "learning_rate": 0.003, + "loss": 4.004, + "step": 22302 + }, + { + "epoch": 0.22303, + "grad_norm": 1.027376413345337, + "learning_rate": 0.003, + "loss": 3.9451, + "step": 22303 + }, + { + "epoch": 0.22304, + "grad_norm": 0.9803479909896851, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 22304 + }, + { + "epoch": 0.22305, + "grad_norm": 0.7982239723205566, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 22305 + }, + { + "epoch": 0.22306, + "grad_norm": 0.6880765557289124, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 22306 + }, + { + "epoch": 0.22307, + "grad_norm": 0.6101251244544983, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 22307 + }, + { + "epoch": 0.22308, + "grad_norm": 0.6390963196754456, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 22308 + }, + { + "epoch": 0.22309, + "grad_norm": 0.6506239175796509, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 22309 + }, + { + "epoch": 0.2231, + "grad_norm": 0.716517984867096, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 22310 + }, + { + "epoch": 0.22311, + "grad_norm": 0.811601996421814, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 22311 + }, + { + "epoch": 0.22312, + "grad_norm": 0.7844415903091431, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 22312 + }, + { + "epoch": 0.22313, + "grad_norm": 0.8179540634155273, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 22313 + }, + { + "epoch": 0.22314, + "grad_norm": 0.9694624543190002, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 22314 + }, + { + "epoch": 0.22315, + "grad_norm": 1.1313526630401611, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 22315 + }, + { + "epoch": 0.22316, + "grad_norm": 0.8822540044784546, + "learning_rate": 0.003, + "loss": 4.006, + "step": 22316 + }, + { + "epoch": 0.22317, + "grad_norm": 0.7677550911903381, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 22317 + }, + { + "epoch": 0.22318, + "grad_norm": 0.8026158213615417, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 22318 + }, + { + "epoch": 0.22319, + "grad_norm": 0.7660064101219177, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 22319 + }, + { + "epoch": 0.2232, + "grad_norm": 0.6075426340103149, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 22320 + }, + { + "epoch": 0.22321, + "grad_norm": 0.5844227075576782, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 22321 + }, + { + "epoch": 0.22322, + "grad_norm": 0.5472224950790405, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 22322 + }, + { + "epoch": 0.22323, + "grad_norm": 0.6714351177215576, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 22323 + }, + { + "epoch": 0.22324, + "grad_norm": 0.8313835263252258, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 22324 + }, + { + "epoch": 0.22325, + "grad_norm": 1.1295771598815918, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 22325 + }, + { + "epoch": 0.22326, + "grad_norm": 1.0392239093780518, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 22326 + }, + { + "epoch": 0.22327, + "grad_norm": 0.930332601070404, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 22327 + }, + { + "epoch": 0.22328, + "grad_norm": 0.9336153864860535, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 22328 + }, + { + "epoch": 0.22329, + "grad_norm": 1.0793565511703491, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 22329 + }, + { + "epoch": 0.2233, + "grad_norm": 0.9663935899734497, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 22330 + }, + { + "epoch": 0.22331, + "grad_norm": 0.9161788821220398, + "learning_rate": 0.003, + "loss": 4.004, + "step": 22331 + }, + { + "epoch": 0.22332, + "grad_norm": 0.8911570906639099, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 22332 + }, + { + "epoch": 0.22333, + "grad_norm": 0.9095597863197327, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 22333 + }, + { + "epoch": 0.22334, + "grad_norm": 0.949296772480011, + "learning_rate": 0.003, + "loss": 4.041, + "step": 22334 + }, + { + "epoch": 0.22335, + "grad_norm": 0.8943361043930054, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 22335 + }, + { + "epoch": 0.22336, + "grad_norm": 0.9249036312103271, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 22336 + }, + { + "epoch": 0.22337, + "grad_norm": 0.9699913859367371, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 22337 + }, + { + "epoch": 0.22338, + "grad_norm": 1.0823391675949097, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 22338 + }, + { + "epoch": 0.22339, + "grad_norm": 1.0763885974884033, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 22339 + }, + { + "epoch": 0.2234, + "grad_norm": 0.8873983025550842, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 22340 + }, + { + "epoch": 0.22341, + "grad_norm": 0.7473699450492859, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 22341 + }, + { + "epoch": 0.22342, + "grad_norm": 0.8520479202270508, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 22342 + }, + { + "epoch": 0.22343, + "grad_norm": 0.9033915996551514, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 22343 + }, + { + "epoch": 0.22344, + "grad_norm": 0.963074803352356, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 22344 + }, + { + "epoch": 0.22345, + "grad_norm": 0.9976442456245422, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 22345 + }, + { + "epoch": 0.22346, + "grad_norm": 0.9095351696014404, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 22346 + }, + { + "epoch": 0.22347, + "grad_norm": 0.7204092741012573, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 22347 + }, + { + "epoch": 0.22348, + "grad_norm": 0.6711597442626953, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 22348 + }, + { + "epoch": 0.22349, + "grad_norm": 0.7238585352897644, + "learning_rate": 0.003, + "loss": 3.994, + "step": 22349 + }, + { + "epoch": 0.2235, + "grad_norm": 0.7849287390708923, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 22350 + }, + { + "epoch": 0.22351, + "grad_norm": 0.8802497386932373, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 22351 + }, + { + "epoch": 0.22352, + "grad_norm": 0.8883559703826904, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 22352 + }, + { + "epoch": 0.22353, + "grad_norm": 0.8620057702064514, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 22353 + }, + { + "epoch": 0.22354, + "grad_norm": 0.7732952833175659, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 22354 + }, + { + "epoch": 0.22355, + "grad_norm": 0.6646769642829895, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 22355 + }, + { + "epoch": 0.22356, + "grad_norm": 0.7639184594154358, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 22356 + }, + { + "epoch": 0.22357, + "grad_norm": 0.7673458456993103, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 22357 + }, + { + "epoch": 0.22358, + "grad_norm": 0.7774982452392578, + "learning_rate": 0.003, + "loss": 4.033, + "step": 22358 + }, + { + "epoch": 0.22359, + "grad_norm": 0.805784285068512, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 22359 + }, + { + "epoch": 0.2236, + "grad_norm": 0.8793980479240417, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 22360 + }, + { + "epoch": 0.22361, + "grad_norm": 0.8804288506507874, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 22361 + }, + { + "epoch": 0.22362, + "grad_norm": 0.8955217599868774, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 22362 + }, + { + "epoch": 0.22363, + "grad_norm": 0.9688836336135864, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 22363 + }, + { + "epoch": 0.22364, + "grad_norm": 1.0069862604141235, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 22364 + }, + { + "epoch": 0.22365, + "grad_norm": 0.9188225269317627, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 22365 + }, + { + "epoch": 0.22366, + "grad_norm": 0.8616890907287598, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 22366 + }, + { + "epoch": 0.22367, + "grad_norm": 0.7603508830070496, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 22367 + }, + { + "epoch": 0.22368, + "grad_norm": 0.7719579935073853, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 22368 + }, + { + "epoch": 0.22369, + "grad_norm": 0.8027116060256958, + "learning_rate": 0.003, + "loss": 3.999, + "step": 22369 + }, + { + "epoch": 0.2237, + "grad_norm": 0.8554940223693848, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 22370 + }, + { + "epoch": 0.22371, + "grad_norm": 1.034635066986084, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 22371 + }, + { + "epoch": 0.22372, + "grad_norm": 1.0838009119033813, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 22372 + }, + { + "epoch": 0.22373, + "grad_norm": 1.0853278636932373, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 22373 + }, + { + "epoch": 0.22374, + "grad_norm": 0.8330978155136108, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 22374 + }, + { + "epoch": 0.22375, + "grad_norm": 0.7772606015205383, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 22375 + }, + { + "epoch": 0.22376, + "grad_norm": 0.7410527467727661, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 22376 + }, + { + "epoch": 0.22377, + "grad_norm": 0.8399563431739807, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 22377 + }, + { + "epoch": 0.22378, + "grad_norm": 0.9800384640693665, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 22378 + }, + { + "epoch": 0.22379, + "grad_norm": 0.9919523000717163, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 22379 + }, + { + "epoch": 0.2238, + "grad_norm": 0.8406188488006592, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 22380 + }, + { + "epoch": 0.22381, + "grad_norm": 0.7718432545661926, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 22381 + }, + { + "epoch": 0.22382, + "grad_norm": 0.7791183590888977, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 22382 + }, + { + "epoch": 0.22383, + "grad_norm": 0.7718709111213684, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 22383 + }, + { + "epoch": 0.22384, + "grad_norm": 0.8588314652442932, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 22384 + }, + { + "epoch": 0.22385, + "grad_norm": 0.9846243262290955, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 22385 + }, + { + "epoch": 0.22386, + "grad_norm": 1.069280743598938, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 22386 + }, + { + "epoch": 0.22387, + "grad_norm": 0.9645265340805054, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 22387 + }, + { + "epoch": 0.22388, + "grad_norm": 0.8739482164382935, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 22388 + }, + { + "epoch": 0.22389, + "grad_norm": 0.8581519722938538, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 22389 + }, + { + "epoch": 0.2239, + "grad_norm": 0.7317993640899658, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 22390 + }, + { + "epoch": 0.22391, + "grad_norm": 0.7070882320404053, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 22391 + }, + { + "epoch": 0.22392, + "grad_norm": 0.7292404770851135, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 22392 + }, + { + "epoch": 0.22393, + "grad_norm": 0.7449306845664978, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 22393 + }, + { + "epoch": 0.22394, + "grad_norm": 0.7657275795936584, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 22394 + }, + { + "epoch": 0.22395, + "grad_norm": 0.8831289410591125, + "learning_rate": 0.003, + "loss": 4.019, + "step": 22395 + }, + { + "epoch": 0.22396, + "grad_norm": 0.9149502515792847, + "learning_rate": 0.003, + "loss": 4.017, + "step": 22396 + }, + { + "epoch": 0.22397, + "grad_norm": 1.0543224811553955, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 22397 + }, + { + "epoch": 0.22398, + "grad_norm": 1.083691120147705, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 22398 + }, + { + "epoch": 0.22399, + "grad_norm": 0.8254730701446533, + "learning_rate": 0.003, + "loss": 4.018, + "step": 22399 + }, + { + "epoch": 0.224, + "grad_norm": 0.750239908695221, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 22400 + }, + { + "epoch": 0.22401, + "grad_norm": 0.7438710927963257, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 22401 + }, + { + "epoch": 0.22402, + "grad_norm": 0.9534478187561035, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 22402 + }, + { + "epoch": 0.22403, + "grad_norm": 0.9826512932777405, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 22403 + }, + { + "epoch": 0.22404, + "grad_norm": 0.9033602476119995, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 22404 + }, + { + "epoch": 0.22405, + "grad_norm": 0.8978239297866821, + "learning_rate": 0.003, + "loss": 3.979, + "step": 22405 + }, + { + "epoch": 0.22406, + "grad_norm": 1.0453298091888428, + "learning_rate": 0.003, + "loss": 4.004, + "step": 22406 + }, + { + "epoch": 0.22407, + "grad_norm": 1.2624454498291016, + "learning_rate": 0.003, + "loss": 3.99, + "step": 22407 + }, + { + "epoch": 0.22408, + "grad_norm": 0.6774479746818542, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 22408 + }, + { + "epoch": 0.22409, + "grad_norm": 0.7257993221282959, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 22409 + }, + { + "epoch": 0.2241, + "grad_norm": 0.9024410843849182, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 22410 + }, + { + "epoch": 0.22411, + "grad_norm": 0.9306569695472717, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 22411 + }, + { + "epoch": 0.22412, + "grad_norm": 0.9208981990814209, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 22412 + }, + { + "epoch": 0.22413, + "grad_norm": 0.8520879149436951, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 22413 + }, + { + "epoch": 0.22414, + "grad_norm": 0.8270443677902222, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 22414 + }, + { + "epoch": 0.22415, + "grad_norm": 0.7050793766975403, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 22415 + }, + { + "epoch": 0.22416, + "grad_norm": 0.5810856223106384, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 22416 + }, + { + "epoch": 0.22417, + "grad_norm": 0.5449209809303284, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 22417 + }, + { + "epoch": 0.22418, + "grad_norm": 0.6007904410362244, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 22418 + }, + { + "epoch": 0.22419, + "grad_norm": 0.6364153623580933, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 22419 + }, + { + "epoch": 0.2242, + "grad_norm": 0.6961277723312378, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 22420 + }, + { + "epoch": 0.22421, + "grad_norm": 0.7028837203979492, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 22421 + }, + { + "epoch": 0.22422, + "grad_norm": 0.6823407411575317, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 22422 + }, + { + "epoch": 0.22423, + "grad_norm": 0.697035551071167, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 22423 + }, + { + "epoch": 0.22424, + "grad_norm": 0.6980616450309753, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 22424 + }, + { + "epoch": 0.22425, + "grad_norm": 0.6847192049026489, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 22425 + }, + { + "epoch": 0.22426, + "grad_norm": 0.781924843788147, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 22426 + }, + { + "epoch": 0.22427, + "grad_norm": 0.8105159401893616, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 22427 + }, + { + "epoch": 0.22428, + "grad_norm": 0.7041856646537781, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 22428 + }, + { + "epoch": 0.22429, + "grad_norm": 0.646094560623169, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 22429 + }, + { + "epoch": 0.2243, + "grad_norm": 0.8782885074615479, + "learning_rate": 0.003, + "loss": 3.979, + "step": 22430 + }, + { + "epoch": 0.22431, + "grad_norm": 1.2158031463623047, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 22431 + }, + { + "epoch": 0.22432, + "grad_norm": 1.2991441488265991, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 22432 + }, + { + "epoch": 0.22433, + "grad_norm": 0.6069093346595764, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 22433 + }, + { + "epoch": 0.22434, + "grad_norm": 0.825773298740387, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 22434 + }, + { + "epoch": 0.22435, + "grad_norm": 1.1204975843429565, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 22435 + }, + { + "epoch": 0.22436, + "grad_norm": 0.842893123626709, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 22436 + }, + { + "epoch": 0.22437, + "grad_norm": 0.760613739490509, + "learning_rate": 0.003, + "loss": 3.988, + "step": 22437 + }, + { + "epoch": 0.22438, + "grad_norm": 0.7518401145935059, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 22438 + }, + { + "epoch": 0.22439, + "grad_norm": 0.7762293815612793, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 22439 + }, + { + "epoch": 0.2244, + "grad_norm": 0.754931628704071, + "learning_rate": 0.003, + "loss": 4.012, + "step": 22440 + }, + { + "epoch": 0.22441, + "grad_norm": 0.8530749082565308, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 22441 + }, + { + "epoch": 0.22442, + "grad_norm": 0.8422187566757202, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 22442 + }, + { + "epoch": 0.22443, + "grad_norm": 0.9789145588874817, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 22443 + }, + { + "epoch": 0.22444, + "grad_norm": 1.1723881959915161, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 22444 + }, + { + "epoch": 0.22445, + "grad_norm": 0.9591441750526428, + "learning_rate": 0.003, + "loss": 3.99, + "step": 22445 + }, + { + "epoch": 0.22446, + "grad_norm": 0.9187237024307251, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 22446 + }, + { + "epoch": 0.22447, + "grad_norm": 0.9891934394836426, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 22447 + }, + { + "epoch": 0.22448, + "grad_norm": 1.0083531141281128, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 22448 + }, + { + "epoch": 0.22449, + "grad_norm": 1.0540077686309814, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 22449 + }, + { + "epoch": 0.2245, + "grad_norm": 1.1320610046386719, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 22450 + }, + { + "epoch": 0.22451, + "grad_norm": 0.8045580983161926, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 22451 + }, + { + "epoch": 0.22452, + "grad_norm": 0.9374115467071533, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 22452 + }, + { + "epoch": 0.22453, + "grad_norm": 1.2033405303955078, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 22453 + }, + { + "epoch": 0.22454, + "grad_norm": 0.7011934518814087, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 22454 + }, + { + "epoch": 0.22455, + "grad_norm": 0.587223470211029, + "learning_rate": 0.003, + "loss": 3.967, + "step": 22455 + }, + { + "epoch": 0.22456, + "grad_norm": 0.6454535126686096, + "learning_rate": 0.003, + "loss": 3.998, + "step": 22456 + }, + { + "epoch": 0.22457, + "grad_norm": 0.7086803317070007, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 22457 + }, + { + "epoch": 0.22458, + "grad_norm": 0.7654173374176025, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 22458 + }, + { + "epoch": 0.22459, + "grad_norm": 0.8126182556152344, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 22459 + }, + { + "epoch": 0.2246, + "grad_norm": 1.0757782459259033, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 22460 + }, + { + "epoch": 0.22461, + "grad_norm": 1.185837745666504, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 22461 + }, + { + "epoch": 0.22462, + "grad_norm": 0.7935544848442078, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 22462 + }, + { + "epoch": 0.22463, + "grad_norm": 0.6941680312156677, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 22463 + }, + { + "epoch": 0.22464, + "grad_norm": 0.6664535999298096, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 22464 + }, + { + "epoch": 0.22465, + "grad_norm": 0.7434986233711243, + "learning_rate": 0.003, + "loss": 4.02, + "step": 22465 + }, + { + "epoch": 0.22466, + "grad_norm": 0.8151493072509766, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 22466 + }, + { + "epoch": 0.22467, + "grad_norm": 0.7463300824165344, + "learning_rate": 0.003, + "loss": 3.987, + "step": 22467 + }, + { + "epoch": 0.22468, + "grad_norm": 0.7675447463989258, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 22468 + }, + { + "epoch": 0.22469, + "grad_norm": 0.7666777968406677, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 22469 + }, + { + "epoch": 0.2247, + "grad_norm": 0.9437169432640076, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 22470 + }, + { + "epoch": 0.22471, + "grad_norm": 1.1254273653030396, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 22471 + }, + { + "epoch": 0.22472, + "grad_norm": 0.8957785964012146, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 22472 + }, + { + "epoch": 0.22473, + "grad_norm": 0.6806485056877136, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 22473 + }, + { + "epoch": 0.22474, + "grad_norm": 0.6468930244445801, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 22474 + }, + { + "epoch": 0.22475, + "grad_norm": 0.8046294450759888, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 22475 + }, + { + "epoch": 0.22476, + "grad_norm": 0.9049405455589294, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 22476 + }, + { + "epoch": 0.22477, + "grad_norm": 1.0358330011367798, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 22477 + }, + { + "epoch": 0.22478, + "grad_norm": 0.9369770884513855, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 22478 + }, + { + "epoch": 0.22479, + "grad_norm": 0.8336724042892456, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 22479 + }, + { + "epoch": 0.2248, + "grad_norm": 0.6661465167999268, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 22480 + }, + { + "epoch": 0.22481, + "grad_norm": 0.6457740068435669, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 22481 + }, + { + "epoch": 0.22482, + "grad_norm": 0.7267902493476868, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 22482 + }, + { + "epoch": 0.22483, + "grad_norm": 0.8097878098487854, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 22483 + }, + { + "epoch": 0.22484, + "grad_norm": 0.8831115365028381, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 22484 + }, + { + "epoch": 0.22485, + "grad_norm": 0.8672139644622803, + "learning_rate": 0.003, + "loss": 3.979, + "step": 22485 + }, + { + "epoch": 0.22486, + "grad_norm": 0.7510951161384583, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 22486 + }, + { + "epoch": 0.22487, + "grad_norm": 0.7795006632804871, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 22487 + }, + { + "epoch": 0.22488, + "grad_norm": 0.7788167595863342, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 22488 + }, + { + "epoch": 0.22489, + "grad_norm": 0.8402526378631592, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 22489 + }, + { + "epoch": 0.2249, + "grad_norm": 1.1594212055206299, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 22490 + }, + { + "epoch": 0.22491, + "grad_norm": 1.1323182582855225, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 22491 + }, + { + "epoch": 0.22492, + "grad_norm": 1.0545185804367065, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 22492 + }, + { + "epoch": 0.22493, + "grad_norm": 0.9478466510772705, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 22493 + }, + { + "epoch": 0.22494, + "grad_norm": 0.8053195476531982, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 22494 + }, + { + "epoch": 0.22495, + "grad_norm": 0.6407778859138489, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 22495 + }, + { + "epoch": 0.22496, + "grad_norm": 0.7275606989860535, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 22496 + }, + { + "epoch": 0.22497, + "grad_norm": 0.8413774371147156, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 22497 + }, + { + "epoch": 0.22498, + "grad_norm": 1.0912706851959229, + "learning_rate": 0.003, + "loss": 3.987, + "step": 22498 + }, + { + "epoch": 0.22499, + "grad_norm": 0.9036959409713745, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 22499 + }, + { + "epoch": 0.225, + "grad_norm": 0.7728074193000793, + "learning_rate": 0.003, + "loss": 3.988, + "step": 22500 + }, + { + "epoch": 0.22501, + "grad_norm": 0.7115880846977234, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 22501 + }, + { + "epoch": 0.22502, + "grad_norm": 0.8135708570480347, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 22502 + }, + { + "epoch": 0.22503, + "grad_norm": 0.945081353187561, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 22503 + }, + { + "epoch": 0.22504, + "grad_norm": 0.859524130821228, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 22504 + }, + { + "epoch": 0.22505, + "grad_norm": 0.7827999591827393, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 22505 + }, + { + "epoch": 0.22506, + "grad_norm": 0.797634482383728, + "learning_rate": 0.003, + "loss": 4.015, + "step": 22506 + }, + { + "epoch": 0.22507, + "grad_norm": 0.8643887042999268, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 22507 + }, + { + "epoch": 0.22508, + "grad_norm": 0.8752967715263367, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 22508 + }, + { + "epoch": 0.22509, + "grad_norm": 0.9563010334968567, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 22509 + }, + { + "epoch": 0.2251, + "grad_norm": 1.0486440658569336, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 22510 + }, + { + "epoch": 0.22511, + "grad_norm": 0.9055473208427429, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 22511 + }, + { + "epoch": 0.22512, + "grad_norm": 0.9977585673332214, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 22512 + }, + { + "epoch": 0.22513, + "grad_norm": 1.1328200101852417, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 22513 + }, + { + "epoch": 0.22514, + "grad_norm": 0.9681284427642822, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 22514 + }, + { + "epoch": 0.22515, + "grad_norm": 0.8296999335289001, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 22515 + }, + { + "epoch": 0.22516, + "grad_norm": 0.7772644758224487, + "learning_rate": 0.003, + "loss": 4.021, + "step": 22516 + }, + { + "epoch": 0.22517, + "grad_norm": 0.7846390604972839, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 22517 + }, + { + "epoch": 0.22518, + "grad_norm": 0.8850115537643433, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 22518 + }, + { + "epoch": 0.22519, + "grad_norm": 0.9973191022872925, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 22519 + }, + { + "epoch": 0.2252, + "grad_norm": 0.9508351683616638, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 22520 + }, + { + "epoch": 0.22521, + "grad_norm": 1.1023129224777222, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 22521 + }, + { + "epoch": 0.22522, + "grad_norm": 0.9027165174484253, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 22522 + }, + { + "epoch": 0.22523, + "grad_norm": 0.8210141658782959, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 22523 + }, + { + "epoch": 0.22524, + "grad_norm": 0.7370973825454712, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 22524 + }, + { + "epoch": 0.22525, + "grad_norm": 0.6930164694786072, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 22525 + }, + { + "epoch": 0.22526, + "grad_norm": 0.7075406908988953, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 22526 + }, + { + "epoch": 0.22527, + "grad_norm": 0.7073801159858704, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 22527 + }, + { + "epoch": 0.22528, + "grad_norm": 0.7473903298377991, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 22528 + }, + { + "epoch": 0.22529, + "grad_norm": 0.8731586337089539, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 22529 + }, + { + "epoch": 0.2253, + "grad_norm": 1.039387822151184, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 22530 + }, + { + "epoch": 0.22531, + "grad_norm": 1.1554207801818848, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 22531 + }, + { + "epoch": 0.22532, + "grad_norm": 0.6320087313652039, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 22532 + }, + { + "epoch": 0.22533, + "grad_norm": 0.6454406380653381, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 22533 + }, + { + "epoch": 0.22534, + "grad_norm": 0.8062059283256531, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 22534 + }, + { + "epoch": 0.22535, + "grad_norm": 0.9918100833892822, + "learning_rate": 0.003, + "loss": 4.011, + "step": 22535 + }, + { + "epoch": 0.22536, + "grad_norm": 1.3157193660736084, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 22536 + }, + { + "epoch": 0.22537, + "grad_norm": 0.5585388541221619, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 22537 + }, + { + "epoch": 0.22538, + "grad_norm": 0.5835403203964233, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 22538 + }, + { + "epoch": 0.22539, + "grad_norm": 0.7463785409927368, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 22539 + }, + { + "epoch": 0.2254, + "grad_norm": 0.948160707950592, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 22540 + }, + { + "epoch": 0.22541, + "grad_norm": 1.2967313528060913, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 22541 + }, + { + "epoch": 0.22542, + "grad_norm": 0.5992338061332703, + "learning_rate": 0.003, + "loss": 3.996, + "step": 22542 + }, + { + "epoch": 0.22543, + "grad_norm": 0.7315712571144104, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 22543 + }, + { + "epoch": 0.22544, + "grad_norm": 1.0605504512786865, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 22544 + }, + { + "epoch": 0.22545, + "grad_norm": 0.7975765466690063, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 22545 + }, + { + "epoch": 0.22546, + "grad_norm": 0.6689954996109009, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 22546 + }, + { + "epoch": 0.22547, + "grad_norm": 0.7156060934066772, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 22547 + }, + { + "epoch": 0.22548, + "grad_norm": 0.806228518486023, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 22548 + }, + { + "epoch": 0.22549, + "grad_norm": 0.9537408351898193, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 22549 + }, + { + "epoch": 0.2255, + "grad_norm": 1.0425735712051392, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 22550 + }, + { + "epoch": 0.22551, + "grad_norm": 1.0739209651947021, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 22551 + }, + { + "epoch": 0.22552, + "grad_norm": 1.0399507284164429, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 22552 + }, + { + "epoch": 0.22553, + "grad_norm": 0.9293482303619385, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 22553 + }, + { + "epoch": 0.22554, + "grad_norm": 0.8368396162986755, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 22554 + }, + { + "epoch": 0.22555, + "grad_norm": 0.9350905418395996, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 22555 + }, + { + "epoch": 0.22556, + "grad_norm": 0.90952467918396, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 22556 + }, + { + "epoch": 0.22557, + "grad_norm": 0.9364259243011475, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 22557 + }, + { + "epoch": 0.22558, + "grad_norm": 0.8886951804161072, + "learning_rate": 0.003, + "loss": 4.02, + "step": 22558 + }, + { + "epoch": 0.22559, + "grad_norm": 1.1178419589996338, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 22559 + }, + { + "epoch": 0.2256, + "grad_norm": 1.3153520822525024, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 22560 + }, + { + "epoch": 0.22561, + "grad_norm": 0.6973604559898376, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 22561 + }, + { + "epoch": 0.22562, + "grad_norm": 0.6641629338264465, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 22562 + }, + { + "epoch": 0.22563, + "grad_norm": 0.8292673826217651, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 22563 + }, + { + "epoch": 0.22564, + "grad_norm": 0.8302044868469238, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 22564 + }, + { + "epoch": 0.22565, + "grad_norm": 0.7561751008033752, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 22565 + }, + { + "epoch": 0.22566, + "grad_norm": 0.7331043481826782, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 22566 + }, + { + "epoch": 0.22567, + "grad_norm": 0.773192822933197, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 22567 + }, + { + "epoch": 0.22568, + "grad_norm": 0.7455373406410217, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 22568 + }, + { + "epoch": 0.22569, + "grad_norm": 0.6378149390220642, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 22569 + }, + { + "epoch": 0.2257, + "grad_norm": 0.5871608257293701, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 22570 + }, + { + "epoch": 0.22571, + "grad_norm": 0.6416295766830444, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 22571 + }, + { + "epoch": 0.22572, + "grad_norm": 0.6621878743171692, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 22572 + }, + { + "epoch": 0.22573, + "grad_norm": 0.7293329238891602, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 22573 + }, + { + "epoch": 0.22574, + "grad_norm": 0.6863351464271545, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 22574 + }, + { + "epoch": 0.22575, + "grad_norm": 0.6455560922622681, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 22575 + }, + { + "epoch": 0.22576, + "grad_norm": 0.7613818049430847, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 22576 + }, + { + "epoch": 0.22577, + "grad_norm": 0.8336321115493774, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 22577 + }, + { + "epoch": 0.22578, + "grad_norm": 0.9948531985282898, + "learning_rate": 0.003, + "loss": 3.992, + "step": 22578 + }, + { + "epoch": 0.22579, + "grad_norm": 1.1955511569976807, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 22579 + }, + { + "epoch": 0.2258, + "grad_norm": 0.9244723320007324, + "learning_rate": 0.003, + "loss": 3.972, + "step": 22580 + }, + { + "epoch": 0.22581, + "grad_norm": 1.098958969116211, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 22581 + }, + { + "epoch": 0.22582, + "grad_norm": 1.1332523822784424, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 22582 + }, + { + "epoch": 0.22583, + "grad_norm": 0.7345426678657532, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 22583 + }, + { + "epoch": 0.22584, + "grad_norm": 0.7212810516357422, + "learning_rate": 0.003, + "loss": 4.014, + "step": 22584 + }, + { + "epoch": 0.22585, + "grad_norm": 0.8274242281913757, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 22585 + }, + { + "epoch": 0.22586, + "grad_norm": 0.9139873385429382, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 22586 + }, + { + "epoch": 0.22587, + "grad_norm": 0.9039983749389648, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 22587 + }, + { + "epoch": 0.22588, + "grad_norm": 0.8439368009567261, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 22588 + }, + { + "epoch": 0.22589, + "grad_norm": 0.799583375453949, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 22589 + }, + { + "epoch": 0.2259, + "grad_norm": 0.7067766785621643, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 22590 + }, + { + "epoch": 0.22591, + "grad_norm": 0.7131248116493225, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 22591 + }, + { + "epoch": 0.22592, + "grad_norm": 0.7867693901062012, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 22592 + }, + { + "epoch": 0.22593, + "grad_norm": 0.7823998332023621, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 22593 + }, + { + "epoch": 0.22594, + "grad_norm": 0.7367347478866577, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 22594 + }, + { + "epoch": 0.22595, + "grad_norm": 0.7267398238182068, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 22595 + }, + { + "epoch": 0.22596, + "grad_norm": 0.8540565967559814, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 22596 + }, + { + "epoch": 0.22597, + "grad_norm": 0.793163001537323, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 22597 + }, + { + "epoch": 0.22598, + "grad_norm": 0.8111885786056519, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 22598 + }, + { + "epoch": 0.22599, + "grad_norm": 0.9221287369728088, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 22599 + }, + { + "epoch": 0.226, + "grad_norm": 0.9441787004470825, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 22600 + }, + { + "epoch": 0.22601, + "grad_norm": 0.914944589138031, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 22601 + }, + { + "epoch": 0.22602, + "grad_norm": 1.1398375034332275, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 22602 + }, + { + "epoch": 0.22603, + "grad_norm": 0.8702055215835571, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 22603 + }, + { + "epoch": 0.22604, + "grad_norm": 0.7077505588531494, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 22604 + }, + { + "epoch": 0.22605, + "grad_norm": 0.7599282264709473, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 22605 + }, + { + "epoch": 0.22606, + "grad_norm": 0.9103191494941711, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 22606 + }, + { + "epoch": 0.22607, + "grad_norm": 1.2925947904586792, + "learning_rate": 0.003, + "loss": 4.016, + "step": 22607 + }, + { + "epoch": 0.22608, + "grad_norm": 1.0025123357772827, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 22608 + }, + { + "epoch": 0.22609, + "grad_norm": 1.0659791231155396, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 22609 + }, + { + "epoch": 0.2261, + "grad_norm": 1.0477756261825562, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 22610 + }, + { + "epoch": 0.22611, + "grad_norm": 0.81010901927948, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 22611 + }, + { + "epoch": 0.22612, + "grad_norm": 0.7477598190307617, + "learning_rate": 0.003, + "loss": 4.001, + "step": 22612 + }, + { + "epoch": 0.22613, + "grad_norm": 0.7600870728492737, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 22613 + }, + { + "epoch": 0.22614, + "grad_norm": 0.7450019717216492, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 22614 + }, + { + "epoch": 0.22615, + "grad_norm": 0.7952964901924133, + "learning_rate": 0.003, + "loss": 3.998, + "step": 22615 + }, + { + "epoch": 0.22616, + "grad_norm": 0.834259033203125, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 22616 + }, + { + "epoch": 0.22617, + "grad_norm": 0.8939469456672668, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 22617 + }, + { + "epoch": 0.22618, + "grad_norm": 1.225689172744751, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 22618 + }, + { + "epoch": 0.22619, + "grad_norm": 0.7489709854125977, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 22619 + }, + { + "epoch": 0.2262, + "grad_norm": 0.6429768204689026, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 22620 + }, + { + "epoch": 0.22621, + "grad_norm": 0.9768158793449402, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 22621 + }, + { + "epoch": 0.22622, + "grad_norm": 1.1879620552062988, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 22622 + }, + { + "epoch": 0.22623, + "grad_norm": 0.9641862511634827, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 22623 + }, + { + "epoch": 0.22624, + "grad_norm": 0.8921433687210083, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 22624 + }, + { + "epoch": 0.22625, + "grad_norm": 0.8252694010734558, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 22625 + }, + { + "epoch": 0.22626, + "grad_norm": 0.8314810991287231, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 22626 + }, + { + "epoch": 0.22627, + "grad_norm": 0.849328339099884, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 22627 + }, + { + "epoch": 0.22628, + "grad_norm": 0.7866566777229309, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 22628 + }, + { + "epoch": 0.22629, + "grad_norm": 0.799117922782898, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 22629 + }, + { + "epoch": 0.2263, + "grad_norm": 0.8188341856002808, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 22630 + }, + { + "epoch": 0.22631, + "grad_norm": 0.8227394819259644, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 22631 + }, + { + "epoch": 0.22632, + "grad_norm": 0.7365055084228516, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 22632 + }, + { + "epoch": 0.22633, + "grad_norm": 0.7293735146522522, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 22633 + }, + { + "epoch": 0.22634, + "grad_norm": 0.912386953830719, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 22634 + }, + { + "epoch": 0.22635, + "grad_norm": 1.0821549892425537, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 22635 + }, + { + "epoch": 0.22636, + "grad_norm": 1.1439697742462158, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 22636 + }, + { + "epoch": 0.22637, + "grad_norm": 0.9225965142250061, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 22637 + }, + { + "epoch": 0.22638, + "grad_norm": 0.9950436353683472, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 22638 + }, + { + "epoch": 0.22639, + "grad_norm": 1.1409218311309814, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 22639 + }, + { + "epoch": 0.2264, + "grad_norm": 0.7746313214302063, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 22640 + }, + { + "epoch": 0.22641, + "grad_norm": 0.7244758009910583, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 22641 + }, + { + "epoch": 0.22642, + "grad_norm": 0.7932890057563782, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 22642 + }, + { + "epoch": 0.22643, + "grad_norm": 0.8649126887321472, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 22643 + }, + { + "epoch": 0.22644, + "grad_norm": 0.9749638438224792, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 22644 + }, + { + "epoch": 0.22645, + "grad_norm": 1.1931370496749878, + "learning_rate": 0.003, + "loss": 3.997, + "step": 22645 + }, + { + "epoch": 0.22646, + "grad_norm": 0.8668730854988098, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 22646 + }, + { + "epoch": 0.22647, + "grad_norm": 0.7980725765228271, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 22647 + }, + { + "epoch": 0.22648, + "grad_norm": 0.7940992712974548, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 22648 + }, + { + "epoch": 0.22649, + "grad_norm": 0.7490081191062927, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 22649 + }, + { + "epoch": 0.2265, + "grad_norm": 0.7087478041648865, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 22650 + }, + { + "epoch": 0.22651, + "grad_norm": 0.733660101890564, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 22651 + }, + { + "epoch": 0.22652, + "grad_norm": 0.7767938375473022, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 22652 + }, + { + "epoch": 0.22653, + "grad_norm": 0.846406877040863, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 22653 + }, + { + "epoch": 0.22654, + "grad_norm": 0.9157037138938904, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 22654 + }, + { + "epoch": 0.22655, + "grad_norm": 1.0050452947616577, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 22655 + }, + { + "epoch": 0.22656, + "grad_norm": 1.1236766576766968, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 22656 + }, + { + "epoch": 0.22657, + "grad_norm": 0.9150348901748657, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 22657 + }, + { + "epoch": 0.22658, + "grad_norm": 0.8400134444236755, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 22658 + }, + { + "epoch": 0.22659, + "grad_norm": 0.8557019233703613, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 22659 + }, + { + "epoch": 0.2266, + "grad_norm": 0.9143447875976562, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 22660 + }, + { + "epoch": 0.22661, + "grad_norm": 0.8481775522232056, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 22661 + }, + { + "epoch": 0.22662, + "grad_norm": 0.8349915146827698, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 22662 + }, + { + "epoch": 0.22663, + "grad_norm": 0.7645879983901978, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 22663 + }, + { + "epoch": 0.22664, + "grad_norm": 0.8313425183296204, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 22664 + }, + { + "epoch": 0.22665, + "grad_norm": 0.8810864686965942, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 22665 + }, + { + "epoch": 0.22666, + "grad_norm": 0.8931402564048767, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 22666 + }, + { + "epoch": 0.22667, + "grad_norm": 1.0830293893814087, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 22667 + }, + { + "epoch": 0.22668, + "grad_norm": 0.94843989610672, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 22668 + }, + { + "epoch": 0.22669, + "grad_norm": 0.7998262047767639, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 22669 + }, + { + "epoch": 0.2267, + "grad_norm": 0.737326443195343, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 22670 + }, + { + "epoch": 0.22671, + "grad_norm": 0.6808015704154968, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 22671 + }, + { + "epoch": 0.22672, + "grad_norm": 0.6738890409469604, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 22672 + }, + { + "epoch": 0.22673, + "grad_norm": 0.8552408218383789, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 22673 + }, + { + "epoch": 0.22674, + "grad_norm": 1.1433689594268799, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 22674 + }, + { + "epoch": 0.22675, + "grad_norm": 0.81049644947052, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 22675 + }, + { + "epoch": 0.22676, + "grad_norm": 0.5778343677520752, + "learning_rate": 0.003, + "loss": 3.978, + "step": 22676 + }, + { + "epoch": 0.22677, + "grad_norm": 0.550416886806488, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 22677 + }, + { + "epoch": 0.22678, + "grad_norm": 0.664819061756134, + "learning_rate": 0.003, + "loss": 3.993, + "step": 22678 + }, + { + "epoch": 0.22679, + "grad_norm": 0.7812240719795227, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 22679 + }, + { + "epoch": 0.2268, + "grad_norm": 0.8896509408950806, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 22680 + }, + { + "epoch": 0.22681, + "grad_norm": 1.0583137273788452, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 22681 + }, + { + "epoch": 0.22682, + "grad_norm": 0.7820581793785095, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 22682 + }, + { + "epoch": 0.22683, + "grad_norm": 0.8588399291038513, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 22683 + }, + { + "epoch": 0.22684, + "grad_norm": 1.0560506582260132, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 22684 + }, + { + "epoch": 0.22685, + "grad_norm": 1.0607281923294067, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 22685 + }, + { + "epoch": 0.22686, + "grad_norm": 1.0641882419586182, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 22686 + }, + { + "epoch": 0.22687, + "grad_norm": 0.8671996593475342, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 22687 + }, + { + "epoch": 0.22688, + "grad_norm": 0.7094112038612366, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 22688 + }, + { + "epoch": 0.22689, + "grad_norm": 0.8635150790214539, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 22689 + }, + { + "epoch": 0.2269, + "grad_norm": 0.8993562459945679, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 22690 + }, + { + "epoch": 0.22691, + "grad_norm": 0.8544433116912842, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 22691 + }, + { + "epoch": 0.22692, + "grad_norm": 0.8454415202140808, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 22692 + }, + { + "epoch": 0.22693, + "grad_norm": 0.7009323835372925, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 22693 + }, + { + "epoch": 0.22694, + "grad_norm": 0.8342335820198059, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 22694 + }, + { + "epoch": 0.22695, + "grad_norm": 0.8785995841026306, + "learning_rate": 0.003, + "loss": 3.99, + "step": 22695 + }, + { + "epoch": 0.22696, + "grad_norm": 0.8515464663505554, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 22696 + }, + { + "epoch": 0.22697, + "grad_norm": 0.9107221961021423, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 22697 + }, + { + "epoch": 0.22698, + "grad_norm": 1.0068321228027344, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 22698 + }, + { + "epoch": 0.22699, + "grad_norm": 1.1902743577957153, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 22699 + }, + { + "epoch": 0.227, + "grad_norm": 0.8712179660797119, + "learning_rate": 0.003, + "loss": 3.988, + "step": 22700 + }, + { + "epoch": 0.22701, + "grad_norm": 0.7888413071632385, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 22701 + }, + { + "epoch": 0.22702, + "grad_norm": 0.8043325543403625, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 22702 + }, + { + "epoch": 0.22703, + "grad_norm": 0.7365092039108276, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 22703 + }, + { + "epoch": 0.22704, + "grad_norm": 0.6359355449676514, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 22704 + }, + { + "epoch": 0.22705, + "grad_norm": 0.6768190860748291, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 22705 + }, + { + "epoch": 0.22706, + "grad_norm": 0.7089597582817078, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 22706 + }, + { + "epoch": 0.22707, + "grad_norm": 0.8444584012031555, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 22707 + }, + { + "epoch": 0.22708, + "grad_norm": 1.0524044036865234, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 22708 + }, + { + "epoch": 0.22709, + "grad_norm": 1.1379387378692627, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 22709 + }, + { + "epoch": 0.2271, + "grad_norm": 0.8187793493270874, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 22710 + }, + { + "epoch": 0.22711, + "grad_norm": 0.7972493767738342, + "learning_rate": 0.003, + "loss": 3.999, + "step": 22711 + }, + { + "epoch": 0.22712, + "grad_norm": 1.0039690732955933, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 22712 + }, + { + "epoch": 0.22713, + "grad_norm": 1.1437420845031738, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 22713 + }, + { + "epoch": 0.22714, + "grad_norm": 0.9905338883399963, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 22714 + }, + { + "epoch": 0.22715, + "grad_norm": 0.9183775186538696, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 22715 + }, + { + "epoch": 0.22716, + "grad_norm": 1.083058476448059, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 22716 + }, + { + "epoch": 0.22717, + "grad_norm": 0.9954180121421814, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 22717 + }, + { + "epoch": 0.22718, + "grad_norm": 1.0120686292648315, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 22718 + }, + { + "epoch": 0.22719, + "grad_norm": 1.0972919464111328, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 22719 + }, + { + "epoch": 0.2272, + "grad_norm": 1.0402734279632568, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 22720 + }, + { + "epoch": 0.22721, + "grad_norm": 1.0250205993652344, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 22721 + }, + { + "epoch": 0.22722, + "grad_norm": 1.0363314151763916, + "learning_rate": 0.003, + "loss": 4.02, + "step": 22722 + }, + { + "epoch": 0.22723, + "grad_norm": 1.1491438150405884, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 22723 + }, + { + "epoch": 0.22724, + "grad_norm": 1.124851942062378, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 22724 + }, + { + "epoch": 0.22725, + "grad_norm": 0.9142146706581116, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 22725 + }, + { + "epoch": 0.22726, + "grad_norm": 1.017228126525879, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 22726 + }, + { + "epoch": 0.22727, + "grad_norm": 1.1621403694152832, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 22727 + }, + { + "epoch": 0.22728, + "grad_norm": 0.7189894318580627, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 22728 + }, + { + "epoch": 0.22729, + "grad_norm": 0.6909321546554565, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 22729 + }, + { + "epoch": 0.2273, + "grad_norm": 0.8472306728363037, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 22730 + }, + { + "epoch": 0.22731, + "grad_norm": 0.8492780923843384, + "learning_rate": 0.003, + "loss": 3.999, + "step": 22731 + }, + { + "epoch": 0.22732, + "grad_norm": 0.8895362019538879, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 22732 + }, + { + "epoch": 0.22733, + "grad_norm": 0.9408416152000427, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 22733 + }, + { + "epoch": 0.22734, + "grad_norm": 0.949094831943512, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 22734 + }, + { + "epoch": 0.22735, + "grad_norm": 0.8157100081443787, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 22735 + }, + { + "epoch": 0.22736, + "grad_norm": 0.7857760787010193, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 22736 + }, + { + "epoch": 0.22737, + "grad_norm": 0.8392558693885803, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 22737 + }, + { + "epoch": 0.22738, + "grad_norm": 0.7167733311653137, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 22738 + }, + { + "epoch": 0.22739, + "grad_norm": 0.6167979836463928, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 22739 + }, + { + "epoch": 0.2274, + "grad_norm": 0.6920311450958252, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 22740 + }, + { + "epoch": 0.22741, + "grad_norm": 0.7541581988334656, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 22741 + }, + { + "epoch": 0.22742, + "grad_norm": 0.8818638324737549, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 22742 + }, + { + "epoch": 0.22743, + "grad_norm": 1.0702123641967773, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 22743 + }, + { + "epoch": 0.22744, + "grad_norm": 0.868956446647644, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 22744 + }, + { + "epoch": 0.22745, + "grad_norm": 0.6465038657188416, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 22745 + }, + { + "epoch": 0.22746, + "grad_norm": 0.5570217967033386, + "learning_rate": 0.003, + "loss": 4.0, + "step": 22746 + }, + { + "epoch": 0.22747, + "grad_norm": 0.6476839780807495, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 22747 + }, + { + "epoch": 0.22748, + "grad_norm": 0.8337581157684326, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 22748 + }, + { + "epoch": 0.22749, + "grad_norm": 0.9499635696411133, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 22749 + }, + { + "epoch": 0.2275, + "grad_norm": 1.0580387115478516, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 22750 + }, + { + "epoch": 0.22751, + "grad_norm": 0.788001537322998, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 22751 + }, + { + "epoch": 0.22752, + "grad_norm": 0.6245743036270142, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 22752 + }, + { + "epoch": 0.22753, + "grad_norm": 0.6570910215377808, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 22753 + }, + { + "epoch": 0.22754, + "grad_norm": 0.7587313055992126, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 22754 + }, + { + "epoch": 0.22755, + "grad_norm": 0.8080280423164368, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 22755 + }, + { + "epoch": 0.22756, + "grad_norm": 0.709997296333313, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 22756 + }, + { + "epoch": 0.22757, + "grad_norm": 0.6425041556358337, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 22757 + }, + { + "epoch": 0.22758, + "grad_norm": 0.6960175633430481, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 22758 + }, + { + "epoch": 0.22759, + "grad_norm": 0.8068851828575134, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 22759 + }, + { + "epoch": 0.2276, + "grad_norm": 0.9725735783576965, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 22760 + }, + { + "epoch": 0.22761, + "grad_norm": 1.104339361190796, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 22761 + }, + { + "epoch": 0.22762, + "grad_norm": 0.912956953048706, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 22762 + }, + { + "epoch": 0.22763, + "grad_norm": 0.8380911946296692, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 22763 + }, + { + "epoch": 0.22764, + "grad_norm": 0.8706415295600891, + "learning_rate": 0.003, + "loss": 4.013, + "step": 22764 + }, + { + "epoch": 0.22765, + "grad_norm": 1.014489769935608, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 22765 + }, + { + "epoch": 0.22766, + "grad_norm": 0.9416565895080566, + "learning_rate": 0.003, + "loss": 4.008, + "step": 22766 + }, + { + "epoch": 0.22767, + "grad_norm": 0.8618533611297607, + "learning_rate": 0.003, + "loss": 3.996, + "step": 22767 + }, + { + "epoch": 0.22768, + "grad_norm": 0.8966159820556641, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 22768 + }, + { + "epoch": 0.22769, + "grad_norm": 1.139390468597412, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 22769 + }, + { + "epoch": 0.2277, + "grad_norm": 1.0194278955459595, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 22770 + }, + { + "epoch": 0.22771, + "grad_norm": 0.9959157109260559, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 22771 + }, + { + "epoch": 0.22772, + "grad_norm": 1.0136313438415527, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 22772 + }, + { + "epoch": 0.22773, + "grad_norm": 0.9768470525741577, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 22773 + }, + { + "epoch": 0.22774, + "grad_norm": 0.9392907023429871, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 22774 + }, + { + "epoch": 0.22775, + "grad_norm": 0.9634270071983337, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 22775 + }, + { + "epoch": 0.22776, + "grad_norm": 1.1525689363479614, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 22776 + }, + { + "epoch": 0.22777, + "grad_norm": 0.8796603679656982, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 22777 + }, + { + "epoch": 0.22778, + "grad_norm": 0.8567058444023132, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 22778 + }, + { + "epoch": 0.22779, + "grad_norm": 0.9193599224090576, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 22779 + }, + { + "epoch": 0.2278, + "grad_norm": 0.915293276309967, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 22780 + }, + { + "epoch": 0.22781, + "grad_norm": 0.8134163022041321, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 22781 + }, + { + "epoch": 0.22782, + "grad_norm": 0.7062757611274719, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 22782 + }, + { + "epoch": 0.22783, + "grad_norm": 0.6965892910957336, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 22783 + }, + { + "epoch": 0.22784, + "grad_norm": 0.7337034940719604, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 22784 + }, + { + "epoch": 0.22785, + "grad_norm": 0.8737407326698303, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 22785 + }, + { + "epoch": 0.22786, + "grad_norm": 0.8701823949813843, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 22786 + }, + { + "epoch": 0.22787, + "grad_norm": 0.8549156188964844, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 22787 + }, + { + "epoch": 0.22788, + "grad_norm": 0.9004518985748291, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 22788 + }, + { + "epoch": 0.22789, + "grad_norm": 0.9349706768989563, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 22789 + }, + { + "epoch": 0.2279, + "grad_norm": 0.7645196318626404, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 22790 + }, + { + "epoch": 0.22791, + "grad_norm": 0.8005019426345825, + "learning_rate": 0.003, + "loss": 4.008, + "step": 22791 + }, + { + "epoch": 0.22792, + "grad_norm": 0.7448391318321228, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 22792 + }, + { + "epoch": 0.22793, + "grad_norm": 0.670491099357605, + "learning_rate": 0.003, + "loss": 3.966, + "step": 22793 + }, + { + "epoch": 0.22794, + "grad_norm": 0.6934297680854797, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 22794 + }, + { + "epoch": 0.22795, + "grad_norm": 0.7491574883460999, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 22795 + }, + { + "epoch": 0.22796, + "grad_norm": 0.8810040950775146, + "learning_rate": 0.003, + "loss": 4.0, + "step": 22796 + }, + { + "epoch": 0.22797, + "grad_norm": 0.9166193604469299, + "learning_rate": 0.003, + "loss": 4.0, + "step": 22797 + }, + { + "epoch": 0.22798, + "grad_norm": 0.9302740693092346, + "learning_rate": 0.003, + "loss": 4.016, + "step": 22798 + }, + { + "epoch": 0.22799, + "grad_norm": 1.1510775089263916, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 22799 + }, + { + "epoch": 0.228, + "grad_norm": 1.1456074714660645, + "learning_rate": 0.003, + "loss": 3.992, + "step": 22800 + }, + { + "epoch": 0.22801, + "grad_norm": 0.7626516819000244, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 22801 + }, + { + "epoch": 0.22802, + "grad_norm": 0.7304949760437012, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 22802 + }, + { + "epoch": 0.22803, + "grad_norm": 0.7148295044898987, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 22803 + }, + { + "epoch": 0.22804, + "grad_norm": 0.8273639678955078, + "learning_rate": 0.003, + "loss": 3.998, + "step": 22804 + }, + { + "epoch": 0.22805, + "grad_norm": 0.754592776298523, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 22805 + }, + { + "epoch": 0.22806, + "grad_norm": 0.6627060174942017, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 22806 + }, + { + "epoch": 0.22807, + "grad_norm": 0.5853172540664673, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 22807 + }, + { + "epoch": 0.22808, + "grad_norm": 0.6546884775161743, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 22808 + }, + { + "epoch": 0.22809, + "grad_norm": 0.6860325336456299, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 22809 + }, + { + "epoch": 0.2281, + "grad_norm": 0.8244639039039612, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 22810 + }, + { + "epoch": 0.22811, + "grad_norm": 0.994722843170166, + "learning_rate": 0.003, + "loss": 3.996, + "step": 22811 + }, + { + "epoch": 0.22812, + "grad_norm": 1.1154168844223022, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 22812 + }, + { + "epoch": 0.22813, + "grad_norm": 0.8840178847312927, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 22813 + }, + { + "epoch": 0.22814, + "grad_norm": 0.737061619758606, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 22814 + }, + { + "epoch": 0.22815, + "grad_norm": 0.6754586696624756, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 22815 + }, + { + "epoch": 0.22816, + "grad_norm": 0.6540216207504272, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 22816 + }, + { + "epoch": 0.22817, + "grad_norm": 0.7404739856719971, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 22817 + }, + { + "epoch": 0.22818, + "grad_norm": 0.8696702122688293, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 22818 + }, + { + "epoch": 0.22819, + "grad_norm": 0.9527407288551331, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 22819 + }, + { + "epoch": 0.2282, + "grad_norm": 1.0084891319274902, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 22820 + }, + { + "epoch": 0.22821, + "grad_norm": 0.9716843962669373, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 22821 + }, + { + "epoch": 0.22822, + "grad_norm": 0.9331386685371399, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 22822 + }, + { + "epoch": 0.22823, + "grad_norm": 0.9431948065757751, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 22823 + }, + { + "epoch": 0.22824, + "grad_norm": 1.048907995223999, + "learning_rate": 0.003, + "loss": 4.005, + "step": 22824 + }, + { + "epoch": 0.22825, + "grad_norm": 1.0735005140304565, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 22825 + }, + { + "epoch": 0.22826, + "grad_norm": 1.1868399381637573, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 22826 + }, + { + "epoch": 0.22827, + "grad_norm": 0.9685457944869995, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 22827 + }, + { + "epoch": 0.22828, + "grad_norm": 0.8449652194976807, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 22828 + }, + { + "epoch": 0.22829, + "grad_norm": 0.8988775014877319, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 22829 + }, + { + "epoch": 0.2283, + "grad_norm": 1.0099470615386963, + "learning_rate": 0.003, + "loss": 4.005, + "step": 22830 + }, + { + "epoch": 0.22831, + "grad_norm": 0.9623905420303345, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 22831 + }, + { + "epoch": 0.22832, + "grad_norm": 0.8183230757713318, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 22832 + }, + { + "epoch": 0.22833, + "grad_norm": 0.7908297777175903, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 22833 + }, + { + "epoch": 0.22834, + "grad_norm": 0.7866353392601013, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 22834 + }, + { + "epoch": 0.22835, + "grad_norm": 0.7895807027816772, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 22835 + }, + { + "epoch": 0.22836, + "grad_norm": 0.7941407561302185, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 22836 + }, + { + "epoch": 0.22837, + "grad_norm": 0.8204095959663391, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 22837 + }, + { + "epoch": 0.22838, + "grad_norm": 0.8785392642021179, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 22838 + }, + { + "epoch": 0.22839, + "grad_norm": 0.8239431977272034, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 22839 + }, + { + "epoch": 0.2284, + "grad_norm": 0.6744070649147034, + "learning_rate": 0.003, + "loss": 4.0, + "step": 22840 + }, + { + "epoch": 0.22841, + "grad_norm": 0.7500600218772888, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 22841 + }, + { + "epoch": 0.22842, + "grad_norm": 0.7995817065238953, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 22842 + }, + { + "epoch": 0.22843, + "grad_norm": 0.9434783458709717, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 22843 + }, + { + "epoch": 0.22844, + "grad_norm": 1.1756136417388916, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 22844 + }, + { + "epoch": 0.22845, + "grad_norm": 0.9171916246414185, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 22845 + }, + { + "epoch": 0.22846, + "grad_norm": 0.879538357257843, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 22846 + }, + { + "epoch": 0.22847, + "grad_norm": 0.9054990410804749, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 22847 + }, + { + "epoch": 0.22848, + "grad_norm": 0.9794584512710571, + "learning_rate": 0.003, + "loss": 4.035, + "step": 22848 + }, + { + "epoch": 0.22849, + "grad_norm": 0.987642228603363, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 22849 + }, + { + "epoch": 0.2285, + "grad_norm": 0.9297903180122375, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 22850 + }, + { + "epoch": 0.22851, + "grad_norm": 0.864916205406189, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 22851 + }, + { + "epoch": 0.22852, + "grad_norm": 0.8236303329467773, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 22852 + }, + { + "epoch": 0.22853, + "grad_norm": 0.8080064058303833, + "learning_rate": 0.003, + "loss": 3.997, + "step": 22853 + }, + { + "epoch": 0.22854, + "grad_norm": 0.7594829797744751, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 22854 + }, + { + "epoch": 0.22855, + "grad_norm": 0.6762077808380127, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 22855 + }, + { + "epoch": 0.22856, + "grad_norm": 0.6480247974395752, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 22856 + }, + { + "epoch": 0.22857, + "grad_norm": 0.7744237780570984, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 22857 + }, + { + "epoch": 0.22858, + "grad_norm": 0.9318351745605469, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 22858 + }, + { + "epoch": 0.22859, + "grad_norm": 1.1502577066421509, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 22859 + }, + { + "epoch": 0.2286, + "grad_norm": 0.9954847097396851, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 22860 + }, + { + "epoch": 0.22861, + "grad_norm": 0.769961416721344, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 22861 + }, + { + "epoch": 0.22862, + "grad_norm": 0.6278361082077026, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 22862 + }, + { + "epoch": 0.22863, + "grad_norm": 0.6613967418670654, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 22863 + }, + { + "epoch": 0.22864, + "grad_norm": 0.7463530898094177, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 22864 + }, + { + "epoch": 0.22865, + "grad_norm": 0.7999584674835205, + "learning_rate": 0.003, + "loss": 3.978, + "step": 22865 + }, + { + "epoch": 0.22866, + "grad_norm": 0.9065727591514587, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 22866 + }, + { + "epoch": 0.22867, + "grad_norm": 0.9896615147590637, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 22867 + }, + { + "epoch": 0.22868, + "grad_norm": 1.0216121673583984, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 22868 + }, + { + "epoch": 0.22869, + "grad_norm": 1.0427114963531494, + "learning_rate": 0.003, + "loss": 4.007, + "step": 22869 + }, + { + "epoch": 0.2287, + "grad_norm": 0.9338259696960449, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 22870 + }, + { + "epoch": 0.22871, + "grad_norm": 0.8644422292709351, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 22871 + }, + { + "epoch": 0.22872, + "grad_norm": 0.7470993399620056, + "learning_rate": 0.003, + "loss": 3.968, + "step": 22872 + }, + { + "epoch": 0.22873, + "grad_norm": 0.6022185683250427, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 22873 + }, + { + "epoch": 0.22874, + "grad_norm": 0.6354525089263916, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 22874 + }, + { + "epoch": 0.22875, + "grad_norm": 0.6285504102706909, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 22875 + }, + { + "epoch": 0.22876, + "grad_norm": 0.6668216586112976, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 22876 + }, + { + "epoch": 0.22877, + "grad_norm": 0.6569263935089111, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 22877 + }, + { + "epoch": 0.22878, + "grad_norm": 0.6664939522743225, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 22878 + }, + { + "epoch": 0.22879, + "grad_norm": 0.6565495729446411, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 22879 + }, + { + "epoch": 0.2288, + "grad_norm": 0.7892429232597351, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 22880 + }, + { + "epoch": 0.22881, + "grad_norm": 0.9769039154052734, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 22881 + }, + { + "epoch": 0.22882, + "grad_norm": 1.111923336982727, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 22882 + }, + { + "epoch": 0.22883, + "grad_norm": 1.0032240152359009, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 22883 + }, + { + "epoch": 0.22884, + "grad_norm": 1.4658493995666504, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 22884 + }, + { + "epoch": 0.22885, + "grad_norm": 0.7581064105033875, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 22885 + }, + { + "epoch": 0.22886, + "grad_norm": 0.6206052303314209, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 22886 + }, + { + "epoch": 0.22887, + "grad_norm": 0.6585862636566162, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 22887 + }, + { + "epoch": 0.22888, + "grad_norm": 0.7490751147270203, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 22888 + }, + { + "epoch": 0.22889, + "grad_norm": 0.9168391227722168, + "learning_rate": 0.003, + "loss": 4.006, + "step": 22889 + }, + { + "epoch": 0.2289, + "grad_norm": 1.071951985359192, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 22890 + }, + { + "epoch": 0.22891, + "grad_norm": 0.7684674859046936, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 22891 + }, + { + "epoch": 0.22892, + "grad_norm": 0.559437096118927, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 22892 + }, + { + "epoch": 0.22893, + "grad_norm": 0.6369199156761169, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 22893 + }, + { + "epoch": 0.22894, + "grad_norm": 0.7140424251556396, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 22894 + }, + { + "epoch": 0.22895, + "grad_norm": 0.7683032751083374, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 22895 + }, + { + "epoch": 0.22896, + "grad_norm": 0.8863502144813538, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 22896 + }, + { + "epoch": 0.22897, + "grad_norm": 1.0554530620574951, + "learning_rate": 0.003, + "loss": 3.963, + "step": 22897 + }, + { + "epoch": 0.22898, + "grad_norm": 0.8525095582008362, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 22898 + }, + { + "epoch": 0.22899, + "grad_norm": 0.7730855941772461, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 22899 + }, + { + "epoch": 0.229, + "grad_norm": 0.9278606176376343, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 22900 + }, + { + "epoch": 0.22901, + "grad_norm": 0.8681670427322388, + "learning_rate": 0.003, + "loss": 3.982, + "step": 22901 + }, + { + "epoch": 0.22902, + "grad_norm": 0.8975440263748169, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 22902 + }, + { + "epoch": 0.22903, + "grad_norm": 0.8295832872390747, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 22903 + }, + { + "epoch": 0.22904, + "grad_norm": 0.7885738611221313, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 22904 + }, + { + "epoch": 0.22905, + "grad_norm": 0.7374867796897888, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 22905 + }, + { + "epoch": 0.22906, + "grad_norm": 0.8003861904144287, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 22906 + }, + { + "epoch": 0.22907, + "grad_norm": 1.015002965927124, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 22907 + }, + { + "epoch": 0.22908, + "grad_norm": 1.2798688411712646, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 22908 + }, + { + "epoch": 0.22909, + "grad_norm": 0.6066967248916626, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 22909 + }, + { + "epoch": 0.2291, + "grad_norm": 0.7699502110481262, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 22910 + }, + { + "epoch": 0.22911, + "grad_norm": 0.9835498332977295, + "learning_rate": 0.003, + "loss": 3.986, + "step": 22911 + }, + { + "epoch": 0.22912, + "grad_norm": 1.0399408340454102, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 22912 + }, + { + "epoch": 0.22913, + "grad_norm": 0.9194453954696655, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 22913 + }, + { + "epoch": 0.22914, + "grad_norm": 0.9842389822006226, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 22914 + }, + { + "epoch": 0.22915, + "grad_norm": 0.9452569484710693, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 22915 + }, + { + "epoch": 0.22916, + "grad_norm": 1.0009175539016724, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 22916 + }, + { + "epoch": 0.22917, + "grad_norm": 1.2031450271606445, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 22917 + }, + { + "epoch": 0.22918, + "grad_norm": 0.7978301644325256, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 22918 + }, + { + "epoch": 0.22919, + "grad_norm": 0.6478793621063232, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 22919 + }, + { + "epoch": 0.2292, + "grad_norm": 0.6479148864746094, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 22920 + }, + { + "epoch": 0.22921, + "grad_norm": 0.6559903621673584, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 22921 + }, + { + "epoch": 0.22922, + "grad_norm": 0.6622194051742554, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 22922 + }, + { + "epoch": 0.22923, + "grad_norm": 0.6873556971549988, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 22923 + }, + { + "epoch": 0.22924, + "grad_norm": 0.6897494196891785, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 22924 + }, + { + "epoch": 0.22925, + "grad_norm": 0.5513507723808289, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 22925 + }, + { + "epoch": 0.22926, + "grad_norm": 0.5068768858909607, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 22926 + }, + { + "epoch": 0.22927, + "grad_norm": 0.4971695840358734, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 22927 + }, + { + "epoch": 0.22928, + "grad_norm": 0.5085691213607788, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 22928 + }, + { + "epoch": 0.22929, + "grad_norm": 0.604036271572113, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 22929 + }, + { + "epoch": 0.2293, + "grad_norm": 0.7915269732475281, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 22930 + }, + { + "epoch": 0.22931, + "grad_norm": 1.1088626384735107, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 22931 + }, + { + "epoch": 0.22932, + "grad_norm": 1.2415170669555664, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 22932 + }, + { + "epoch": 0.22933, + "grad_norm": 0.7159534692764282, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 22933 + }, + { + "epoch": 0.22934, + "grad_norm": 0.6303723454475403, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 22934 + }, + { + "epoch": 0.22935, + "grad_norm": 0.7068621516227722, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 22935 + }, + { + "epoch": 0.22936, + "grad_norm": 0.918329656124115, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 22936 + }, + { + "epoch": 0.22937, + "grad_norm": 1.1570148468017578, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 22937 + }, + { + "epoch": 0.22938, + "grad_norm": 1.1038559675216675, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 22938 + }, + { + "epoch": 0.22939, + "grad_norm": 0.8516200184822083, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 22939 + }, + { + "epoch": 0.2294, + "grad_norm": 0.7503311634063721, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 22940 + }, + { + "epoch": 0.22941, + "grad_norm": 0.8286864757537842, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 22941 + }, + { + "epoch": 0.22942, + "grad_norm": 0.9053454995155334, + "learning_rate": 0.003, + "loss": 3.957, + "step": 22942 + }, + { + "epoch": 0.22943, + "grad_norm": 1.110000491142273, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 22943 + }, + { + "epoch": 0.22944, + "grad_norm": 0.8126286864280701, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 22944 + }, + { + "epoch": 0.22945, + "grad_norm": 0.6649305820465088, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 22945 + }, + { + "epoch": 0.22946, + "grad_norm": 0.7056620121002197, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 22946 + }, + { + "epoch": 0.22947, + "grad_norm": 0.6882185339927673, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 22947 + }, + { + "epoch": 0.22948, + "grad_norm": 0.7334792613983154, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 22948 + }, + { + "epoch": 0.22949, + "grad_norm": 0.8187629580497742, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 22949 + }, + { + "epoch": 0.2295, + "grad_norm": 0.8649359345436096, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 22950 + }, + { + "epoch": 0.22951, + "grad_norm": 0.8981301784515381, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 22951 + }, + { + "epoch": 0.22952, + "grad_norm": 0.9364953637123108, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 22952 + }, + { + "epoch": 0.22953, + "grad_norm": 0.9543210864067078, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 22953 + }, + { + "epoch": 0.22954, + "grad_norm": 0.939450204372406, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 22954 + }, + { + "epoch": 0.22955, + "grad_norm": 0.8398085236549377, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 22955 + }, + { + "epoch": 0.22956, + "grad_norm": 0.8824459314346313, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 22956 + }, + { + "epoch": 0.22957, + "grad_norm": 1.0881253480911255, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 22957 + }, + { + "epoch": 0.22958, + "grad_norm": 1.4197942018508911, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 22958 + }, + { + "epoch": 0.22959, + "grad_norm": 0.7080182433128357, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 22959 + }, + { + "epoch": 0.2296, + "grad_norm": 0.7154512405395508, + "learning_rate": 0.003, + "loss": 3.9463, + "step": 22960 + }, + { + "epoch": 0.22961, + "grad_norm": 0.8434643149375916, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 22961 + }, + { + "epoch": 0.22962, + "grad_norm": 1.0162034034729004, + "learning_rate": 0.003, + "loss": 4.019, + "step": 22962 + }, + { + "epoch": 0.22963, + "grad_norm": 1.2151683568954468, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 22963 + }, + { + "epoch": 0.22964, + "grad_norm": 0.9526553153991699, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 22964 + }, + { + "epoch": 0.22965, + "grad_norm": 0.8533622622489929, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 22965 + }, + { + "epoch": 0.22966, + "grad_norm": 0.8567134141921997, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 22966 + }, + { + "epoch": 0.22967, + "grad_norm": 1.0433229207992554, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 22967 + }, + { + "epoch": 0.22968, + "grad_norm": 1.082944393157959, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 22968 + }, + { + "epoch": 0.22969, + "grad_norm": 0.9346987009048462, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 22969 + }, + { + "epoch": 0.2297, + "grad_norm": 0.9211357235908508, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 22970 + }, + { + "epoch": 0.22971, + "grad_norm": 0.8701499104499817, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 22971 + }, + { + "epoch": 0.22972, + "grad_norm": 0.8525499105453491, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 22972 + }, + { + "epoch": 0.22973, + "grad_norm": 0.920916736125946, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 22973 + }, + { + "epoch": 0.22974, + "grad_norm": 1.0573474168777466, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 22974 + }, + { + "epoch": 0.22975, + "grad_norm": 0.8372592926025391, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 22975 + }, + { + "epoch": 0.22976, + "grad_norm": 0.9291589856147766, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 22976 + }, + { + "epoch": 0.22977, + "grad_norm": 0.9978088736534119, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 22977 + }, + { + "epoch": 0.22978, + "grad_norm": 1.2472418546676636, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 22978 + }, + { + "epoch": 0.22979, + "grad_norm": 0.6479096412658691, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 22979 + }, + { + "epoch": 0.2298, + "grad_norm": 0.6380423903465271, + "learning_rate": 0.003, + "loss": 4.015, + "step": 22980 + }, + { + "epoch": 0.22981, + "grad_norm": 0.6899369955062866, + "learning_rate": 0.003, + "loss": 4.027, + "step": 22981 + }, + { + "epoch": 0.22982, + "grad_norm": 0.7871737480163574, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 22982 + }, + { + "epoch": 0.22983, + "grad_norm": 0.855445921421051, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 22983 + }, + { + "epoch": 0.22984, + "grad_norm": 0.9433035850524902, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 22984 + }, + { + "epoch": 0.22985, + "grad_norm": 1.1320658922195435, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 22985 + }, + { + "epoch": 0.22986, + "grad_norm": 1.0430923700332642, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 22986 + }, + { + "epoch": 0.22987, + "grad_norm": 0.8788118958473206, + "learning_rate": 0.003, + "loss": 4.005, + "step": 22987 + }, + { + "epoch": 0.22988, + "grad_norm": 0.7646600604057312, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 22988 + }, + { + "epoch": 0.22989, + "grad_norm": 0.6964078545570374, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 22989 + }, + { + "epoch": 0.2299, + "grad_norm": 0.7160056829452515, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 22990 + }, + { + "epoch": 0.22991, + "grad_norm": 0.7959632873535156, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 22991 + }, + { + "epoch": 0.22992, + "grad_norm": 1.1392194032669067, + "learning_rate": 0.003, + "loss": 3.987, + "step": 22992 + }, + { + "epoch": 0.22993, + "grad_norm": 0.948128342628479, + "learning_rate": 0.003, + "loss": 4.034, + "step": 22993 + }, + { + "epoch": 0.22994, + "grad_norm": 0.7246876358985901, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 22994 + }, + { + "epoch": 0.22995, + "grad_norm": 0.5793988704681396, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 22995 + }, + { + "epoch": 0.22996, + "grad_norm": 0.6376774907112122, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 22996 + }, + { + "epoch": 0.22997, + "grad_norm": 0.660190224647522, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 22997 + }, + { + "epoch": 0.22998, + "grad_norm": 0.6963244676589966, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 22998 + }, + { + "epoch": 0.22999, + "grad_norm": 0.7412117123603821, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 22999 + }, + { + "epoch": 0.23, + "grad_norm": 0.8455507159233093, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 23000 + }, + { + "epoch": 0.23001, + "grad_norm": 1.042688250541687, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 23001 + }, + { + "epoch": 0.23002, + "grad_norm": 0.8959544897079468, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 23002 + }, + { + "epoch": 0.23003, + "grad_norm": 0.720615029335022, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 23003 + }, + { + "epoch": 0.23004, + "grad_norm": 0.7091113328933716, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 23004 + }, + { + "epoch": 0.23005, + "grad_norm": 0.6828072667121887, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 23005 + }, + { + "epoch": 0.23006, + "grad_norm": 0.7758764624595642, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 23006 + }, + { + "epoch": 0.23007, + "grad_norm": 0.708320677280426, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 23007 + }, + { + "epoch": 0.23008, + "grad_norm": 0.7290263772010803, + "learning_rate": 0.003, + "loss": 3.996, + "step": 23008 + }, + { + "epoch": 0.23009, + "grad_norm": 0.6381104588508606, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 23009 + }, + { + "epoch": 0.2301, + "grad_norm": 0.6442878246307373, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 23010 + }, + { + "epoch": 0.23011, + "grad_norm": 0.7125610709190369, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 23011 + }, + { + "epoch": 0.23012, + "grad_norm": 0.7132952809333801, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 23012 + }, + { + "epoch": 0.23013, + "grad_norm": 0.9006332159042358, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 23013 + }, + { + "epoch": 0.23014, + "grad_norm": 1.1828340291976929, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 23014 + }, + { + "epoch": 0.23015, + "grad_norm": 1.0224099159240723, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 23015 + }, + { + "epoch": 0.23016, + "grad_norm": 1.089600920677185, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 23016 + }, + { + "epoch": 0.23017, + "grad_norm": 1.0581661462783813, + "learning_rate": 0.003, + "loss": 4.012, + "step": 23017 + }, + { + "epoch": 0.23018, + "grad_norm": 0.862858235836029, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 23018 + }, + { + "epoch": 0.23019, + "grad_norm": 0.7663451433181763, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 23019 + }, + { + "epoch": 0.2302, + "grad_norm": 0.8006938695907593, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 23020 + }, + { + "epoch": 0.23021, + "grad_norm": 1.0573605298995972, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 23021 + }, + { + "epoch": 0.23022, + "grad_norm": 1.3617256879806519, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 23022 + }, + { + "epoch": 0.23023, + "grad_norm": 0.732063353061676, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 23023 + }, + { + "epoch": 0.23024, + "grad_norm": 0.7377094030380249, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 23024 + }, + { + "epoch": 0.23025, + "grad_norm": 0.8654947876930237, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 23025 + }, + { + "epoch": 0.23026, + "grad_norm": 0.9887498617172241, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 23026 + }, + { + "epoch": 0.23027, + "grad_norm": 1.171537160873413, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 23027 + }, + { + "epoch": 0.23028, + "grad_norm": 0.8718369603157043, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 23028 + }, + { + "epoch": 0.23029, + "grad_norm": 0.8021326065063477, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 23029 + }, + { + "epoch": 0.2303, + "grad_norm": 0.7239749431610107, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 23030 + }, + { + "epoch": 0.23031, + "grad_norm": 0.925020694732666, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 23031 + }, + { + "epoch": 0.23032, + "grad_norm": 1.0521376132965088, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 23032 + }, + { + "epoch": 0.23033, + "grad_norm": 1.0607941150665283, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 23033 + }, + { + "epoch": 0.23034, + "grad_norm": 1.0004092454910278, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 23034 + }, + { + "epoch": 0.23035, + "grad_norm": 1.1764265298843384, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 23035 + }, + { + "epoch": 0.23036, + "grad_norm": 0.7950125336647034, + "learning_rate": 0.003, + "loss": 3.971, + "step": 23036 + }, + { + "epoch": 0.23037, + "grad_norm": 0.6976262331008911, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 23037 + }, + { + "epoch": 0.23038, + "grad_norm": 0.7908344268798828, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 23038 + }, + { + "epoch": 0.23039, + "grad_norm": 0.8579150438308716, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 23039 + }, + { + "epoch": 0.2304, + "grad_norm": 0.9021633267402649, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 23040 + }, + { + "epoch": 0.23041, + "grad_norm": 0.816045343875885, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 23041 + }, + { + "epoch": 0.23042, + "grad_norm": 0.791937530040741, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 23042 + }, + { + "epoch": 0.23043, + "grad_norm": 0.9895734786987305, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 23043 + }, + { + "epoch": 0.23044, + "grad_norm": 1.2826098203659058, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 23044 + }, + { + "epoch": 0.23045, + "grad_norm": 0.9061729311943054, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 23045 + }, + { + "epoch": 0.23046, + "grad_norm": 0.8642900586128235, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 23046 + }, + { + "epoch": 0.23047, + "grad_norm": 0.8523795008659363, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 23047 + }, + { + "epoch": 0.23048, + "grad_norm": 0.9773958921432495, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 23048 + }, + { + "epoch": 0.23049, + "grad_norm": 0.9777716994285583, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 23049 + }, + { + "epoch": 0.2305, + "grad_norm": 0.8990427851676941, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 23050 + }, + { + "epoch": 0.23051, + "grad_norm": 0.8751457333564758, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 23051 + }, + { + "epoch": 0.23052, + "grad_norm": 0.8882004618644714, + "learning_rate": 0.003, + "loss": 4.009, + "step": 23052 + }, + { + "epoch": 0.23053, + "grad_norm": 1.0927101373672485, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 23053 + }, + { + "epoch": 0.23054, + "grad_norm": 1.002712607383728, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 23054 + }, + { + "epoch": 0.23055, + "grad_norm": 1.0374126434326172, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 23055 + }, + { + "epoch": 0.23056, + "grad_norm": 1.047447919845581, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 23056 + }, + { + "epoch": 0.23057, + "grad_norm": 0.7151628732681274, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 23057 + }, + { + "epoch": 0.23058, + "grad_norm": 0.6960196495056152, + "learning_rate": 0.003, + "loss": 3.963, + "step": 23058 + }, + { + "epoch": 0.23059, + "grad_norm": 0.784977912902832, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 23059 + }, + { + "epoch": 0.2306, + "grad_norm": 0.8704864978790283, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 23060 + }, + { + "epoch": 0.23061, + "grad_norm": 1.0772920846939087, + "learning_rate": 0.003, + "loss": 4.034, + "step": 23061 + }, + { + "epoch": 0.23062, + "grad_norm": 0.8620952367782593, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 23062 + }, + { + "epoch": 0.23063, + "grad_norm": 0.7458146214485168, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 23063 + }, + { + "epoch": 0.23064, + "grad_norm": 0.7019989490509033, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 23064 + }, + { + "epoch": 0.23065, + "grad_norm": 0.7365975379943848, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 23065 + }, + { + "epoch": 0.23066, + "grad_norm": 0.7908383011817932, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 23066 + }, + { + "epoch": 0.23067, + "grad_norm": 0.8031426072120667, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 23067 + }, + { + "epoch": 0.23068, + "grad_norm": 0.7227094769477844, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 23068 + }, + { + "epoch": 0.23069, + "grad_norm": 0.701834499835968, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 23069 + }, + { + "epoch": 0.2307, + "grad_norm": 0.7857080101966858, + "learning_rate": 0.003, + "loss": 3.999, + "step": 23070 + }, + { + "epoch": 0.23071, + "grad_norm": 0.8926976919174194, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 23071 + }, + { + "epoch": 0.23072, + "grad_norm": 1.0815329551696777, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 23072 + }, + { + "epoch": 0.23073, + "grad_norm": 0.9465259909629822, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 23073 + }, + { + "epoch": 0.23074, + "grad_norm": 0.9945594072341919, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 23074 + }, + { + "epoch": 0.23075, + "grad_norm": 1.0083024501800537, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 23075 + }, + { + "epoch": 0.23076, + "grad_norm": 0.9037533402442932, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 23076 + }, + { + "epoch": 0.23077, + "grad_norm": 0.7912936806678772, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 23077 + }, + { + "epoch": 0.23078, + "grad_norm": 0.7787169218063354, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 23078 + }, + { + "epoch": 0.23079, + "grad_norm": 0.7152718901634216, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 23079 + }, + { + "epoch": 0.2308, + "grad_norm": 0.6530928611755371, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 23080 + }, + { + "epoch": 0.23081, + "grad_norm": 0.6435402631759644, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 23081 + }, + { + "epoch": 0.23082, + "grad_norm": 0.7087308168411255, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 23082 + }, + { + "epoch": 0.23083, + "grad_norm": 0.6906288862228394, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 23083 + }, + { + "epoch": 0.23084, + "grad_norm": 0.6898327469825745, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 23084 + }, + { + "epoch": 0.23085, + "grad_norm": 0.7034412026405334, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 23085 + }, + { + "epoch": 0.23086, + "grad_norm": 0.8162875771522522, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 23086 + }, + { + "epoch": 0.23087, + "grad_norm": 1.2827194929122925, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 23087 + }, + { + "epoch": 0.23088, + "grad_norm": 0.7645619511604309, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 23088 + }, + { + "epoch": 0.23089, + "grad_norm": 0.5765758752822876, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 23089 + }, + { + "epoch": 0.2309, + "grad_norm": 0.6566334366798401, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 23090 + }, + { + "epoch": 0.23091, + "grad_norm": 0.626915454864502, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 23091 + }, + { + "epoch": 0.23092, + "grad_norm": 0.6524862051010132, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 23092 + }, + { + "epoch": 0.23093, + "grad_norm": 0.6352232098579407, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 23093 + }, + { + "epoch": 0.23094, + "grad_norm": 0.6740639209747314, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 23094 + }, + { + "epoch": 0.23095, + "grad_norm": 0.6913418769836426, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 23095 + }, + { + "epoch": 0.23096, + "grad_norm": 0.7570476531982422, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 23096 + }, + { + "epoch": 0.23097, + "grad_norm": 0.9425538778305054, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 23097 + }, + { + "epoch": 0.23098, + "grad_norm": 1.2895063161849976, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 23098 + }, + { + "epoch": 0.23099, + "grad_norm": 1.0214767456054688, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 23099 + }, + { + "epoch": 0.231, + "grad_norm": 1.0568567514419556, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 23100 + }, + { + "epoch": 0.23101, + "grad_norm": 0.9886485934257507, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 23101 + }, + { + "epoch": 0.23102, + "grad_norm": 0.9668920040130615, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 23102 + }, + { + "epoch": 0.23103, + "grad_norm": 0.9179192781448364, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 23103 + }, + { + "epoch": 0.23104, + "grad_norm": 0.988434374332428, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 23104 + }, + { + "epoch": 0.23105, + "grad_norm": 1.0827306509017944, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 23105 + }, + { + "epoch": 0.23106, + "grad_norm": 1.0776193141937256, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 23106 + }, + { + "epoch": 0.23107, + "grad_norm": 0.8165929317474365, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 23107 + }, + { + "epoch": 0.23108, + "grad_norm": 0.8577418327331543, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 23108 + }, + { + "epoch": 0.23109, + "grad_norm": 1.0895438194274902, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 23109 + }, + { + "epoch": 0.2311, + "grad_norm": 1.047911286354065, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 23110 + }, + { + "epoch": 0.23111, + "grad_norm": 0.8592563271522522, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 23111 + }, + { + "epoch": 0.23112, + "grad_norm": 1.0875024795532227, + "learning_rate": 0.003, + "loss": 4.051, + "step": 23112 + }, + { + "epoch": 0.23113, + "grad_norm": 1.0649064779281616, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 23113 + }, + { + "epoch": 0.23114, + "grad_norm": 0.9362956285476685, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 23114 + }, + { + "epoch": 0.23115, + "grad_norm": 1.0652841329574585, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 23115 + }, + { + "epoch": 0.23116, + "grad_norm": 0.9305871725082397, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 23116 + }, + { + "epoch": 0.23117, + "grad_norm": 0.9791569113731384, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 23117 + }, + { + "epoch": 0.23118, + "grad_norm": 0.9458165764808655, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 23118 + }, + { + "epoch": 0.23119, + "grad_norm": 0.8208984136581421, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 23119 + }, + { + "epoch": 0.2312, + "grad_norm": 0.9451833963394165, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 23120 + }, + { + "epoch": 0.23121, + "grad_norm": 1.2403461933135986, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 23121 + }, + { + "epoch": 0.23122, + "grad_norm": 0.7685925960540771, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 23122 + }, + { + "epoch": 0.23123, + "grad_norm": 0.6764114499092102, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 23123 + }, + { + "epoch": 0.23124, + "grad_norm": 0.7103784680366516, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 23124 + }, + { + "epoch": 0.23125, + "grad_norm": 0.9381802678108215, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 23125 + }, + { + "epoch": 0.23126, + "grad_norm": 1.0787749290466309, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 23126 + }, + { + "epoch": 0.23127, + "grad_norm": 0.9687568545341492, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 23127 + }, + { + "epoch": 0.23128, + "grad_norm": 0.9377732872962952, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 23128 + }, + { + "epoch": 0.23129, + "grad_norm": 0.9301403760910034, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 23129 + }, + { + "epoch": 0.2313, + "grad_norm": 1.0157380104064941, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 23130 + }, + { + "epoch": 0.23131, + "grad_norm": 1.027206540107727, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 23131 + }, + { + "epoch": 0.23132, + "grad_norm": 0.9323575496673584, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 23132 + }, + { + "epoch": 0.23133, + "grad_norm": 0.7760303616523743, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 23133 + }, + { + "epoch": 0.23134, + "grad_norm": 0.8039150238037109, + "learning_rate": 0.003, + "loss": 3.994, + "step": 23134 + }, + { + "epoch": 0.23135, + "grad_norm": 0.8443168997764587, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 23135 + }, + { + "epoch": 0.23136, + "grad_norm": 0.7079364061355591, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 23136 + }, + { + "epoch": 0.23137, + "grad_norm": 0.6657370328903198, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 23137 + }, + { + "epoch": 0.23138, + "grad_norm": 0.5412644147872925, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 23138 + }, + { + "epoch": 0.23139, + "grad_norm": 0.5117861032485962, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 23139 + }, + { + "epoch": 0.2314, + "grad_norm": 0.5636317133903503, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 23140 + }, + { + "epoch": 0.23141, + "grad_norm": 0.602226197719574, + "learning_rate": 0.003, + "loss": 3.981, + "step": 23141 + }, + { + "epoch": 0.23142, + "grad_norm": 0.6775985956192017, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 23142 + }, + { + "epoch": 0.23143, + "grad_norm": 0.6760642528533936, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 23143 + }, + { + "epoch": 0.23144, + "grad_norm": 0.5938423871994019, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 23144 + }, + { + "epoch": 0.23145, + "grad_norm": 0.9121343493461609, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 23145 + }, + { + "epoch": 0.23146, + "grad_norm": 1.3758800029754639, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 23146 + }, + { + "epoch": 0.23147, + "grad_norm": 0.6953003406524658, + "learning_rate": 0.003, + "loss": 4.004, + "step": 23147 + }, + { + "epoch": 0.23148, + "grad_norm": 0.6517739295959473, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 23148 + }, + { + "epoch": 0.23149, + "grad_norm": 0.8872220516204834, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 23149 + }, + { + "epoch": 0.2315, + "grad_norm": 0.9447194933891296, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 23150 + }, + { + "epoch": 0.23151, + "grad_norm": 0.8151304721832275, + "learning_rate": 0.003, + "loss": 3.993, + "step": 23151 + }, + { + "epoch": 0.23152, + "grad_norm": 0.6569218635559082, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 23152 + }, + { + "epoch": 0.23153, + "grad_norm": 0.6699762344360352, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 23153 + }, + { + "epoch": 0.23154, + "grad_norm": 0.7029743194580078, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 23154 + }, + { + "epoch": 0.23155, + "grad_norm": 0.7596902251243591, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 23155 + }, + { + "epoch": 0.23156, + "grad_norm": 0.7759253978729248, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 23156 + }, + { + "epoch": 0.23157, + "grad_norm": 0.7212099432945251, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 23157 + }, + { + "epoch": 0.23158, + "grad_norm": 0.6254028677940369, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 23158 + }, + { + "epoch": 0.23159, + "grad_norm": 0.692669689655304, + "learning_rate": 0.003, + "loss": 3.988, + "step": 23159 + }, + { + "epoch": 0.2316, + "grad_norm": 0.7579753994941711, + "learning_rate": 0.003, + "loss": 3.989, + "step": 23160 + }, + { + "epoch": 0.23161, + "grad_norm": 0.8415212631225586, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 23161 + }, + { + "epoch": 0.23162, + "grad_norm": 0.8660801649093628, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 23162 + }, + { + "epoch": 0.23163, + "grad_norm": 1.0686578750610352, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 23163 + }, + { + "epoch": 0.23164, + "grad_norm": 1.3484357595443726, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 23164 + }, + { + "epoch": 0.23165, + "grad_norm": 0.5678802728652954, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 23165 + }, + { + "epoch": 0.23166, + "grad_norm": 0.7133793830871582, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 23166 + }, + { + "epoch": 0.23167, + "grad_norm": 0.9378290176391602, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 23167 + }, + { + "epoch": 0.23168, + "grad_norm": 1.0485478639602661, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 23168 + }, + { + "epoch": 0.23169, + "grad_norm": 1.1269663572311401, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 23169 + }, + { + "epoch": 0.2317, + "grad_norm": 0.8737238645553589, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 23170 + }, + { + "epoch": 0.23171, + "grad_norm": 1.0013928413391113, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 23171 + }, + { + "epoch": 0.23172, + "grad_norm": 1.1904878616333008, + "learning_rate": 0.003, + "loss": 4.015, + "step": 23172 + }, + { + "epoch": 0.23173, + "grad_norm": 1.1590728759765625, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 23173 + }, + { + "epoch": 0.23174, + "grad_norm": 1.023846983909607, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 23174 + }, + { + "epoch": 0.23175, + "grad_norm": 0.9026808738708496, + "learning_rate": 0.003, + "loss": 4.012, + "step": 23175 + }, + { + "epoch": 0.23176, + "grad_norm": 0.7075026631355286, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 23176 + }, + { + "epoch": 0.23177, + "grad_norm": 0.8209516406059265, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 23177 + }, + { + "epoch": 0.23178, + "grad_norm": 1.0743259191513062, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 23178 + }, + { + "epoch": 0.23179, + "grad_norm": 1.1427286863327026, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 23179 + }, + { + "epoch": 0.2318, + "grad_norm": 1.0803613662719727, + "learning_rate": 0.003, + "loss": 4.021, + "step": 23180 + }, + { + "epoch": 0.23181, + "grad_norm": 0.908909797668457, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 23181 + }, + { + "epoch": 0.23182, + "grad_norm": 0.9240555763244629, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 23182 + }, + { + "epoch": 0.23183, + "grad_norm": 0.9652478098869324, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 23183 + }, + { + "epoch": 0.23184, + "grad_norm": 1.0944721698760986, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 23184 + }, + { + "epoch": 0.23185, + "grad_norm": 0.8901162147521973, + "learning_rate": 0.003, + "loss": 4.021, + "step": 23185 + }, + { + "epoch": 0.23186, + "grad_norm": 0.711893618106842, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 23186 + }, + { + "epoch": 0.23187, + "grad_norm": 0.7592436075210571, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 23187 + }, + { + "epoch": 0.23188, + "grad_norm": 0.8489288687705994, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 23188 + }, + { + "epoch": 0.23189, + "grad_norm": 0.8202161192893982, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 23189 + }, + { + "epoch": 0.2319, + "grad_norm": 0.9616050124168396, + "learning_rate": 0.003, + "loss": 3.982, + "step": 23190 + }, + { + "epoch": 0.23191, + "grad_norm": 1.1884790658950806, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 23191 + }, + { + "epoch": 0.23192, + "grad_norm": 0.9715366363525391, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 23192 + }, + { + "epoch": 0.23193, + "grad_norm": 0.786625862121582, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 23193 + }, + { + "epoch": 0.23194, + "grad_norm": 0.6661534905433655, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 23194 + }, + { + "epoch": 0.23195, + "grad_norm": 0.7163292169570923, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 23195 + }, + { + "epoch": 0.23196, + "grad_norm": 0.6712885499000549, + "learning_rate": 0.003, + "loss": 4.02, + "step": 23196 + }, + { + "epoch": 0.23197, + "grad_norm": 0.7473501563072205, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 23197 + }, + { + "epoch": 0.23198, + "grad_norm": 0.7254409790039062, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 23198 + }, + { + "epoch": 0.23199, + "grad_norm": 0.5941123366355896, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 23199 + }, + { + "epoch": 0.232, + "grad_norm": 0.611175537109375, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 23200 + }, + { + "epoch": 0.23201, + "grad_norm": 0.6366611123085022, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 23201 + }, + { + "epoch": 0.23202, + "grad_norm": 0.712209165096283, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 23202 + }, + { + "epoch": 0.23203, + "grad_norm": 0.8324995636940002, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 23203 + }, + { + "epoch": 0.23204, + "grad_norm": 1.0045596361160278, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 23204 + }, + { + "epoch": 0.23205, + "grad_norm": 1.1900713443756104, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 23205 + }, + { + "epoch": 0.23206, + "grad_norm": 0.7876248359680176, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 23206 + }, + { + "epoch": 0.23207, + "grad_norm": 0.7939680218696594, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 23207 + }, + { + "epoch": 0.23208, + "grad_norm": 0.8914170265197754, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 23208 + }, + { + "epoch": 0.23209, + "grad_norm": 1.0042827129364014, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 23209 + }, + { + "epoch": 0.2321, + "grad_norm": 0.8809590935707092, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 23210 + }, + { + "epoch": 0.23211, + "grad_norm": 0.751802384853363, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 23211 + }, + { + "epoch": 0.23212, + "grad_norm": 0.7041476368904114, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 23212 + }, + { + "epoch": 0.23213, + "grad_norm": 0.6990548968315125, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 23213 + }, + { + "epoch": 0.23214, + "grad_norm": 0.8522382974624634, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 23214 + }, + { + "epoch": 0.23215, + "grad_norm": 1.1369420289993286, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 23215 + }, + { + "epoch": 0.23216, + "grad_norm": 1.2524713277816772, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 23216 + }, + { + "epoch": 0.23217, + "grad_norm": 0.6062924861907959, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 23217 + }, + { + "epoch": 0.23218, + "grad_norm": 0.7876224517822266, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 23218 + }, + { + "epoch": 0.23219, + "grad_norm": 1.0908962488174438, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 23219 + }, + { + "epoch": 0.2322, + "grad_norm": 0.9003163576126099, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 23220 + }, + { + "epoch": 0.23221, + "grad_norm": 0.739678144454956, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 23221 + }, + { + "epoch": 0.23222, + "grad_norm": 0.6994432210922241, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 23222 + }, + { + "epoch": 0.23223, + "grad_norm": 0.6228587031364441, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 23223 + }, + { + "epoch": 0.23224, + "grad_norm": 0.7365922331809998, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 23224 + }, + { + "epoch": 0.23225, + "grad_norm": 0.7307228446006775, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 23225 + }, + { + "epoch": 0.23226, + "grad_norm": 0.818894624710083, + "learning_rate": 0.003, + "loss": 3.971, + "step": 23226 + }, + { + "epoch": 0.23227, + "grad_norm": 0.936927080154419, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 23227 + }, + { + "epoch": 0.23228, + "grad_norm": 1.097700834274292, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 23228 + }, + { + "epoch": 0.23229, + "grad_norm": 0.8098295331001282, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 23229 + }, + { + "epoch": 0.2323, + "grad_norm": 0.7045392990112305, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 23230 + }, + { + "epoch": 0.23231, + "grad_norm": 0.7731865644454956, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 23231 + }, + { + "epoch": 0.23232, + "grad_norm": 0.8143264055252075, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 23232 + }, + { + "epoch": 0.23233, + "grad_norm": 0.8811879754066467, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 23233 + }, + { + "epoch": 0.23234, + "grad_norm": 1.0242512226104736, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 23234 + }, + { + "epoch": 0.23235, + "grad_norm": 0.8928782343864441, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 23235 + }, + { + "epoch": 0.23236, + "grad_norm": 0.8371671438217163, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 23236 + }, + { + "epoch": 0.23237, + "grad_norm": 0.8973673582077026, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 23237 + }, + { + "epoch": 0.23238, + "grad_norm": 0.9465650916099548, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 23238 + }, + { + "epoch": 0.23239, + "grad_norm": 1.039687991142273, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 23239 + }, + { + "epoch": 0.2324, + "grad_norm": 1.244040846824646, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 23240 + }, + { + "epoch": 0.23241, + "grad_norm": 0.927910327911377, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 23241 + }, + { + "epoch": 0.23242, + "grad_norm": 0.8296650052070618, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 23242 + }, + { + "epoch": 0.23243, + "grad_norm": 0.8977550864219666, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 23243 + }, + { + "epoch": 0.23244, + "grad_norm": 0.8344443440437317, + "learning_rate": 0.003, + "loss": 4.001, + "step": 23244 + }, + { + "epoch": 0.23245, + "grad_norm": 0.8173187375068665, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 23245 + }, + { + "epoch": 0.23246, + "grad_norm": 0.8630403280258179, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 23246 + }, + { + "epoch": 0.23247, + "grad_norm": 0.9840368032455444, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 23247 + }, + { + "epoch": 0.23248, + "grad_norm": 1.0217231512069702, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 23248 + }, + { + "epoch": 0.23249, + "grad_norm": 1.1772297620773315, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 23249 + }, + { + "epoch": 0.2325, + "grad_norm": 0.9188389182090759, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 23250 + }, + { + "epoch": 0.23251, + "grad_norm": 0.7768462896347046, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 23251 + }, + { + "epoch": 0.23252, + "grad_norm": 0.7858250141143799, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 23252 + }, + { + "epoch": 0.23253, + "grad_norm": 0.8975591659545898, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 23253 + }, + { + "epoch": 0.23254, + "grad_norm": 0.8841562867164612, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 23254 + }, + { + "epoch": 0.23255, + "grad_norm": 0.8217187523841858, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 23255 + }, + { + "epoch": 0.23256, + "grad_norm": 0.7890130281448364, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 23256 + }, + { + "epoch": 0.23257, + "grad_norm": 0.8426260352134705, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 23257 + }, + { + "epoch": 0.23258, + "grad_norm": 0.754285991191864, + "learning_rate": 0.003, + "loss": 3.955, + "step": 23258 + }, + { + "epoch": 0.23259, + "grad_norm": 0.764949381351471, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 23259 + }, + { + "epoch": 0.2326, + "grad_norm": 0.7114693522453308, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 23260 + }, + { + "epoch": 0.23261, + "grad_norm": 0.7044433355331421, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 23261 + }, + { + "epoch": 0.23262, + "grad_norm": 0.594560444355011, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 23262 + }, + { + "epoch": 0.23263, + "grad_norm": 0.5621315240859985, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 23263 + }, + { + "epoch": 0.23264, + "grad_norm": 0.6395027041435242, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 23264 + }, + { + "epoch": 0.23265, + "grad_norm": 0.8939841389656067, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 23265 + }, + { + "epoch": 0.23266, + "grad_norm": 1.2718225717544556, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 23266 + }, + { + "epoch": 0.23267, + "grad_norm": 1.0413259267807007, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 23267 + }, + { + "epoch": 0.23268, + "grad_norm": 0.7865359783172607, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 23268 + }, + { + "epoch": 0.23269, + "grad_norm": 0.6124400496482849, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 23269 + }, + { + "epoch": 0.2327, + "grad_norm": 0.5689623355865479, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 23270 + }, + { + "epoch": 0.23271, + "grad_norm": 0.8033220171928406, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 23271 + }, + { + "epoch": 0.23272, + "grad_norm": 1.0399224758148193, + "learning_rate": 0.003, + "loss": 4.007, + "step": 23272 + }, + { + "epoch": 0.23273, + "grad_norm": 0.9858866333961487, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 23273 + }, + { + "epoch": 0.23274, + "grad_norm": 0.9755273461341858, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 23274 + }, + { + "epoch": 0.23275, + "grad_norm": 0.8104730248451233, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 23275 + }, + { + "epoch": 0.23276, + "grad_norm": 0.6555654406547546, + "learning_rate": 0.003, + "loss": 4.017, + "step": 23276 + }, + { + "epoch": 0.23277, + "grad_norm": 0.6013662815093994, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 23277 + }, + { + "epoch": 0.23278, + "grad_norm": 0.6152356863021851, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 23278 + }, + { + "epoch": 0.23279, + "grad_norm": 0.5727058053016663, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 23279 + }, + { + "epoch": 0.2328, + "grad_norm": 0.7068799734115601, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 23280 + }, + { + "epoch": 0.23281, + "grad_norm": 0.7504807114601135, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 23281 + }, + { + "epoch": 0.23282, + "grad_norm": 0.8526983261108398, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 23282 + }, + { + "epoch": 0.23283, + "grad_norm": 1.1225813627243042, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 23283 + }, + { + "epoch": 0.23284, + "grad_norm": 1.102518916130066, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 23284 + }, + { + "epoch": 0.23285, + "grad_norm": 0.8981680274009705, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 23285 + }, + { + "epoch": 0.23286, + "grad_norm": 0.8418294787406921, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 23286 + }, + { + "epoch": 0.23287, + "grad_norm": 0.8913272023200989, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 23287 + }, + { + "epoch": 0.23288, + "grad_norm": 1.0730458498001099, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 23288 + }, + { + "epoch": 0.23289, + "grad_norm": 0.907920777797699, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 23289 + }, + { + "epoch": 0.2329, + "grad_norm": 1.004433035850525, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 23290 + }, + { + "epoch": 0.23291, + "grad_norm": 1.0563387870788574, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 23291 + }, + { + "epoch": 0.23292, + "grad_norm": 0.8693835139274597, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 23292 + }, + { + "epoch": 0.23293, + "grad_norm": 0.8308678865432739, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 23293 + }, + { + "epoch": 0.23294, + "grad_norm": 0.9750970005989075, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 23294 + }, + { + "epoch": 0.23295, + "grad_norm": 1.058716893196106, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 23295 + }, + { + "epoch": 0.23296, + "grad_norm": 1.011573314666748, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 23296 + }, + { + "epoch": 0.23297, + "grad_norm": 1.131119728088379, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 23297 + }, + { + "epoch": 0.23298, + "grad_norm": 1.099518895149231, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 23298 + }, + { + "epoch": 0.23299, + "grad_norm": 0.9257704615592957, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 23299 + }, + { + "epoch": 0.233, + "grad_norm": 1.0638272762298584, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 23300 + }, + { + "epoch": 0.23301, + "grad_norm": 0.9316908717155457, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 23301 + }, + { + "epoch": 0.23302, + "grad_norm": 0.8141741156578064, + "learning_rate": 0.003, + "loss": 4.012, + "step": 23302 + }, + { + "epoch": 0.23303, + "grad_norm": 0.6842164993286133, + "learning_rate": 0.003, + "loss": 3.954, + "step": 23303 + }, + { + "epoch": 0.23304, + "grad_norm": 0.6948202848434448, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 23304 + }, + { + "epoch": 0.23305, + "grad_norm": 0.6842511296272278, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 23305 + }, + { + "epoch": 0.23306, + "grad_norm": 0.6752930283546448, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 23306 + }, + { + "epoch": 0.23307, + "grad_norm": 0.6374004483222961, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 23307 + }, + { + "epoch": 0.23308, + "grad_norm": 0.6382659077644348, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 23308 + }, + { + "epoch": 0.23309, + "grad_norm": 0.7026656270027161, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 23309 + }, + { + "epoch": 0.2331, + "grad_norm": 0.7003085613250732, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 23310 + }, + { + "epoch": 0.23311, + "grad_norm": 0.8024913668632507, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 23311 + }, + { + "epoch": 0.23312, + "grad_norm": 0.857912003993988, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 23312 + }, + { + "epoch": 0.23313, + "grad_norm": 0.8871775269508362, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 23313 + }, + { + "epoch": 0.23314, + "grad_norm": 1.0320825576782227, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 23314 + }, + { + "epoch": 0.23315, + "grad_norm": 1.0750254392623901, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 23315 + }, + { + "epoch": 0.23316, + "grad_norm": 0.7807958722114563, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 23316 + }, + { + "epoch": 0.23317, + "grad_norm": 0.655523419380188, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 23317 + }, + { + "epoch": 0.23318, + "grad_norm": 0.6345911622047424, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 23318 + }, + { + "epoch": 0.23319, + "grad_norm": 0.5786257386207581, + "learning_rate": 0.003, + "loss": 3.974, + "step": 23319 + }, + { + "epoch": 0.2332, + "grad_norm": 0.6055142283439636, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 23320 + }, + { + "epoch": 0.23321, + "grad_norm": 0.6612492203712463, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 23321 + }, + { + "epoch": 0.23322, + "grad_norm": 0.6529426574707031, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 23322 + }, + { + "epoch": 0.23323, + "grad_norm": 0.7285264134407043, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 23323 + }, + { + "epoch": 0.23324, + "grad_norm": 0.841522753238678, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 23324 + }, + { + "epoch": 0.23325, + "grad_norm": 1.1446211338043213, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 23325 + }, + { + "epoch": 0.23326, + "grad_norm": 0.9618659615516663, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 23326 + }, + { + "epoch": 0.23327, + "grad_norm": 0.9662145972251892, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 23327 + }, + { + "epoch": 0.23328, + "grad_norm": 0.9876563549041748, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 23328 + }, + { + "epoch": 0.23329, + "grad_norm": 0.8884350061416626, + "learning_rate": 0.003, + "loss": 3.991, + "step": 23329 + }, + { + "epoch": 0.2333, + "grad_norm": 0.8458871841430664, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 23330 + }, + { + "epoch": 0.23331, + "grad_norm": 0.9960213303565979, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 23331 + }, + { + "epoch": 0.23332, + "grad_norm": 1.0989124774932861, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 23332 + }, + { + "epoch": 0.23333, + "grad_norm": 0.9135029911994934, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 23333 + }, + { + "epoch": 0.23334, + "grad_norm": 0.8467676639556885, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 23334 + }, + { + "epoch": 0.23335, + "grad_norm": 0.9009972810745239, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 23335 + }, + { + "epoch": 0.23336, + "grad_norm": 1.0182749032974243, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 23336 + }, + { + "epoch": 0.23337, + "grad_norm": 0.9587500691413879, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 23337 + }, + { + "epoch": 0.23338, + "grad_norm": 1.0142676830291748, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 23338 + }, + { + "epoch": 0.23339, + "grad_norm": 1.204971194267273, + "learning_rate": 0.003, + "loss": 4.015, + "step": 23339 + }, + { + "epoch": 0.2334, + "grad_norm": 1.1356176137924194, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 23340 + }, + { + "epoch": 0.23341, + "grad_norm": 0.8286142945289612, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 23341 + }, + { + "epoch": 0.23342, + "grad_norm": 0.769892692565918, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 23342 + }, + { + "epoch": 0.23343, + "grad_norm": 0.7292924523353577, + "learning_rate": 0.003, + "loss": 4.012, + "step": 23343 + }, + { + "epoch": 0.23344, + "grad_norm": 0.6526293754577637, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 23344 + }, + { + "epoch": 0.23345, + "grad_norm": 0.6981724500656128, + "learning_rate": 0.003, + "loss": 3.99, + "step": 23345 + }, + { + "epoch": 0.23346, + "grad_norm": 0.6546510457992554, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 23346 + }, + { + "epoch": 0.23347, + "grad_norm": 0.6182064414024353, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 23347 + }, + { + "epoch": 0.23348, + "grad_norm": 0.674395740032196, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 23348 + }, + { + "epoch": 0.23349, + "grad_norm": 0.7991765141487122, + "learning_rate": 0.003, + "loss": 3.956, + "step": 23349 + }, + { + "epoch": 0.2335, + "grad_norm": 0.8852419853210449, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 23350 + }, + { + "epoch": 0.23351, + "grad_norm": 1.0975373983383179, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 23351 + }, + { + "epoch": 0.23352, + "grad_norm": 1.008069396018982, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 23352 + }, + { + "epoch": 0.23353, + "grad_norm": 0.9211266040802002, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 23353 + }, + { + "epoch": 0.23354, + "grad_norm": 0.6970239877700806, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 23354 + }, + { + "epoch": 0.23355, + "grad_norm": 0.634482741355896, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 23355 + }, + { + "epoch": 0.23356, + "grad_norm": 0.6545979976654053, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 23356 + }, + { + "epoch": 0.23357, + "grad_norm": 0.6891321539878845, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 23357 + }, + { + "epoch": 0.23358, + "grad_norm": 0.9773377776145935, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 23358 + }, + { + "epoch": 0.23359, + "grad_norm": 1.3792848587036133, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 23359 + }, + { + "epoch": 0.2336, + "grad_norm": 0.7976160049438477, + "learning_rate": 0.003, + "loss": 4.009, + "step": 23360 + }, + { + "epoch": 0.23361, + "grad_norm": 0.7308042645454407, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 23361 + }, + { + "epoch": 0.23362, + "grad_norm": 0.8197801113128662, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 23362 + }, + { + "epoch": 0.23363, + "grad_norm": 0.7918054461479187, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 23363 + }, + { + "epoch": 0.23364, + "grad_norm": 0.7224875092506409, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 23364 + }, + { + "epoch": 0.23365, + "grad_norm": 0.7249277234077454, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 23365 + }, + { + "epoch": 0.23366, + "grad_norm": 0.7207806706428528, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 23366 + }, + { + "epoch": 0.23367, + "grad_norm": 0.8110926151275635, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 23367 + }, + { + "epoch": 0.23368, + "grad_norm": 0.882693886756897, + "learning_rate": 0.003, + "loss": 3.983, + "step": 23368 + }, + { + "epoch": 0.23369, + "grad_norm": 0.757243812084198, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 23369 + }, + { + "epoch": 0.2337, + "grad_norm": 0.7440449595451355, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 23370 + }, + { + "epoch": 0.23371, + "grad_norm": 0.9055160880088806, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 23371 + }, + { + "epoch": 0.23372, + "grad_norm": 0.9307541251182556, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 23372 + }, + { + "epoch": 0.23373, + "grad_norm": 1.0206472873687744, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 23373 + }, + { + "epoch": 0.23374, + "grad_norm": 1.16123366355896, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 23374 + }, + { + "epoch": 0.23375, + "grad_norm": 1.0054970979690552, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 23375 + }, + { + "epoch": 0.23376, + "grad_norm": 1.0778231620788574, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 23376 + }, + { + "epoch": 0.23377, + "grad_norm": 0.945652961730957, + "learning_rate": 0.003, + "loss": 4.007, + "step": 23377 + }, + { + "epoch": 0.23378, + "grad_norm": 0.9255437850952148, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 23378 + }, + { + "epoch": 0.23379, + "grad_norm": 0.930460512638092, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 23379 + }, + { + "epoch": 0.2338, + "grad_norm": 1.0094643831253052, + "learning_rate": 0.003, + "loss": 4.018, + "step": 23380 + }, + { + "epoch": 0.23381, + "grad_norm": 1.1004003286361694, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 23381 + }, + { + "epoch": 0.23382, + "grad_norm": 0.8471861481666565, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 23382 + }, + { + "epoch": 0.23383, + "grad_norm": 0.8207311630249023, + "learning_rate": 0.003, + "loss": 3.987, + "step": 23383 + }, + { + "epoch": 0.23384, + "grad_norm": 0.8808017373085022, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 23384 + }, + { + "epoch": 0.23385, + "grad_norm": 0.8175532221794128, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 23385 + }, + { + "epoch": 0.23386, + "grad_norm": 0.8900026082992554, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 23386 + }, + { + "epoch": 0.23387, + "grad_norm": 0.9093499183654785, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 23387 + }, + { + "epoch": 0.23388, + "grad_norm": 0.7498447299003601, + "learning_rate": 0.003, + "loss": 3.996, + "step": 23388 + }, + { + "epoch": 0.23389, + "grad_norm": 0.7937158346176147, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 23389 + }, + { + "epoch": 0.2339, + "grad_norm": 0.8181902766227722, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 23390 + }, + { + "epoch": 0.23391, + "grad_norm": 0.8318588137626648, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 23391 + }, + { + "epoch": 0.23392, + "grad_norm": 1.0605374574661255, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 23392 + }, + { + "epoch": 0.23393, + "grad_norm": 1.166043758392334, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 23393 + }, + { + "epoch": 0.23394, + "grad_norm": 0.854134202003479, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 23394 + }, + { + "epoch": 0.23395, + "grad_norm": 0.8396636247634888, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 23395 + }, + { + "epoch": 0.23396, + "grad_norm": 0.7680071592330933, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 23396 + }, + { + "epoch": 0.23397, + "grad_norm": 0.8411863446235657, + "learning_rate": 0.003, + "loss": 3.999, + "step": 23397 + }, + { + "epoch": 0.23398, + "grad_norm": 0.9183284044265747, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 23398 + }, + { + "epoch": 0.23399, + "grad_norm": 1.0573327541351318, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 23399 + }, + { + "epoch": 0.234, + "grad_norm": 1.076817274093628, + "learning_rate": 0.003, + "loss": 4.007, + "step": 23400 + }, + { + "epoch": 0.23401, + "grad_norm": 0.9518916010856628, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 23401 + }, + { + "epoch": 0.23402, + "grad_norm": 0.9694162607192993, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 23402 + }, + { + "epoch": 0.23403, + "grad_norm": 0.9257795810699463, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 23403 + }, + { + "epoch": 0.23404, + "grad_norm": 0.7687200307846069, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 23404 + }, + { + "epoch": 0.23405, + "grad_norm": 0.7162486910820007, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 23405 + }, + { + "epoch": 0.23406, + "grad_norm": 0.6962159276008606, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 23406 + }, + { + "epoch": 0.23407, + "grad_norm": 0.6228445172309875, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 23407 + }, + { + "epoch": 0.23408, + "grad_norm": 0.6473580002784729, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 23408 + }, + { + "epoch": 0.23409, + "grad_norm": 0.7118189334869385, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 23409 + }, + { + "epoch": 0.2341, + "grad_norm": 0.7790793776512146, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 23410 + }, + { + "epoch": 0.23411, + "grad_norm": 0.8359365463256836, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 23411 + }, + { + "epoch": 0.23412, + "grad_norm": 0.8378966450691223, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 23412 + }, + { + "epoch": 0.23413, + "grad_norm": 0.8740316033363342, + "learning_rate": 0.003, + "loss": 3.992, + "step": 23413 + }, + { + "epoch": 0.23414, + "grad_norm": 1.0481330156326294, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 23414 + }, + { + "epoch": 0.23415, + "grad_norm": 0.900974452495575, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 23415 + }, + { + "epoch": 0.23416, + "grad_norm": 0.6630672216415405, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 23416 + }, + { + "epoch": 0.23417, + "grad_norm": 0.7171139717102051, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 23417 + }, + { + "epoch": 0.23418, + "grad_norm": 0.8322917819023132, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 23418 + }, + { + "epoch": 0.23419, + "grad_norm": 0.8701794743537903, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 23419 + }, + { + "epoch": 0.2342, + "grad_norm": 0.7523330450057983, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 23420 + }, + { + "epoch": 0.23421, + "grad_norm": 0.7523906230926514, + "learning_rate": 0.003, + "loss": 3.99, + "step": 23421 + }, + { + "epoch": 0.23422, + "grad_norm": 0.8647992610931396, + "learning_rate": 0.003, + "loss": 3.983, + "step": 23422 + }, + { + "epoch": 0.23423, + "grad_norm": 0.9035136103630066, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 23423 + }, + { + "epoch": 0.23424, + "grad_norm": 0.9456508159637451, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 23424 + }, + { + "epoch": 0.23425, + "grad_norm": 0.981437623500824, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 23425 + }, + { + "epoch": 0.23426, + "grad_norm": 0.789150595664978, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 23426 + }, + { + "epoch": 0.23427, + "grad_norm": 0.7201241850852966, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 23427 + }, + { + "epoch": 0.23428, + "grad_norm": 0.9542236328125, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 23428 + }, + { + "epoch": 0.23429, + "grad_norm": 1.4219156503677368, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 23429 + }, + { + "epoch": 0.2343, + "grad_norm": 0.5906811952590942, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 23430 + }, + { + "epoch": 0.23431, + "grad_norm": 0.7598404288291931, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 23431 + }, + { + "epoch": 0.23432, + "grad_norm": 0.8713325262069702, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 23432 + }, + { + "epoch": 0.23433, + "grad_norm": 0.8605551719665527, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 23433 + }, + { + "epoch": 0.23434, + "grad_norm": 0.811640202999115, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 23434 + }, + { + "epoch": 0.23435, + "grad_norm": 1.0931929349899292, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 23435 + }, + { + "epoch": 0.23436, + "grad_norm": 1.3259621858596802, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 23436 + }, + { + "epoch": 0.23437, + "grad_norm": 0.6889467239379883, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 23437 + }, + { + "epoch": 0.23438, + "grad_norm": 0.6901146173477173, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 23438 + }, + { + "epoch": 0.23439, + "grad_norm": 0.7982661128044128, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 23439 + }, + { + "epoch": 0.2344, + "grad_norm": 0.7621657848358154, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 23440 + }, + { + "epoch": 0.23441, + "grad_norm": 0.7077150940895081, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 23441 + }, + { + "epoch": 0.23442, + "grad_norm": 0.6702290177345276, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 23442 + }, + { + "epoch": 0.23443, + "grad_norm": 0.7219530344009399, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 23443 + }, + { + "epoch": 0.23444, + "grad_norm": 0.7062245607376099, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 23444 + }, + { + "epoch": 0.23445, + "grad_norm": 0.665654718875885, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 23445 + }, + { + "epoch": 0.23446, + "grad_norm": 0.7844275236129761, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 23446 + }, + { + "epoch": 0.23447, + "grad_norm": 0.8807093501091003, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 23447 + }, + { + "epoch": 0.23448, + "grad_norm": 0.8727428913116455, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 23448 + }, + { + "epoch": 0.23449, + "grad_norm": 0.8580114245414734, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 23449 + }, + { + "epoch": 0.2345, + "grad_norm": 0.9003655910491943, + "learning_rate": 0.003, + "loss": 4.003, + "step": 23450 + }, + { + "epoch": 0.23451, + "grad_norm": 1.1044108867645264, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 23451 + }, + { + "epoch": 0.23452, + "grad_norm": 0.9227287173271179, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 23452 + }, + { + "epoch": 0.23453, + "grad_norm": 1.0362731218338013, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 23453 + }, + { + "epoch": 0.23454, + "grad_norm": 1.0237703323364258, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 23454 + }, + { + "epoch": 0.23455, + "grad_norm": 0.9875350594520569, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 23455 + }, + { + "epoch": 0.23456, + "grad_norm": 0.9987820386886597, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 23456 + }, + { + "epoch": 0.23457, + "grad_norm": 0.863436222076416, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 23457 + }, + { + "epoch": 0.23458, + "grad_norm": 0.7844811677932739, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 23458 + }, + { + "epoch": 0.23459, + "grad_norm": 0.9037507772445679, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 23459 + }, + { + "epoch": 0.2346, + "grad_norm": 1.0604289770126343, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 23460 + }, + { + "epoch": 0.23461, + "grad_norm": 1.041627287864685, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 23461 + }, + { + "epoch": 0.23462, + "grad_norm": 0.8328134417533875, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 23462 + }, + { + "epoch": 0.23463, + "grad_norm": 0.7294815182685852, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 23463 + }, + { + "epoch": 0.23464, + "grad_norm": 0.7260693311691284, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 23464 + }, + { + "epoch": 0.23465, + "grad_norm": 0.7134116291999817, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 23465 + }, + { + "epoch": 0.23466, + "grad_norm": 0.811359167098999, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 23466 + }, + { + "epoch": 0.23467, + "grad_norm": 0.9486062526702881, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 23467 + }, + { + "epoch": 0.23468, + "grad_norm": 1.0555917024612427, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 23468 + }, + { + "epoch": 0.23469, + "grad_norm": 0.9738951921463013, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 23469 + }, + { + "epoch": 0.2347, + "grad_norm": 0.9290216565132141, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 23470 + }, + { + "epoch": 0.23471, + "grad_norm": 0.9628164768218994, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 23471 + }, + { + "epoch": 0.23472, + "grad_norm": 1.088392734527588, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 23472 + }, + { + "epoch": 0.23473, + "grad_norm": 0.7942327857017517, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 23473 + }, + { + "epoch": 0.23474, + "grad_norm": 0.7042145133018494, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 23474 + }, + { + "epoch": 0.23475, + "grad_norm": 0.773266613483429, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 23475 + }, + { + "epoch": 0.23476, + "grad_norm": 0.9011462330818176, + "learning_rate": 0.003, + "loss": 3.999, + "step": 23476 + }, + { + "epoch": 0.23477, + "grad_norm": 1.0302375555038452, + "learning_rate": 0.003, + "loss": 4.002, + "step": 23477 + }, + { + "epoch": 0.23478, + "grad_norm": 0.973683774471283, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 23478 + }, + { + "epoch": 0.23479, + "grad_norm": 0.8824337124824524, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 23479 + }, + { + "epoch": 0.2348, + "grad_norm": 0.8241666555404663, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 23480 + }, + { + "epoch": 0.23481, + "grad_norm": 0.7533800601959229, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 23481 + }, + { + "epoch": 0.23482, + "grad_norm": 0.8882649540901184, + "learning_rate": 0.003, + "loss": 4.002, + "step": 23482 + }, + { + "epoch": 0.23483, + "grad_norm": 0.9837201833724976, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 23483 + }, + { + "epoch": 0.23484, + "grad_norm": 1.1338051557540894, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 23484 + }, + { + "epoch": 0.23485, + "grad_norm": 0.740549623966217, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 23485 + }, + { + "epoch": 0.23486, + "grad_norm": 0.6906768083572388, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 23486 + }, + { + "epoch": 0.23487, + "grad_norm": 0.8047739267349243, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 23487 + }, + { + "epoch": 0.23488, + "grad_norm": 0.9584016799926758, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 23488 + }, + { + "epoch": 0.23489, + "grad_norm": 1.1385374069213867, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 23489 + }, + { + "epoch": 0.2349, + "grad_norm": 0.7839148640632629, + "learning_rate": 0.003, + "loss": 4.019, + "step": 23490 + }, + { + "epoch": 0.23491, + "grad_norm": 0.6853298544883728, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 23491 + }, + { + "epoch": 0.23492, + "grad_norm": 0.6120742559432983, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 23492 + }, + { + "epoch": 0.23493, + "grad_norm": 0.7567932605743408, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 23493 + }, + { + "epoch": 0.23494, + "grad_norm": 0.8830358386039734, + "learning_rate": 0.003, + "loss": 4.004, + "step": 23494 + }, + { + "epoch": 0.23495, + "grad_norm": 0.9580311179161072, + "learning_rate": 0.003, + "loss": 4.023, + "step": 23495 + }, + { + "epoch": 0.23496, + "grad_norm": 1.0361665487289429, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 23496 + }, + { + "epoch": 0.23497, + "grad_norm": 1.1188985109329224, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 23497 + }, + { + "epoch": 0.23498, + "grad_norm": 1.041815996170044, + "learning_rate": 0.003, + "loss": 3.985, + "step": 23498 + }, + { + "epoch": 0.23499, + "grad_norm": 0.7883170247077942, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 23499 + }, + { + "epoch": 0.235, + "grad_norm": 0.6815652251243591, + "learning_rate": 0.003, + "loss": 4.019, + "step": 23500 + }, + { + "epoch": 0.23501, + "grad_norm": 0.704131007194519, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 23501 + }, + { + "epoch": 0.23502, + "grad_norm": 0.7115319967269897, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 23502 + }, + { + "epoch": 0.23503, + "grad_norm": 0.7279991507530212, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 23503 + }, + { + "epoch": 0.23504, + "grad_norm": 0.7242021560668945, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 23504 + }, + { + "epoch": 0.23505, + "grad_norm": 0.6775855422019958, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 23505 + }, + { + "epoch": 0.23506, + "grad_norm": 0.6182830333709717, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 23506 + }, + { + "epoch": 0.23507, + "grad_norm": 0.5431252121925354, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 23507 + }, + { + "epoch": 0.23508, + "grad_norm": 0.6067520976066589, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 23508 + }, + { + "epoch": 0.23509, + "grad_norm": 0.6552537083625793, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 23509 + }, + { + "epoch": 0.2351, + "grad_norm": 0.7031375169754028, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 23510 + }, + { + "epoch": 0.23511, + "grad_norm": 0.9384838938713074, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 23511 + }, + { + "epoch": 0.23512, + "grad_norm": 1.3321774005889893, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 23512 + }, + { + "epoch": 0.23513, + "grad_norm": 0.8902077674865723, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 23513 + }, + { + "epoch": 0.23514, + "grad_norm": 0.8608815670013428, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 23514 + }, + { + "epoch": 0.23515, + "grad_norm": 0.8887969255447388, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 23515 + }, + { + "epoch": 0.23516, + "grad_norm": 0.9530085921287537, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 23516 + }, + { + "epoch": 0.23517, + "grad_norm": 0.8865307569503784, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 23517 + }, + { + "epoch": 0.23518, + "grad_norm": 0.8597562909126282, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 23518 + }, + { + "epoch": 0.23519, + "grad_norm": 0.8724894523620605, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 23519 + }, + { + "epoch": 0.2352, + "grad_norm": 0.8962650895118713, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 23520 + }, + { + "epoch": 0.23521, + "grad_norm": 0.9189071655273438, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 23521 + }, + { + "epoch": 0.23522, + "grad_norm": 0.9055014252662659, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 23522 + }, + { + "epoch": 0.23523, + "grad_norm": 0.8857681155204773, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 23523 + }, + { + "epoch": 0.23524, + "grad_norm": 0.9173164963722229, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 23524 + }, + { + "epoch": 0.23525, + "grad_norm": 0.9316970705986023, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 23525 + }, + { + "epoch": 0.23526, + "grad_norm": 1.0325679779052734, + "learning_rate": 0.003, + "loss": 4.01, + "step": 23526 + }, + { + "epoch": 0.23527, + "grad_norm": 0.9668265581130981, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 23527 + }, + { + "epoch": 0.23528, + "grad_norm": 0.89481121301651, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 23528 + }, + { + "epoch": 0.23529, + "grad_norm": 0.9330294132232666, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 23529 + }, + { + "epoch": 0.2353, + "grad_norm": 1.0114556550979614, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 23530 + }, + { + "epoch": 0.23531, + "grad_norm": 0.9982279539108276, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 23531 + }, + { + "epoch": 0.23532, + "grad_norm": 1.1483893394470215, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 23532 + }, + { + "epoch": 0.23533, + "grad_norm": 1.0811078548431396, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 23533 + }, + { + "epoch": 0.23534, + "grad_norm": 1.3236910104751587, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 23534 + }, + { + "epoch": 0.23535, + "grad_norm": 0.8918457627296448, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 23535 + }, + { + "epoch": 0.23536, + "grad_norm": 0.7635582685470581, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 23536 + }, + { + "epoch": 0.23537, + "grad_norm": 0.7888791561126709, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 23537 + }, + { + "epoch": 0.23538, + "grad_norm": 0.8811279535293579, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 23538 + }, + { + "epoch": 0.23539, + "grad_norm": 1.0202007293701172, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 23539 + }, + { + "epoch": 0.2354, + "grad_norm": 1.380194902420044, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 23540 + }, + { + "epoch": 0.23541, + "grad_norm": 0.7049074769020081, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 23541 + }, + { + "epoch": 0.23542, + "grad_norm": 0.7330548763275146, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 23542 + }, + { + "epoch": 0.23543, + "grad_norm": 0.8423771262168884, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 23543 + }, + { + "epoch": 0.23544, + "grad_norm": 0.919819712638855, + "learning_rate": 0.003, + "loss": 3.989, + "step": 23544 + }, + { + "epoch": 0.23545, + "grad_norm": 0.9021384119987488, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 23545 + }, + { + "epoch": 0.23546, + "grad_norm": 0.834520697593689, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 23546 + }, + { + "epoch": 0.23547, + "grad_norm": 0.7668264508247375, + "learning_rate": 0.003, + "loss": 4.002, + "step": 23547 + }, + { + "epoch": 0.23548, + "grad_norm": 0.710982620716095, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 23548 + }, + { + "epoch": 0.23549, + "grad_norm": 0.707950234413147, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 23549 + }, + { + "epoch": 0.2355, + "grad_norm": 0.820759117603302, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 23550 + }, + { + "epoch": 0.23551, + "grad_norm": 1.0242942571640015, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 23551 + }, + { + "epoch": 0.23552, + "grad_norm": 0.9243529438972473, + "learning_rate": 0.003, + "loss": 3.997, + "step": 23552 + }, + { + "epoch": 0.23553, + "grad_norm": 0.7698137760162354, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 23553 + }, + { + "epoch": 0.23554, + "grad_norm": 0.6879913210868835, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 23554 + }, + { + "epoch": 0.23555, + "grad_norm": 0.7060564756393433, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 23555 + }, + { + "epoch": 0.23556, + "grad_norm": 0.7667385935783386, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 23556 + }, + { + "epoch": 0.23557, + "grad_norm": 0.7614707350730896, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 23557 + }, + { + "epoch": 0.23558, + "grad_norm": 0.7831237316131592, + "learning_rate": 0.003, + "loss": 3.995, + "step": 23558 + }, + { + "epoch": 0.23559, + "grad_norm": 0.801593005657196, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 23559 + }, + { + "epoch": 0.2356, + "grad_norm": 0.9975457787513733, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 23560 + }, + { + "epoch": 0.23561, + "grad_norm": 1.1727756261825562, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 23561 + }, + { + "epoch": 0.23562, + "grad_norm": 0.8991934657096863, + "learning_rate": 0.003, + "loss": 3.988, + "step": 23562 + }, + { + "epoch": 0.23563, + "grad_norm": 0.8032999634742737, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 23563 + }, + { + "epoch": 0.23564, + "grad_norm": 0.80031818151474, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 23564 + }, + { + "epoch": 0.23565, + "grad_norm": 0.721490740776062, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 23565 + }, + { + "epoch": 0.23566, + "grad_norm": 0.7022581100463867, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 23566 + }, + { + "epoch": 0.23567, + "grad_norm": 0.8682166934013367, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 23567 + }, + { + "epoch": 0.23568, + "grad_norm": 0.940638542175293, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 23568 + }, + { + "epoch": 0.23569, + "grad_norm": 1.1110423803329468, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 23569 + }, + { + "epoch": 0.2357, + "grad_norm": 0.8911081552505493, + "learning_rate": 0.003, + "loss": 3.994, + "step": 23570 + }, + { + "epoch": 0.23571, + "grad_norm": 0.7177302241325378, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 23571 + }, + { + "epoch": 0.23572, + "grad_norm": 0.6688079237937927, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 23572 + }, + { + "epoch": 0.23573, + "grad_norm": 0.6945911049842834, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 23573 + }, + { + "epoch": 0.23574, + "grad_norm": 0.7250649929046631, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 23574 + }, + { + "epoch": 0.23575, + "grad_norm": 0.8393401503562927, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 23575 + }, + { + "epoch": 0.23576, + "grad_norm": 1.0022213459014893, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 23576 + }, + { + "epoch": 0.23577, + "grad_norm": 1.0669891834259033, + "learning_rate": 0.003, + "loss": 3.999, + "step": 23577 + }, + { + "epoch": 0.23578, + "grad_norm": 0.8827222585678101, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 23578 + }, + { + "epoch": 0.23579, + "grad_norm": 0.7493778467178345, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 23579 + }, + { + "epoch": 0.2358, + "grad_norm": 0.7480388879776001, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 23580 + }, + { + "epoch": 0.23581, + "grad_norm": 0.9012417197227478, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 23581 + }, + { + "epoch": 0.23582, + "grad_norm": 1.190975308418274, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 23582 + }, + { + "epoch": 0.23583, + "grad_norm": 0.8099722266197205, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 23583 + }, + { + "epoch": 0.23584, + "grad_norm": 0.5927342772483826, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 23584 + }, + { + "epoch": 0.23585, + "grad_norm": 0.6614275574684143, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 23585 + }, + { + "epoch": 0.23586, + "grad_norm": 0.7221860885620117, + "learning_rate": 0.003, + "loss": 3.971, + "step": 23586 + }, + { + "epoch": 0.23587, + "grad_norm": 0.7274786233901978, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 23587 + }, + { + "epoch": 0.23588, + "grad_norm": 0.8059537410736084, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 23588 + }, + { + "epoch": 0.23589, + "grad_norm": 0.8644411563873291, + "learning_rate": 0.003, + "loss": 3.998, + "step": 23589 + }, + { + "epoch": 0.2359, + "grad_norm": 1.0203708410263062, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 23590 + }, + { + "epoch": 0.23591, + "grad_norm": 1.226479411125183, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 23591 + }, + { + "epoch": 0.23592, + "grad_norm": 0.8996565937995911, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 23592 + }, + { + "epoch": 0.23593, + "grad_norm": 0.766966700553894, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 23593 + }, + { + "epoch": 0.23594, + "grad_norm": 0.7482280135154724, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 23594 + }, + { + "epoch": 0.23595, + "grad_norm": 0.7280831336975098, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 23595 + }, + { + "epoch": 0.23596, + "grad_norm": 0.8845469951629639, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 23596 + }, + { + "epoch": 0.23597, + "grad_norm": 0.9400532245635986, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 23597 + }, + { + "epoch": 0.23598, + "grad_norm": 1.0305627584457397, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 23598 + }, + { + "epoch": 0.23599, + "grad_norm": 1.094538688659668, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 23599 + }, + { + "epoch": 0.236, + "grad_norm": 0.8124465346336365, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 23600 + }, + { + "epoch": 0.23601, + "grad_norm": 0.9502313733100891, + "learning_rate": 0.003, + "loss": 3.983, + "step": 23601 + }, + { + "epoch": 0.23602, + "grad_norm": 1.0658084154129028, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 23602 + }, + { + "epoch": 0.23603, + "grad_norm": 0.9293392896652222, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 23603 + }, + { + "epoch": 0.23604, + "grad_norm": 0.7132201194763184, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 23604 + }, + { + "epoch": 0.23605, + "grad_norm": 0.6419816017150879, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 23605 + }, + { + "epoch": 0.23606, + "grad_norm": 0.6787327527999878, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 23606 + }, + { + "epoch": 0.23607, + "grad_norm": 0.8710855841636658, + "learning_rate": 0.003, + "loss": 4.013, + "step": 23607 + }, + { + "epoch": 0.23608, + "grad_norm": 1.1132811307907104, + "learning_rate": 0.003, + "loss": 3.975, + "step": 23608 + }, + { + "epoch": 0.23609, + "grad_norm": 1.1425007581710815, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 23609 + }, + { + "epoch": 0.2361, + "grad_norm": 0.8493500351905823, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 23610 + }, + { + "epoch": 0.23611, + "grad_norm": 0.7826309204101562, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 23611 + }, + { + "epoch": 0.23612, + "grad_norm": 0.7571776509284973, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 23612 + }, + { + "epoch": 0.23613, + "grad_norm": 0.7528085112571716, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 23613 + }, + { + "epoch": 0.23614, + "grad_norm": 0.7031254768371582, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 23614 + }, + { + "epoch": 0.23615, + "grad_norm": 0.7295961976051331, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 23615 + }, + { + "epoch": 0.23616, + "grad_norm": 0.7955227494239807, + "learning_rate": 0.003, + "loss": 3.978, + "step": 23616 + }, + { + "epoch": 0.23617, + "grad_norm": 0.7214534878730774, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 23617 + }, + { + "epoch": 0.23618, + "grad_norm": 0.7763646245002747, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 23618 + }, + { + "epoch": 0.23619, + "grad_norm": 0.7673700451850891, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 23619 + }, + { + "epoch": 0.2362, + "grad_norm": 0.707784116268158, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 23620 + }, + { + "epoch": 0.23621, + "grad_norm": 0.8049967288970947, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 23621 + }, + { + "epoch": 0.23622, + "grad_norm": 0.8459908962249756, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 23622 + }, + { + "epoch": 0.23623, + "grad_norm": 0.745107889175415, + "learning_rate": 0.003, + "loss": 3.957, + "step": 23623 + }, + { + "epoch": 0.23624, + "grad_norm": 0.6760168075561523, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 23624 + }, + { + "epoch": 0.23625, + "grad_norm": 0.6307828426361084, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 23625 + }, + { + "epoch": 0.23626, + "grad_norm": 0.6267635226249695, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 23626 + }, + { + "epoch": 0.23627, + "grad_norm": 0.7318971753120422, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 23627 + }, + { + "epoch": 0.23628, + "grad_norm": 0.9262677431106567, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 23628 + }, + { + "epoch": 0.23629, + "grad_norm": 1.2221388816833496, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 23629 + }, + { + "epoch": 0.2363, + "grad_norm": 1.2877801656723022, + "learning_rate": 0.003, + "loss": 4.015, + "step": 23630 + }, + { + "epoch": 0.23631, + "grad_norm": 0.7645659446716309, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 23631 + }, + { + "epoch": 0.23632, + "grad_norm": 0.6646572351455688, + "learning_rate": 0.003, + "loss": 3.983, + "step": 23632 + }, + { + "epoch": 0.23633, + "grad_norm": 0.6736848950386047, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 23633 + }, + { + "epoch": 0.23634, + "grad_norm": 0.7057893872261047, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 23634 + }, + { + "epoch": 0.23635, + "grad_norm": 0.7268689870834351, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 23635 + }, + { + "epoch": 0.23636, + "grad_norm": 0.85333251953125, + "learning_rate": 0.003, + "loss": 3.993, + "step": 23636 + }, + { + "epoch": 0.23637, + "grad_norm": 1.004431962966919, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 23637 + }, + { + "epoch": 0.23638, + "grad_norm": 1.206244707107544, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 23638 + }, + { + "epoch": 0.23639, + "grad_norm": 0.8943990468978882, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 23639 + }, + { + "epoch": 0.2364, + "grad_norm": 0.8737308979034424, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 23640 + }, + { + "epoch": 0.23641, + "grad_norm": 0.9814113974571228, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 23641 + }, + { + "epoch": 0.23642, + "grad_norm": 1.0602644681930542, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 23642 + }, + { + "epoch": 0.23643, + "grad_norm": 0.8849413990974426, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 23643 + }, + { + "epoch": 0.23644, + "grad_norm": 0.8219353556632996, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 23644 + }, + { + "epoch": 0.23645, + "grad_norm": 0.8834403157234192, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 23645 + }, + { + "epoch": 0.23646, + "grad_norm": 0.9699289798736572, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 23646 + }, + { + "epoch": 0.23647, + "grad_norm": 0.940680980682373, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 23647 + }, + { + "epoch": 0.23648, + "grad_norm": 0.8660580515861511, + "learning_rate": 0.003, + "loss": 3.995, + "step": 23648 + }, + { + "epoch": 0.23649, + "grad_norm": 0.8483887314796448, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 23649 + }, + { + "epoch": 0.2365, + "grad_norm": 0.8241614103317261, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 23650 + }, + { + "epoch": 0.23651, + "grad_norm": 0.9246995449066162, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 23651 + }, + { + "epoch": 0.23652, + "grad_norm": 0.8324962258338928, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 23652 + }, + { + "epoch": 0.23653, + "grad_norm": 0.7726301550865173, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 23653 + }, + { + "epoch": 0.23654, + "grad_norm": 0.9132267236709595, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 23654 + }, + { + "epoch": 0.23655, + "grad_norm": 0.9978023171424866, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 23655 + }, + { + "epoch": 0.23656, + "grad_norm": 1.0363349914550781, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 23656 + }, + { + "epoch": 0.23657, + "grad_norm": 1.357732892036438, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 23657 + }, + { + "epoch": 0.23658, + "grad_norm": 1.0219706296920776, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 23658 + }, + { + "epoch": 0.23659, + "grad_norm": 0.86076420545578, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 23659 + }, + { + "epoch": 0.2366, + "grad_norm": 0.7942736744880676, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 23660 + }, + { + "epoch": 0.23661, + "grad_norm": 0.6746869683265686, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 23661 + }, + { + "epoch": 0.23662, + "grad_norm": 0.7790113091468811, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 23662 + }, + { + "epoch": 0.23663, + "grad_norm": 0.885614275932312, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 23663 + }, + { + "epoch": 0.23664, + "grad_norm": 1.004623293876648, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 23664 + }, + { + "epoch": 0.23665, + "grad_norm": 1.0761792659759521, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 23665 + }, + { + "epoch": 0.23666, + "grad_norm": 1.0219818353652954, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 23666 + }, + { + "epoch": 0.23667, + "grad_norm": 1.087648868560791, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 23667 + }, + { + "epoch": 0.23668, + "grad_norm": 0.9711748361587524, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 23668 + }, + { + "epoch": 0.23669, + "grad_norm": 0.9047676920890808, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 23669 + }, + { + "epoch": 0.2367, + "grad_norm": 0.9903032183647156, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 23670 + }, + { + "epoch": 0.23671, + "grad_norm": 1.0302881002426147, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 23671 + }, + { + "epoch": 0.23672, + "grad_norm": 0.9264562129974365, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 23672 + }, + { + "epoch": 0.23673, + "grad_norm": 0.7811916470527649, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 23673 + }, + { + "epoch": 0.23674, + "grad_norm": 0.6379555463790894, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 23674 + }, + { + "epoch": 0.23675, + "grad_norm": 0.6818837523460388, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 23675 + }, + { + "epoch": 0.23676, + "grad_norm": 0.7744259834289551, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 23676 + }, + { + "epoch": 0.23677, + "grad_norm": 0.7656677961349487, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 23677 + }, + { + "epoch": 0.23678, + "grad_norm": 0.7243162393569946, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 23678 + }, + { + "epoch": 0.23679, + "grad_norm": 0.7910912036895752, + "learning_rate": 0.003, + "loss": 3.991, + "step": 23679 + }, + { + "epoch": 0.2368, + "grad_norm": 0.8631349802017212, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 23680 + }, + { + "epoch": 0.23681, + "grad_norm": 1.0868220329284668, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 23681 + }, + { + "epoch": 0.23682, + "grad_norm": 1.1420514583587646, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 23682 + }, + { + "epoch": 0.23683, + "grad_norm": 0.6462451219558716, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 23683 + }, + { + "epoch": 0.23684, + "grad_norm": 0.5510276556015015, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 23684 + }, + { + "epoch": 0.23685, + "grad_norm": 0.6520527601242065, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 23685 + }, + { + "epoch": 0.23686, + "grad_norm": 0.5974164605140686, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 23686 + }, + { + "epoch": 0.23687, + "grad_norm": 0.6888923048973083, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 23687 + }, + { + "epoch": 0.23688, + "grad_norm": 0.8432384729385376, + "learning_rate": 0.003, + "loss": 3.998, + "step": 23688 + }, + { + "epoch": 0.23689, + "grad_norm": 1.0016157627105713, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 23689 + }, + { + "epoch": 0.2369, + "grad_norm": 1.0528491735458374, + "learning_rate": 0.003, + "loss": 4.019, + "step": 23690 + }, + { + "epoch": 0.23691, + "grad_norm": 0.908726155757904, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 23691 + }, + { + "epoch": 0.23692, + "grad_norm": 0.7770360112190247, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 23692 + }, + { + "epoch": 0.23693, + "grad_norm": 0.7631237506866455, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 23693 + }, + { + "epoch": 0.23694, + "grad_norm": 0.8512758016586304, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 23694 + }, + { + "epoch": 0.23695, + "grad_norm": 0.8578788638114929, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 23695 + }, + { + "epoch": 0.23696, + "grad_norm": 0.9126070737838745, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 23696 + }, + { + "epoch": 0.23697, + "grad_norm": 0.9172295928001404, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 23697 + }, + { + "epoch": 0.23698, + "grad_norm": 1.0368863344192505, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 23698 + }, + { + "epoch": 0.23699, + "grad_norm": 0.816555380821228, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 23699 + }, + { + "epoch": 0.237, + "grad_norm": 0.7659728527069092, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 23700 + }, + { + "epoch": 0.23701, + "grad_norm": 0.7375189065933228, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 23701 + }, + { + "epoch": 0.23702, + "grad_norm": 0.9682885408401489, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 23702 + }, + { + "epoch": 0.23703, + "grad_norm": 0.9783527255058289, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 23703 + }, + { + "epoch": 0.23704, + "grad_norm": 0.847342848777771, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 23704 + }, + { + "epoch": 0.23705, + "grad_norm": 0.7268058061599731, + "learning_rate": 0.003, + "loss": 3.968, + "step": 23705 + }, + { + "epoch": 0.23706, + "grad_norm": 0.7453255653381348, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 23706 + }, + { + "epoch": 0.23707, + "grad_norm": 0.6553342342376709, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 23707 + }, + { + "epoch": 0.23708, + "grad_norm": 0.6369809508323669, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 23708 + }, + { + "epoch": 0.23709, + "grad_norm": 0.749583899974823, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 23709 + }, + { + "epoch": 0.2371, + "grad_norm": 0.7410933375358582, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 23710 + }, + { + "epoch": 0.23711, + "grad_norm": 0.7737413048744202, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 23711 + }, + { + "epoch": 0.23712, + "grad_norm": 0.843481183052063, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 23712 + }, + { + "epoch": 0.23713, + "grad_norm": 0.8842707276344299, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 23713 + }, + { + "epoch": 0.23714, + "grad_norm": 1.0029457807540894, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 23714 + }, + { + "epoch": 0.23715, + "grad_norm": 1.2244681119918823, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 23715 + }, + { + "epoch": 0.23716, + "grad_norm": 0.9904945492744446, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 23716 + }, + { + "epoch": 0.23717, + "grad_norm": 1.1552927494049072, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 23717 + }, + { + "epoch": 0.23718, + "grad_norm": 0.970582127571106, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 23718 + }, + { + "epoch": 0.23719, + "grad_norm": 1.020983338356018, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 23719 + }, + { + "epoch": 0.2372, + "grad_norm": 1.1324379444122314, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 23720 + }, + { + "epoch": 0.23721, + "grad_norm": 0.8088292479515076, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 23721 + }, + { + "epoch": 0.23722, + "grad_norm": 0.6515463590621948, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 23722 + }, + { + "epoch": 0.23723, + "grad_norm": 0.6463795900344849, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 23723 + }, + { + "epoch": 0.23724, + "grad_norm": 0.6604560017585754, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 23724 + }, + { + "epoch": 0.23725, + "grad_norm": 0.6887049674987793, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 23725 + }, + { + "epoch": 0.23726, + "grad_norm": 0.7006163597106934, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 23726 + }, + { + "epoch": 0.23727, + "grad_norm": 0.8166240453720093, + "learning_rate": 0.003, + "loss": 3.998, + "step": 23727 + }, + { + "epoch": 0.23728, + "grad_norm": 1.0572659969329834, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 23728 + }, + { + "epoch": 0.23729, + "grad_norm": 1.0111148357391357, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 23729 + }, + { + "epoch": 0.2373, + "grad_norm": 0.8641826510429382, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 23730 + }, + { + "epoch": 0.23731, + "grad_norm": 0.8890543580055237, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 23731 + }, + { + "epoch": 0.23732, + "grad_norm": 0.9322623610496521, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 23732 + }, + { + "epoch": 0.23733, + "grad_norm": 0.8937170505523682, + "learning_rate": 0.003, + "loss": 4.018, + "step": 23733 + }, + { + "epoch": 0.23734, + "grad_norm": 0.7896233797073364, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 23734 + }, + { + "epoch": 0.23735, + "grad_norm": 0.842714786529541, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 23735 + }, + { + "epoch": 0.23736, + "grad_norm": 0.8911498188972473, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 23736 + }, + { + "epoch": 0.23737, + "grad_norm": 1.139247179031372, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 23737 + }, + { + "epoch": 0.23738, + "grad_norm": 0.9641310572624207, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 23738 + }, + { + "epoch": 0.23739, + "grad_norm": 0.9660186767578125, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 23739 + }, + { + "epoch": 0.2374, + "grad_norm": 0.9905936121940613, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 23740 + }, + { + "epoch": 0.23741, + "grad_norm": 0.963261604309082, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 23741 + }, + { + "epoch": 0.23742, + "grad_norm": 0.8250371813774109, + "learning_rate": 0.003, + "loss": 3.992, + "step": 23742 + }, + { + "epoch": 0.23743, + "grad_norm": 0.7518148422241211, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 23743 + }, + { + "epoch": 0.23744, + "grad_norm": 0.7974687218666077, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 23744 + }, + { + "epoch": 0.23745, + "grad_norm": 0.7960631251335144, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 23745 + }, + { + "epoch": 0.23746, + "grad_norm": 0.8922495245933533, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 23746 + }, + { + "epoch": 0.23747, + "grad_norm": 0.9672883749008179, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 23747 + }, + { + "epoch": 0.23748, + "grad_norm": 0.870411217212677, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 23748 + }, + { + "epoch": 0.23749, + "grad_norm": 0.8624403476715088, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 23749 + }, + { + "epoch": 0.2375, + "grad_norm": 0.8989052176475525, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 23750 + }, + { + "epoch": 0.23751, + "grad_norm": 0.8660366535186768, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 23751 + }, + { + "epoch": 0.23752, + "grad_norm": 1.0035899877548218, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 23752 + }, + { + "epoch": 0.23753, + "grad_norm": 1.062063217163086, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 23753 + }, + { + "epoch": 0.23754, + "grad_norm": 0.8284898996353149, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 23754 + }, + { + "epoch": 0.23755, + "grad_norm": 0.8423393964767456, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 23755 + }, + { + "epoch": 0.23756, + "grad_norm": 0.9343189597129822, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 23756 + }, + { + "epoch": 0.23757, + "grad_norm": 1.1121056079864502, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 23757 + }, + { + "epoch": 0.23758, + "grad_norm": 0.8807165622711182, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 23758 + }, + { + "epoch": 0.23759, + "grad_norm": 0.8449285626411438, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 23759 + }, + { + "epoch": 0.2376, + "grad_norm": 0.963933527469635, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 23760 + }, + { + "epoch": 0.23761, + "grad_norm": 1.0363460779190063, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 23761 + }, + { + "epoch": 0.23762, + "grad_norm": 0.952482283115387, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 23762 + }, + { + "epoch": 0.23763, + "grad_norm": 0.8761563897132874, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 23763 + }, + { + "epoch": 0.23764, + "grad_norm": 0.7540985345840454, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 23764 + }, + { + "epoch": 0.23765, + "grad_norm": 0.7391047477722168, + "learning_rate": 0.003, + "loss": 3.997, + "step": 23765 + }, + { + "epoch": 0.23766, + "grad_norm": 0.6441829204559326, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 23766 + }, + { + "epoch": 0.23767, + "grad_norm": 0.5878599882125854, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 23767 + }, + { + "epoch": 0.23768, + "grad_norm": 0.6302638649940491, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 23768 + }, + { + "epoch": 0.23769, + "grad_norm": 0.6156554222106934, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 23769 + }, + { + "epoch": 0.2377, + "grad_norm": 0.6710278391838074, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 23770 + }, + { + "epoch": 0.23771, + "grad_norm": 0.7719600796699524, + "learning_rate": 0.003, + "loss": 3.973, + "step": 23771 + }, + { + "epoch": 0.23772, + "grad_norm": 1.015855312347412, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 23772 + }, + { + "epoch": 0.23773, + "grad_norm": 1.182342290878296, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 23773 + }, + { + "epoch": 0.23774, + "grad_norm": 0.6234725713729858, + "learning_rate": 0.003, + "loss": 3.99, + "step": 23774 + }, + { + "epoch": 0.23775, + "grad_norm": 0.5669736266136169, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 23775 + }, + { + "epoch": 0.23776, + "grad_norm": 0.7421907782554626, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 23776 + }, + { + "epoch": 0.23777, + "grad_norm": 0.8842352032661438, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 23777 + }, + { + "epoch": 0.23778, + "grad_norm": 1.0764153003692627, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 23778 + }, + { + "epoch": 0.23779, + "grad_norm": 0.8120744824409485, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 23779 + }, + { + "epoch": 0.2378, + "grad_norm": 0.6154041886329651, + "learning_rate": 0.003, + "loss": 3.9403, + "step": 23780 + }, + { + "epoch": 0.23781, + "grad_norm": 0.6963661909103394, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 23781 + }, + { + "epoch": 0.23782, + "grad_norm": 0.7057574391365051, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 23782 + }, + { + "epoch": 0.23783, + "grad_norm": 0.8826019763946533, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 23783 + }, + { + "epoch": 0.23784, + "grad_norm": 1.1662051677703857, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 23784 + }, + { + "epoch": 0.23785, + "grad_norm": 1.0353947877883911, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 23785 + }, + { + "epoch": 0.23786, + "grad_norm": 0.871455192565918, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 23786 + }, + { + "epoch": 0.23787, + "grad_norm": 0.8273305296897888, + "learning_rate": 0.003, + "loss": 3.995, + "step": 23787 + }, + { + "epoch": 0.23788, + "grad_norm": 0.9897767901420593, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 23788 + }, + { + "epoch": 0.23789, + "grad_norm": 0.9893234968185425, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 23789 + }, + { + "epoch": 0.2379, + "grad_norm": 1.0915952920913696, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 23790 + }, + { + "epoch": 0.23791, + "grad_norm": 1.067588448524475, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 23791 + }, + { + "epoch": 0.23792, + "grad_norm": 0.897010087966919, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 23792 + }, + { + "epoch": 0.23793, + "grad_norm": 0.9177330732345581, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 23793 + }, + { + "epoch": 0.23794, + "grad_norm": 0.9199936985969543, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 23794 + }, + { + "epoch": 0.23795, + "grad_norm": 0.9021243453025818, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 23795 + }, + { + "epoch": 0.23796, + "grad_norm": 0.9108153581619263, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 23796 + }, + { + "epoch": 0.23797, + "grad_norm": 0.9284119009971619, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 23797 + }, + { + "epoch": 0.23798, + "grad_norm": 1.0781867504119873, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 23798 + }, + { + "epoch": 0.23799, + "grad_norm": 1.1488069295883179, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 23799 + }, + { + "epoch": 0.238, + "grad_norm": 0.9300514459609985, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 23800 + }, + { + "epoch": 0.23801, + "grad_norm": 0.8805649280548096, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 23801 + }, + { + "epoch": 0.23802, + "grad_norm": 0.8955497145652771, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 23802 + }, + { + "epoch": 0.23803, + "grad_norm": 1.0028194189071655, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 23803 + }, + { + "epoch": 0.23804, + "grad_norm": 0.9778033494949341, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 23804 + }, + { + "epoch": 0.23805, + "grad_norm": 0.9503707885742188, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 23805 + }, + { + "epoch": 0.23806, + "grad_norm": 1.02264404296875, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 23806 + }, + { + "epoch": 0.23807, + "grad_norm": 1.0590420961380005, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 23807 + }, + { + "epoch": 0.23808, + "grad_norm": 0.9413914084434509, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 23808 + }, + { + "epoch": 0.23809, + "grad_norm": 1.075340747833252, + "learning_rate": 0.003, + "loss": 4.009, + "step": 23809 + }, + { + "epoch": 0.2381, + "grad_norm": 0.8629778027534485, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 23810 + }, + { + "epoch": 0.23811, + "grad_norm": 0.752785325050354, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 23811 + }, + { + "epoch": 0.23812, + "grad_norm": 0.6868883371353149, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 23812 + }, + { + "epoch": 0.23813, + "grad_norm": 0.6632366180419922, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 23813 + }, + { + "epoch": 0.23814, + "grad_norm": 0.7330594658851624, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 23814 + }, + { + "epoch": 0.23815, + "grad_norm": 0.7809762358665466, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 23815 + }, + { + "epoch": 0.23816, + "grad_norm": 0.8607485890388489, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 23816 + }, + { + "epoch": 0.23817, + "grad_norm": 0.9762835502624512, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 23817 + }, + { + "epoch": 0.23818, + "grad_norm": 0.9260298609733582, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 23818 + }, + { + "epoch": 0.23819, + "grad_norm": 1.0767014026641846, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 23819 + }, + { + "epoch": 0.2382, + "grad_norm": 0.9877113103866577, + "learning_rate": 0.003, + "loss": 3.999, + "step": 23820 + }, + { + "epoch": 0.23821, + "grad_norm": 0.9498253464698792, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 23821 + }, + { + "epoch": 0.23822, + "grad_norm": 0.8559045195579529, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 23822 + }, + { + "epoch": 0.23823, + "grad_norm": 0.7979329228401184, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 23823 + }, + { + "epoch": 0.23824, + "grad_norm": 0.8390047550201416, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 23824 + }, + { + "epoch": 0.23825, + "grad_norm": 0.932802140712738, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 23825 + }, + { + "epoch": 0.23826, + "grad_norm": 0.8875012993812561, + "learning_rate": 0.003, + "loss": 3.971, + "step": 23826 + }, + { + "epoch": 0.23827, + "grad_norm": 0.728776752948761, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 23827 + }, + { + "epoch": 0.23828, + "grad_norm": 0.7667000889778137, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 23828 + }, + { + "epoch": 0.23829, + "grad_norm": 0.6944807767868042, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 23829 + }, + { + "epoch": 0.2383, + "grad_norm": 0.6778382062911987, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 23830 + }, + { + "epoch": 0.23831, + "grad_norm": 0.7424401640892029, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 23831 + }, + { + "epoch": 0.23832, + "grad_norm": 0.8382511734962463, + "learning_rate": 0.003, + "loss": 3.992, + "step": 23832 + }, + { + "epoch": 0.23833, + "grad_norm": 0.9459271430969238, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 23833 + }, + { + "epoch": 0.23834, + "grad_norm": 0.9476693272590637, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 23834 + }, + { + "epoch": 0.23835, + "grad_norm": 0.7799937129020691, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 23835 + }, + { + "epoch": 0.23836, + "grad_norm": 0.6064723134040833, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 23836 + }, + { + "epoch": 0.23837, + "grad_norm": 0.581373393535614, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 23837 + }, + { + "epoch": 0.23838, + "grad_norm": 0.64168381690979, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 23838 + }, + { + "epoch": 0.23839, + "grad_norm": 0.7018834352493286, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 23839 + }, + { + "epoch": 0.2384, + "grad_norm": 0.7506743669509888, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 23840 + }, + { + "epoch": 0.23841, + "grad_norm": 0.8276138305664062, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 23841 + }, + { + "epoch": 0.23842, + "grad_norm": 0.9328670501708984, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 23842 + }, + { + "epoch": 0.23843, + "grad_norm": 1.0450549125671387, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 23843 + }, + { + "epoch": 0.23844, + "grad_norm": 0.8692760467529297, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 23844 + }, + { + "epoch": 0.23845, + "grad_norm": 0.8682751655578613, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 23845 + }, + { + "epoch": 0.23846, + "grad_norm": 0.9090334177017212, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 23846 + }, + { + "epoch": 0.23847, + "grad_norm": 0.9573761820793152, + "learning_rate": 0.003, + "loss": 3.993, + "step": 23847 + }, + { + "epoch": 0.23848, + "grad_norm": 0.9086956977844238, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 23848 + }, + { + "epoch": 0.23849, + "grad_norm": 0.9204684495925903, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 23849 + }, + { + "epoch": 0.2385, + "grad_norm": 0.8970891833305359, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 23850 + }, + { + "epoch": 0.23851, + "grad_norm": 0.8092312812805176, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 23851 + }, + { + "epoch": 0.23852, + "grad_norm": 0.7760641574859619, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 23852 + }, + { + "epoch": 0.23853, + "grad_norm": 0.8049925565719604, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 23853 + }, + { + "epoch": 0.23854, + "grad_norm": 0.9391586184501648, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 23854 + }, + { + "epoch": 0.23855, + "grad_norm": 1.073109745979309, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 23855 + }, + { + "epoch": 0.23856, + "grad_norm": 0.9554718732833862, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 23856 + }, + { + "epoch": 0.23857, + "grad_norm": 0.9953920245170593, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 23857 + }, + { + "epoch": 0.23858, + "grad_norm": 1.0308576822280884, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 23858 + }, + { + "epoch": 0.23859, + "grad_norm": 0.831104576587677, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 23859 + }, + { + "epoch": 0.2386, + "grad_norm": 0.7946001887321472, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 23860 + }, + { + "epoch": 0.23861, + "grad_norm": 0.7132794260978699, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 23861 + }, + { + "epoch": 0.23862, + "grad_norm": 0.612296998500824, + "learning_rate": 0.003, + "loss": 3.977, + "step": 23862 + }, + { + "epoch": 0.23863, + "grad_norm": 0.6082380414009094, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 23863 + }, + { + "epoch": 0.23864, + "grad_norm": 0.6919535994529724, + "learning_rate": 0.003, + "loss": 3.986, + "step": 23864 + }, + { + "epoch": 0.23865, + "grad_norm": 0.8079025745391846, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 23865 + }, + { + "epoch": 0.23866, + "grad_norm": 1.0395094156265259, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 23866 + }, + { + "epoch": 0.23867, + "grad_norm": 1.4624147415161133, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 23867 + }, + { + "epoch": 0.23868, + "grad_norm": 0.8247044086456299, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 23868 + }, + { + "epoch": 0.23869, + "grad_norm": 0.6952419281005859, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 23869 + }, + { + "epoch": 0.2387, + "grad_norm": 0.6948174238204956, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 23870 + }, + { + "epoch": 0.23871, + "grad_norm": 0.6716610789299011, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 23871 + }, + { + "epoch": 0.23872, + "grad_norm": 0.7647894620895386, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 23872 + }, + { + "epoch": 0.23873, + "grad_norm": 0.7906863689422607, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 23873 + }, + { + "epoch": 0.23874, + "grad_norm": 0.7799371480941772, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 23874 + }, + { + "epoch": 0.23875, + "grad_norm": 0.8083313703536987, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 23875 + }, + { + "epoch": 0.23876, + "grad_norm": 1.06098210811615, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 23876 + }, + { + "epoch": 0.23877, + "grad_norm": 1.115160584449768, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 23877 + }, + { + "epoch": 0.23878, + "grad_norm": 0.8913708329200745, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 23878 + }, + { + "epoch": 0.23879, + "grad_norm": 0.7398980259895325, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 23879 + }, + { + "epoch": 0.2388, + "grad_norm": 0.7704770565032959, + "learning_rate": 0.003, + "loss": 3.993, + "step": 23880 + }, + { + "epoch": 0.23881, + "grad_norm": 0.8696398138999939, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 23881 + }, + { + "epoch": 0.23882, + "grad_norm": 0.8584381341934204, + "learning_rate": 0.003, + "loss": 3.985, + "step": 23882 + }, + { + "epoch": 0.23883, + "grad_norm": 0.9413838386535645, + "learning_rate": 0.003, + "loss": 4.026, + "step": 23883 + }, + { + "epoch": 0.23884, + "grad_norm": 0.921601414680481, + "learning_rate": 0.003, + "loss": 3.998, + "step": 23884 + }, + { + "epoch": 0.23885, + "grad_norm": 0.9314766526222229, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 23885 + }, + { + "epoch": 0.23886, + "grad_norm": 0.8797277212142944, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 23886 + }, + { + "epoch": 0.23887, + "grad_norm": 0.8425590991973877, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 23887 + }, + { + "epoch": 0.23888, + "grad_norm": 0.9020876884460449, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 23888 + }, + { + "epoch": 0.23889, + "grad_norm": 0.9022796154022217, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 23889 + }, + { + "epoch": 0.2389, + "grad_norm": 1.0433920621871948, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 23890 + }, + { + "epoch": 0.23891, + "grad_norm": 0.8587849140167236, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 23891 + }, + { + "epoch": 0.23892, + "grad_norm": 0.7801781892776489, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 23892 + }, + { + "epoch": 0.23893, + "grad_norm": 0.7614020109176636, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 23893 + }, + { + "epoch": 0.23894, + "grad_norm": 0.8398752212524414, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 23894 + }, + { + "epoch": 0.23895, + "grad_norm": 0.7848305702209473, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 23895 + }, + { + "epoch": 0.23896, + "grad_norm": 0.7347824573516846, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 23896 + }, + { + "epoch": 0.23897, + "grad_norm": 0.7922550439834595, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 23897 + }, + { + "epoch": 0.23898, + "grad_norm": 0.9099634885787964, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 23898 + }, + { + "epoch": 0.23899, + "grad_norm": 1.1431479454040527, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 23899 + }, + { + "epoch": 0.239, + "grad_norm": 0.9483999013900757, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 23900 + }, + { + "epoch": 0.23901, + "grad_norm": 0.9486942291259766, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 23901 + }, + { + "epoch": 0.23902, + "grad_norm": 0.8391926884651184, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 23902 + }, + { + "epoch": 0.23903, + "grad_norm": 0.842767596244812, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 23903 + }, + { + "epoch": 0.23904, + "grad_norm": 1.0417910814285278, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 23904 + }, + { + "epoch": 0.23905, + "grad_norm": 1.0539915561676025, + "learning_rate": 0.003, + "loss": 3.996, + "step": 23905 + }, + { + "epoch": 0.23906, + "grad_norm": 0.8192107081413269, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 23906 + }, + { + "epoch": 0.23907, + "grad_norm": 0.911364734172821, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 23907 + }, + { + "epoch": 0.23908, + "grad_norm": 0.9505627751350403, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 23908 + }, + { + "epoch": 0.23909, + "grad_norm": 1.1416152715682983, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 23909 + }, + { + "epoch": 0.2391, + "grad_norm": 0.9701867699623108, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 23910 + }, + { + "epoch": 0.23911, + "grad_norm": 1.0680506229400635, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 23911 + }, + { + "epoch": 0.23912, + "grad_norm": 0.9237954020500183, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 23912 + }, + { + "epoch": 0.23913, + "grad_norm": 0.6470546722412109, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 23913 + }, + { + "epoch": 0.23914, + "grad_norm": 0.6414022445678711, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 23914 + }, + { + "epoch": 0.23915, + "grad_norm": 0.7368568778038025, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 23915 + }, + { + "epoch": 0.23916, + "grad_norm": 0.8634165525436401, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 23916 + }, + { + "epoch": 0.23917, + "grad_norm": 0.7824901938438416, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 23917 + }, + { + "epoch": 0.23918, + "grad_norm": 0.8770608305931091, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 23918 + }, + { + "epoch": 0.23919, + "grad_norm": 0.9526880383491516, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 23919 + }, + { + "epoch": 0.2392, + "grad_norm": 0.9670585989952087, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 23920 + }, + { + "epoch": 0.23921, + "grad_norm": 0.7598057985305786, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 23921 + }, + { + "epoch": 0.23922, + "grad_norm": 0.586688220500946, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 23922 + }, + { + "epoch": 0.23923, + "grad_norm": 0.6842291355133057, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 23923 + }, + { + "epoch": 0.23924, + "grad_norm": 0.7313446402549744, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 23924 + }, + { + "epoch": 0.23925, + "grad_norm": 0.7644690275192261, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 23925 + }, + { + "epoch": 0.23926, + "grad_norm": 0.8702702522277832, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 23926 + }, + { + "epoch": 0.23927, + "grad_norm": 0.7801819443702698, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 23927 + }, + { + "epoch": 0.23928, + "grad_norm": 0.8061171770095825, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 23928 + }, + { + "epoch": 0.23929, + "grad_norm": 0.994542121887207, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 23929 + }, + { + "epoch": 0.2393, + "grad_norm": 1.2291271686553955, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 23930 + }, + { + "epoch": 0.23931, + "grad_norm": 0.8223946690559387, + "learning_rate": 0.003, + "loss": 4.023, + "step": 23931 + }, + { + "epoch": 0.23932, + "grad_norm": 0.772325336933136, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 23932 + }, + { + "epoch": 0.23933, + "grad_norm": 0.7452566623687744, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 23933 + }, + { + "epoch": 0.23934, + "grad_norm": 0.8208703398704529, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 23934 + }, + { + "epoch": 0.23935, + "grad_norm": 0.9261417388916016, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 23935 + }, + { + "epoch": 0.23936, + "grad_norm": 1.0116227865219116, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 23936 + }, + { + "epoch": 0.23937, + "grad_norm": 0.9992405772209167, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 23937 + }, + { + "epoch": 0.23938, + "grad_norm": 0.9463914036750793, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 23938 + }, + { + "epoch": 0.23939, + "grad_norm": 0.8754844069480896, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 23939 + }, + { + "epoch": 0.2394, + "grad_norm": 0.8468733429908752, + "learning_rate": 0.003, + "loss": 4.01, + "step": 23940 + }, + { + "epoch": 0.23941, + "grad_norm": 0.8293908834457397, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 23941 + }, + { + "epoch": 0.23942, + "grad_norm": 0.8099701404571533, + "learning_rate": 0.003, + "loss": 3.997, + "step": 23942 + }, + { + "epoch": 0.23943, + "grad_norm": 0.7749500274658203, + "learning_rate": 0.003, + "loss": 3.968, + "step": 23943 + }, + { + "epoch": 0.23944, + "grad_norm": 0.8789747357368469, + "learning_rate": 0.003, + "loss": 3.987, + "step": 23944 + }, + { + "epoch": 0.23945, + "grad_norm": 1.0388861894607544, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 23945 + }, + { + "epoch": 0.23946, + "grad_norm": 1.0592159032821655, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 23946 + }, + { + "epoch": 0.23947, + "grad_norm": 0.8729231357574463, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 23947 + }, + { + "epoch": 0.23948, + "grad_norm": 0.9557076096534729, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 23948 + }, + { + "epoch": 0.23949, + "grad_norm": 1.2300409078598022, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 23949 + }, + { + "epoch": 0.2395, + "grad_norm": 0.9815405011177063, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 23950 + }, + { + "epoch": 0.23951, + "grad_norm": 1.1076850891113281, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 23951 + }, + { + "epoch": 0.23952, + "grad_norm": 1.0483007431030273, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 23952 + }, + { + "epoch": 0.23953, + "grad_norm": 1.0386093854904175, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 23953 + }, + { + "epoch": 0.23954, + "grad_norm": 0.8717804551124573, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 23954 + }, + { + "epoch": 0.23955, + "grad_norm": 0.7528671622276306, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 23955 + }, + { + "epoch": 0.23956, + "grad_norm": 0.7613261938095093, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 23956 + }, + { + "epoch": 0.23957, + "grad_norm": 1.032530426979065, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 23957 + }, + { + "epoch": 0.23958, + "grad_norm": 1.0866122245788574, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 23958 + }, + { + "epoch": 0.23959, + "grad_norm": 0.6510752439498901, + "learning_rate": 0.003, + "loss": 4.004, + "step": 23959 + }, + { + "epoch": 0.2396, + "grad_norm": 0.6565369367599487, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 23960 + }, + { + "epoch": 0.23961, + "grad_norm": 0.9323978424072266, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 23961 + }, + { + "epoch": 0.23962, + "grad_norm": 1.010831356048584, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 23962 + }, + { + "epoch": 0.23963, + "grad_norm": 0.9497255682945251, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 23963 + }, + { + "epoch": 0.23964, + "grad_norm": 0.7668091058731079, + "learning_rate": 0.003, + "loss": 3.992, + "step": 23964 + }, + { + "epoch": 0.23965, + "grad_norm": 0.6088720560073853, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 23965 + }, + { + "epoch": 0.23966, + "grad_norm": 0.601540207862854, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 23966 + }, + { + "epoch": 0.23967, + "grad_norm": 0.5463195443153381, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 23967 + }, + { + "epoch": 0.23968, + "grad_norm": 0.5215985178947449, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 23968 + }, + { + "epoch": 0.23969, + "grad_norm": 0.6764963269233704, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 23969 + }, + { + "epoch": 0.2397, + "grad_norm": 0.7411446571350098, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 23970 + }, + { + "epoch": 0.23971, + "grad_norm": 0.7534075379371643, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 23971 + }, + { + "epoch": 0.23972, + "grad_norm": 0.693107545375824, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 23972 + }, + { + "epoch": 0.23973, + "grad_norm": 0.6980293393135071, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 23973 + }, + { + "epoch": 0.23974, + "grad_norm": 0.7093387246131897, + "learning_rate": 0.003, + "loss": 3.999, + "step": 23974 + }, + { + "epoch": 0.23975, + "grad_norm": 0.8891940712928772, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 23975 + }, + { + "epoch": 0.23976, + "grad_norm": 0.9738562703132629, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 23976 + }, + { + "epoch": 0.23977, + "grad_norm": 0.995661735534668, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 23977 + }, + { + "epoch": 0.23978, + "grad_norm": 0.9638500213623047, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 23978 + }, + { + "epoch": 0.23979, + "grad_norm": 0.8609791398048401, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 23979 + }, + { + "epoch": 0.2398, + "grad_norm": 0.793429970741272, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 23980 + }, + { + "epoch": 0.23981, + "grad_norm": 0.8712732195854187, + "learning_rate": 0.003, + "loss": 4.028, + "step": 23981 + }, + { + "epoch": 0.23982, + "grad_norm": 0.8926950693130493, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 23982 + }, + { + "epoch": 0.23983, + "grad_norm": 0.8730477690696716, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 23983 + }, + { + "epoch": 0.23984, + "grad_norm": 0.8065358996391296, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 23984 + }, + { + "epoch": 0.23985, + "grad_norm": 0.823911726474762, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 23985 + }, + { + "epoch": 0.23986, + "grad_norm": 0.7764955759048462, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 23986 + }, + { + "epoch": 0.23987, + "grad_norm": 0.8108323216438293, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 23987 + }, + { + "epoch": 0.23988, + "grad_norm": 1.1004838943481445, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 23988 + }, + { + "epoch": 0.23989, + "grad_norm": 0.9425547122955322, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 23989 + }, + { + "epoch": 0.2399, + "grad_norm": 0.9099467396736145, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 23990 + }, + { + "epoch": 0.23991, + "grad_norm": 1.002200961112976, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 23991 + }, + { + "epoch": 0.23992, + "grad_norm": 1.1405901908874512, + "learning_rate": 0.003, + "loss": 3.99, + "step": 23992 + }, + { + "epoch": 0.23993, + "grad_norm": 1.141911506652832, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 23993 + }, + { + "epoch": 0.23994, + "grad_norm": 0.6744820475578308, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 23994 + }, + { + "epoch": 0.23995, + "grad_norm": 0.6444469690322876, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 23995 + }, + { + "epoch": 0.23996, + "grad_norm": 0.9729512929916382, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 23996 + }, + { + "epoch": 0.23997, + "grad_norm": 1.05567467212677, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 23997 + }, + { + "epoch": 0.23998, + "grad_norm": 0.8706008195877075, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 23998 + }, + { + "epoch": 0.23999, + "grad_norm": 0.7566577196121216, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 23999 + }, + { + "epoch": 0.24, + "grad_norm": 0.7624983787536621, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 24000 + }, + { + "epoch": 0.24001, + "grad_norm": 0.8067689538002014, + "learning_rate": 0.003, + "loss": 3.982, + "step": 24001 + }, + { + "epoch": 0.24002, + "grad_norm": 0.8898313641548157, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 24002 + }, + { + "epoch": 0.24003, + "grad_norm": 0.8800652623176575, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 24003 + }, + { + "epoch": 0.24004, + "grad_norm": 1.0004292726516724, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 24004 + }, + { + "epoch": 0.24005, + "grad_norm": 0.9803778529167175, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 24005 + }, + { + "epoch": 0.24006, + "grad_norm": 0.9501041173934937, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 24006 + }, + { + "epoch": 0.24007, + "grad_norm": 1.0287295579910278, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 24007 + }, + { + "epoch": 0.24008, + "grad_norm": 1.072988510131836, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 24008 + }, + { + "epoch": 0.24009, + "grad_norm": 1.013957142829895, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 24009 + }, + { + "epoch": 0.2401, + "grad_norm": 1.072562336921692, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 24010 + }, + { + "epoch": 0.24011, + "grad_norm": 0.8773704767227173, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 24011 + }, + { + "epoch": 0.24012, + "grad_norm": 0.6834995746612549, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 24012 + }, + { + "epoch": 0.24013, + "grad_norm": 0.753741979598999, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 24013 + }, + { + "epoch": 0.24014, + "grad_norm": 0.8028086423873901, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 24014 + }, + { + "epoch": 0.24015, + "grad_norm": 0.7823895215988159, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 24015 + }, + { + "epoch": 0.24016, + "grad_norm": 0.7242679595947266, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 24016 + }, + { + "epoch": 0.24017, + "grad_norm": 0.6720256805419922, + "learning_rate": 0.003, + "loss": 3.997, + "step": 24017 + }, + { + "epoch": 0.24018, + "grad_norm": 0.6365830302238464, + "learning_rate": 0.003, + "loss": 3.997, + "step": 24018 + }, + { + "epoch": 0.24019, + "grad_norm": 0.7252905368804932, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 24019 + }, + { + "epoch": 0.2402, + "grad_norm": 0.8351150155067444, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 24020 + }, + { + "epoch": 0.24021, + "grad_norm": 0.9882925152778625, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 24021 + }, + { + "epoch": 0.24022, + "grad_norm": 1.2000802755355835, + "learning_rate": 0.003, + "loss": 4.027, + "step": 24022 + }, + { + "epoch": 0.24023, + "grad_norm": 0.7530702948570251, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 24023 + }, + { + "epoch": 0.24024, + "grad_norm": 0.6786104440689087, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 24024 + }, + { + "epoch": 0.24025, + "grad_norm": 0.6752521991729736, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 24025 + }, + { + "epoch": 0.24026, + "grad_norm": 0.6589224338531494, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 24026 + }, + { + "epoch": 0.24027, + "grad_norm": 0.788747251033783, + "learning_rate": 0.003, + "loss": 4.023, + "step": 24027 + }, + { + "epoch": 0.24028, + "grad_norm": 1.0930339097976685, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 24028 + }, + { + "epoch": 0.24029, + "grad_norm": 1.0810792446136475, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 24029 + }, + { + "epoch": 0.2403, + "grad_norm": 0.7615997791290283, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 24030 + }, + { + "epoch": 0.24031, + "grad_norm": 0.6934864521026611, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 24031 + }, + { + "epoch": 0.24032, + "grad_norm": 0.6373262405395508, + "learning_rate": 0.003, + "loss": 3.9471, + "step": 24032 + }, + { + "epoch": 0.24033, + "grad_norm": 0.5571783781051636, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 24033 + }, + { + "epoch": 0.24034, + "grad_norm": 0.5228264331817627, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 24034 + }, + { + "epoch": 0.24035, + "grad_norm": 0.6412609219551086, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 24035 + }, + { + "epoch": 0.24036, + "grad_norm": 0.6968461871147156, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 24036 + }, + { + "epoch": 0.24037, + "grad_norm": 0.8000894784927368, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 24037 + }, + { + "epoch": 0.24038, + "grad_norm": 0.889763355255127, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 24038 + }, + { + "epoch": 0.24039, + "grad_norm": 0.9583657383918762, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 24039 + }, + { + "epoch": 0.2404, + "grad_norm": 0.9983492493629456, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 24040 + }, + { + "epoch": 0.24041, + "grad_norm": 1.0740101337432861, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 24041 + }, + { + "epoch": 0.24042, + "grad_norm": 0.9051121473312378, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 24042 + }, + { + "epoch": 0.24043, + "grad_norm": 0.814412534236908, + "learning_rate": 0.003, + "loss": 4.019, + "step": 24043 + }, + { + "epoch": 0.24044, + "grad_norm": 0.8128172755241394, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 24044 + }, + { + "epoch": 0.24045, + "grad_norm": 0.8057212233543396, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 24045 + }, + { + "epoch": 0.24046, + "grad_norm": 0.7995898127555847, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 24046 + }, + { + "epoch": 0.24047, + "grad_norm": 0.8585660457611084, + "learning_rate": 0.003, + "loss": 4.01, + "step": 24047 + }, + { + "epoch": 0.24048, + "grad_norm": 1.0174612998962402, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 24048 + }, + { + "epoch": 0.24049, + "grad_norm": 1.105061411857605, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 24049 + }, + { + "epoch": 0.2405, + "grad_norm": 1.0302680730819702, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 24050 + }, + { + "epoch": 0.24051, + "grad_norm": 1.0031523704528809, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 24051 + }, + { + "epoch": 0.24052, + "grad_norm": 0.9524077773094177, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 24052 + }, + { + "epoch": 0.24053, + "grad_norm": 0.8197669386863708, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 24053 + }, + { + "epoch": 0.24054, + "grad_norm": 0.8970304131507874, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 24054 + }, + { + "epoch": 0.24055, + "grad_norm": 1.1677584648132324, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 24055 + }, + { + "epoch": 0.24056, + "grad_norm": 0.8615006804466248, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 24056 + }, + { + "epoch": 0.24057, + "grad_norm": 0.8732965588569641, + "learning_rate": 0.003, + "loss": 3.969, + "step": 24057 + }, + { + "epoch": 0.24058, + "grad_norm": 0.9552995562553406, + "learning_rate": 0.003, + "loss": 3.989, + "step": 24058 + }, + { + "epoch": 0.24059, + "grad_norm": 1.088826060295105, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 24059 + }, + { + "epoch": 0.2406, + "grad_norm": 0.7690457105636597, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 24060 + }, + { + "epoch": 0.24061, + "grad_norm": 0.639744222164154, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 24061 + }, + { + "epoch": 0.24062, + "grad_norm": 0.6786643266677856, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 24062 + }, + { + "epoch": 0.24063, + "grad_norm": 0.5991512537002563, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 24063 + }, + { + "epoch": 0.24064, + "grad_norm": 0.635001540184021, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 24064 + }, + { + "epoch": 0.24065, + "grad_norm": 0.7283830046653748, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 24065 + }, + { + "epoch": 0.24066, + "grad_norm": 0.7195993661880493, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 24066 + }, + { + "epoch": 0.24067, + "grad_norm": 0.7994078993797302, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 24067 + }, + { + "epoch": 0.24068, + "grad_norm": 0.83152836561203, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 24068 + }, + { + "epoch": 0.24069, + "grad_norm": 0.7601258754730225, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 24069 + }, + { + "epoch": 0.2407, + "grad_norm": 0.8376691937446594, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 24070 + }, + { + "epoch": 0.24071, + "grad_norm": 1.0589216947555542, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 24071 + }, + { + "epoch": 0.24072, + "grad_norm": 1.0876210927963257, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 24072 + }, + { + "epoch": 0.24073, + "grad_norm": 0.9772271513938904, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 24073 + }, + { + "epoch": 0.24074, + "grad_norm": 1.0784342288970947, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 24074 + }, + { + "epoch": 0.24075, + "grad_norm": 0.7942222356796265, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 24075 + }, + { + "epoch": 0.24076, + "grad_norm": 0.6856736540794373, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 24076 + }, + { + "epoch": 0.24077, + "grad_norm": 0.7023267149925232, + "learning_rate": 0.003, + "loss": 3.985, + "step": 24077 + }, + { + "epoch": 0.24078, + "grad_norm": 0.7600565552711487, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 24078 + }, + { + "epoch": 0.24079, + "grad_norm": 0.7559943795204163, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 24079 + }, + { + "epoch": 0.2408, + "grad_norm": 0.7524386048316956, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 24080 + }, + { + "epoch": 0.24081, + "grad_norm": 0.8650981187820435, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 24081 + }, + { + "epoch": 0.24082, + "grad_norm": 1.0360424518585205, + "learning_rate": 0.003, + "loss": 3.988, + "step": 24082 + }, + { + "epoch": 0.24083, + "grad_norm": 1.0921028852462769, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 24083 + }, + { + "epoch": 0.24084, + "grad_norm": 0.9492068290710449, + "learning_rate": 0.003, + "loss": 4.038, + "step": 24084 + }, + { + "epoch": 0.24085, + "grad_norm": 0.9898362755775452, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 24085 + }, + { + "epoch": 0.24086, + "grad_norm": 1.1069024801254272, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 24086 + }, + { + "epoch": 0.24087, + "grad_norm": 0.7375279664993286, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 24087 + }, + { + "epoch": 0.24088, + "grad_norm": 0.7496459484100342, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 24088 + }, + { + "epoch": 0.24089, + "grad_norm": 0.8298799395561218, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 24089 + }, + { + "epoch": 0.2409, + "grad_norm": 0.9120739102363586, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 24090 + }, + { + "epoch": 0.24091, + "grad_norm": 0.984941303730011, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 24091 + }, + { + "epoch": 0.24092, + "grad_norm": 1.06369149684906, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 24092 + }, + { + "epoch": 0.24093, + "grad_norm": 0.9703496694564819, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 24093 + }, + { + "epoch": 0.24094, + "grad_norm": 0.8562248945236206, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 24094 + }, + { + "epoch": 0.24095, + "grad_norm": 0.8310874104499817, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 24095 + }, + { + "epoch": 0.24096, + "grad_norm": 0.7521524429321289, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 24096 + }, + { + "epoch": 0.24097, + "grad_norm": 0.6825215220451355, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 24097 + }, + { + "epoch": 0.24098, + "grad_norm": 0.7346031665802002, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 24098 + }, + { + "epoch": 0.24099, + "grad_norm": 0.9496220946311951, + "learning_rate": 0.003, + "loss": 4.002, + "step": 24099 + }, + { + "epoch": 0.241, + "grad_norm": 1.2669556140899658, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 24100 + }, + { + "epoch": 0.24101, + "grad_norm": 0.7179388999938965, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 24101 + }, + { + "epoch": 0.24102, + "grad_norm": 0.6274367570877075, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 24102 + }, + { + "epoch": 0.24103, + "grad_norm": 0.6840031743049622, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 24103 + }, + { + "epoch": 0.24104, + "grad_norm": 0.7888739109039307, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 24104 + }, + { + "epoch": 0.24105, + "grad_norm": 0.6246904134750366, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 24105 + }, + { + "epoch": 0.24106, + "grad_norm": 0.6284182071685791, + "learning_rate": 0.003, + "loss": 4.021, + "step": 24106 + }, + { + "epoch": 0.24107, + "grad_norm": 0.6604494452476501, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 24107 + }, + { + "epoch": 0.24108, + "grad_norm": 0.8342052698135376, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 24108 + }, + { + "epoch": 0.24109, + "grad_norm": 0.9433430433273315, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 24109 + }, + { + "epoch": 0.2411, + "grad_norm": 0.8320357203483582, + "learning_rate": 0.003, + "loss": 4.008, + "step": 24110 + }, + { + "epoch": 0.24111, + "grad_norm": 0.7370584011077881, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 24111 + }, + { + "epoch": 0.24112, + "grad_norm": 0.8017516732215881, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 24112 + }, + { + "epoch": 0.24113, + "grad_norm": 1.0253359079360962, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 24113 + }, + { + "epoch": 0.24114, + "grad_norm": 1.1755030155181885, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 24114 + }, + { + "epoch": 0.24115, + "grad_norm": 1.1060174703598022, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 24115 + }, + { + "epoch": 0.24116, + "grad_norm": 0.9710884094238281, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 24116 + }, + { + "epoch": 0.24117, + "grad_norm": 0.9340130686759949, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 24117 + }, + { + "epoch": 0.24118, + "grad_norm": 0.8979191184043884, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 24118 + }, + { + "epoch": 0.24119, + "grad_norm": 0.8649385571479797, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 24119 + }, + { + "epoch": 0.2412, + "grad_norm": 0.907820463180542, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 24120 + }, + { + "epoch": 0.24121, + "grad_norm": 0.962208092212677, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 24121 + }, + { + "epoch": 0.24122, + "grad_norm": 0.894923746585846, + "learning_rate": 0.003, + "loss": 3.956, + "step": 24122 + }, + { + "epoch": 0.24123, + "grad_norm": 0.77435302734375, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 24123 + }, + { + "epoch": 0.24124, + "grad_norm": 0.8093234300613403, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 24124 + }, + { + "epoch": 0.24125, + "grad_norm": 1.1143426895141602, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 24125 + }, + { + "epoch": 0.24126, + "grad_norm": 0.9138668775558472, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 24126 + }, + { + "epoch": 0.24127, + "grad_norm": 0.8198195099830627, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 24127 + }, + { + "epoch": 0.24128, + "grad_norm": 0.76114422082901, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 24128 + }, + { + "epoch": 0.24129, + "grad_norm": 0.7072831988334656, + "learning_rate": 0.003, + "loss": 4.008, + "step": 24129 + }, + { + "epoch": 0.2413, + "grad_norm": 0.6897961497306824, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 24130 + }, + { + "epoch": 0.24131, + "grad_norm": 0.6996879577636719, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 24131 + }, + { + "epoch": 0.24132, + "grad_norm": 0.7060827016830444, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 24132 + }, + { + "epoch": 0.24133, + "grad_norm": 0.7298224568367004, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 24133 + }, + { + "epoch": 0.24134, + "grad_norm": 0.7347943186759949, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 24134 + }, + { + "epoch": 0.24135, + "grad_norm": 0.812019407749176, + "learning_rate": 0.003, + "loss": 3.997, + "step": 24135 + }, + { + "epoch": 0.24136, + "grad_norm": 0.9119550585746765, + "learning_rate": 0.003, + "loss": 4.0, + "step": 24136 + }, + { + "epoch": 0.24137, + "grad_norm": 0.8857710957527161, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 24137 + }, + { + "epoch": 0.24138, + "grad_norm": 0.8409410715103149, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 24138 + }, + { + "epoch": 0.24139, + "grad_norm": 0.8665996193885803, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 24139 + }, + { + "epoch": 0.2414, + "grad_norm": 0.9897549152374268, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 24140 + }, + { + "epoch": 0.24141, + "grad_norm": 1.1662150621414185, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 24141 + }, + { + "epoch": 0.24142, + "grad_norm": 1.074530005455017, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 24142 + }, + { + "epoch": 0.24143, + "grad_norm": 0.9430781006813049, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 24143 + }, + { + "epoch": 0.24144, + "grad_norm": 0.9876217246055603, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 24144 + }, + { + "epoch": 0.24145, + "grad_norm": 1.3440173864364624, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 24145 + }, + { + "epoch": 0.24146, + "grad_norm": 0.6672643423080444, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 24146 + }, + { + "epoch": 0.24147, + "grad_norm": 0.7157257795333862, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 24147 + }, + { + "epoch": 0.24148, + "grad_norm": 0.7720828056335449, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 24148 + }, + { + "epoch": 0.24149, + "grad_norm": 0.6639391779899597, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 24149 + }, + { + "epoch": 0.2415, + "grad_norm": 0.78180330991745, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 24150 + }, + { + "epoch": 0.24151, + "grad_norm": 0.8597205877304077, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 24151 + }, + { + "epoch": 0.24152, + "grad_norm": 1.015058994293213, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 24152 + }, + { + "epoch": 0.24153, + "grad_norm": 1.3773125410079956, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 24153 + }, + { + "epoch": 0.24154, + "grad_norm": 0.5857715010643005, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 24154 + }, + { + "epoch": 0.24155, + "grad_norm": 0.7540925741195679, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 24155 + }, + { + "epoch": 0.24156, + "grad_norm": 0.89894700050354, + "learning_rate": 0.003, + "loss": 3.999, + "step": 24156 + }, + { + "epoch": 0.24157, + "grad_norm": 1.0340477228164673, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 24157 + }, + { + "epoch": 0.24158, + "grad_norm": 1.2010512351989746, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 24158 + }, + { + "epoch": 0.24159, + "grad_norm": 0.5984905958175659, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 24159 + }, + { + "epoch": 0.2416, + "grad_norm": 0.721567690372467, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 24160 + }, + { + "epoch": 0.24161, + "grad_norm": 0.7833555340766907, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 24161 + }, + { + "epoch": 0.24162, + "grad_norm": 0.7727726101875305, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 24162 + }, + { + "epoch": 0.24163, + "grad_norm": 0.9688026309013367, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 24163 + }, + { + "epoch": 0.24164, + "grad_norm": 0.948297917842865, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 24164 + }, + { + "epoch": 0.24165, + "grad_norm": 0.7932071089744568, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 24165 + }, + { + "epoch": 0.24166, + "grad_norm": 0.818384051322937, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 24166 + }, + { + "epoch": 0.24167, + "grad_norm": 0.8786194920539856, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 24167 + }, + { + "epoch": 0.24168, + "grad_norm": 0.9941129088401794, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 24168 + }, + { + "epoch": 0.24169, + "grad_norm": 1.1686395406723022, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 24169 + }, + { + "epoch": 0.2417, + "grad_norm": 0.842773973941803, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 24170 + }, + { + "epoch": 0.24171, + "grad_norm": 0.8219358325004578, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 24171 + }, + { + "epoch": 0.24172, + "grad_norm": 0.9586386680603027, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 24172 + }, + { + "epoch": 0.24173, + "grad_norm": 0.9807718992233276, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 24173 + }, + { + "epoch": 0.24174, + "grad_norm": 0.9369253516197205, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 24174 + }, + { + "epoch": 0.24175, + "grad_norm": 1.0876275300979614, + "learning_rate": 0.003, + "loss": 4.006, + "step": 24175 + }, + { + "epoch": 0.24176, + "grad_norm": 1.0466890335083008, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 24176 + }, + { + "epoch": 0.24177, + "grad_norm": 0.8735606670379639, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 24177 + }, + { + "epoch": 0.24178, + "grad_norm": 0.8297070264816284, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 24178 + }, + { + "epoch": 0.24179, + "grad_norm": 0.8231585621833801, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 24179 + }, + { + "epoch": 0.2418, + "grad_norm": 0.9158185124397278, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 24180 + }, + { + "epoch": 0.24181, + "grad_norm": 1.1067228317260742, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 24181 + }, + { + "epoch": 0.24182, + "grad_norm": 0.9517270922660828, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 24182 + }, + { + "epoch": 0.24183, + "grad_norm": 0.776860773563385, + "learning_rate": 0.003, + "loss": 3.976, + "step": 24183 + }, + { + "epoch": 0.24184, + "grad_norm": 0.7430579662322998, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 24184 + }, + { + "epoch": 0.24185, + "grad_norm": 0.7413129210472107, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 24185 + }, + { + "epoch": 0.24186, + "grad_norm": 0.7758918404579163, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 24186 + }, + { + "epoch": 0.24187, + "grad_norm": 0.6804990768432617, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 24187 + }, + { + "epoch": 0.24188, + "grad_norm": 0.5927160382270813, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 24188 + }, + { + "epoch": 0.24189, + "grad_norm": 0.6047726273536682, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 24189 + }, + { + "epoch": 0.2419, + "grad_norm": 0.7194401621818542, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 24190 + }, + { + "epoch": 0.24191, + "grad_norm": 0.6403484344482422, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 24191 + }, + { + "epoch": 0.24192, + "grad_norm": 0.5371664762496948, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 24192 + }, + { + "epoch": 0.24193, + "grad_norm": 0.6039645671844482, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 24193 + }, + { + "epoch": 0.24194, + "grad_norm": 0.7431378364562988, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 24194 + }, + { + "epoch": 0.24195, + "grad_norm": 1.1407995223999023, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 24195 + }, + { + "epoch": 0.24196, + "grad_norm": 1.2284046411514282, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 24196 + }, + { + "epoch": 0.24197, + "grad_norm": 0.6798164248466492, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 24197 + }, + { + "epoch": 0.24198, + "grad_norm": 0.6801814436912537, + "learning_rate": 0.003, + "loss": 3.975, + "step": 24198 + }, + { + "epoch": 0.24199, + "grad_norm": 0.6405549049377441, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 24199 + }, + { + "epoch": 0.242, + "grad_norm": 0.6176456212997437, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 24200 + }, + { + "epoch": 0.24201, + "grad_norm": 0.7574557065963745, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 24201 + }, + { + "epoch": 0.24202, + "grad_norm": 0.8952157497406006, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 24202 + }, + { + "epoch": 0.24203, + "grad_norm": 0.9731083512306213, + "learning_rate": 0.003, + "loss": 3.98, + "step": 24203 + }, + { + "epoch": 0.24204, + "grad_norm": 1.1066205501556396, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 24204 + }, + { + "epoch": 0.24205, + "grad_norm": 0.9378203749656677, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 24205 + }, + { + "epoch": 0.24206, + "grad_norm": 0.6474951505661011, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 24206 + }, + { + "epoch": 0.24207, + "grad_norm": 0.6557741761207581, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 24207 + }, + { + "epoch": 0.24208, + "grad_norm": 0.8852442502975464, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 24208 + }, + { + "epoch": 0.24209, + "grad_norm": 1.191359043121338, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 24209 + }, + { + "epoch": 0.2421, + "grad_norm": 0.8946442604064941, + "learning_rate": 0.003, + "loss": 4.001, + "step": 24210 + }, + { + "epoch": 0.24211, + "grad_norm": 0.7831708788871765, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 24211 + }, + { + "epoch": 0.24212, + "grad_norm": 0.8748942613601685, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 24212 + }, + { + "epoch": 0.24213, + "grad_norm": 0.8469408750534058, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 24213 + }, + { + "epoch": 0.24214, + "grad_norm": 0.8735719919204712, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 24214 + }, + { + "epoch": 0.24215, + "grad_norm": 0.993291437625885, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 24215 + }, + { + "epoch": 0.24216, + "grad_norm": 1.08012056350708, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 24216 + }, + { + "epoch": 0.24217, + "grad_norm": 0.8592896461486816, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 24217 + }, + { + "epoch": 0.24218, + "grad_norm": 0.8167977929115295, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 24218 + }, + { + "epoch": 0.24219, + "grad_norm": 0.8976151943206787, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 24219 + }, + { + "epoch": 0.2422, + "grad_norm": 1.1303269863128662, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 24220 + }, + { + "epoch": 0.24221, + "grad_norm": 0.9404118657112122, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 24221 + }, + { + "epoch": 0.24222, + "grad_norm": 0.9237927794456482, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 24222 + }, + { + "epoch": 0.24223, + "grad_norm": 0.9288538098335266, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 24223 + }, + { + "epoch": 0.24224, + "grad_norm": 0.9267579913139343, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 24224 + }, + { + "epoch": 0.24225, + "grad_norm": 1.0558236837387085, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 24225 + }, + { + "epoch": 0.24226, + "grad_norm": 1.0452866554260254, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 24226 + }, + { + "epoch": 0.24227, + "grad_norm": 0.7189522981643677, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 24227 + }, + { + "epoch": 0.24228, + "grad_norm": 0.7157479524612427, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 24228 + }, + { + "epoch": 0.24229, + "grad_norm": 0.8164672255516052, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 24229 + }, + { + "epoch": 0.2423, + "grad_norm": 0.9145386815071106, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 24230 + }, + { + "epoch": 0.24231, + "grad_norm": 0.9043770432472229, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 24231 + }, + { + "epoch": 0.24232, + "grad_norm": 0.9058763384819031, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 24232 + }, + { + "epoch": 0.24233, + "grad_norm": 1.0408285856246948, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 24233 + }, + { + "epoch": 0.24234, + "grad_norm": 1.0608769655227661, + "learning_rate": 0.003, + "loss": 4.029, + "step": 24234 + }, + { + "epoch": 0.24235, + "grad_norm": 1.1708331108093262, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 24235 + }, + { + "epoch": 0.24236, + "grad_norm": 0.8797174096107483, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 24236 + }, + { + "epoch": 0.24237, + "grad_norm": 0.9085307717323303, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 24237 + }, + { + "epoch": 0.24238, + "grad_norm": 0.8104785084724426, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 24238 + }, + { + "epoch": 0.24239, + "grad_norm": 0.7317141890525818, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 24239 + }, + { + "epoch": 0.2424, + "grad_norm": 0.773259162902832, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 24240 + }, + { + "epoch": 0.24241, + "grad_norm": 0.6973122358322144, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 24241 + }, + { + "epoch": 0.24242, + "grad_norm": 0.7163370251655579, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 24242 + }, + { + "epoch": 0.24243, + "grad_norm": 0.6233422160148621, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 24243 + }, + { + "epoch": 0.24244, + "grad_norm": 0.7451505661010742, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 24244 + }, + { + "epoch": 0.24245, + "grad_norm": 1.319509506225586, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 24245 + }, + { + "epoch": 0.24246, + "grad_norm": 0.987000048160553, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 24246 + }, + { + "epoch": 0.24247, + "grad_norm": 0.7025144100189209, + "learning_rate": 0.003, + "loss": 4.003, + "step": 24247 + }, + { + "epoch": 0.24248, + "grad_norm": 0.6538518071174622, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 24248 + }, + { + "epoch": 0.24249, + "grad_norm": 0.7193374037742615, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 24249 + }, + { + "epoch": 0.2425, + "grad_norm": 0.7582658529281616, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 24250 + }, + { + "epoch": 0.24251, + "grad_norm": 0.6614522933959961, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 24251 + }, + { + "epoch": 0.24252, + "grad_norm": 0.7000927329063416, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 24252 + }, + { + "epoch": 0.24253, + "grad_norm": 0.7912187576293945, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 24253 + }, + { + "epoch": 0.24254, + "grad_norm": 0.9036010503768921, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 24254 + }, + { + "epoch": 0.24255, + "grad_norm": 0.9756733179092407, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 24255 + }, + { + "epoch": 0.24256, + "grad_norm": 1.0480183362960815, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 24256 + }, + { + "epoch": 0.24257, + "grad_norm": 0.8163262009620667, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 24257 + }, + { + "epoch": 0.24258, + "grad_norm": 0.6424911618232727, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 24258 + }, + { + "epoch": 0.24259, + "grad_norm": 0.6515970230102539, + "learning_rate": 0.003, + "loss": 3.9478, + "step": 24259 + }, + { + "epoch": 0.2426, + "grad_norm": 0.6862676739692688, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 24260 + }, + { + "epoch": 0.24261, + "grad_norm": 0.7388269305229187, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 24261 + }, + { + "epoch": 0.24262, + "grad_norm": 0.8990063071250916, + "learning_rate": 0.003, + "loss": 4.002, + "step": 24262 + }, + { + "epoch": 0.24263, + "grad_norm": 1.0756982564926147, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 24263 + }, + { + "epoch": 0.24264, + "grad_norm": 1.1378021240234375, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 24264 + }, + { + "epoch": 0.24265, + "grad_norm": 0.9093742966651917, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 24265 + }, + { + "epoch": 0.24266, + "grad_norm": 0.8484373092651367, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 24266 + }, + { + "epoch": 0.24267, + "grad_norm": 0.8649173378944397, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 24267 + }, + { + "epoch": 0.24268, + "grad_norm": 0.8747851252555847, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 24268 + }, + { + "epoch": 0.24269, + "grad_norm": 0.9463253617286682, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 24269 + }, + { + "epoch": 0.2427, + "grad_norm": 1.0161617994308472, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 24270 + }, + { + "epoch": 0.24271, + "grad_norm": 1.1006355285644531, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 24271 + }, + { + "epoch": 0.24272, + "grad_norm": 0.842169463634491, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 24272 + }, + { + "epoch": 0.24273, + "grad_norm": 0.6309932470321655, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 24273 + }, + { + "epoch": 0.24274, + "grad_norm": 0.5599721670150757, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 24274 + }, + { + "epoch": 0.24275, + "grad_norm": 0.5426430106163025, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 24275 + }, + { + "epoch": 0.24276, + "grad_norm": 0.5850679874420166, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 24276 + }, + { + "epoch": 0.24277, + "grad_norm": 0.5796646475791931, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 24277 + }, + { + "epoch": 0.24278, + "grad_norm": 0.6452776193618774, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 24278 + }, + { + "epoch": 0.24279, + "grad_norm": 0.7937586307525635, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 24279 + }, + { + "epoch": 0.2428, + "grad_norm": 0.88665771484375, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 24280 + }, + { + "epoch": 0.24281, + "grad_norm": 0.9205717444419861, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 24281 + }, + { + "epoch": 0.24282, + "grad_norm": 0.9923974275588989, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 24282 + }, + { + "epoch": 0.24283, + "grad_norm": 1.0312395095825195, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 24283 + }, + { + "epoch": 0.24284, + "grad_norm": 0.9431099891662598, + "learning_rate": 0.003, + "loss": 4.003, + "step": 24284 + }, + { + "epoch": 0.24285, + "grad_norm": 1.133252739906311, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 24285 + }, + { + "epoch": 0.24286, + "grad_norm": 0.8795227408409119, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 24286 + }, + { + "epoch": 0.24287, + "grad_norm": 0.6642279624938965, + "learning_rate": 0.003, + "loss": 3.982, + "step": 24287 + }, + { + "epoch": 0.24288, + "grad_norm": 0.7265197038650513, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 24288 + }, + { + "epoch": 0.24289, + "grad_norm": 0.7638937830924988, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 24289 + }, + { + "epoch": 0.2429, + "grad_norm": 0.7722568511962891, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 24290 + }, + { + "epoch": 0.24291, + "grad_norm": 0.8117692470550537, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 24291 + }, + { + "epoch": 0.24292, + "grad_norm": 0.9055297374725342, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 24292 + }, + { + "epoch": 0.24293, + "grad_norm": 1.1438698768615723, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 24293 + }, + { + "epoch": 0.24294, + "grad_norm": 1.162084698677063, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 24294 + }, + { + "epoch": 0.24295, + "grad_norm": 1.1040618419647217, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 24295 + }, + { + "epoch": 0.24296, + "grad_norm": 0.7663518786430359, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 24296 + }, + { + "epoch": 0.24297, + "grad_norm": 0.7392738461494446, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 24297 + }, + { + "epoch": 0.24298, + "grad_norm": 0.782947301864624, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 24298 + }, + { + "epoch": 0.24299, + "grad_norm": 0.762938380241394, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 24299 + }, + { + "epoch": 0.243, + "grad_norm": 0.7952916026115417, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 24300 + }, + { + "epoch": 0.24301, + "grad_norm": 0.8860375285148621, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 24301 + }, + { + "epoch": 0.24302, + "grad_norm": 1.0110918283462524, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 24302 + }, + { + "epoch": 0.24303, + "grad_norm": 1.037291407585144, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 24303 + }, + { + "epoch": 0.24304, + "grad_norm": 0.8967602252960205, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 24304 + }, + { + "epoch": 0.24305, + "grad_norm": 0.8323214054107666, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 24305 + }, + { + "epoch": 0.24306, + "grad_norm": 0.899467945098877, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 24306 + }, + { + "epoch": 0.24307, + "grad_norm": 0.9835702180862427, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 24307 + }, + { + "epoch": 0.24308, + "grad_norm": 1.070469617843628, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 24308 + }, + { + "epoch": 0.24309, + "grad_norm": 0.9148688316345215, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 24309 + }, + { + "epoch": 0.2431, + "grad_norm": 0.9679651260375977, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 24310 + }, + { + "epoch": 0.24311, + "grad_norm": 1.179229497909546, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 24311 + }, + { + "epoch": 0.24312, + "grad_norm": 0.8881710171699524, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 24312 + }, + { + "epoch": 0.24313, + "grad_norm": 0.7553343772888184, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 24313 + }, + { + "epoch": 0.24314, + "grad_norm": 0.822820782661438, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 24314 + }, + { + "epoch": 0.24315, + "grad_norm": 0.9005582332611084, + "learning_rate": 0.003, + "loss": 4.002, + "step": 24315 + }, + { + "epoch": 0.24316, + "grad_norm": 0.9052172303199768, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 24316 + }, + { + "epoch": 0.24317, + "grad_norm": 0.9219019412994385, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 24317 + }, + { + "epoch": 0.24318, + "grad_norm": 0.8720504641532898, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 24318 + }, + { + "epoch": 0.24319, + "grad_norm": 0.7494379878044128, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 24319 + }, + { + "epoch": 0.2432, + "grad_norm": 0.7328191995620728, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 24320 + }, + { + "epoch": 0.24321, + "grad_norm": 0.7547847032546997, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 24321 + }, + { + "epoch": 0.24322, + "grad_norm": 0.7140090465545654, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 24322 + }, + { + "epoch": 0.24323, + "grad_norm": 0.6971606612205505, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 24323 + }, + { + "epoch": 0.24324, + "grad_norm": 0.650955319404602, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 24324 + }, + { + "epoch": 0.24325, + "grad_norm": 0.7189648151397705, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 24325 + }, + { + "epoch": 0.24326, + "grad_norm": 0.7265533208847046, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 24326 + }, + { + "epoch": 0.24327, + "grad_norm": 0.8301386833190918, + "learning_rate": 0.003, + "loss": 3.999, + "step": 24327 + }, + { + "epoch": 0.24328, + "grad_norm": 0.9678142666816711, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 24328 + }, + { + "epoch": 0.24329, + "grad_norm": 1.2379730939865112, + "learning_rate": 0.003, + "loss": 4.009, + "step": 24329 + }, + { + "epoch": 0.2433, + "grad_norm": 0.8067208528518677, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 24330 + }, + { + "epoch": 0.24331, + "grad_norm": 0.7881233096122742, + "learning_rate": 0.003, + "loss": 3.971, + "step": 24331 + }, + { + "epoch": 0.24332, + "grad_norm": 0.9155017137527466, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 24332 + }, + { + "epoch": 0.24333, + "grad_norm": 1.1484644412994385, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 24333 + }, + { + "epoch": 0.24334, + "grad_norm": 0.8496889472007751, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 24334 + }, + { + "epoch": 0.24335, + "grad_norm": 0.7134077548980713, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 24335 + }, + { + "epoch": 0.24336, + "grad_norm": 0.6998562812805176, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 24336 + }, + { + "epoch": 0.24337, + "grad_norm": 0.7047627568244934, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 24337 + }, + { + "epoch": 0.24338, + "grad_norm": 0.6584394574165344, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 24338 + }, + { + "epoch": 0.24339, + "grad_norm": 0.6794366240501404, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 24339 + }, + { + "epoch": 0.2434, + "grad_norm": 0.5890007019042969, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 24340 + }, + { + "epoch": 0.24341, + "grad_norm": 0.5139788389205933, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 24341 + }, + { + "epoch": 0.24342, + "grad_norm": 0.6266186833381653, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 24342 + }, + { + "epoch": 0.24343, + "grad_norm": 0.7824134826660156, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 24343 + }, + { + "epoch": 0.24344, + "grad_norm": 1.0562269687652588, + "learning_rate": 0.003, + "loss": 4.003, + "step": 24344 + }, + { + "epoch": 0.24345, + "grad_norm": 1.1253581047058105, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 24345 + }, + { + "epoch": 0.24346, + "grad_norm": 0.879284143447876, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 24346 + }, + { + "epoch": 0.24347, + "grad_norm": 0.849868655204773, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 24347 + }, + { + "epoch": 0.24348, + "grad_norm": 0.8444521427154541, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 24348 + }, + { + "epoch": 0.24349, + "grad_norm": 0.7563900351524353, + "learning_rate": 0.003, + "loss": 4.0, + "step": 24349 + }, + { + "epoch": 0.2435, + "grad_norm": 0.7423490285873413, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 24350 + }, + { + "epoch": 0.24351, + "grad_norm": 0.8140950202941895, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 24351 + }, + { + "epoch": 0.24352, + "grad_norm": 0.8714144825935364, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 24352 + }, + { + "epoch": 0.24353, + "grad_norm": 0.9874817132949829, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 24353 + }, + { + "epoch": 0.24354, + "grad_norm": 1.0559970140457153, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 24354 + }, + { + "epoch": 0.24355, + "grad_norm": 0.9778796434402466, + "learning_rate": 0.003, + "loss": 4.011, + "step": 24355 + }, + { + "epoch": 0.24356, + "grad_norm": 1.1194885969161987, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 24356 + }, + { + "epoch": 0.24357, + "grad_norm": 1.0187160968780518, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 24357 + }, + { + "epoch": 0.24358, + "grad_norm": 1.0746526718139648, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 24358 + }, + { + "epoch": 0.24359, + "grad_norm": 1.0508475303649902, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 24359 + }, + { + "epoch": 0.2436, + "grad_norm": 1.1922521591186523, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 24360 + }, + { + "epoch": 0.24361, + "grad_norm": 1.0380300283432007, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 24361 + }, + { + "epoch": 0.24362, + "grad_norm": 1.007228970527649, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 24362 + }, + { + "epoch": 0.24363, + "grad_norm": 1.0945950746536255, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 24363 + }, + { + "epoch": 0.24364, + "grad_norm": 0.9768579602241516, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 24364 + }, + { + "epoch": 0.24365, + "grad_norm": 1.1311932802200317, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 24365 + }, + { + "epoch": 0.24366, + "grad_norm": 0.9079676270484924, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 24366 + }, + { + "epoch": 0.24367, + "grad_norm": 0.8820545077323914, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 24367 + }, + { + "epoch": 0.24368, + "grad_norm": 0.9336348176002502, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 24368 + }, + { + "epoch": 0.24369, + "grad_norm": 0.9338898658752441, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 24369 + }, + { + "epoch": 0.2437, + "grad_norm": 1.1141010522842407, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 24370 + }, + { + "epoch": 0.24371, + "grad_norm": 1.1137514114379883, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 24371 + }, + { + "epoch": 0.24372, + "grad_norm": 0.9232408404350281, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 24372 + }, + { + "epoch": 0.24373, + "grad_norm": 0.9463221430778503, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 24373 + }, + { + "epoch": 0.24374, + "grad_norm": 0.9682784080505371, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 24374 + }, + { + "epoch": 0.24375, + "grad_norm": 0.9680209159851074, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 24375 + }, + { + "epoch": 0.24376, + "grad_norm": 0.8425778150558472, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 24376 + }, + { + "epoch": 0.24377, + "grad_norm": 0.7373186945915222, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 24377 + }, + { + "epoch": 0.24378, + "grad_norm": 0.8012325763702393, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 24378 + }, + { + "epoch": 0.24379, + "grad_norm": 0.8191224932670593, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 24379 + }, + { + "epoch": 0.2438, + "grad_norm": 0.7719004154205322, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 24380 + }, + { + "epoch": 0.24381, + "grad_norm": 0.7410956621170044, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 24381 + }, + { + "epoch": 0.24382, + "grad_norm": 0.6663102507591248, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 24382 + }, + { + "epoch": 0.24383, + "grad_norm": 0.6674877405166626, + "learning_rate": 0.003, + "loss": 3.977, + "step": 24383 + }, + { + "epoch": 0.24384, + "grad_norm": 0.7758035063743591, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 24384 + }, + { + "epoch": 0.24385, + "grad_norm": 0.7862753868103027, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 24385 + }, + { + "epoch": 0.24386, + "grad_norm": 0.7202532291412354, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 24386 + }, + { + "epoch": 0.24387, + "grad_norm": 0.7088238596916199, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 24387 + }, + { + "epoch": 0.24388, + "grad_norm": 0.6416760683059692, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 24388 + }, + { + "epoch": 0.24389, + "grad_norm": 0.5948203206062317, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 24389 + }, + { + "epoch": 0.2439, + "grad_norm": 0.5630350708961487, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 24390 + }, + { + "epoch": 0.24391, + "grad_norm": 0.6018035411834717, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 24391 + }, + { + "epoch": 0.24392, + "grad_norm": 0.6622597575187683, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 24392 + }, + { + "epoch": 0.24393, + "grad_norm": 0.6130614876747131, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 24393 + }, + { + "epoch": 0.24394, + "grad_norm": 0.5697718262672424, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 24394 + }, + { + "epoch": 0.24395, + "grad_norm": 0.9074701070785522, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 24395 + }, + { + "epoch": 0.24396, + "grad_norm": 1.5890822410583496, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 24396 + }, + { + "epoch": 0.24397, + "grad_norm": 0.48694777488708496, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 24397 + }, + { + "epoch": 0.24398, + "grad_norm": 1.0338784456253052, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 24398 + }, + { + "epoch": 0.24399, + "grad_norm": 1.3032265901565552, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 24399 + }, + { + "epoch": 0.244, + "grad_norm": 0.5795349478721619, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 24400 + }, + { + "epoch": 0.24401, + "grad_norm": 0.6832484602928162, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 24401 + }, + { + "epoch": 0.24402, + "grad_norm": 0.744670569896698, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 24402 + }, + { + "epoch": 0.24403, + "grad_norm": 0.7396913170814514, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 24403 + }, + { + "epoch": 0.24404, + "grad_norm": 0.8493564128875732, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 24404 + }, + { + "epoch": 0.24405, + "grad_norm": 0.9573858976364136, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 24405 + }, + { + "epoch": 0.24406, + "grad_norm": 1.0892409086227417, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 24406 + }, + { + "epoch": 0.24407, + "grad_norm": 0.9571846723556519, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 24407 + }, + { + "epoch": 0.24408, + "grad_norm": 1.1677541732788086, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 24408 + }, + { + "epoch": 0.24409, + "grad_norm": 0.9839617609977722, + "learning_rate": 0.003, + "loss": 3.989, + "step": 24409 + }, + { + "epoch": 0.2441, + "grad_norm": 0.9990457892417908, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 24410 + }, + { + "epoch": 0.24411, + "grad_norm": 0.9626514315605164, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 24411 + }, + { + "epoch": 0.24412, + "grad_norm": 1.1488869190216064, + "learning_rate": 0.003, + "loss": 3.995, + "step": 24412 + }, + { + "epoch": 0.24413, + "grad_norm": 0.7928274273872375, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 24413 + }, + { + "epoch": 0.24414, + "grad_norm": 0.8261349201202393, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 24414 + }, + { + "epoch": 0.24415, + "grad_norm": 0.8000668883323669, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 24415 + }, + { + "epoch": 0.24416, + "grad_norm": 0.8791115283966064, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 24416 + }, + { + "epoch": 0.24417, + "grad_norm": 0.8411887884140015, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 24417 + }, + { + "epoch": 0.24418, + "grad_norm": 0.9706425070762634, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 24418 + }, + { + "epoch": 0.24419, + "grad_norm": 0.9877373576164246, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 24419 + }, + { + "epoch": 0.2442, + "grad_norm": 0.9917827844619751, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 24420 + }, + { + "epoch": 0.24421, + "grad_norm": 1.0765424966812134, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 24421 + }, + { + "epoch": 0.24422, + "grad_norm": 1.008981704711914, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 24422 + }, + { + "epoch": 0.24423, + "grad_norm": 1.0158170461654663, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 24423 + }, + { + "epoch": 0.24424, + "grad_norm": 0.9564645290374756, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 24424 + }, + { + "epoch": 0.24425, + "grad_norm": 1.0663225650787354, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 24425 + }, + { + "epoch": 0.24426, + "grad_norm": 0.8671601414680481, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 24426 + }, + { + "epoch": 0.24427, + "grad_norm": 0.8171334266662598, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 24427 + }, + { + "epoch": 0.24428, + "grad_norm": 0.8455215692520142, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 24428 + }, + { + "epoch": 0.24429, + "grad_norm": 0.7950472235679626, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 24429 + }, + { + "epoch": 0.2443, + "grad_norm": 0.9936681389808655, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 24430 + }, + { + "epoch": 0.24431, + "grad_norm": 1.6029963493347168, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 24431 + }, + { + "epoch": 0.24432, + "grad_norm": 0.7407926321029663, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 24432 + }, + { + "epoch": 0.24433, + "grad_norm": 0.7223249077796936, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 24433 + }, + { + "epoch": 0.24434, + "grad_norm": 0.7443443536758423, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 24434 + }, + { + "epoch": 0.24435, + "grad_norm": 0.8511462211608887, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 24435 + }, + { + "epoch": 0.24436, + "grad_norm": 0.956386148929596, + "learning_rate": 0.003, + "loss": 4.005, + "step": 24436 + }, + { + "epoch": 0.24437, + "grad_norm": 1.1090093851089478, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 24437 + }, + { + "epoch": 0.24438, + "grad_norm": 0.7701601982116699, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 24438 + }, + { + "epoch": 0.24439, + "grad_norm": 0.6127216815948486, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 24439 + }, + { + "epoch": 0.2444, + "grad_norm": 0.7334783673286438, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 24440 + }, + { + "epoch": 0.24441, + "grad_norm": 1.02713143825531, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 24441 + }, + { + "epoch": 0.24442, + "grad_norm": 1.3430263996124268, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 24442 + }, + { + "epoch": 0.24443, + "grad_norm": 0.5718302726745605, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 24443 + }, + { + "epoch": 0.24444, + "grad_norm": 0.7819834351539612, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 24444 + }, + { + "epoch": 0.24445, + "grad_norm": 0.8183718919754028, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 24445 + }, + { + "epoch": 0.24446, + "grad_norm": 0.7424532175064087, + "learning_rate": 0.003, + "loss": 4.001, + "step": 24446 + }, + { + "epoch": 0.24447, + "grad_norm": 0.8671339154243469, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 24447 + }, + { + "epoch": 0.24448, + "grad_norm": 0.9028390645980835, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 24448 + }, + { + "epoch": 0.24449, + "grad_norm": 0.9230623245239258, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 24449 + }, + { + "epoch": 0.2445, + "grad_norm": 0.7987543344497681, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 24450 + }, + { + "epoch": 0.24451, + "grad_norm": 0.6923510432243347, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 24451 + }, + { + "epoch": 0.24452, + "grad_norm": 0.731844961643219, + "learning_rate": 0.003, + "loss": 3.995, + "step": 24452 + }, + { + "epoch": 0.24453, + "grad_norm": 0.729561984539032, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 24453 + }, + { + "epoch": 0.24454, + "grad_norm": 0.6925594210624695, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 24454 + }, + { + "epoch": 0.24455, + "grad_norm": 0.7086461782455444, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 24455 + }, + { + "epoch": 0.24456, + "grad_norm": 0.6785542964935303, + "learning_rate": 0.003, + "loss": 3.995, + "step": 24456 + }, + { + "epoch": 0.24457, + "grad_norm": 0.6417614817619324, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 24457 + }, + { + "epoch": 0.24458, + "grad_norm": 0.6091236472129822, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 24458 + }, + { + "epoch": 0.24459, + "grad_norm": 0.714839518070221, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 24459 + }, + { + "epoch": 0.2446, + "grad_norm": 0.8169758319854736, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 24460 + }, + { + "epoch": 0.24461, + "grad_norm": 0.8879427313804626, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 24461 + }, + { + "epoch": 0.24462, + "grad_norm": 0.8160597681999207, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 24462 + }, + { + "epoch": 0.24463, + "grad_norm": 0.9406057596206665, + "learning_rate": 0.003, + "loss": 3.999, + "step": 24463 + }, + { + "epoch": 0.24464, + "grad_norm": 1.0898197889328003, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 24464 + }, + { + "epoch": 0.24465, + "grad_norm": 1.1369131803512573, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 24465 + }, + { + "epoch": 0.24466, + "grad_norm": 1.1583081483840942, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 24466 + }, + { + "epoch": 0.24467, + "grad_norm": 0.7063309550285339, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 24467 + }, + { + "epoch": 0.24468, + "grad_norm": 0.6708469390869141, + "learning_rate": 0.003, + "loss": 3.988, + "step": 24468 + }, + { + "epoch": 0.24469, + "grad_norm": 0.6918720006942749, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 24469 + }, + { + "epoch": 0.2447, + "grad_norm": 0.6936019062995911, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 24470 + }, + { + "epoch": 0.24471, + "grad_norm": 0.7153578996658325, + "learning_rate": 0.003, + "loss": 3.977, + "step": 24471 + }, + { + "epoch": 0.24472, + "grad_norm": 0.7084479928016663, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 24472 + }, + { + "epoch": 0.24473, + "grad_norm": 0.6705135107040405, + "learning_rate": 0.003, + "loss": 4.005, + "step": 24473 + }, + { + "epoch": 0.24474, + "grad_norm": 0.7231819033622742, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 24474 + }, + { + "epoch": 0.24475, + "grad_norm": 0.723834216594696, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 24475 + }, + { + "epoch": 0.24476, + "grad_norm": 0.7275248765945435, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 24476 + }, + { + "epoch": 0.24477, + "grad_norm": 0.793616771697998, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 24477 + }, + { + "epoch": 0.24478, + "grad_norm": 0.9256736040115356, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 24478 + }, + { + "epoch": 0.24479, + "grad_norm": 1.3068212270736694, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 24479 + }, + { + "epoch": 0.2448, + "grad_norm": 0.8406258225440979, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 24480 + }, + { + "epoch": 0.24481, + "grad_norm": 0.9424657225608826, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 24481 + }, + { + "epoch": 0.24482, + "grad_norm": 1.4036558866500854, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 24482 + }, + { + "epoch": 0.24483, + "grad_norm": 0.8460255861282349, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 24483 + }, + { + "epoch": 0.24484, + "grad_norm": 0.6961652636528015, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 24484 + }, + { + "epoch": 0.24485, + "grad_norm": 0.6791902780532837, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 24485 + }, + { + "epoch": 0.24486, + "grad_norm": 0.7373426556587219, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 24486 + }, + { + "epoch": 0.24487, + "grad_norm": 0.7842598557472229, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 24487 + }, + { + "epoch": 0.24488, + "grad_norm": 0.9167132377624512, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 24488 + }, + { + "epoch": 0.24489, + "grad_norm": 1.0499416589736938, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 24489 + }, + { + "epoch": 0.2449, + "grad_norm": 0.9714927077293396, + "learning_rate": 0.003, + "loss": 4.007, + "step": 24490 + }, + { + "epoch": 0.24491, + "grad_norm": 0.7824407815933228, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 24491 + }, + { + "epoch": 0.24492, + "grad_norm": 0.8762727975845337, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 24492 + }, + { + "epoch": 0.24493, + "grad_norm": 1.1781975030899048, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 24493 + }, + { + "epoch": 0.24494, + "grad_norm": 0.9320746064186096, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 24494 + }, + { + "epoch": 0.24495, + "grad_norm": 0.96622633934021, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 24495 + }, + { + "epoch": 0.24496, + "grad_norm": 0.8392488360404968, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 24496 + }, + { + "epoch": 0.24497, + "grad_norm": 0.7100114226341248, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 24497 + }, + { + "epoch": 0.24498, + "grad_norm": 0.6612435579299927, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 24498 + }, + { + "epoch": 0.24499, + "grad_norm": 0.8289709687232971, + "learning_rate": 0.003, + "loss": 3.965, + "step": 24499 + }, + { + "epoch": 0.245, + "grad_norm": 1.0422425270080566, + "learning_rate": 0.003, + "loss": 3.986, + "step": 24500 + }, + { + "epoch": 0.24501, + "grad_norm": 0.954322338104248, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 24501 + }, + { + "epoch": 0.24502, + "grad_norm": 0.8213654160499573, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 24502 + }, + { + "epoch": 0.24503, + "grad_norm": 0.8856193423271179, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 24503 + }, + { + "epoch": 0.24504, + "grad_norm": 0.8652067184448242, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 24504 + }, + { + "epoch": 0.24505, + "grad_norm": 0.8557260632514954, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 24505 + }, + { + "epoch": 0.24506, + "grad_norm": 1.0337462425231934, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 24506 + }, + { + "epoch": 0.24507, + "grad_norm": 1.0695793628692627, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 24507 + }, + { + "epoch": 0.24508, + "grad_norm": 0.8464872241020203, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 24508 + }, + { + "epoch": 0.24509, + "grad_norm": 0.761448085308075, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 24509 + }, + { + "epoch": 0.2451, + "grad_norm": 0.8186254501342773, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 24510 + }, + { + "epoch": 0.24511, + "grad_norm": 0.8206101059913635, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 24511 + }, + { + "epoch": 0.24512, + "grad_norm": 0.7922770380973816, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 24512 + }, + { + "epoch": 0.24513, + "grad_norm": 0.8868371248245239, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 24513 + }, + { + "epoch": 0.24514, + "grad_norm": 0.8534619808197021, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 24514 + }, + { + "epoch": 0.24515, + "grad_norm": 1.0945661067962646, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 24515 + }, + { + "epoch": 0.24516, + "grad_norm": 1.2560889720916748, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 24516 + }, + { + "epoch": 0.24517, + "grad_norm": 0.8513187170028687, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 24517 + }, + { + "epoch": 0.24518, + "grad_norm": 0.7543923258781433, + "learning_rate": 0.003, + "loss": 3.999, + "step": 24518 + }, + { + "epoch": 0.24519, + "grad_norm": 0.7318226099014282, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 24519 + }, + { + "epoch": 0.2452, + "grad_norm": 0.7642946839332581, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 24520 + }, + { + "epoch": 0.24521, + "grad_norm": 0.8515164256095886, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 24521 + }, + { + "epoch": 0.24522, + "grad_norm": 0.8021851778030396, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 24522 + }, + { + "epoch": 0.24523, + "grad_norm": 0.8050875663757324, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 24523 + }, + { + "epoch": 0.24524, + "grad_norm": 0.8655991554260254, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 24524 + }, + { + "epoch": 0.24525, + "grad_norm": 0.9964656829833984, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 24525 + }, + { + "epoch": 0.24526, + "grad_norm": 1.083990216255188, + "learning_rate": 0.003, + "loss": 3.988, + "step": 24526 + }, + { + "epoch": 0.24527, + "grad_norm": 0.986695408821106, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 24527 + }, + { + "epoch": 0.24528, + "grad_norm": 0.8350574970245361, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 24528 + }, + { + "epoch": 0.24529, + "grad_norm": 0.7125077247619629, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 24529 + }, + { + "epoch": 0.2453, + "grad_norm": 0.7098248600959778, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 24530 + }, + { + "epoch": 0.24531, + "grad_norm": 0.795041024684906, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 24531 + }, + { + "epoch": 0.24532, + "grad_norm": 0.7796959280967712, + "learning_rate": 0.003, + "loss": 3.999, + "step": 24532 + }, + { + "epoch": 0.24533, + "grad_norm": 0.9860004186630249, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 24533 + }, + { + "epoch": 0.24534, + "grad_norm": 1.4677226543426514, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 24534 + }, + { + "epoch": 0.24535, + "grad_norm": 0.7187691330909729, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 24535 + }, + { + "epoch": 0.24536, + "grad_norm": 0.7061315178871155, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 24536 + }, + { + "epoch": 0.24537, + "grad_norm": 0.7225534915924072, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 24537 + }, + { + "epoch": 0.24538, + "grad_norm": 0.6294336915016174, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 24538 + }, + { + "epoch": 0.24539, + "grad_norm": 0.647836446762085, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 24539 + }, + { + "epoch": 0.2454, + "grad_norm": 0.7019044160842896, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 24540 + }, + { + "epoch": 0.24541, + "grad_norm": 0.8210511803627014, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 24541 + }, + { + "epoch": 0.24542, + "grad_norm": 0.9146199822425842, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 24542 + }, + { + "epoch": 0.24543, + "grad_norm": 1.116714358329773, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 24543 + }, + { + "epoch": 0.24544, + "grad_norm": 0.9303229451179504, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 24544 + }, + { + "epoch": 0.24545, + "grad_norm": 0.9178400039672852, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 24545 + }, + { + "epoch": 0.24546, + "grad_norm": 0.8506669998168945, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 24546 + }, + { + "epoch": 0.24547, + "grad_norm": 0.6559498310089111, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 24547 + }, + { + "epoch": 0.24548, + "grad_norm": 0.6241545677185059, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 24548 + }, + { + "epoch": 0.24549, + "grad_norm": 0.6625649333000183, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 24549 + }, + { + "epoch": 0.2455, + "grad_norm": 0.7543411254882812, + "learning_rate": 0.003, + "loss": 3.979, + "step": 24550 + }, + { + "epoch": 0.24551, + "grad_norm": 0.8398850560188293, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 24551 + }, + { + "epoch": 0.24552, + "grad_norm": 0.7694680094718933, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 24552 + }, + { + "epoch": 0.24553, + "grad_norm": 0.6425963044166565, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 24553 + }, + { + "epoch": 0.24554, + "grad_norm": 0.6388164162635803, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 24554 + }, + { + "epoch": 0.24555, + "grad_norm": 0.7068902850151062, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 24555 + }, + { + "epoch": 0.24556, + "grad_norm": 0.7672874331474304, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 24556 + }, + { + "epoch": 0.24557, + "grad_norm": 0.6834967136383057, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 24557 + }, + { + "epoch": 0.24558, + "grad_norm": 0.6754685640335083, + "learning_rate": 0.003, + "loss": 4.01, + "step": 24558 + }, + { + "epoch": 0.24559, + "grad_norm": 0.8095713257789612, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 24559 + }, + { + "epoch": 0.2456, + "grad_norm": 1.0592540502548218, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 24560 + }, + { + "epoch": 0.24561, + "grad_norm": 1.2447470426559448, + "learning_rate": 0.003, + "loss": 3.972, + "step": 24561 + }, + { + "epoch": 0.24562, + "grad_norm": 0.969610333442688, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 24562 + }, + { + "epoch": 0.24563, + "grad_norm": 0.9348779916763306, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 24563 + }, + { + "epoch": 0.24564, + "grad_norm": 0.7918995022773743, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 24564 + }, + { + "epoch": 0.24565, + "grad_norm": 0.895946204662323, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 24565 + }, + { + "epoch": 0.24566, + "grad_norm": 1.2126567363739014, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 24566 + }, + { + "epoch": 0.24567, + "grad_norm": 1.2151145935058594, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 24567 + }, + { + "epoch": 0.24568, + "grad_norm": 0.826689600944519, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 24568 + }, + { + "epoch": 0.24569, + "grad_norm": 0.6880024075508118, + "learning_rate": 0.003, + "loss": 3.964, + "step": 24569 + }, + { + "epoch": 0.2457, + "grad_norm": 0.7377886176109314, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 24570 + }, + { + "epoch": 0.24571, + "grad_norm": 0.7851471900939941, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 24571 + }, + { + "epoch": 0.24572, + "grad_norm": 0.8951870203018188, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 24572 + }, + { + "epoch": 0.24573, + "grad_norm": 0.9546773433685303, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 24573 + }, + { + "epoch": 0.24574, + "grad_norm": 0.987238347530365, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 24574 + }, + { + "epoch": 0.24575, + "grad_norm": 0.9911940693855286, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 24575 + }, + { + "epoch": 0.24576, + "grad_norm": 0.9408009052276611, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 24576 + }, + { + "epoch": 0.24577, + "grad_norm": 0.9311189651489258, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 24577 + }, + { + "epoch": 0.24578, + "grad_norm": 0.7981687784194946, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 24578 + }, + { + "epoch": 0.24579, + "grad_norm": 0.8425568342208862, + "learning_rate": 0.003, + "loss": 3.998, + "step": 24579 + }, + { + "epoch": 0.2458, + "grad_norm": 0.8311449289321899, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 24580 + }, + { + "epoch": 0.24581, + "grad_norm": 0.8978705406188965, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 24581 + }, + { + "epoch": 0.24582, + "grad_norm": 1.1327588558197021, + "learning_rate": 0.003, + "loss": 4.046, + "step": 24582 + }, + { + "epoch": 0.24583, + "grad_norm": 0.901011049747467, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 24583 + }, + { + "epoch": 0.24584, + "grad_norm": 0.7642689347267151, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 24584 + }, + { + "epoch": 0.24585, + "grad_norm": 0.6698705554008484, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 24585 + }, + { + "epoch": 0.24586, + "grad_norm": 0.5179368853569031, + "learning_rate": 0.003, + "loss": 3.998, + "step": 24586 + }, + { + "epoch": 0.24587, + "grad_norm": 0.4540056884288788, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 24587 + }, + { + "epoch": 0.24588, + "grad_norm": 0.5180246829986572, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 24588 + }, + { + "epoch": 0.24589, + "grad_norm": 0.550722599029541, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 24589 + }, + { + "epoch": 0.2459, + "grad_norm": 0.5376285910606384, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 24590 + }, + { + "epoch": 0.24591, + "grad_norm": 0.5396836996078491, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 24591 + }, + { + "epoch": 0.24592, + "grad_norm": 0.5385528802871704, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 24592 + }, + { + "epoch": 0.24593, + "grad_norm": 0.6246936321258545, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 24593 + }, + { + "epoch": 0.24594, + "grad_norm": 0.7900716662406921, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 24594 + }, + { + "epoch": 0.24595, + "grad_norm": 1.119105339050293, + "learning_rate": 0.003, + "loss": 3.985, + "step": 24595 + }, + { + "epoch": 0.24596, + "grad_norm": 1.3152092695236206, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 24596 + }, + { + "epoch": 0.24597, + "grad_norm": 0.7165833115577698, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 24597 + }, + { + "epoch": 0.24598, + "grad_norm": 0.7387587428092957, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 24598 + }, + { + "epoch": 0.24599, + "grad_norm": 0.7698293328285217, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 24599 + }, + { + "epoch": 0.246, + "grad_norm": 0.8914187550544739, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 24600 + }, + { + "epoch": 0.24601, + "grad_norm": 1.1730300188064575, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 24601 + }, + { + "epoch": 0.24602, + "grad_norm": 0.8330344557762146, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 24602 + }, + { + "epoch": 0.24603, + "grad_norm": 0.7794854044914246, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 24603 + }, + { + "epoch": 0.24604, + "grad_norm": 0.6019008159637451, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 24604 + }, + { + "epoch": 0.24605, + "grad_norm": 0.6435580849647522, + "learning_rate": 0.003, + "loss": 3.998, + "step": 24605 + }, + { + "epoch": 0.24606, + "grad_norm": 0.7953606843948364, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 24606 + }, + { + "epoch": 0.24607, + "grad_norm": 0.9429084062576294, + "learning_rate": 0.003, + "loss": 4.014, + "step": 24607 + }, + { + "epoch": 0.24608, + "grad_norm": 1.0044176578521729, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 24608 + }, + { + "epoch": 0.24609, + "grad_norm": 1.0098133087158203, + "learning_rate": 0.003, + "loss": 3.984, + "step": 24609 + }, + { + "epoch": 0.2461, + "grad_norm": 1.0126692056655884, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 24610 + }, + { + "epoch": 0.24611, + "grad_norm": 1.025146245956421, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 24611 + }, + { + "epoch": 0.24612, + "grad_norm": 1.225239634513855, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 24612 + }, + { + "epoch": 0.24613, + "grad_norm": 1.2699867486953735, + "learning_rate": 0.003, + "loss": 4.016, + "step": 24613 + }, + { + "epoch": 0.24614, + "grad_norm": 0.6255699992179871, + "learning_rate": 0.003, + "loss": 3.989, + "step": 24614 + }, + { + "epoch": 0.24615, + "grad_norm": 0.7257640361785889, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 24615 + }, + { + "epoch": 0.24616, + "grad_norm": 0.9838553071022034, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 24616 + }, + { + "epoch": 0.24617, + "grad_norm": 1.1333922147750854, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 24617 + }, + { + "epoch": 0.24618, + "grad_norm": 1.1869831085205078, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 24618 + }, + { + "epoch": 0.24619, + "grad_norm": 0.7528954148292542, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 24619 + }, + { + "epoch": 0.2462, + "grad_norm": 0.7268115878105164, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 24620 + }, + { + "epoch": 0.24621, + "grad_norm": 0.7651877403259277, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 24621 + }, + { + "epoch": 0.24622, + "grad_norm": 0.6624248623847961, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 24622 + }, + { + "epoch": 0.24623, + "grad_norm": 0.7681838870048523, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 24623 + }, + { + "epoch": 0.24624, + "grad_norm": 0.7746779918670654, + "learning_rate": 0.003, + "loss": 3.989, + "step": 24624 + }, + { + "epoch": 0.24625, + "grad_norm": 0.8453778624534607, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 24625 + }, + { + "epoch": 0.24626, + "grad_norm": 0.9177544116973877, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 24626 + }, + { + "epoch": 0.24627, + "grad_norm": 0.9964888095855713, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 24627 + }, + { + "epoch": 0.24628, + "grad_norm": 1.199730396270752, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 24628 + }, + { + "epoch": 0.24629, + "grad_norm": 0.7351564168930054, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 24629 + }, + { + "epoch": 0.2463, + "grad_norm": 0.5996708273887634, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 24630 + }, + { + "epoch": 0.24631, + "grad_norm": 0.6411522030830383, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 24631 + }, + { + "epoch": 0.24632, + "grad_norm": 0.7440872192382812, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 24632 + }, + { + "epoch": 0.24633, + "grad_norm": 0.828597366809845, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 24633 + }, + { + "epoch": 0.24634, + "grad_norm": 0.9964712262153625, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 24634 + }, + { + "epoch": 0.24635, + "grad_norm": 1.0856482982635498, + "learning_rate": 0.003, + "loss": 4.023, + "step": 24635 + }, + { + "epoch": 0.24636, + "grad_norm": 0.825675904750824, + "learning_rate": 0.003, + "loss": 4.001, + "step": 24636 + }, + { + "epoch": 0.24637, + "grad_norm": 1.0052566528320312, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 24637 + }, + { + "epoch": 0.24638, + "grad_norm": 1.0923922061920166, + "learning_rate": 0.003, + "loss": 3.999, + "step": 24638 + }, + { + "epoch": 0.24639, + "grad_norm": 0.9277573227882385, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 24639 + }, + { + "epoch": 0.2464, + "grad_norm": 0.9405017495155334, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 24640 + }, + { + "epoch": 0.24641, + "grad_norm": 1.0012842416763306, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 24641 + }, + { + "epoch": 0.24642, + "grad_norm": 1.2281357049942017, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 24642 + }, + { + "epoch": 0.24643, + "grad_norm": 1.0370917320251465, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 24643 + }, + { + "epoch": 0.24644, + "grad_norm": 0.9260424375534058, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 24644 + }, + { + "epoch": 0.24645, + "grad_norm": 0.8959420323371887, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 24645 + }, + { + "epoch": 0.24646, + "grad_norm": 1.0576906204223633, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 24646 + }, + { + "epoch": 0.24647, + "grad_norm": 0.8320174813270569, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 24647 + }, + { + "epoch": 0.24648, + "grad_norm": 0.8743922114372253, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 24648 + }, + { + "epoch": 0.24649, + "grad_norm": 0.8123977184295654, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 24649 + }, + { + "epoch": 0.2465, + "grad_norm": 0.8017212152481079, + "learning_rate": 0.003, + "loss": 4.0, + "step": 24650 + }, + { + "epoch": 0.24651, + "grad_norm": 0.7856309413909912, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 24651 + }, + { + "epoch": 0.24652, + "grad_norm": 0.8401914238929749, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 24652 + }, + { + "epoch": 0.24653, + "grad_norm": 0.8806281089782715, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 24653 + }, + { + "epoch": 0.24654, + "grad_norm": 0.9998926520347595, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 24654 + }, + { + "epoch": 0.24655, + "grad_norm": 1.3150643110275269, + "learning_rate": 0.003, + "loss": 4.047, + "step": 24655 + }, + { + "epoch": 0.24656, + "grad_norm": 0.7087311744689941, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 24656 + }, + { + "epoch": 0.24657, + "grad_norm": 0.6015285849571228, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 24657 + }, + { + "epoch": 0.24658, + "grad_norm": 0.7103121876716614, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 24658 + }, + { + "epoch": 0.24659, + "grad_norm": 0.763187050819397, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 24659 + }, + { + "epoch": 0.2466, + "grad_norm": 1.0595413446426392, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 24660 + }, + { + "epoch": 0.24661, + "grad_norm": 1.1412715911865234, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 24661 + }, + { + "epoch": 0.24662, + "grad_norm": 0.6147118806838989, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 24662 + }, + { + "epoch": 0.24663, + "grad_norm": 0.6184364557266235, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 24663 + }, + { + "epoch": 0.24664, + "grad_norm": 0.6470197439193726, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 24664 + }, + { + "epoch": 0.24665, + "grad_norm": 0.700421929359436, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 24665 + }, + { + "epoch": 0.24666, + "grad_norm": 0.8017335534095764, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 24666 + }, + { + "epoch": 0.24667, + "grad_norm": 0.7758115530014038, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 24667 + }, + { + "epoch": 0.24668, + "grad_norm": 0.7210459113121033, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 24668 + }, + { + "epoch": 0.24669, + "grad_norm": 0.7445424199104309, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 24669 + }, + { + "epoch": 0.2467, + "grad_norm": 0.8884041905403137, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 24670 + }, + { + "epoch": 0.24671, + "grad_norm": 1.2713096141815186, + "learning_rate": 0.003, + "loss": 3.989, + "step": 24671 + }, + { + "epoch": 0.24672, + "grad_norm": 0.951755940914154, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 24672 + }, + { + "epoch": 0.24673, + "grad_norm": 0.8798564672470093, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 24673 + }, + { + "epoch": 0.24674, + "grad_norm": 0.7937695980072021, + "learning_rate": 0.003, + "loss": 3.995, + "step": 24674 + }, + { + "epoch": 0.24675, + "grad_norm": 0.6838546395301819, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 24675 + }, + { + "epoch": 0.24676, + "grad_norm": 0.6999340057373047, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 24676 + }, + { + "epoch": 0.24677, + "grad_norm": 0.6817321181297302, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 24677 + }, + { + "epoch": 0.24678, + "grad_norm": 0.863294243812561, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 24678 + }, + { + "epoch": 0.24679, + "grad_norm": 1.2580363750457764, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 24679 + }, + { + "epoch": 0.2468, + "grad_norm": 0.93878573179245, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 24680 + }, + { + "epoch": 0.24681, + "grad_norm": 0.8864484429359436, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 24681 + }, + { + "epoch": 0.24682, + "grad_norm": 0.960409939289093, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 24682 + }, + { + "epoch": 0.24683, + "grad_norm": 0.9043546915054321, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 24683 + }, + { + "epoch": 0.24684, + "grad_norm": 1.0016083717346191, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 24684 + }, + { + "epoch": 0.24685, + "grad_norm": 1.0362223386764526, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 24685 + }, + { + "epoch": 0.24686, + "grad_norm": 0.9994737505912781, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 24686 + }, + { + "epoch": 0.24687, + "grad_norm": 1.0364867448806763, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 24687 + }, + { + "epoch": 0.24688, + "grad_norm": 1.018257737159729, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 24688 + }, + { + "epoch": 0.24689, + "grad_norm": 0.833156168460846, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 24689 + }, + { + "epoch": 0.2469, + "grad_norm": 0.7173692584037781, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 24690 + }, + { + "epoch": 0.24691, + "grad_norm": 0.6706775426864624, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 24691 + }, + { + "epoch": 0.24692, + "grad_norm": 0.8207137584686279, + "learning_rate": 0.003, + "loss": 3.978, + "step": 24692 + }, + { + "epoch": 0.24693, + "grad_norm": 1.0563616752624512, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 24693 + }, + { + "epoch": 0.24694, + "grad_norm": 1.2570785284042358, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 24694 + }, + { + "epoch": 0.24695, + "grad_norm": 0.6506404280662537, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 24695 + }, + { + "epoch": 0.24696, + "grad_norm": 0.7469118237495422, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 24696 + }, + { + "epoch": 0.24697, + "grad_norm": 0.9335440993309021, + "learning_rate": 0.003, + "loss": 3.988, + "step": 24697 + }, + { + "epoch": 0.24698, + "grad_norm": 0.8773851990699768, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 24698 + }, + { + "epoch": 0.24699, + "grad_norm": 0.7988091111183167, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 24699 + }, + { + "epoch": 0.247, + "grad_norm": 0.6964368224143982, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 24700 + }, + { + "epoch": 0.24701, + "grad_norm": 0.6985162496566772, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 24701 + }, + { + "epoch": 0.24702, + "grad_norm": 0.7271296977996826, + "learning_rate": 0.003, + "loss": 3.979, + "step": 24702 + }, + { + "epoch": 0.24703, + "grad_norm": 0.7886885404586792, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 24703 + }, + { + "epoch": 0.24704, + "grad_norm": 1.028521180152893, + "learning_rate": 0.003, + "loss": 4.011, + "step": 24704 + }, + { + "epoch": 0.24705, + "grad_norm": 1.2982841730117798, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 24705 + }, + { + "epoch": 0.24706, + "grad_norm": 0.72129225730896, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 24706 + }, + { + "epoch": 0.24707, + "grad_norm": 0.8212591409683228, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 24707 + }, + { + "epoch": 0.24708, + "grad_norm": 0.8618593811988831, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 24708 + }, + { + "epoch": 0.24709, + "grad_norm": 0.77595454454422, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 24709 + }, + { + "epoch": 0.2471, + "grad_norm": 0.6629289984703064, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 24710 + }, + { + "epoch": 0.24711, + "grad_norm": 0.6858501434326172, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 24711 + }, + { + "epoch": 0.24712, + "grad_norm": 0.7351704239845276, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 24712 + }, + { + "epoch": 0.24713, + "grad_norm": 0.8099431395530701, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 24713 + }, + { + "epoch": 0.24714, + "grad_norm": 0.8665048480033875, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 24714 + }, + { + "epoch": 0.24715, + "grad_norm": 0.8870483636856079, + "learning_rate": 0.003, + "loss": 4.0, + "step": 24715 + }, + { + "epoch": 0.24716, + "grad_norm": 0.9643385410308838, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 24716 + }, + { + "epoch": 0.24717, + "grad_norm": 0.899376630783081, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 24717 + }, + { + "epoch": 0.24718, + "grad_norm": 1.0726062059402466, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 24718 + }, + { + "epoch": 0.24719, + "grad_norm": 1.0510318279266357, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 24719 + }, + { + "epoch": 0.2472, + "grad_norm": 1.1290702819824219, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 24720 + }, + { + "epoch": 0.24721, + "grad_norm": 1.210442066192627, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 24721 + }, + { + "epoch": 0.24722, + "grad_norm": 0.8648130297660828, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 24722 + }, + { + "epoch": 0.24723, + "grad_norm": 0.6202104687690735, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 24723 + }, + { + "epoch": 0.24724, + "grad_norm": 0.6014062762260437, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 24724 + }, + { + "epoch": 0.24725, + "grad_norm": 0.5872074365615845, + "learning_rate": 0.003, + "loss": 3.972, + "step": 24725 + }, + { + "epoch": 0.24726, + "grad_norm": 0.613756000995636, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 24726 + }, + { + "epoch": 0.24727, + "grad_norm": 0.5461459159851074, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 24727 + }, + { + "epoch": 0.24728, + "grad_norm": 0.5478624105453491, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 24728 + }, + { + "epoch": 0.24729, + "grad_norm": 0.6532526612281799, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 24729 + }, + { + "epoch": 0.2473, + "grad_norm": 0.78339684009552, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 24730 + }, + { + "epoch": 0.24731, + "grad_norm": 0.828387975692749, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 24731 + }, + { + "epoch": 0.24732, + "grad_norm": 1.0356923341751099, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 24732 + }, + { + "epoch": 0.24733, + "grad_norm": 1.4180331230163574, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 24733 + }, + { + "epoch": 0.24734, + "grad_norm": 0.532661497592926, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 24734 + }, + { + "epoch": 0.24735, + "grad_norm": 0.8910018801689148, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 24735 + }, + { + "epoch": 0.24736, + "grad_norm": 1.145992398262024, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 24736 + }, + { + "epoch": 0.24737, + "grad_norm": 0.785019040107727, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 24737 + }, + { + "epoch": 0.24738, + "grad_norm": 0.6694530248641968, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 24738 + }, + { + "epoch": 0.24739, + "grad_norm": 0.627246081829071, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 24739 + }, + { + "epoch": 0.2474, + "grad_norm": 0.7213286757469177, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 24740 + }, + { + "epoch": 0.24741, + "grad_norm": 0.7510721683502197, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 24741 + }, + { + "epoch": 0.24742, + "grad_norm": 0.7680305242538452, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 24742 + }, + { + "epoch": 0.24743, + "grad_norm": 0.8497138023376465, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 24743 + }, + { + "epoch": 0.24744, + "grad_norm": 0.9213382601737976, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 24744 + }, + { + "epoch": 0.24745, + "grad_norm": 1.1705552339553833, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 24745 + }, + { + "epoch": 0.24746, + "grad_norm": 0.9279419183731079, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 24746 + }, + { + "epoch": 0.24747, + "grad_norm": 0.8532731533050537, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 24747 + }, + { + "epoch": 0.24748, + "grad_norm": 0.9292433261871338, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 24748 + }, + { + "epoch": 0.24749, + "grad_norm": 1.1180248260498047, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 24749 + }, + { + "epoch": 0.2475, + "grad_norm": 0.828274667263031, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 24750 + }, + { + "epoch": 0.24751, + "grad_norm": 0.7856637239456177, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 24751 + }, + { + "epoch": 0.24752, + "grad_norm": 0.8178903460502625, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 24752 + }, + { + "epoch": 0.24753, + "grad_norm": 1.0824123620986938, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 24753 + }, + { + "epoch": 0.24754, + "grad_norm": 0.9845330715179443, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 24754 + }, + { + "epoch": 0.24755, + "grad_norm": 0.9138570427894592, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 24755 + }, + { + "epoch": 0.24756, + "grad_norm": 0.9604371190071106, + "learning_rate": 0.003, + "loss": 4.003, + "step": 24756 + }, + { + "epoch": 0.24757, + "grad_norm": 0.9706985950469971, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 24757 + }, + { + "epoch": 0.24758, + "grad_norm": 0.9941304922103882, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 24758 + }, + { + "epoch": 0.24759, + "grad_norm": 0.9728471040725708, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 24759 + }, + { + "epoch": 0.2476, + "grad_norm": 0.947729229927063, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 24760 + }, + { + "epoch": 0.24761, + "grad_norm": 1.2029330730438232, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 24761 + }, + { + "epoch": 0.24762, + "grad_norm": 0.8686008453369141, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 24762 + }, + { + "epoch": 0.24763, + "grad_norm": 0.9104997515678406, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 24763 + }, + { + "epoch": 0.24764, + "grad_norm": 1.0400750637054443, + "learning_rate": 0.003, + "loss": 4.025, + "step": 24764 + }, + { + "epoch": 0.24765, + "grad_norm": 1.1206459999084473, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 24765 + }, + { + "epoch": 0.24766, + "grad_norm": 0.7800242900848389, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 24766 + }, + { + "epoch": 0.24767, + "grad_norm": 0.7925225496292114, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 24767 + }, + { + "epoch": 0.24768, + "grad_norm": 0.8124421834945679, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 24768 + }, + { + "epoch": 0.24769, + "grad_norm": 0.7310168147087097, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 24769 + }, + { + "epoch": 0.2477, + "grad_norm": 0.6278213262557983, + "learning_rate": 0.003, + "loss": 4.013, + "step": 24770 + }, + { + "epoch": 0.24771, + "grad_norm": 0.6168035268783569, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 24771 + }, + { + "epoch": 0.24772, + "grad_norm": 0.6934630274772644, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 24772 + }, + { + "epoch": 0.24773, + "grad_norm": 0.6442554593086243, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 24773 + }, + { + "epoch": 0.24774, + "grad_norm": 0.8078383207321167, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 24774 + }, + { + "epoch": 0.24775, + "grad_norm": 1.223679780960083, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 24775 + }, + { + "epoch": 0.24776, + "grad_norm": 1.0514804124832153, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 24776 + }, + { + "epoch": 0.24777, + "grad_norm": 0.8552996516227722, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 24777 + }, + { + "epoch": 0.24778, + "grad_norm": 0.8119766712188721, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 24778 + }, + { + "epoch": 0.24779, + "grad_norm": 0.9220768809318542, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 24779 + }, + { + "epoch": 0.2478, + "grad_norm": 0.9574283957481384, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 24780 + }, + { + "epoch": 0.24781, + "grad_norm": 0.8165639042854309, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 24781 + }, + { + "epoch": 0.24782, + "grad_norm": 0.8203229904174805, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 24782 + }, + { + "epoch": 0.24783, + "grad_norm": 0.9229028224945068, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 24783 + }, + { + "epoch": 0.24784, + "grad_norm": 0.9933738708496094, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 24784 + }, + { + "epoch": 0.24785, + "grad_norm": 1.0953911542892456, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 24785 + }, + { + "epoch": 0.24786, + "grad_norm": 1.235741376876831, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 24786 + }, + { + "epoch": 0.24787, + "grad_norm": 1.0051509141921997, + "learning_rate": 0.003, + "loss": 4.01, + "step": 24787 + }, + { + "epoch": 0.24788, + "grad_norm": 1.0165199041366577, + "learning_rate": 0.003, + "loss": 3.993, + "step": 24788 + }, + { + "epoch": 0.24789, + "grad_norm": 1.1195025444030762, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 24789 + }, + { + "epoch": 0.2479, + "grad_norm": 0.8793075680732727, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 24790 + }, + { + "epoch": 0.24791, + "grad_norm": 0.993976891040802, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 24791 + }, + { + "epoch": 0.24792, + "grad_norm": 1.2091047763824463, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 24792 + }, + { + "epoch": 0.24793, + "grad_norm": 0.6669012308120728, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 24793 + }, + { + "epoch": 0.24794, + "grad_norm": 0.6312685012817383, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 24794 + }, + { + "epoch": 0.24795, + "grad_norm": 0.6631552577018738, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 24795 + }, + { + "epoch": 0.24796, + "grad_norm": 0.9534623026847839, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 24796 + }, + { + "epoch": 0.24797, + "grad_norm": 1.0199733972549438, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 24797 + }, + { + "epoch": 0.24798, + "grad_norm": 0.8658064007759094, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 24798 + }, + { + "epoch": 0.24799, + "grad_norm": 0.7652483582496643, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 24799 + }, + { + "epoch": 0.248, + "grad_norm": 0.7063347101211548, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 24800 + }, + { + "epoch": 0.24801, + "grad_norm": 0.7836357355117798, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 24801 + }, + { + "epoch": 0.24802, + "grad_norm": 0.7500140070915222, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 24802 + }, + { + "epoch": 0.24803, + "grad_norm": 0.7520449161529541, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 24803 + }, + { + "epoch": 0.24804, + "grad_norm": 0.8072378039360046, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 24804 + }, + { + "epoch": 0.24805, + "grad_norm": 0.8734562993049622, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 24805 + }, + { + "epoch": 0.24806, + "grad_norm": 0.9341596364974976, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 24806 + }, + { + "epoch": 0.24807, + "grad_norm": 1.0741232633590698, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 24807 + }, + { + "epoch": 0.24808, + "grad_norm": 1.0526961088180542, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 24808 + }, + { + "epoch": 0.24809, + "grad_norm": 0.8061981201171875, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 24809 + }, + { + "epoch": 0.2481, + "grad_norm": 0.6738079190254211, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 24810 + }, + { + "epoch": 0.24811, + "grad_norm": 0.6337065100669861, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 24811 + }, + { + "epoch": 0.24812, + "grad_norm": 0.5607208609580994, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 24812 + }, + { + "epoch": 0.24813, + "grad_norm": 0.6865894794464111, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 24813 + }, + { + "epoch": 0.24814, + "grad_norm": 0.8439019918441772, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 24814 + }, + { + "epoch": 0.24815, + "grad_norm": 1.004426121711731, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 24815 + }, + { + "epoch": 0.24816, + "grad_norm": 0.9761579632759094, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 24816 + }, + { + "epoch": 0.24817, + "grad_norm": 1.0125516653060913, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 24817 + }, + { + "epoch": 0.24818, + "grad_norm": 0.9964504241943359, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 24818 + }, + { + "epoch": 0.24819, + "grad_norm": 0.8722450137138367, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 24819 + }, + { + "epoch": 0.2482, + "grad_norm": 0.8917888402938843, + "learning_rate": 0.003, + "loss": 4.029, + "step": 24820 + }, + { + "epoch": 0.24821, + "grad_norm": 0.9461867213249207, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 24821 + }, + { + "epoch": 0.24822, + "grad_norm": 0.8376356959342957, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 24822 + }, + { + "epoch": 0.24823, + "grad_norm": 0.8976319432258606, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 24823 + }, + { + "epoch": 0.24824, + "grad_norm": 1.1950767040252686, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 24824 + }, + { + "epoch": 0.24825, + "grad_norm": 1.0443089008331299, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 24825 + }, + { + "epoch": 0.24826, + "grad_norm": 0.9099676609039307, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 24826 + }, + { + "epoch": 0.24827, + "grad_norm": 0.8096210956573486, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 24827 + }, + { + "epoch": 0.24828, + "grad_norm": 0.7644968628883362, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 24828 + }, + { + "epoch": 0.24829, + "grad_norm": 0.9758970141410828, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 24829 + }, + { + "epoch": 0.2483, + "grad_norm": 1.0365173816680908, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 24830 + }, + { + "epoch": 0.24831, + "grad_norm": 0.932898223400116, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 24831 + }, + { + "epoch": 0.24832, + "grad_norm": 0.9754995703697205, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 24832 + }, + { + "epoch": 0.24833, + "grad_norm": 0.9336578845977783, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 24833 + }, + { + "epoch": 0.24834, + "grad_norm": 0.907413899898529, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 24834 + }, + { + "epoch": 0.24835, + "grad_norm": 0.8196460604667664, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 24835 + }, + { + "epoch": 0.24836, + "grad_norm": 0.8799903988838196, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 24836 + }, + { + "epoch": 0.24837, + "grad_norm": 1.0383399724960327, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 24837 + }, + { + "epoch": 0.24838, + "grad_norm": 1.1219711303710938, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 24838 + }, + { + "epoch": 0.24839, + "grad_norm": 0.8553099036216736, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 24839 + }, + { + "epoch": 0.2484, + "grad_norm": 0.6803898811340332, + "learning_rate": 0.003, + "loss": 4.026, + "step": 24840 + }, + { + "epoch": 0.24841, + "grad_norm": 0.7031388282775879, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 24841 + }, + { + "epoch": 0.24842, + "grad_norm": 0.7037783265113831, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 24842 + }, + { + "epoch": 0.24843, + "grad_norm": 0.7645953893661499, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 24843 + }, + { + "epoch": 0.24844, + "grad_norm": 0.7612361311912537, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 24844 + }, + { + "epoch": 0.24845, + "grad_norm": 0.667593240737915, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 24845 + }, + { + "epoch": 0.24846, + "grad_norm": 0.5542976260185242, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 24846 + }, + { + "epoch": 0.24847, + "grad_norm": 0.8434585332870483, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 24847 + }, + { + "epoch": 0.24848, + "grad_norm": 1.0403292179107666, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 24848 + }, + { + "epoch": 0.24849, + "grad_norm": 1.172052264213562, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 24849 + }, + { + "epoch": 0.2485, + "grad_norm": 0.7472240328788757, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 24850 + }, + { + "epoch": 0.24851, + "grad_norm": 0.6084393858909607, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 24851 + }, + { + "epoch": 0.24852, + "grad_norm": 0.6105056405067444, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 24852 + }, + { + "epoch": 0.24853, + "grad_norm": 0.5973578691482544, + "learning_rate": 0.003, + "loss": 3.985, + "step": 24853 + }, + { + "epoch": 0.24854, + "grad_norm": 0.6493803262710571, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 24854 + }, + { + "epoch": 0.24855, + "grad_norm": 0.6333584189414978, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 24855 + }, + { + "epoch": 0.24856, + "grad_norm": 0.6767736077308655, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 24856 + }, + { + "epoch": 0.24857, + "grad_norm": 0.8103201985359192, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 24857 + }, + { + "epoch": 0.24858, + "grad_norm": 1.0185002088546753, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 24858 + }, + { + "epoch": 0.24859, + "grad_norm": 1.2535626888275146, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 24859 + }, + { + "epoch": 0.2486, + "grad_norm": 0.9250200986862183, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 24860 + }, + { + "epoch": 0.24861, + "grad_norm": 0.827885091304779, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 24861 + }, + { + "epoch": 0.24862, + "grad_norm": 0.7812620401382446, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 24862 + }, + { + "epoch": 0.24863, + "grad_norm": 0.7248989939689636, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 24863 + }, + { + "epoch": 0.24864, + "grad_norm": 0.6979221105575562, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 24864 + }, + { + "epoch": 0.24865, + "grad_norm": 0.7078675627708435, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 24865 + }, + { + "epoch": 0.24866, + "grad_norm": 0.7209116220474243, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 24866 + }, + { + "epoch": 0.24867, + "grad_norm": 0.6925181746482849, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 24867 + }, + { + "epoch": 0.24868, + "grad_norm": 0.6849539875984192, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 24868 + }, + { + "epoch": 0.24869, + "grad_norm": 0.8978288769721985, + "learning_rate": 0.003, + "loss": 3.972, + "step": 24869 + }, + { + "epoch": 0.2487, + "grad_norm": 1.0007386207580566, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 24870 + }, + { + "epoch": 0.24871, + "grad_norm": 1.1908923387527466, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 24871 + }, + { + "epoch": 0.24872, + "grad_norm": 0.7878948450088501, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 24872 + }, + { + "epoch": 0.24873, + "grad_norm": 0.7913317680358887, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 24873 + }, + { + "epoch": 0.24874, + "grad_norm": 0.8740472793579102, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 24874 + }, + { + "epoch": 0.24875, + "grad_norm": 0.9708170294761658, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 24875 + }, + { + "epoch": 0.24876, + "grad_norm": 1.0536351203918457, + "learning_rate": 0.003, + "loss": 4.001, + "step": 24876 + }, + { + "epoch": 0.24877, + "grad_norm": 1.1638703346252441, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 24877 + }, + { + "epoch": 0.24878, + "grad_norm": 1.1455599069595337, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 24878 + }, + { + "epoch": 0.24879, + "grad_norm": 0.8190259337425232, + "learning_rate": 0.003, + "loss": 4.004, + "step": 24879 + }, + { + "epoch": 0.2488, + "grad_norm": 0.894655704498291, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 24880 + }, + { + "epoch": 0.24881, + "grad_norm": 1.2307565212249756, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 24881 + }, + { + "epoch": 0.24882, + "grad_norm": 0.7912026643753052, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 24882 + }, + { + "epoch": 0.24883, + "grad_norm": 0.5952629446983337, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 24883 + }, + { + "epoch": 0.24884, + "grad_norm": 0.7497758865356445, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 24884 + }, + { + "epoch": 0.24885, + "grad_norm": 0.7522904276847839, + "learning_rate": 0.003, + "loss": 4.015, + "step": 24885 + }, + { + "epoch": 0.24886, + "grad_norm": 0.8197332620620728, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 24886 + }, + { + "epoch": 0.24887, + "grad_norm": 0.8685560822486877, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 24887 + }, + { + "epoch": 0.24888, + "grad_norm": 0.9106157422065735, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 24888 + }, + { + "epoch": 0.24889, + "grad_norm": 0.786583662033081, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 24889 + }, + { + "epoch": 0.2489, + "grad_norm": 0.7639341950416565, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 24890 + }, + { + "epoch": 0.24891, + "grad_norm": 0.8201864361763, + "learning_rate": 0.003, + "loss": 4.01, + "step": 24891 + }, + { + "epoch": 0.24892, + "grad_norm": 0.8059186339378357, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 24892 + }, + { + "epoch": 0.24893, + "grad_norm": 0.7123501896858215, + "learning_rate": 0.003, + "loss": 3.996, + "step": 24893 + }, + { + "epoch": 0.24894, + "grad_norm": 0.7638243436813354, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 24894 + }, + { + "epoch": 0.24895, + "grad_norm": 0.7964179515838623, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 24895 + }, + { + "epoch": 0.24896, + "grad_norm": 0.7235068678855896, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 24896 + }, + { + "epoch": 0.24897, + "grad_norm": 0.853526771068573, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 24897 + }, + { + "epoch": 0.24898, + "grad_norm": 1.2153451442718506, + "learning_rate": 0.003, + "loss": 4.008, + "step": 24898 + }, + { + "epoch": 0.24899, + "grad_norm": 0.9904343485832214, + "learning_rate": 0.003, + "loss": 3.982, + "step": 24899 + }, + { + "epoch": 0.249, + "grad_norm": 1.3399964570999146, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 24900 + }, + { + "epoch": 0.24901, + "grad_norm": 0.7212924361228943, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 24901 + }, + { + "epoch": 0.24902, + "grad_norm": 0.5745200514793396, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 24902 + }, + { + "epoch": 0.24903, + "grad_norm": 0.6382490992546082, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 24903 + }, + { + "epoch": 0.24904, + "grad_norm": 0.6826664805412292, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 24904 + }, + { + "epoch": 0.24905, + "grad_norm": 0.7203302383422852, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 24905 + }, + { + "epoch": 0.24906, + "grad_norm": 0.8856229782104492, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 24906 + }, + { + "epoch": 0.24907, + "grad_norm": 0.9316160082817078, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 24907 + }, + { + "epoch": 0.24908, + "grad_norm": 0.9750249981880188, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 24908 + }, + { + "epoch": 0.24909, + "grad_norm": 0.9319607019424438, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 24909 + }, + { + "epoch": 0.2491, + "grad_norm": 0.8210610151290894, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 24910 + }, + { + "epoch": 0.24911, + "grad_norm": 0.7790420651435852, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 24911 + }, + { + "epoch": 0.24912, + "grad_norm": 0.7690111398696899, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 24912 + }, + { + "epoch": 0.24913, + "grad_norm": 0.8793837428092957, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 24913 + }, + { + "epoch": 0.24914, + "grad_norm": 1.1589597463607788, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 24914 + }, + { + "epoch": 0.24915, + "grad_norm": 1.1635655164718628, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 24915 + }, + { + "epoch": 0.24916, + "grad_norm": 0.9549830555915833, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 24916 + }, + { + "epoch": 0.24917, + "grad_norm": 0.9169217944145203, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 24917 + }, + { + "epoch": 0.24918, + "grad_norm": 0.839543879032135, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 24918 + }, + { + "epoch": 0.24919, + "grad_norm": 0.8051316738128662, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 24919 + }, + { + "epoch": 0.2492, + "grad_norm": 0.7818151116371155, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 24920 + }, + { + "epoch": 0.24921, + "grad_norm": 0.8598986268043518, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 24921 + }, + { + "epoch": 0.24922, + "grad_norm": 1.0425068140029907, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 24922 + }, + { + "epoch": 0.24923, + "grad_norm": 0.9475182294845581, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 24923 + }, + { + "epoch": 0.24924, + "grad_norm": 1.0884451866149902, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 24924 + }, + { + "epoch": 0.24925, + "grad_norm": 0.9664582014083862, + "learning_rate": 0.003, + "loss": 4.001, + "step": 24925 + }, + { + "epoch": 0.24926, + "grad_norm": 0.8767949342727661, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 24926 + }, + { + "epoch": 0.24927, + "grad_norm": 0.9679237008094788, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 24927 + }, + { + "epoch": 0.24928, + "grad_norm": 1.2628779411315918, + "learning_rate": 0.003, + "loss": 4.017, + "step": 24928 + }, + { + "epoch": 0.24929, + "grad_norm": 1.1055541038513184, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 24929 + }, + { + "epoch": 0.2493, + "grad_norm": 0.783417820930481, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 24930 + }, + { + "epoch": 0.24931, + "grad_norm": 0.624814510345459, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 24931 + }, + { + "epoch": 0.24932, + "grad_norm": 0.5978679656982422, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 24932 + }, + { + "epoch": 0.24933, + "grad_norm": 0.6002329587936401, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 24933 + }, + { + "epoch": 0.24934, + "grad_norm": 0.6914220452308655, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 24934 + }, + { + "epoch": 0.24935, + "grad_norm": 0.6727545857429504, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 24935 + }, + { + "epoch": 0.24936, + "grad_norm": 0.659959077835083, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 24936 + }, + { + "epoch": 0.24937, + "grad_norm": 0.8038156032562256, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 24937 + }, + { + "epoch": 0.24938, + "grad_norm": 0.9119524955749512, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 24938 + }, + { + "epoch": 0.24939, + "grad_norm": 1.0471304655075073, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 24939 + }, + { + "epoch": 0.2494, + "grad_norm": 1.0342525243759155, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 24940 + }, + { + "epoch": 0.24941, + "grad_norm": 1.1073507070541382, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 24941 + }, + { + "epoch": 0.24942, + "grad_norm": 0.9822791218757629, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 24942 + }, + { + "epoch": 0.24943, + "grad_norm": 1.0233731269836426, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 24943 + }, + { + "epoch": 0.24944, + "grad_norm": 0.946275532245636, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 24944 + }, + { + "epoch": 0.24945, + "grad_norm": 0.8050229549407959, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 24945 + }, + { + "epoch": 0.24946, + "grad_norm": 0.7565122842788696, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 24946 + }, + { + "epoch": 0.24947, + "grad_norm": 0.7792460322380066, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 24947 + }, + { + "epoch": 0.24948, + "grad_norm": 0.7781099677085876, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 24948 + }, + { + "epoch": 0.24949, + "grad_norm": 0.644450843334198, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 24949 + }, + { + "epoch": 0.2495, + "grad_norm": 0.7191748023033142, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 24950 + }, + { + "epoch": 0.24951, + "grad_norm": 0.6668493747711182, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 24951 + }, + { + "epoch": 0.24952, + "grad_norm": 0.5238726735115051, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 24952 + }, + { + "epoch": 0.24953, + "grad_norm": 0.7102416157722473, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 24953 + }, + { + "epoch": 0.24954, + "grad_norm": 0.8396332263946533, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 24954 + }, + { + "epoch": 0.24955, + "grad_norm": 1.1628581285476685, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 24955 + }, + { + "epoch": 0.24956, + "grad_norm": 0.9007691144943237, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 24956 + }, + { + "epoch": 0.24957, + "grad_norm": 0.7600373029708862, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 24957 + }, + { + "epoch": 0.24958, + "grad_norm": 0.7599941492080688, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 24958 + }, + { + "epoch": 0.24959, + "grad_norm": 0.9637325406074524, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 24959 + }, + { + "epoch": 0.2496, + "grad_norm": 1.0238664150238037, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 24960 + }, + { + "epoch": 0.24961, + "grad_norm": 0.9374799728393555, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 24961 + }, + { + "epoch": 0.24962, + "grad_norm": 1.0772109031677246, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 24962 + }, + { + "epoch": 0.24963, + "grad_norm": 1.140130877494812, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 24963 + }, + { + "epoch": 0.24964, + "grad_norm": 0.716785192489624, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 24964 + }, + { + "epoch": 0.24965, + "grad_norm": 0.6555307507514954, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 24965 + }, + { + "epoch": 0.24966, + "grad_norm": 0.7031381130218506, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 24966 + }, + { + "epoch": 0.24967, + "grad_norm": 0.8908666372299194, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 24967 + }, + { + "epoch": 0.24968, + "grad_norm": 1.062209129333496, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 24968 + }, + { + "epoch": 0.24969, + "grad_norm": 0.9823837876319885, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 24969 + }, + { + "epoch": 0.2497, + "grad_norm": 1.0286146402359009, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 24970 + }, + { + "epoch": 0.24971, + "grad_norm": 0.9185134172439575, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 24971 + }, + { + "epoch": 0.24972, + "grad_norm": 0.9895694255828857, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 24972 + }, + { + "epoch": 0.24973, + "grad_norm": 0.8451065421104431, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 24973 + }, + { + "epoch": 0.24974, + "grad_norm": 0.9605506062507629, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 24974 + }, + { + "epoch": 0.24975, + "grad_norm": 1.2551090717315674, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 24975 + }, + { + "epoch": 0.24976, + "grad_norm": 1.0171700716018677, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 24976 + }, + { + "epoch": 0.24977, + "grad_norm": 0.9530161619186401, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 24977 + }, + { + "epoch": 0.24978, + "grad_norm": 0.7558655142784119, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 24978 + }, + { + "epoch": 0.24979, + "grad_norm": 0.7506139278411865, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 24979 + }, + { + "epoch": 0.2498, + "grad_norm": 0.7247182726860046, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 24980 + }, + { + "epoch": 0.24981, + "grad_norm": 0.8856322169303894, + "learning_rate": 0.003, + "loss": 3.997, + "step": 24981 + }, + { + "epoch": 0.24982, + "grad_norm": 0.997178316116333, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 24982 + }, + { + "epoch": 0.24983, + "grad_norm": 1.4340121746063232, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 24983 + }, + { + "epoch": 0.24984, + "grad_norm": 0.6305608749389648, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 24984 + }, + { + "epoch": 0.24985, + "grad_norm": 0.8140807747840881, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 24985 + }, + { + "epoch": 0.24986, + "grad_norm": 0.9802111983299255, + "learning_rate": 0.003, + "loss": 3.997, + "step": 24986 + }, + { + "epoch": 0.24987, + "grad_norm": 1.0682618618011475, + "learning_rate": 0.003, + "loss": 4.006, + "step": 24987 + }, + { + "epoch": 0.24988, + "grad_norm": 0.9249688386917114, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 24988 + }, + { + "epoch": 0.24989, + "grad_norm": 0.867851197719574, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 24989 + }, + { + "epoch": 0.2499, + "grad_norm": 0.8419514894485474, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 24990 + }, + { + "epoch": 0.24991, + "grad_norm": 0.9869087934494019, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 24991 + }, + { + "epoch": 0.24992, + "grad_norm": 1.353484869003296, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 24992 + }, + { + "epoch": 0.24993, + "grad_norm": 0.7531378269195557, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 24993 + }, + { + "epoch": 0.24994, + "grad_norm": 0.6890097856521606, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 24994 + }, + { + "epoch": 0.24995, + "grad_norm": 0.7659112215042114, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 24995 + }, + { + "epoch": 0.24996, + "grad_norm": 0.7793158292770386, + "learning_rate": 0.003, + "loss": 4.001, + "step": 24996 + }, + { + "epoch": 0.24997, + "grad_norm": 0.7181977033615112, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 24997 + }, + { + "epoch": 0.24998, + "grad_norm": 0.6397198438644409, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 24998 + }, + { + "epoch": 0.24999, + "grad_norm": 0.6205217242240906, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 24999 + }, + { + "epoch": 0.25, + "grad_norm": 0.5569931268692017, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 25000 + }, + { + "epoch": 0.25001, + "grad_norm": 0.6187779307365417, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 25001 + }, + { + "epoch": 0.25002, + "grad_norm": 0.5893850922584534, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 25002 + }, + { + "epoch": 0.25003, + "grad_norm": 0.6980180740356445, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 25003 + }, + { + "epoch": 0.25004, + "grad_norm": 0.783515453338623, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 25004 + }, + { + "epoch": 0.25005, + "grad_norm": 1.0048247575759888, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 25005 + }, + { + "epoch": 0.25006, + "grad_norm": 1.366137981414795, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 25006 + }, + { + "epoch": 0.25007, + "grad_norm": 0.5171303153038025, + "learning_rate": 0.003, + "loss": 3.953, + "step": 25007 + }, + { + "epoch": 0.25008, + "grad_norm": 0.8088678121566772, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 25008 + }, + { + "epoch": 0.25009, + "grad_norm": 1.180821418762207, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 25009 + }, + { + "epoch": 0.2501, + "grad_norm": 0.6644198894500732, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 25010 + }, + { + "epoch": 0.25011, + "grad_norm": 0.5600188970565796, + "learning_rate": 0.003, + "loss": 4.003, + "step": 25011 + }, + { + "epoch": 0.25012, + "grad_norm": 0.5553941130638123, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 25012 + }, + { + "epoch": 0.25013, + "grad_norm": 0.7056133151054382, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 25013 + }, + { + "epoch": 0.25014, + "grad_norm": 0.8655219674110413, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 25014 + }, + { + "epoch": 0.25015, + "grad_norm": 1.029463768005371, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 25015 + }, + { + "epoch": 0.25016, + "grad_norm": 1.0934683084487915, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 25016 + }, + { + "epoch": 0.25017, + "grad_norm": 0.9624761939048767, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 25017 + }, + { + "epoch": 0.25018, + "grad_norm": 0.8333072066307068, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 25018 + }, + { + "epoch": 0.25019, + "grad_norm": 0.7767633199691772, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 25019 + }, + { + "epoch": 0.2502, + "grad_norm": 0.7589020133018494, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 25020 + }, + { + "epoch": 0.25021, + "grad_norm": 0.9120230078697205, + "learning_rate": 0.003, + "loss": 4.015, + "step": 25021 + }, + { + "epoch": 0.25022, + "grad_norm": 1.0409542322158813, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 25022 + }, + { + "epoch": 0.25023, + "grad_norm": 0.8931517601013184, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 25023 + }, + { + "epoch": 0.25024, + "grad_norm": 0.9246020317077637, + "learning_rate": 0.003, + "loss": 3.985, + "step": 25024 + }, + { + "epoch": 0.25025, + "grad_norm": 1.0397624969482422, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 25025 + }, + { + "epoch": 0.25026, + "grad_norm": 1.110180139541626, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 25026 + }, + { + "epoch": 0.25027, + "grad_norm": 1.1033974885940552, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 25027 + }, + { + "epoch": 0.25028, + "grad_norm": 0.9354603886604309, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 25028 + }, + { + "epoch": 0.25029, + "grad_norm": 0.7728162407875061, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 25029 + }, + { + "epoch": 0.2503, + "grad_norm": 0.7347375750541687, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 25030 + }, + { + "epoch": 0.25031, + "grad_norm": 0.7553921341896057, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 25031 + }, + { + "epoch": 0.25032, + "grad_norm": 0.786914050579071, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 25032 + }, + { + "epoch": 0.25033, + "grad_norm": 0.7837997674942017, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 25033 + }, + { + "epoch": 0.25034, + "grad_norm": 0.9106943011283875, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 25034 + }, + { + "epoch": 0.25035, + "grad_norm": 0.9290919303894043, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 25035 + }, + { + "epoch": 0.25036, + "grad_norm": 0.7321922183036804, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 25036 + }, + { + "epoch": 0.25037, + "grad_norm": 0.7229407429695129, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 25037 + }, + { + "epoch": 0.25038, + "grad_norm": 0.849196195602417, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 25038 + }, + { + "epoch": 0.25039, + "grad_norm": 0.8887252807617188, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 25039 + }, + { + "epoch": 0.2504, + "grad_norm": 1.007759690284729, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 25040 + }, + { + "epoch": 0.25041, + "grad_norm": 1.2617206573486328, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 25041 + }, + { + "epoch": 0.25042, + "grad_norm": 0.8610366582870483, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 25042 + }, + { + "epoch": 0.25043, + "grad_norm": 0.8770789504051208, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 25043 + }, + { + "epoch": 0.25044, + "grad_norm": 0.8545438051223755, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 25044 + }, + { + "epoch": 0.25045, + "grad_norm": 0.7912514209747314, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 25045 + }, + { + "epoch": 0.25046, + "grad_norm": 0.6824645400047302, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 25046 + }, + { + "epoch": 0.25047, + "grad_norm": 0.7741122245788574, + "learning_rate": 0.003, + "loss": 4.006, + "step": 25047 + }, + { + "epoch": 0.25048, + "grad_norm": 0.7578386664390564, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 25048 + }, + { + "epoch": 0.25049, + "grad_norm": 0.7173295617103577, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 25049 + }, + { + "epoch": 0.2505, + "grad_norm": 0.7746570706367493, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 25050 + }, + { + "epoch": 0.25051, + "grad_norm": 0.7619470357894897, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 25051 + }, + { + "epoch": 0.25052, + "grad_norm": 0.828438401222229, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 25052 + }, + { + "epoch": 0.25053, + "grad_norm": 1.0055310726165771, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 25053 + }, + { + "epoch": 0.25054, + "grad_norm": 1.1854537725448608, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 25054 + }, + { + "epoch": 0.25055, + "grad_norm": 0.9107411503791809, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 25055 + }, + { + "epoch": 0.25056, + "grad_norm": 1.0010607242584229, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 25056 + }, + { + "epoch": 0.25057, + "grad_norm": 1.1279263496398926, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 25057 + }, + { + "epoch": 0.25058, + "grad_norm": 0.7654117941856384, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 25058 + }, + { + "epoch": 0.25059, + "grad_norm": 0.7461522817611694, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 25059 + }, + { + "epoch": 0.2506, + "grad_norm": 0.8360660672187805, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 25060 + }, + { + "epoch": 0.25061, + "grad_norm": 0.8594798445701599, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 25061 + }, + { + "epoch": 0.25062, + "grad_norm": 0.9523767232894897, + "learning_rate": 0.003, + "loss": 4.0, + "step": 25062 + }, + { + "epoch": 0.25063, + "grad_norm": 0.9855654239654541, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 25063 + }, + { + "epoch": 0.25064, + "grad_norm": 1.1024210453033447, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 25064 + }, + { + "epoch": 0.25065, + "grad_norm": 0.8705532550811768, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 25065 + }, + { + "epoch": 0.25066, + "grad_norm": 0.7434629797935486, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 25066 + }, + { + "epoch": 0.25067, + "grad_norm": 0.6619343757629395, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 25067 + }, + { + "epoch": 0.25068, + "grad_norm": 0.7158844470977783, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 25068 + }, + { + "epoch": 0.25069, + "grad_norm": 0.6880189180374146, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 25069 + }, + { + "epoch": 0.2507, + "grad_norm": 0.7689472436904907, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 25070 + }, + { + "epoch": 0.25071, + "grad_norm": 0.9669755101203918, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 25071 + }, + { + "epoch": 0.25072, + "grad_norm": 1.0685898065567017, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 25072 + }, + { + "epoch": 0.25073, + "grad_norm": 0.9331470727920532, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 25073 + }, + { + "epoch": 0.25074, + "grad_norm": 0.8514257073402405, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 25074 + }, + { + "epoch": 0.25075, + "grad_norm": 0.927262544631958, + "learning_rate": 0.003, + "loss": 3.958, + "step": 25075 + }, + { + "epoch": 0.25076, + "grad_norm": 0.9881807565689087, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 25076 + }, + { + "epoch": 0.25077, + "grad_norm": 1.0732923746109009, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 25077 + }, + { + "epoch": 0.25078, + "grad_norm": 0.8447161316871643, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 25078 + }, + { + "epoch": 0.25079, + "grad_norm": 0.8596442341804504, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 25079 + }, + { + "epoch": 0.2508, + "grad_norm": 0.8244372010231018, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 25080 + }, + { + "epoch": 0.25081, + "grad_norm": 0.738349437713623, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 25081 + }, + { + "epoch": 0.25082, + "grad_norm": 0.7469916939735413, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 25082 + }, + { + "epoch": 0.25083, + "grad_norm": 0.7897987961769104, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 25083 + }, + { + "epoch": 0.25084, + "grad_norm": 0.8438945412635803, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 25084 + }, + { + "epoch": 0.25085, + "grad_norm": 0.826300859451294, + "learning_rate": 0.003, + "loss": 3.994, + "step": 25085 + }, + { + "epoch": 0.25086, + "grad_norm": 0.8091835975646973, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 25086 + }, + { + "epoch": 0.25087, + "grad_norm": 0.7727853059768677, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 25087 + }, + { + "epoch": 0.25088, + "grad_norm": 0.8168857097625732, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 25088 + }, + { + "epoch": 0.25089, + "grad_norm": 0.9348554611206055, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 25089 + }, + { + "epoch": 0.2509, + "grad_norm": 0.8839529752731323, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 25090 + }, + { + "epoch": 0.25091, + "grad_norm": 0.9132164120674133, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 25091 + }, + { + "epoch": 0.25092, + "grad_norm": 1.0684548616409302, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 25092 + }, + { + "epoch": 0.25093, + "grad_norm": 0.8897774815559387, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 25093 + }, + { + "epoch": 0.25094, + "grad_norm": 0.960533082485199, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 25094 + }, + { + "epoch": 0.25095, + "grad_norm": 1.222136378288269, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 25095 + }, + { + "epoch": 0.25096, + "grad_norm": 0.7742778658866882, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 25096 + }, + { + "epoch": 0.25097, + "grad_norm": 0.8425857424736023, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 25097 + }, + { + "epoch": 0.25098, + "grad_norm": 0.8879294395446777, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 25098 + }, + { + "epoch": 0.25099, + "grad_norm": 1.0282862186431885, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 25099 + }, + { + "epoch": 0.251, + "grad_norm": 1.061036467552185, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 25100 + }, + { + "epoch": 0.25101, + "grad_norm": 1.1084964275360107, + "learning_rate": 0.003, + "loss": 4.004, + "step": 25101 + }, + { + "epoch": 0.25102, + "grad_norm": 0.9957682490348816, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 25102 + }, + { + "epoch": 0.25103, + "grad_norm": 0.9039728045463562, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 25103 + }, + { + "epoch": 0.25104, + "grad_norm": 0.7985256314277649, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 25104 + }, + { + "epoch": 0.25105, + "grad_norm": 0.9282469749450684, + "learning_rate": 0.003, + "loss": 4.016, + "step": 25105 + }, + { + "epoch": 0.25106, + "grad_norm": 1.0768885612487793, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 25106 + }, + { + "epoch": 0.25107, + "grad_norm": 0.8836683630943298, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 25107 + }, + { + "epoch": 0.25108, + "grad_norm": 0.771129310131073, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 25108 + }, + { + "epoch": 0.25109, + "grad_norm": 0.7034766674041748, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 25109 + }, + { + "epoch": 0.2511, + "grad_norm": 0.5936494469642639, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 25110 + }, + { + "epoch": 0.25111, + "grad_norm": 0.6851494908332825, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 25111 + }, + { + "epoch": 0.25112, + "grad_norm": 0.743125855922699, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 25112 + }, + { + "epoch": 0.25113, + "grad_norm": 0.7603784799575806, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 25113 + }, + { + "epoch": 0.25114, + "grad_norm": 0.8774493932723999, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 25114 + }, + { + "epoch": 0.25115, + "grad_norm": 0.9673449993133545, + "learning_rate": 0.003, + "loss": 3.969, + "step": 25115 + }, + { + "epoch": 0.25116, + "grad_norm": 1.1485332250595093, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 25116 + }, + { + "epoch": 0.25117, + "grad_norm": 0.8631411194801331, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 25117 + }, + { + "epoch": 0.25118, + "grad_norm": 0.676422119140625, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 25118 + }, + { + "epoch": 0.25119, + "grad_norm": 0.6931309103965759, + "learning_rate": 0.003, + "loss": 3.9417, + "step": 25119 + }, + { + "epoch": 0.2512, + "grad_norm": 0.7168868780136108, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 25120 + }, + { + "epoch": 0.25121, + "grad_norm": 0.7025383114814758, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 25121 + }, + { + "epoch": 0.25122, + "grad_norm": 0.7758429646492004, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 25122 + }, + { + "epoch": 0.25123, + "grad_norm": 0.9265406727790833, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 25123 + }, + { + "epoch": 0.25124, + "grad_norm": 1.111802101135254, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 25124 + }, + { + "epoch": 0.25125, + "grad_norm": 0.7745994925498962, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 25125 + }, + { + "epoch": 0.25126, + "grad_norm": 0.8004317283630371, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 25126 + }, + { + "epoch": 0.25127, + "grad_norm": 0.8511545062065125, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 25127 + }, + { + "epoch": 0.25128, + "grad_norm": 1.0500389337539673, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 25128 + }, + { + "epoch": 0.25129, + "grad_norm": 0.8517551422119141, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 25129 + }, + { + "epoch": 0.2513, + "grad_norm": 0.7472000122070312, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 25130 + }, + { + "epoch": 0.25131, + "grad_norm": 0.7896560430526733, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 25131 + }, + { + "epoch": 0.25132, + "grad_norm": 0.8965511322021484, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 25132 + }, + { + "epoch": 0.25133, + "grad_norm": 1.0698798894882202, + "learning_rate": 0.003, + "loss": 3.988, + "step": 25133 + }, + { + "epoch": 0.25134, + "grad_norm": 1.0101839303970337, + "learning_rate": 0.003, + "loss": 3.98, + "step": 25134 + }, + { + "epoch": 0.25135, + "grad_norm": 1.1580923795700073, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 25135 + }, + { + "epoch": 0.25136, + "grad_norm": 0.8121195435523987, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 25136 + }, + { + "epoch": 0.25137, + "grad_norm": 0.8070211410522461, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 25137 + }, + { + "epoch": 0.25138, + "grad_norm": 0.9668810963630676, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 25138 + }, + { + "epoch": 0.25139, + "grad_norm": 1.0144202709197998, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 25139 + }, + { + "epoch": 0.2514, + "grad_norm": 1.1003320217132568, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 25140 + }, + { + "epoch": 0.25141, + "grad_norm": 1.3073439598083496, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 25141 + }, + { + "epoch": 0.25142, + "grad_norm": 0.6527990102767944, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 25142 + }, + { + "epoch": 0.25143, + "grad_norm": 0.8274737000465393, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 25143 + }, + { + "epoch": 0.25144, + "grad_norm": 0.8745467066764832, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 25144 + }, + { + "epoch": 0.25145, + "grad_norm": 0.8899051547050476, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 25145 + }, + { + "epoch": 0.25146, + "grad_norm": 0.8374033570289612, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 25146 + }, + { + "epoch": 0.25147, + "grad_norm": 0.7839404940605164, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 25147 + }, + { + "epoch": 0.25148, + "grad_norm": 0.7930657267570496, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 25148 + }, + { + "epoch": 0.25149, + "grad_norm": 0.8661915063858032, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 25149 + }, + { + "epoch": 0.2515, + "grad_norm": 1.0485119819641113, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 25150 + }, + { + "epoch": 0.25151, + "grad_norm": 1.0781011581420898, + "learning_rate": 0.003, + "loss": 4.003, + "step": 25151 + }, + { + "epoch": 0.25152, + "grad_norm": 0.9348540902137756, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 25152 + }, + { + "epoch": 0.25153, + "grad_norm": 0.9212420582771301, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 25153 + }, + { + "epoch": 0.25154, + "grad_norm": 0.9350091218948364, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 25154 + }, + { + "epoch": 0.25155, + "grad_norm": 0.8026881814002991, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 25155 + }, + { + "epoch": 0.25156, + "grad_norm": 0.667753279209137, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 25156 + }, + { + "epoch": 0.25157, + "grad_norm": 0.626608669757843, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 25157 + }, + { + "epoch": 0.25158, + "grad_norm": 0.5969347953796387, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 25158 + }, + { + "epoch": 0.25159, + "grad_norm": 0.6101970672607422, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 25159 + }, + { + "epoch": 0.2516, + "grad_norm": 0.7858369946479797, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 25160 + }, + { + "epoch": 0.25161, + "grad_norm": 1.0118991136550903, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 25161 + }, + { + "epoch": 0.25162, + "grad_norm": 1.023619294166565, + "learning_rate": 0.003, + "loss": 3.983, + "step": 25162 + }, + { + "epoch": 0.25163, + "grad_norm": 0.8542295098304749, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 25163 + }, + { + "epoch": 0.25164, + "grad_norm": 0.7593318819999695, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 25164 + }, + { + "epoch": 0.25165, + "grad_norm": 0.8390832543373108, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 25165 + }, + { + "epoch": 0.25166, + "grad_norm": 1.1227067708969116, + "learning_rate": 0.003, + "loss": 3.997, + "step": 25166 + }, + { + "epoch": 0.25167, + "grad_norm": 0.9222919940948486, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 25167 + }, + { + "epoch": 0.25168, + "grad_norm": 1.0676531791687012, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 25168 + }, + { + "epoch": 0.25169, + "grad_norm": 1.408017635345459, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 25169 + }, + { + "epoch": 0.2517, + "grad_norm": 0.612867534160614, + "learning_rate": 0.003, + "loss": 3.99, + "step": 25170 + }, + { + "epoch": 0.25171, + "grad_norm": 0.8297423124313354, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 25171 + }, + { + "epoch": 0.25172, + "grad_norm": 1.0563338994979858, + "learning_rate": 0.003, + "loss": 3.995, + "step": 25172 + }, + { + "epoch": 0.25173, + "grad_norm": 0.9544999003410339, + "learning_rate": 0.003, + "loss": 3.9467, + "step": 25173 + }, + { + "epoch": 0.25174, + "grad_norm": 0.8660947680473328, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 25174 + }, + { + "epoch": 0.25175, + "grad_norm": 0.7947140336036682, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 25175 + }, + { + "epoch": 0.25176, + "grad_norm": 0.817126989364624, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 25176 + }, + { + "epoch": 0.25177, + "grad_norm": 0.8278509974479675, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 25177 + }, + { + "epoch": 0.25178, + "grad_norm": 0.9430961012840271, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 25178 + }, + { + "epoch": 0.25179, + "grad_norm": 1.007959008216858, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 25179 + }, + { + "epoch": 0.2518, + "grad_norm": 1.0358967781066895, + "learning_rate": 0.003, + "loss": 4.033, + "step": 25180 + }, + { + "epoch": 0.25181, + "grad_norm": 1.198190450668335, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 25181 + }, + { + "epoch": 0.25182, + "grad_norm": 0.948131799697876, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 25182 + }, + { + "epoch": 0.25183, + "grad_norm": 0.9359175562858582, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 25183 + }, + { + "epoch": 0.25184, + "grad_norm": 0.926392674446106, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 25184 + }, + { + "epoch": 0.25185, + "grad_norm": 0.9128230810165405, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 25185 + }, + { + "epoch": 0.25186, + "grad_norm": 1.043452262878418, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 25186 + }, + { + "epoch": 0.25187, + "grad_norm": 0.984515368938446, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 25187 + }, + { + "epoch": 0.25188, + "grad_norm": 0.8300612568855286, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 25188 + }, + { + "epoch": 0.25189, + "grad_norm": 0.7784908413887024, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 25189 + }, + { + "epoch": 0.2519, + "grad_norm": 0.6959231495857239, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 25190 + }, + { + "epoch": 0.25191, + "grad_norm": 0.7799877524375916, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 25191 + }, + { + "epoch": 0.25192, + "grad_norm": 0.8365493416786194, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 25192 + }, + { + "epoch": 0.25193, + "grad_norm": 0.9837875962257385, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 25193 + }, + { + "epoch": 0.25194, + "grad_norm": 1.1105976104736328, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 25194 + }, + { + "epoch": 0.25195, + "grad_norm": 0.8467714190483093, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 25195 + }, + { + "epoch": 0.25196, + "grad_norm": 0.7649572491645813, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 25196 + }, + { + "epoch": 0.25197, + "grad_norm": 0.725572943687439, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 25197 + }, + { + "epoch": 0.25198, + "grad_norm": 0.7367475628852844, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 25198 + }, + { + "epoch": 0.25199, + "grad_norm": 0.6560453176498413, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 25199 + }, + { + "epoch": 0.252, + "grad_norm": 0.5661042332649231, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 25200 + }, + { + "epoch": 0.25201, + "grad_norm": 0.5603251457214355, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 25201 + }, + { + "epoch": 0.25202, + "grad_norm": 0.5925635695457458, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 25202 + }, + { + "epoch": 0.25203, + "grad_norm": 0.6885330080986023, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 25203 + }, + { + "epoch": 0.25204, + "grad_norm": 0.7040902376174927, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 25204 + }, + { + "epoch": 0.25205, + "grad_norm": 0.8820661306381226, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 25205 + }, + { + "epoch": 0.25206, + "grad_norm": 1.1011674404144287, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 25206 + }, + { + "epoch": 0.25207, + "grad_norm": 0.8092846274375916, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 25207 + }, + { + "epoch": 0.25208, + "grad_norm": 0.7047312259674072, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 25208 + }, + { + "epoch": 0.25209, + "grad_norm": 0.750191330909729, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 25209 + }, + { + "epoch": 0.2521, + "grad_norm": 0.8552727103233337, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 25210 + }, + { + "epoch": 0.25211, + "grad_norm": 1.104125738143921, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 25211 + }, + { + "epoch": 0.25212, + "grad_norm": 1.0075645446777344, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 25212 + }, + { + "epoch": 0.25213, + "grad_norm": 1.1650869846343994, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 25213 + }, + { + "epoch": 0.25214, + "grad_norm": 0.6930332183837891, + "learning_rate": 0.003, + "loss": 3.963, + "step": 25214 + }, + { + "epoch": 0.25215, + "grad_norm": 0.6869567036628723, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 25215 + }, + { + "epoch": 0.25216, + "grad_norm": 0.8022205829620361, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 25216 + }, + { + "epoch": 0.25217, + "grad_norm": 0.8286408185958862, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 25217 + }, + { + "epoch": 0.25218, + "grad_norm": 0.8597841262817383, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 25218 + }, + { + "epoch": 0.25219, + "grad_norm": 0.8524219989776611, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 25219 + }, + { + "epoch": 0.2522, + "grad_norm": 1.0035943984985352, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 25220 + }, + { + "epoch": 0.25221, + "grad_norm": 1.0161306858062744, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 25221 + }, + { + "epoch": 0.25222, + "grad_norm": 0.830512523651123, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 25222 + }, + { + "epoch": 0.25223, + "grad_norm": 0.6771985292434692, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 25223 + }, + { + "epoch": 0.25224, + "grad_norm": 0.848779022693634, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 25224 + }, + { + "epoch": 0.25225, + "grad_norm": 1.0159986019134521, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 25225 + }, + { + "epoch": 0.25226, + "grad_norm": 1.1670485734939575, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 25226 + }, + { + "epoch": 0.25227, + "grad_norm": 0.9731599688529968, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 25227 + }, + { + "epoch": 0.25228, + "grad_norm": 0.8670048117637634, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 25228 + }, + { + "epoch": 0.25229, + "grad_norm": 0.9772502183914185, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 25229 + }, + { + "epoch": 0.2523, + "grad_norm": 1.1289979219436646, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 25230 + }, + { + "epoch": 0.25231, + "grad_norm": 0.9109286069869995, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 25231 + }, + { + "epoch": 0.25232, + "grad_norm": 0.8200125098228455, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 25232 + }, + { + "epoch": 0.25233, + "grad_norm": 0.9367732405662537, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 25233 + }, + { + "epoch": 0.25234, + "grad_norm": 0.9262071847915649, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 25234 + }, + { + "epoch": 0.25235, + "grad_norm": 0.7810326814651489, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 25235 + }, + { + "epoch": 0.25236, + "grad_norm": 0.7905666828155518, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 25236 + }, + { + "epoch": 0.25237, + "grad_norm": 0.9070234894752502, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 25237 + }, + { + "epoch": 0.25238, + "grad_norm": 1.0886428356170654, + "learning_rate": 0.003, + "loss": 3.976, + "step": 25238 + }, + { + "epoch": 0.25239, + "grad_norm": 1.0077956914901733, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 25239 + }, + { + "epoch": 0.2524, + "grad_norm": 0.8804937601089478, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 25240 + }, + { + "epoch": 0.25241, + "grad_norm": 0.6442847847938538, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 25241 + }, + { + "epoch": 0.25242, + "grad_norm": 0.6657389402389526, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 25242 + }, + { + "epoch": 0.25243, + "grad_norm": 0.653376579284668, + "learning_rate": 0.003, + "loss": 3.991, + "step": 25243 + }, + { + "epoch": 0.25244, + "grad_norm": 0.6389239430427551, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 25244 + }, + { + "epoch": 0.25245, + "grad_norm": 0.6903284788131714, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 25245 + }, + { + "epoch": 0.25246, + "grad_norm": 0.7405405044555664, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 25246 + }, + { + "epoch": 0.25247, + "grad_norm": 0.7096135020256042, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 25247 + }, + { + "epoch": 0.25248, + "grad_norm": 0.7235273122787476, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 25248 + }, + { + "epoch": 0.25249, + "grad_norm": 0.9115482568740845, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 25249 + }, + { + "epoch": 0.2525, + "grad_norm": 1.2157124280929565, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 25250 + }, + { + "epoch": 0.25251, + "grad_norm": 0.872018575668335, + "learning_rate": 0.003, + "loss": 4.006, + "step": 25251 + }, + { + "epoch": 0.25252, + "grad_norm": 0.8405318856239319, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 25252 + }, + { + "epoch": 0.25253, + "grad_norm": 0.8718681335449219, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 25253 + }, + { + "epoch": 0.25254, + "grad_norm": 1.16887366771698, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 25254 + }, + { + "epoch": 0.25255, + "grad_norm": 0.9562259912490845, + "learning_rate": 0.003, + "loss": 4.004, + "step": 25255 + }, + { + "epoch": 0.25256, + "grad_norm": 0.8612145781517029, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 25256 + }, + { + "epoch": 0.25257, + "grad_norm": 0.8411399722099304, + "learning_rate": 0.003, + "loss": 3.998, + "step": 25257 + }, + { + "epoch": 0.25258, + "grad_norm": 0.833142876625061, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 25258 + }, + { + "epoch": 0.25259, + "grad_norm": 0.9270251989364624, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 25259 + }, + { + "epoch": 0.2526, + "grad_norm": 0.9544125199317932, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 25260 + }, + { + "epoch": 0.25261, + "grad_norm": 0.7873496413230896, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 25261 + }, + { + "epoch": 0.25262, + "grad_norm": 0.7537788152694702, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 25262 + }, + { + "epoch": 0.25263, + "grad_norm": 0.7527837157249451, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 25263 + }, + { + "epoch": 0.25264, + "grad_norm": 0.7070215940475464, + "learning_rate": 0.003, + "loss": 3.998, + "step": 25264 + }, + { + "epoch": 0.25265, + "grad_norm": 0.6420490145683289, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 25265 + }, + { + "epoch": 0.25266, + "grad_norm": 0.7169657945632935, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 25266 + }, + { + "epoch": 0.25267, + "grad_norm": 0.8521289229393005, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 25267 + }, + { + "epoch": 0.25268, + "grad_norm": 0.9185377359390259, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 25268 + }, + { + "epoch": 0.25269, + "grad_norm": 1.1827312707901, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 25269 + }, + { + "epoch": 0.2527, + "grad_norm": 1.049242377281189, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 25270 + }, + { + "epoch": 0.25271, + "grad_norm": 0.8898597359657288, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 25271 + }, + { + "epoch": 0.25272, + "grad_norm": 0.7429059743881226, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 25272 + }, + { + "epoch": 0.25273, + "grad_norm": 0.5976107120513916, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 25273 + }, + { + "epoch": 0.25274, + "grad_norm": 0.6746332049369812, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 25274 + }, + { + "epoch": 0.25275, + "grad_norm": 0.8419301509857178, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 25275 + }, + { + "epoch": 0.25276, + "grad_norm": 0.8890442252159119, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 25276 + }, + { + "epoch": 0.25277, + "grad_norm": 0.9122182130813599, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 25277 + }, + { + "epoch": 0.25278, + "grad_norm": 0.8056383728981018, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 25278 + }, + { + "epoch": 0.25279, + "grad_norm": 0.7088527083396912, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 25279 + }, + { + "epoch": 0.2528, + "grad_norm": 0.6857345700263977, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 25280 + }, + { + "epoch": 0.25281, + "grad_norm": 0.7871561050415039, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 25281 + }, + { + "epoch": 0.25282, + "grad_norm": 0.9268015623092651, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 25282 + }, + { + "epoch": 0.25283, + "grad_norm": 0.9879458546638489, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 25283 + }, + { + "epoch": 0.25284, + "grad_norm": 0.9000425934791565, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 25284 + }, + { + "epoch": 0.25285, + "grad_norm": 0.7387357950210571, + "learning_rate": 0.003, + "loss": 4.004, + "step": 25285 + }, + { + "epoch": 0.25286, + "grad_norm": 0.791253924369812, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 25286 + }, + { + "epoch": 0.25287, + "grad_norm": 0.9312100410461426, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 25287 + }, + { + "epoch": 0.25288, + "grad_norm": 0.8851292133331299, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 25288 + }, + { + "epoch": 0.25289, + "grad_norm": 0.9118655323982239, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 25289 + }, + { + "epoch": 0.2529, + "grad_norm": 0.9758418202400208, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 25290 + }, + { + "epoch": 0.25291, + "grad_norm": 1.0434300899505615, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 25291 + }, + { + "epoch": 0.25292, + "grad_norm": 0.9237330555915833, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 25292 + }, + { + "epoch": 0.25293, + "grad_norm": 0.929424524307251, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 25293 + }, + { + "epoch": 0.25294, + "grad_norm": 0.9513487219810486, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 25294 + }, + { + "epoch": 0.25295, + "grad_norm": 1.0712589025497437, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 25295 + }, + { + "epoch": 0.25296, + "grad_norm": 1.0989584922790527, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 25296 + }, + { + "epoch": 0.25297, + "grad_norm": 0.9379088282585144, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 25297 + }, + { + "epoch": 0.25298, + "grad_norm": 0.8891896605491638, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 25298 + }, + { + "epoch": 0.25299, + "grad_norm": 0.806808352470398, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 25299 + }, + { + "epoch": 0.253, + "grad_norm": 0.9123344421386719, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 25300 + }, + { + "epoch": 0.25301, + "grad_norm": 1.260708212852478, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 25301 + }, + { + "epoch": 0.25302, + "grad_norm": 1.0133821964263916, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 25302 + }, + { + "epoch": 0.25303, + "grad_norm": 0.8792011737823486, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 25303 + }, + { + "epoch": 0.25304, + "grad_norm": 0.7952749133110046, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 25304 + }, + { + "epoch": 0.25305, + "grad_norm": 0.8979367613792419, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 25305 + }, + { + "epoch": 0.25306, + "grad_norm": 1.0293011665344238, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 25306 + }, + { + "epoch": 0.25307, + "grad_norm": 1.1463531255722046, + "learning_rate": 0.003, + "loss": 3.983, + "step": 25307 + }, + { + "epoch": 0.25308, + "grad_norm": 0.9921964406967163, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 25308 + }, + { + "epoch": 0.25309, + "grad_norm": 0.8270585536956787, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 25309 + }, + { + "epoch": 0.2531, + "grad_norm": 0.7510336637496948, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 25310 + }, + { + "epoch": 0.25311, + "grad_norm": 0.7684364318847656, + "learning_rate": 0.003, + "loss": 4.009, + "step": 25311 + }, + { + "epoch": 0.25312, + "grad_norm": 0.7563621401786804, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 25312 + }, + { + "epoch": 0.25313, + "grad_norm": 0.6822839379310608, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 25313 + }, + { + "epoch": 0.25314, + "grad_norm": 0.6856615543365479, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 25314 + }, + { + "epoch": 0.25315, + "grad_norm": 0.7082751989364624, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 25315 + }, + { + "epoch": 0.25316, + "grad_norm": 0.871953010559082, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 25316 + }, + { + "epoch": 0.25317, + "grad_norm": 1.2558739185333252, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 25317 + }, + { + "epoch": 0.25318, + "grad_norm": 0.8089984655380249, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 25318 + }, + { + "epoch": 0.25319, + "grad_norm": 0.621935248374939, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 25319 + }, + { + "epoch": 0.2532, + "grad_norm": 0.6283798217773438, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 25320 + }, + { + "epoch": 0.25321, + "grad_norm": 0.7331159710884094, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 25321 + }, + { + "epoch": 0.25322, + "grad_norm": 0.7436754703521729, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 25322 + }, + { + "epoch": 0.25323, + "grad_norm": 0.8280391693115234, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 25323 + }, + { + "epoch": 0.25324, + "grad_norm": 0.9346519708633423, + "learning_rate": 0.003, + "loss": 4.003, + "step": 25324 + }, + { + "epoch": 0.25325, + "grad_norm": 0.8750914335250854, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 25325 + }, + { + "epoch": 0.25326, + "grad_norm": 0.7104587554931641, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 25326 + }, + { + "epoch": 0.25327, + "grad_norm": 0.70856112241745, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 25327 + }, + { + "epoch": 0.25328, + "grad_norm": 0.7070035338401794, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 25328 + }, + { + "epoch": 0.25329, + "grad_norm": 0.7218126654624939, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 25329 + }, + { + "epoch": 0.2533, + "grad_norm": 0.8824271559715271, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 25330 + }, + { + "epoch": 0.25331, + "grad_norm": 1.0440199375152588, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 25331 + }, + { + "epoch": 0.25332, + "grad_norm": 1.0068455934524536, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 25332 + }, + { + "epoch": 0.25333, + "grad_norm": 1.0302997827529907, + "learning_rate": 0.003, + "loss": 4.023, + "step": 25333 + }, + { + "epoch": 0.25334, + "grad_norm": 0.9188385009765625, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 25334 + }, + { + "epoch": 0.25335, + "grad_norm": 1.070172667503357, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 25335 + }, + { + "epoch": 0.25336, + "grad_norm": 0.9234619736671448, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 25336 + }, + { + "epoch": 0.25337, + "grad_norm": 0.8606792092323303, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 25337 + }, + { + "epoch": 0.25338, + "grad_norm": 0.8394186496734619, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 25338 + }, + { + "epoch": 0.25339, + "grad_norm": 0.7884788513183594, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 25339 + }, + { + "epoch": 0.2534, + "grad_norm": 1.124137282371521, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 25340 + }, + { + "epoch": 0.25341, + "grad_norm": 1.1272714138031006, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 25341 + }, + { + "epoch": 0.25342, + "grad_norm": 0.8192625045776367, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 25342 + }, + { + "epoch": 0.25343, + "grad_norm": 0.7474991083145142, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 25343 + }, + { + "epoch": 0.25344, + "grad_norm": 0.8224437832832336, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 25344 + }, + { + "epoch": 0.25345, + "grad_norm": 0.8154651522636414, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 25345 + }, + { + "epoch": 0.25346, + "grad_norm": 0.75994873046875, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 25346 + }, + { + "epoch": 0.25347, + "grad_norm": 0.8152437210083008, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 25347 + }, + { + "epoch": 0.25348, + "grad_norm": 0.7575472593307495, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 25348 + }, + { + "epoch": 0.25349, + "grad_norm": 0.6524654030799866, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 25349 + }, + { + "epoch": 0.2535, + "grad_norm": 0.6341437101364136, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 25350 + }, + { + "epoch": 0.25351, + "grad_norm": 0.6234481334686279, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 25351 + }, + { + "epoch": 0.25352, + "grad_norm": 0.7691951990127563, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 25352 + }, + { + "epoch": 0.25353, + "grad_norm": 0.9696035981178284, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 25353 + }, + { + "epoch": 0.25354, + "grad_norm": 0.9889363646507263, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 25354 + }, + { + "epoch": 0.25355, + "grad_norm": 0.9389688372612, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 25355 + }, + { + "epoch": 0.25356, + "grad_norm": 0.8848605155944824, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 25356 + }, + { + "epoch": 0.25357, + "grad_norm": 0.9570282697677612, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 25357 + }, + { + "epoch": 0.25358, + "grad_norm": 0.8902246952056885, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 25358 + }, + { + "epoch": 0.25359, + "grad_norm": 0.9999766945838928, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 25359 + }, + { + "epoch": 0.2536, + "grad_norm": 1.2634363174438477, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 25360 + }, + { + "epoch": 0.25361, + "grad_norm": 0.7996293306350708, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 25361 + }, + { + "epoch": 0.25362, + "grad_norm": 0.6470975875854492, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 25362 + }, + { + "epoch": 0.25363, + "grad_norm": 0.5889624953269958, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 25363 + }, + { + "epoch": 0.25364, + "grad_norm": 0.544593870639801, + "learning_rate": 0.003, + "loss": 3.996, + "step": 25364 + }, + { + "epoch": 0.25365, + "grad_norm": 0.5370072722434998, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 25365 + }, + { + "epoch": 0.25366, + "grad_norm": 0.5163022875785828, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 25366 + }, + { + "epoch": 0.25367, + "grad_norm": 0.7484766840934753, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 25367 + }, + { + "epoch": 0.25368, + "grad_norm": 1.1880559921264648, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 25368 + }, + { + "epoch": 0.25369, + "grad_norm": 1.0466206073760986, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 25369 + }, + { + "epoch": 0.2537, + "grad_norm": 1.0875577926635742, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 25370 + }, + { + "epoch": 0.25371, + "grad_norm": 0.8084703087806702, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 25371 + }, + { + "epoch": 0.25372, + "grad_norm": 0.6412041783332825, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 25372 + }, + { + "epoch": 0.25373, + "grad_norm": 0.7521411776542664, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 25373 + }, + { + "epoch": 0.25374, + "grad_norm": 1.028550386428833, + "learning_rate": 0.003, + "loss": 3.986, + "step": 25374 + }, + { + "epoch": 0.25375, + "grad_norm": 1.4204189777374268, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 25375 + }, + { + "epoch": 0.25376, + "grad_norm": 0.6238331198692322, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 25376 + }, + { + "epoch": 0.25377, + "grad_norm": 0.8386913537979126, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 25377 + }, + { + "epoch": 0.25378, + "grad_norm": 0.9614189267158508, + "learning_rate": 0.003, + "loss": 3.957, + "step": 25378 + }, + { + "epoch": 0.25379, + "grad_norm": 0.88890141248703, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 25379 + }, + { + "epoch": 0.2538, + "grad_norm": 0.846735954284668, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 25380 + }, + { + "epoch": 0.25381, + "grad_norm": 0.82142573595047, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 25381 + }, + { + "epoch": 0.25382, + "grad_norm": 0.9290993213653564, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 25382 + }, + { + "epoch": 0.25383, + "grad_norm": 0.7896751165390015, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 25383 + }, + { + "epoch": 0.25384, + "grad_norm": 0.7335056066513062, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 25384 + }, + { + "epoch": 0.25385, + "grad_norm": 0.7370272874832153, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 25385 + }, + { + "epoch": 0.25386, + "grad_norm": 0.7688499093055725, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 25386 + }, + { + "epoch": 0.25387, + "grad_norm": 0.7585076689720154, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 25387 + }, + { + "epoch": 0.25388, + "grad_norm": 0.7362576723098755, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 25388 + }, + { + "epoch": 0.25389, + "grad_norm": 0.8627962470054626, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 25389 + }, + { + "epoch": 0.2539, + "grad_norm": 0.9398807883262634, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 25390 + }, + { + "epoch": 0.25391, + "grad_norm": 1.1245635747909546, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 25391 + }, + { + "epoch": 0.25392, + "grad_norm": 0.982549250125885, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 25392 + }, + { + "epoch": 0.25393, + "grad_norm": 1.0276422500610352, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 25393 + }, + { + "epoch": 0.25394, + "grad_norm": 1.3761544227600098, + "learning_rate": 0.003, + "loss": 4.063, + "step": 25394 + }, + { + "epoch": 0.25395, + "grad_norm": 1.0812209844589233, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 25395 + }, + { + "epoch": 0.25396, + "grad_norm": 0.8826557397842407, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 25396 + }, + { + "epoch": 0.25397, + "grad_norm": 0.9402980804443359, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 25397 + }, + { + "epoch": 0.25398, + "grad_norm": 1.145925760269165, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 25398 + }, + { + "epoch": 0.25399, + "grad_norm": 1.0287882089614868, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 25399 + }, + { + "epoch": 0.254, + "grad_norm": 0.8527262806892395, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 25400 + }, + { + "epoch": 0.25401, + "grad_norm": 1.0013691186904907, + "learning_rate": 0.003, + "loss": 3.99, + "step": 25401 + }, + { + "epoch": 0.25402, + "grad_norm": 1.2570923566818237, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 25402 + }, + { + "epoch": 0.25403, + "grad_norm": 0.8947070240974426, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 25403 + }, + { + "epoch": 0.25404, + "grad_norm": 0.8004720211029053, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 25404 + }, + { + "epoch": 0.25405, + "grad_norm": 0.7401588559150696, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 25405 + }, + { + "epoch": 0.25406, + "grad_norm": 0.8396176099777222, + "learning_rate": 0.003, + "loss": 3.977, + "step": 25406 + }, + { + "epoch": 0.25407, + "grad_norm": 0.8619393706321716, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 25407 + }, + { + "epoch": 0.25408, + "grad_norm": 0.858450710773468, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 25408 + }, + { + "epoch": 0.25409, + "grad_norm": 0.9479227066040039, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 25409 + }, + { + "epoch": 0.2541, + "grad_norm": 1.069057583808899, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 25410 + }, + { + "epoch": 0.25411, + "grad_norm": 0.8253480792045593, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 25411 + }, + { + "epoch": 0.25412, + "grad_norm": 0.8592541813850403, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 25412 + }, + { + "epoch": 0.25413, + "grad_norm": 0.9435031414031982, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 25413 + }, + { + "epoch": 0.25414, + "grad_norm": 0.9143722653388977, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 25414 + }, + { + "epoch": 0.25415, + "grad_norm": 0.6358144879341125, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 25415 + }, + { + "epoch": 0.25416, + "grad_norm": 0.48888853192329407, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 25416 + }, + { + "epoch": 0.25417, + "grad_norm": 0.4963091313838959, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 25417 + }, + { + "epoch": 0.25418, + "grad_norm": 0.58965003490448, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 25418 + }, + { + "epoch": 0.25419, + "grad_norm": 0.6037325859069824, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 25419 + }, + { + "epoch": 0.2542, + "grad_norm": 0.659218430519104, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 25420 + }, + { + "epoch": 0.25421, + "grad_norm": 0.5973354578018188, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 25421 + }, + { + "epoch": 0.25422, + "grad_norm": 0.5672264099121094, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 25422 + }, + { + "epoch": 0.25423, + "grad_norm": 0.6756817102432251, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 25423 + }, + { + "epoch": 0.25424, + "grad_norm": 0.8520943522453308, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 25424 + }, + { + "epoch": 0.25425, + "grad_norm": 1.0058972835540771, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 25425 + }, + { + "epoch": 0.25426, + "grad_norm": 1.029213547706604, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 25426 + }, + { + "epoch": 0.25427, + "grad_norm": 0.8365218043327332, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 25427 + }, + { + "epoch": 0.25428, + "grad_norm": 0.7784957885742188, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 25428 + }, + { + "epoch": 0.25429, + "grad_norm": 0.8872884511947632, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 25429 + }, + { + "epoch": 0.2543, + "grad_norm": 0.8973841071128845, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 25430 + }, + { + "epoch": 0.25431, + "grad_norm": 0.9257317185401917, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 25431 + }, + { + "epoch": 0.25432, + "grad_norm": 0.9848815202713013, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 25432 + }, + { + "epoch": 0.25433, + "grad_norm": 0.9644402861595154, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 25433 + }, + { + "epoch": 0.25434, + "grad_norm": 1.0608144998550415, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 25434 + }, + { + "epoch": 0.25435, + "grad_norm": 1.141629934310913, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 25435 + }, + { + "epoch": 0.25436, + "grad_norm": 0.837598979473114, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 25436 + }, + { + "epoch": 0.25437, + "grad_norm": 0.7553626298904419, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 25437 + }, + { + "epoch": 0.25438, + "grad_norm": 0.6472252607345581, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 25438 + }, + { + "epoch": 0.25439, + "grad_norm": 0.7367480993270874, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 25439 + }, + { + "epoch": 0.2544, + "grad_norm": 0.8513838052749634, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 25440 + }, + { + "epoch": 0.25441, + "grad_norm": 1.142270803451538, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 25441 + }, + { + "epoch": 0.25442, + "grad_norm": 1.1129038333892822, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 25442 + }, + { + "epoch": 0.25443, + "grad_norm": 1.1086269617080688, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 25443 + }, + { + "epoch": 0.25444, + "grad_norm": 0.8637166619300842, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 25444 + }, + { + "epoch": 0.25445, + "grad_norm": 0.6939029693603516, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 25445 + }, + { + "epoch": 0.25446, + "grad_norm": 0.7775781154632568, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 25446 + }, + { + "epoch": 0.25447, + "grad_norm": 0.7565057873725891, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 25447 + }, + { + "epoch": 0.25448, + "grad_norm": 0.7247198224067688, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 25448 + }, + { + "epoch": 0.25449, + "grad_norm": 0.7751010656356812, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 25449 + }, + { + "epoch": 0.2545, + "grad_norm": 0.9341700673103333, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 25450 + }, + { + "epoch": 0.25451, + "grad_norm": 0.8940555453300476, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 25451 + }, + { + "epoch": 0.25452, + "grad_norm": 1.0609402656555176, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 25452 + }, + { + "epoch": 0.25453, + "grad_norm": 1.1629769802093506, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 25453 + }, + { + "epoch": 0.25454, + "grad_norm": 0.6977956295013428, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 25454 + }, + { + "epoch": 0.25455, + "grad_norm": 0.6019397377967834, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 25455 + }, + { + "epoch": 0.25456, + "grad_norm": 0.6770892143249512, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 25456 + }, + { + "epoch": 0.25457, + "grad_norm": 0.7285329103469849, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 25457 + }, + { + "epoch": 0.25458, + "grad_norm": 0.6561967134475708, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 25458 + }, + { + "epoch": 0.25459, + "grad_norm": 0.7178798317909241, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 25459 + }, + { + "epoch": 0.2546, + "grad_norm": 1.0086385011672974, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 25460 + }, + { + "epoch": 0.25461, + "grad_norm": 1.3030850887298584, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 25461 + }, + { + "epoch": 0.25462, + "grad_norm": 0.7980507612228394, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 25462 + }, + { + "epoch": 0.25463, + "grad_norm": 0.7207521796226501, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 25463 + }, + { + "epoch": 0.25464, + "grad_norm": 0.8761317729949951, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 25464 + }, + { + "epoch": 0.25465, + "grad_norm": 0.9619604349136353, + "learning_rate": 0.003, + "loss": 4.004, + "step": 25465 + }, + { + "epoch": 0.25466, + "grad_norm": 1.0856311321258545, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 25466 + }, + { + "epoch": 0.25467, + "grad_norm": 1.215704321861267, + "learning_rate": 0.003, + "loss": 3.99, + "step": 25467 + }, + { + "epoch": 0.25468, + "grad_norm": 0.7138857841491699, + "learning_rate": 0.003, + "loss": 3.978, + "step": 25468 + }, + { + "epoch": 0.25469, + "grad_norm": 0.7002999186515808, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 25469 + }, + { + "epoch": 0.2547, + "grad_norm": 0.808220624923706, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 25470 + }, + { + "epoch": 0.25471, + "grad_norm": 0.7727131247520447, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 25471 + }, + { + "epoch": 0.25472, + "grad_norm": 1.0147082805633545, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 25472 + }, + { + "epoch": 0.25473, + "grad_norm": 1.1899192333221436, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 25473 + }, + { + "epoch": 0.25474, + "grad_norm": 0.6791627407073975, + "learning_rate": 0.003, + "loss": 3.99, + "step": 25474 + }, + { + "epoch": 0.25475, + "grad_norm": 0.8122314214706421, + "learning_rate": 0.003, + "loss": 4.004, + "step": 25475 + }, + { + "epoch": 0.25476, + "grad_norm": 0.8469472527503967, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 25476 + }, + { + "epoch": 0.25477, + "grad_norm": 0.9387763738632202, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 25477 + }, + { + "epoch": 0.25478, + "grad_norm": 1.1349236965179443, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 25478 + }, + { + "epoch": 0.25479, + "grad_norm": 0.9780484437942505, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 25479 + }, + { + "epoch": 0.2548, + "grad_norm": 1.2319815158843994, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 25480 + }, + { + "epoch": 0.25481, + "grad_norm": 0.6850935220718384, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 25481 + }, + { + "epoch": 0.25482, + "grad_norm": 0.8410962820053101, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 25482 + }, + { + "epoch": 0.25483, + "grad_norm": 1.1613019704818726, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 25483 + }, + { + "epoch": 0.25484, + "grad_norm": 1.2396098375320435, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 25484 + }, + { + "epoch": 0.25485, + "grad_norm": 0.7179574370384216, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 25485 + }, + { + "epoch": 0.25486, + "grad_norm": 0.6861761808395386, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 25486 + }, + { + "epoch": 0.25487, + "grad_norm": 0.6550556421279907, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 25487 + }, + { + "epoch": 0.25488, + "grad_norm": 0.6810204982757568, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 25488 + }, + { + "epoch": 0.25489, + "grad_norm": 0.7210050821304321, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 25489 + }, + { + "epoch": 0.2549, + "grad_norm": 0.7716299891471863, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 25490 + }, + { + "epoch": 0.25491, + "grad_norm": 0.8009316325187683, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 25491 + }, + { + "epoch": 0.25492, + "grad_norm": 0.8196008205413818, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 25492 + }, + { + "epoch": 0.25493, + "grad_norm": 0.9082884192466736, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 25493 + }, + { + "epoch": 0.25494, + "grad_norm": 0.693109393119812, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 25494 + }, + { + "epoch": 0.25495, + "grad_norm": 0.6025292873382568, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 25495 + }, + { + "epoch": 0.25496, + "grad_norm": 0.6472448706626892, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 25496 + }, + { + "epoch": 0.25497, + "grad_norm": 0.6174609661102295, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 25497 + }, + { + "epoch": 0.25498, + "grad_norm": 0.6097441911697388, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 25498 + }, + { + "epoch": 0.25499, + "grad_norm": 0.7000227570533752, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 25499 + }, + { + "epoch": 0.255, + "grad_norm": 0.7814828753471375, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 25500 + }, + { + "epoch": 0.25501, + "grad_norm": 0.8005951642990112, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 25501 + }, + { + "epoch": 0.25502, + "grad_norm": 0.7958718538284302, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 25502 + }, + { + "epoch": 0.25503, + "grad_norm": 0.9614733457565308, + "learning_rate": 0.003, + "loss": 4.003, + "step": 25503 + }, + { + "epoch": 0.25504, + "grad_norm": 1.1178629398345947, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 25504 + }, + { + "epoch": 0.25505, + "grad_norm": 0.9258359670639038, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 25505 + }, + { + "epoch": 0.25506, + "grad_norm": 1.0665048360824585, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 25506 + }, + { + "epoch": 0.25507, + "grad_norm": 1.0402182340621948, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 25507 + }, + { + "epoch": 0.25508, + "grad_norm": 0.863735020160675, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 25508 + }, + { + "epoch": 0.25509, + "grad_norm": 0.8930239677429199, + "learning_rate": 0.003, + "loss": 3.981, + "step": 25509 + }, + { + "epoch": 0.2551, + "grad_norm": 0.9425313472747803, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 25510 + }, + { + "epoch": 0.25511, + "grad_norm": 1.0143622159957886, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 25511 + }, + { + "epoch": 0.25512, + "grad_norm": 0.9015200138092041, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 25512 + }, + { + "epoch": 0.25513, + "grad_norm": 0.8790751695632935, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 25513 + }, + { + "epoch": 0.25514, + "grad_norm": 1.068734884262085, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 25514 + }, + { + "epoch": 0.25515, + "grad_norm": 0.989291787147522, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 25515 + }, + { + "epoch": 0.25516, + "grad_norm": 1.0783238410949707, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 25516 + }, + { + "epoch": 0.25517, + "grad_norm": 1.0260382890701294, + "learning_rate": 0.003, + "loss": 4.009, + "step": 25517 + }, + { + "epoch": 0.25518, + "grad_norm": 0.9706366658210754, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 25518 + }, + { + "epoch": 0.25519, + "grad_norm": 1.2016223669052124, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 25519 + }, + { + "epoch": 0.2552, + "grad_norm": 0.8670606017112732, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 25520 + }, + { + "epoch": 0.25521, + "grad_norm": 0.8759750723838806, + "learning_rate": 0.003, + "loss": 4.01, + "step": 25521 + }, + { + "epoch": 0.25522, + "grad_norm": 1.0426381826400757, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 25522 + }, + { + "epoch": 0.25523, + "grad_norm": 0.984062671661377, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 25523 + }, + { + "epoch": 0.25524, + "grad_norm": 0.9192453622817993, + "learning_rate": 0.003, + "loss": 3.992, + "step": 25524 + }, + { + "epoch": 0.25525, + "grad_norm": 1.0823553800582886, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 25525 + }, + { + "epoch": 0.25526, + "grad_norm": 1.2047532796859741, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 25526 + }, + { + "epoch": 0.25527, + "grad_norm": 0.7943227291107178, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 25527 + }, + { + "epoch": 0.25528, + "grad_norm": 0.7982091903686523, + "learning_rate": 0.003, + "loss": 3.984, + "step": 25528 + }, + { + "epoch": 0.25529, + "grad_norm": 0.7377675771713257, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 25529 + }, + { + "epoch": 0.2553, + "grad_norm": 0.6916780471801758, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 25530 + }, + { + "epoch": 0.25531, + "grad_norm": 0.6055135130882263, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 25531 + }, + { + "epoch": 0.25532, + "grad_norm": 0.6628902554512024, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 25532 + }, + { + "epoch": 0.25533, + "grad_norm": 0.7166333794593811, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 25533 + }, + { + "epoch": 0.25534, + "grad_norm": 0.9727382659912109, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 25534 + }, + { + "epoch": 0.25535, + "grad_norm": 1.1182013750076294, + "learning_rate": 0.003, + "loss": 3.97, + "step": 25535 + }, + { + "epoch": 0.25536, + "grad_norm": 0.9283283352851868, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 25536 + }, + { + "epoch": 0.25537, + "grad_norm": 0.8819513916969299, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 25537 + }, + { + "epoch": 0.25538, + "grad_norm": 0.8026654124259949, + "learning_rate": 0.003, + "loss": 3.98, + "step": 25538 + }, + { + "epoch": 0.25539, + "grad_norm": 0.7695189118385315, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 25539 + }, + { + "epoch": 0.2554, + "grad_norm": 0.774635374546051, + "learning_rate": 0.003, + "loss": 3.979, + "step": 25540 + }, + { + "epoch": 0.25541, + "grad_norm": 0.7193542718887329, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 25541 + }, + { + "epoch": 0.25542, + "grad_norm": 0.6456955075263977, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 25542 + }, + { + "epoch": 0.25543, + "grad_norm": 0.638724684715271, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 25543 + }, + { + "epoch": 0.25544, + "grad_norm": 0.6032834053039551, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 25544 + }, + { + "epoch": 0.25545, + "grad_norm": 0.4977414608001709, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 25545 + }, + { + "epoch": 0.25546, + "grad_norm": 0.5021317601203918, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 25546 + }, + { + "epoch": 0.25547, + "grad_norm": 0.6310142874717712, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 25547 + }, + { + "epoch": 0.25548, + "grad_norm": 0.7784028649330139, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 25548 + }, + { + "epoch": 0.25549, + "grad_norm": 0.9501920342445374, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 25549 + }, + { + "epoch": 0.2555, + "grad_norm": 0.9717650413513184, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 25550 + }, + { + "epoch": 0.25551, + "grad_norm": 1.0108182430267334, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 25551 + }, + { + "epoch": 0.25552, + "grad_norm": 1.0795376300811768, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 25552 + }, + { + "epoch": 0.25553, + "grad_norm": 0.7465780973434448, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 25553 + }, + { + "epoch": 0.25554, + "grad_norm": 0.6689134836196899, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 25554 + }, + { + "epoch": 0.25555, + "grad_norm": 0.8653697967529297, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 25555 + }, + { + "epoch": 0.25556, + "grad_norm": 1.0609649419784546, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 25556 + }, + { + "epoch": 0.25557, + "grad_norm": 1.1333070993423462, + "learning_rate": 0.003, + "loss": 4.033, + "step": 25557 + }, + { + "epoch": 0.25558, + "grad_norm": 0.9175860285758972, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 25558 + }, + { + "epoch": 0.25559, + "grad_norm": 0.9913454055786133, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 25559 + }, + { + "epoch": 0.2556, + "grad_norm": 1.030421257019043, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 25560 + }, + { + "epoch": 0.25561, + "grad_norm": 0.9107171297073364, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 25561 + }, + { + "epoch": 0.25562, + "grad_norm": 1.028190016746521, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 25562 + }, + { + "epoch": 0.25563, + "grad_norm": 1.1429566144943237, + "learning_rate": 0.003, + "loss": 4.02, + "step": 25563 + }, + { + "epoch": 0.25564, + "grad_norm": 1.1805790662765503, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 25564 + }, + { + "epoch": 0.25565, + "grad_norm": 1.0524487495422363, + "learning_rate": 0.003, + "loss": 4.005, + "step": 25565 + }, + { + "epoch": 0.25566, + "grad_norm": 1.1039851903915405, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 25566 + }, + { + "epoch": 0.25567, + "grad_norm": 0.8213357925415039, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 25567 + }, + { + "epoch": 0.25568, + "grad_norm": 0.8332287669181824, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 25568 + }, + { + "epoch": 0.25569, + "grad_norm": 1.0496737957000732, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 25569 + }, + { + "epoch": 0.2557, + "grad_norm": 1.2635596990585327, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 25570 + }, + { + "epoch": 0.25571, + "grad_norm": 0.6185611486434937, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 25571 + }, + { + "epoch": 0.25572, + "grad_norm": 0.5796725749969482, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 25572 + }, + { + "epoch": 0.25573, + "grad_norm": 0.5010154843330383, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 25573 + }, + { + "epoch": 0.25574, + "grad_norm": 0.5279836058616638, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 25574 + }, + { + "epoch": 0.25575, + "grad_norm": 0.5594817996025085, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 25575 + }, + { + "epoch": 0.25576, + "grad_norm": 0.6432504653930664, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 25576 + }, + { + "epoch": 0.25577, + "grad_norm": 0.8998452425003052, + "learning_rate": 0.003, + "loss": 3.9364, + "step": 25577 + }, + { + "epoch": 0.25578, + "grad_norm": 1.3321211338043213, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 25578 + }, + { + "epoch": 0.25579, + "grad_norm": 0.7304202914237976, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 25579 + }, + { + "epoch": 0.2558, + "grad_norm": 0.6465408205986023, + "learning_rate": 0.003, + "loss": 3.982, + "step": 25580 + }, + { + "epoch": 0.25581, + "grad_norm": 0.7294855117797852, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 25581 + }, + { + "epoch": 0.25582, + "grad_norm": 0.8117374181747437, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 25582 + }, + { + "epoch": 0.25583, + "grad_norm": 0.8186072111129761, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 25583 + }, + { + "epoch": 0.25584, + "grad_norm": 0.8085074424743652, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 25584 + }, + { + "epoch": 0.25585, + "grad_norm": 0.8650086522102356, + "learning_rate": 0.003, + "loss": 3.99, + "step": 25585 + }, + { + "epoch": 0.25586, + "grad_norm": 0.988926887512207, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 25586 + }, + { + "epoch": 0.25587, + "grad_norm": 1.1431291103363037, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 25587 + }, + { + "epoch": 0.25588, + "grad_norm": 0.7870814204216003, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 25588 + }, + { + "epoch": 0.25589, + "grad_norm": 0.7668725848197937, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 25589 + }, + { + "epoch": 0.2559, + "grad_norm": 0.8283587098121643, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 25590 + }, + { + "epoch": 0.25591, + "grad_norm": 0.8693947196006775, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 25591 + }, + { + "epoch": 0.25592, + "grad_norm": 0.8229303359985352, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 25592 + }, + { + "epoch": 0.25593, + "grad_norm": 0.8583589792251587, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 25593 + }, + { + "epoch": 0.25594, + "grad_norm": 0.9614081382751465, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 25594 + }, + { + "epoch": 0.25595, + "grad_norm": 1.1066399812698364, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 25595 + }, + { + "epoch": 0.25596, + "grad_norm": 0.9215958714485168, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 25596 + }, + { + "epoch": 0.25597, + "grad_norm": 0.913477897644043, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 25597 + }, + { + "epoch": 0.25598, + "grad_norm": 0.9150570034980774, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 25598 + }, + { + "epoch": 0.25599, + "grad_norm": 0.7768281698226929, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 25599 + }, + { + "epoch": 0.256, + "grad_norm": 0.8707113862037659, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 25600 + }, + { + "epoch": 0.25601, + "grad_norm": 1.0236819982528687, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 25601 + }, + { + "epoch": 0.25602, + "grad_norm": 1.0134625434875488, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 25602 + }, + { + "epoch": 0.25603, + "grad_norm": 0.9022558331489563, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 25603 + }, + { + "epoch": 0.25604, + "grad_norm": 0.9508315324783325, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 25604 + }, + { + "epoch": 0.25605, + "grad_norm": 1.0495057106018066, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 25605 + }, + { + "epoch": 0.25606, + "grad_norm": 0.9192548394203186, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 25606 + }, + { + "epoch": 0.25607, + "grad_norm": 1.1854605674743652, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 25607 + }, + { + "epoch": 0.25608, + "grad_norm": 0.7854573726654053, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 25608 + }, + { + "epoch": 0.25609, + "grad_norm": 0.7310811281204224, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 25609 + }, + { + "epoch": 0.2561, + "grad_norm": 0.9624173045158386, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 25610 + }, + { + "epoch": 0.25611, + "grad_norm": 1.1283050775527954, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 25611 + }, + { + "epoch": 0.25612, + "grad_norm": 0.9982665181159973, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 25612 + }, + { + "epoch": 0.25613, + "grad_norm": 1.0667476654052734, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 25613 + }, + { + "epoch": 0.25614, + "grad_norm": 0.8773823380470276, + "learning_rate": 0.003, + "loss": 4.009, + "step": 25614 + }, + { + "epoch": 0.25615, + "grad_norm": 0.828546404838562, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 25615 + }, + { + "epoch": 0.25616, + "grad_norm": 0.7489756345748901, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 25616 + }, + { + "epoch": 0.25617, + "grad_norm": 0.7292338013648987, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 25617 + }, + { + "epoch": 0.25618, + "grad_norm": 0.7057722210884094, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 25618 + }, + { + "epoch": 0.25619, + "grad_norm": 0.6449081301689148, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 25619 + }, + { + "epoch": 0.2562, + "grad_norm": 0.6578495502471924, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 25620 + }, + { + "epoch": 0.25621, + "grad_norm": 0.7867523431777954, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 25621 + }, + { + "epoch": 0.25622, + "grad_norm": 0.8097425699234009, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 25622 + }, + { + "epoch": 0.25623, + "grad_norm": 0.9001954793930054, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 25623 + }, + { + "epoch": 0.25624, + "grad_norm": 0.919344961643219, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 25624 + }, + { + "epoch": 0.25625, + "grad_norm": 0.9139094948768616, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 25625 + }, + { + "epoch": 0.25626, + "grad_norm": 0.7944126129150391, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 25626 + }, + { + "epoch": 0.25627, + "grad_norm": 0.5725326538085938, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 25627 + }, + { + "epoch": 0.25628, + "grad_norm": 0.651527464389801, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 25628 + }, + { + "epoch": 0.25629, + "grad_norm": 0.6480515003204346, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 25629 + }, + { + "epoch": 0.2563, + "grad_norm": 0.7027403712272644, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 25630 + }, + { + "epoch": 0.25631, + "grad_norm": 0.8354399800300598, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 25631 + }, + { + "epoch": 0.25632, + "grad_norm": 0.9531663060188293, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 25632 + }, + { + "epoch": 0.25633, + "grad_norm": 0.9255345463752747, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 25633 + }, + { + "epoch": 0.25634, + "grad_norm": 0.8664073944091797, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 25634 + }, + { + "epoch": 0.25635, + "grad_norm": 1.020575761795044, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 25635 + }, + { + "epoch": 0.25636, + "grad_norm": 1.1445543766021729, + "learning_rate": 0.003, + "loss": 3.989, + "step": 25636 + }, + { + "epoch": 0.25637, + "grad_norm": 0.8264719247817993, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 25637 + }, + { + "epoch": 0.25638, + "grad_norm": 0.9629614949226379, + "learning_rate": 0.003, + "loss": 4.016, + "step": 25638 + }, + { + "epoch": 0.25639, + "grad_norm": 1.0674118995666504, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 25639 + }, + { + "epoch": 0.2564, + "grad_norm": 0.808101236820221, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 25640 + }, + { + "epoch": 0.25641, + "grad_norm": 0.7473341822624207, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 25641 + }, + { + "epoch": 0.25642, + "grad_norm": 0.6567700505256653, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 25642 + }, + { + "epoch": 0.25643, + "grad_norm": 0.5816453695297241, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 25643 + }, + { + "epoch": 0.25644, + "grad_norm": 0.6331604719161987, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 25644 + }, + { + "epoch": 0.25645, + "grad_norm": 0.661094069480896, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 25645 + }, + { + "epoch": 0.25646, + "grad_norm": 0.7354774475097656, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 25646 + }, + { + "epoch": 0.25647, + "grad_norm": 0.7638130784034729, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 25647 + }, + { + "epoch": 0.25648, + "grad_norm": 0.8097441792488098, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 25648 + }, + { + "epoch": 0.25649, + "grad_norm": 0.9556152820587158, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 25649 + }, + { + "epoch": 0.2565, + "grad_norm": 1.1501636505126953, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 25650 + }, + { + "epoch": 0.25651, + "grad_norm": 0.9393385052680969, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 25651 + }, + { + "epoch": 0.25652, + "grad_norm": 0.9081689715385437, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 25652 + }, + { + "epoch": 0.25653, + "grad_norm": 0.9198027849197388, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 25653 + }, + { + "epoch": 0.25654, + "grad_norm": 0.8570000529289246, + "learning_rate": 0.003, + "loss": 4.001, + "step": 25654 + }, + { + "epoch": 0.25655, + "grad_norm": 0.9247301816940308, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 25655 + }, + { + "epoch": 0.25656, + "grad_norm": 1.0351008176803589, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 25656 + }, + { + "epoch": 0.25657, + "grad_norm": 1.1421496868133545, + "learning_rate": 0.003, + "loss": 3.999, + "step": 25657 + }, + { + "epoch": 0.25658, + "grad_norm": 1.04594087600708, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 25658 + }, + { + "epoch": 0.25659, + "grad_norm": 1.0169708728790283, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 25659 + }, + { + "epoch": 0.2566, + "grad_norm": 1.0681848526000977, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 25660 + }, + { + "epoch": 0.25661, + "grad_norm": 0.9712782502174377, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 25661 + }, + { + "epoch": 0.25662, + "grad_norm": 0.902762234210968, + "learning_rate": 0.003, + "loss": 3.984, + "step": 25662 + }, + { + "epoch": 0.25663, + "grad_norm": 0.8772555589675903, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 25663 + }, + { + "epoch": 0.25664, + "grad_norm": 0.7809064984321594, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 25664 + }, + { + "epoch": 0.25665, + "grad_norm": 0.8452991247177124, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 25665 + }, + { + "epoch": 0.25666, + "grad_norm": 1.0374464988708496, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 25666 + }, + { + "epoch": 0.25667, + "grad_norm": 1.0521528720855713, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 25667 + }, + { + "epoch": 0.25668, + "grad_norm": 0.7410352826118469, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 25668 + }, + { + "epoch": 0.25669, + "grad_norm": 0.6520779728889465, + "learning_rate": 0.003, + "loss": 4.015, + "step": 25669 + }, + { + "epoch": 0.2567, + "grad_norm": 0.7949902415275574, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 25670 + }, + { + "epoch": 0.25671, + "grad_norm": 0.8990834355354309, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 25671 + }, + { + "epoch": 0.25672, + "grad_norm": 0.8768930435180664, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 25672 + }, + { + "epoch": 0.25673, + "grad_norm": 0.8376901745796204, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 25673 + }, + { + "epoch": 0.25674, + "grad_norm": 0.8828285932540894, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 25674 + }, + { + "epoch": 0.25675, + "grad_norm": 1.0205416679382324, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 25675 + }, + { + "epoch": 0.25676, + "grad_norm": 0.9979845285415649, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 25676 + }, + { + "epoch": 0.25677, + "grad_norm": 1.0755410194396973, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 25677 + }, + { + "epoch": 0.25678, + "grad_norm": 0.8406369686126709, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 25678 + }, + { + "epoch": 0.25679, + "grad_norm": 0.9956262111663818, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 25679 + }, + { + "epoch": 0.2568, + "grad_norm": 1.1033521890640259, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 25680 + }, + { + "epoch": 0.25681, + "grad_norm": 0.9014893174171448, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 25681 + }, + { + "epoch": 0.25682, + "grad_norm": 1.2062877416610718, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 25682 + }, + { + "epoch": 0.25683, + "grad_norm": 0.9071988463401794, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 25683 + }, + { + "epoch": 0.25684, + "grad_norm": 0.6014238595962524, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 25684 + }, + { + "epoch": 0.25685, + "grad_norm": 0.6106876134872437, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 25685 + }, + { + "epoch": 0.25686, + "grad_norm": 0.575837254524231, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 25686 + }, + { + "epoch": 0.25687, + "grad_norm": 0.616398274898529, + "learning_rate": 0.003, + "loss": 3.972, + "step": 25687 + }, + { + "epoch": 0.25688, + "grad_norm": 0.5601612329483032, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 25688 + }, + { + "epoch": 0.25689, + "grad_norm": 0.5544983744621277, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 25689 + }, + { + "epoch": 0.2569, + "grad_norm": 0.6466320157051086, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 25690 + }, + { + "epoch": 0.25691, + "grad_norm": 0.7297272086143494, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 25691 + }, + { + "epoch": 0.25692, + "grad_norm": 0.9646700024604797, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 25692 + }, + { + "epoch": 0.25693, + "grad_norm": 1.4470372200012207, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 25693 + }, + { + "epoch": 0.25694, + "grad_norm": 0.6371110677719116, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 25694 + }, + { + "epoch": 0.25695, + "grad_norm": 0.7558156847953796, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 25695 + }, + { + "epoch": 0.25696, + "grad_norm": 0.806054949760437, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 25696 + }, + { + "epoch": 0.25697, + "grad_norm": 0.8211185932159424, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 25697 + }, + { + "epoch": 0.25698, + "grad_norm": 1.0635143518447876, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 25698 + }, + { + "epoch": 0.25699, + "grad_norm": 1.0025088787078857, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 25699 + }, + { + "epoch": 0.257, + "grad_norm": 0.9220603704452515, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 25700 + }, + { + "epoch": 0.25701, + "grad_norm": 0.8524656295776367, + "learning_rate": 0.003, + "loss": 3.974, + "step": 25701 + }, + { + "epoch": 0.25702, + "grad_norm": 0.8048814535140991, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 25702 + }, + { + "epoch": 0.25703, + "grad_norm": 0.6970020532608032, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 25703 + }, + { + "epoch": 0.25704, + "grad_norm": 0.7735284566879272, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 25704 + }, + { + "epoch": 0.25705, + "grad_norm": 0.6717774271965027, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 25705 + }, + { + "epoch": 0.25706, + "grad_norm": 0.6669677495956421, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 25706 + }, + { + "epoch": 0.25707, + "grad_norm": 0.6299806833267212, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 25707 + }, + { + "epoch": 0.25708, + "grad_norm": 0.6569746136665344, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 25708 + }, + { + "epoch": 0.25709, + "grad_norm": 0.696101188659668, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 25709 + }, + { + "epoch": 0.2571, + "grad_norm": 0.7977653741836548, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 25710 + }, + { + "epoch": 0.25711, + "grad_norm": 0.969910740852356, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 25711 + }, + { + "epoch": 0.25712, + "grad_norm": 1.1720906496047974, + "learning_rate": 0.003, + "loss": 4.002, + "step": 25712 + }, + { + "epoch": 0.25713, + "grad_norm": 1.297849178314209, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 25713 + }, + { + "epoch": 0.25714, + "grad_norm": 0.9015264511108398, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 25714 + }, + { + "epoch": 0.25715, + "grad_norm": 0.8058189749717712, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 25715 + }, + { + "epoch": 0.25716, + "grad_norm": 1.0180714130401611, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 25716 + }, + { + "epoch": 0.25717, + "grad_norm": 1.1883944272994995, + "learning_rate": 0.003, + "loss": 4.005, + "step": 25717 + }, + { + "epoch": 0.25718, + "grad_norm": 0.8843818306922913, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 25718 + }, + { + "epoch": 0.25719, + "grad_norm": 0.7992709279060364, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 25719 + }, + { + "epoch": 0.2572, + "grad_norm": 0.8805596828460693, + "learning_rate": 0.003, + "loss": 4.019, + "step": 25720 + }, + { + "epoch": 0.25721, + "grad_norm": 0.8954070210456848, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 25721 + }, + { + "epoch": 0.25722, + "grad_norm": 0.9395135045051575, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 25722 + }, + { + "epoch": 0.25723, + "grad_norm": 1.0868067741394043, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 25723 + }, + { + "epoch": 0.25724, + "grad_norm": 1.0281603336334229, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 25724 + }, + { + "epoch": 0.25725, + "grad_norm": 1.063785195350647, + "learning_rate": 0.003, + "loss": 3.998, + "step": 25725 + }, + { + "epoch": 0.25726, + "grad_norm": 1.0560740232467651, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 25726 + }, + { + "epoch": 0.25727, + "grad_norm": 1.0341615676879883, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 25727 + }, + { + "epoch": 0.25728, + "grad_norm": 0.8206321001052856, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 25728 + }, + { + "epoch": 0.25729, + "grad_norm": 0.6672325730323792, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 25729 + }, + { + "epoch": 0.2573, + "grad_norm": 0.6796841621398926, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 25730 + }, + { + "epoch": 0.25731, + "grad_norm": 0.6778923869132996, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 25731 + }, + { + "epoch": 0.25732, + "grad_norm": 0.5899429321289062, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 25732 + }, + { + "epoch": 0.25733, + "grad_norm": 0.6159203052520752, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 25733 + }, + { + "epoch": 0.25734, + "grad_norm": 0.7262720465660095, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 25734 + }, + { + "epoch": 0.25735, + "grad_norm": 0.9055342674255371, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 25735 + }, + { + "epoch": 0.25736, + "grad_norm": 1.0009503364562988, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 25736 + }, + { + "epoch": 0.25737, + "grad_norm": 1.0735852718353271, + "learning_rate": 0.003, + "loss": 4.002, + "step": 25737 + }, + { + "epoch": 0.25738, + "grad_norm": 0.7525038123130798, + "learning_rate": 0.003, + "loss": 3.9395, + "step": 25738 + }, + { + "epoch": 0.25739, + "grad_norm": 0.6372727155685425, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 25739 + }, + { + "epoch": 0.2574, + "grad_norm": 0.6900188326835632, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 25740 + }, + { + "epoch": 0.25741, + "grad_norm": 0.8024082183837891, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 25741 + }, + { + "epoch": 0.25742, + "grad_norm": 0.7860822081565857, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 25742 + }, + { + "epoch": 0.25743, + "grad_norm": 0.8611558675765991, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 25743 + }, + { + "epoch": 0.25744, + "grad_norm": 0.8777139186859131, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 25744 + }, + { + "epoch": 0.25745, + "grad_norm": 0.839604914188385, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 25745 + }, + { + "epoch": 0.25746, + "grad_norm": 0.8636003732681274, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 25746 + }, + { + "epoch": 0.25747, + "grad_norm": 0.8681747913360596, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 25747 + }, + { + "epoch": 0.25748, + "grad_norm": 0.8294050693511963, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 25748 + }, + { + "epoch": 0.25749, + "grad_norm": 0.9757318496704102, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 25749 + }, + { + "epoch": 0.2575, + "grad_norm": 0.9670356512069702, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 25750 + }, + { + "epoch": 0.25751, + "grad_norm": 1.0403915643692017, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 25751 + }, + { + "epoch": 0.25752, + "grad_norm": 1.2003679275512695, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 25752 + }, + { + "epoch": 0.25753, + "grad_norm": 0.899823009967804, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 25753 + }, + { + "epoch": 0.25754, + "grad_norm": 0.8040340542793274, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 25754 + }, + { + "epoch": 0.25755, + "grad_norm": 0.8864023685455322, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 25755 + }, + { + "epoch": 0.25756, + "grad_norm": 0.868986189365387, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 25756 + }, + { + "epoch": 0.25757, + "grad_norm": 0.8223856687545776, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 25757 + }, + { + "epoch": 0.25758, + "grad_norm": 0.7024949193000793, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 25758 + }, + { + "epoch": 0.25759, + "grad_norm": 0.6695616841316223, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 25759 + }, + { + "epoch": 0.2576, + "grad_norm": 0.6841169595718384, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 25760 + }, + { + "epoch": 0.25761, + "grad_norm": 1.0102027654647827, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 25761 + }, + { + "epoch": 0.25762, + "grad_norm": 1.148611307144165, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 25762 + }, + { + "epoch": 0.25763, + "grad_norm": 0.8752733469009399, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 25763 + }, + { + "epoch": 0.25764, + "grad_norm": 0.942444384098053, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 25764 + }, + { + "epoch": 0.25765, + "grad_norm": 0.9313087463378906, + "learning_rate": 0.003, + "loss": 4.009, + "step": 25765 + }, + { + "epoch": 0.25766, + "grad_norm": 0.8997340798377991, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 25766 + }, + { + "epoch": 0.25767, + "grad_norm": 0.8184183835983276, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 25767 + }, + { + "epoch": 0.25768, + "grad_norm": 0.7136496901512146, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 25768 + }, + { + "epoch": 0.25769, + "grad_norm": 0.6068182587623596, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 25769 + }, + { + "epoch": 0.2577, + "grad_norm": 0.6249021887779236, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 25770 + }, + { + "epoch": 0.25771, + "grad_norm": 0.6377755403518677, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 25771 + }, + { + "epoch": 0.25772, + "grad_norm": 0.7788415551185608, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 25772 + }, + { + "epoch": 0.25773, + "grad_norm": 0.9147919416427612, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 25773 + }, + { + "epoch": 0.25774, + "grad_norm": 1.079099178314209, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 25774 + }, + { + "epoch": 0.25775, + "grad_norm": 1.0160892009735107, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 25775 + }, + { + "epoch": 0.25776, + "grad_norm": 0.8718315958976746, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 25776 + }, + { + "epoch": 0.25777, + "grad_norm": 0.7896316051483154, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 25777 + }, + { + "epoch": 0.25778, + "grad_norm": 0.8253551721572876, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 25778 + }, + { + "epoch": 0.25779, + "grad_norm": 0.8716263175010681, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 25779 + }, + { + "epoch": 0.2578, + "grad_norm": 0.7554570436477661, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 25780 + }, + { + "epoch": 0.25781, + "grad_norm": 0.7188590168952942, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 25781 + }, + { + "epoch": 0.25782, + "grad_norm": 0.7618316411972046, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 25782 + }, + { + "epoch": 0.25783, + "grad_norm": 0.6961883902549744, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 25783 + }, + { + "epoch": 0.25784, + "grad_norm": 0.6778168082237244, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 25784 + }, + { + "epoch": 0.25785, + "grad_norm": 0.6886443495750427, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 25785 + }, + { + "epoch": 0.25786, + "grad_norm": 0.7822475433349609, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 25786 + }, + { + "epoch": 0.25787, + "grad_norm": 1.0905755758285522, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 25787 + }, + { + "epoch": 0.25788, + "grad_norm": 1.3049473762512207, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 25788 + }, + { + "epoch": 0.25789, + "grad_norm": 0.8604804873466492, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 25789 + }, + { + "epoch": 0.2579, + "grad_norm": 0.9013181924819946, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 25790 + }, + { + "epoch": 0.25791, + "grad_norm": 0.9492705464363098, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 25791 + }, + { + "epoch": 0.25792, + "grad_norm": 0.9128889441490173, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 25792 + }, + { + "epoch": 0.25793, + "grad_norm": 0.7963501214981079, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 25793 + }, + { + "epoch": 0.25794, + "grad_norm": 0.682733416557312, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 25794 + }, + { + "epoch": 0.25795, + "grad_norm": 0.7519292831420898, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 25795 + }, + { + "epoch": 0.25796, + "grad_norm": 0.8289951682090759, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 25796 + }, + { + "epoch": 0.25797, + "grad_norm": 1.0555105209350586, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 25797 + }, + { + "epoch": 0.25798, + "grad_norm": 1.0788732767105103, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 25798 + }, + { + "epoch": 0.25799, + "grad_norm": 0.8763403296470642, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 25799 + }, + { + "epoch": 0.258, + "grad_norm": 0.824902355670929, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 25800 + }, + { + "epoch": 0.25801, + "grad_norm": 0.8434444665908813, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 25801 + }, + { + "epoch": 0.25802, + "grad_norm": 0.8424826860427856, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 25802 + }, + { + "epoch": 0.25803, + "grad_norm": 0.8501095175743103, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 25803 + }, + { + "epoch": 0.25804, + "grad_norm": 0.9422181844711304, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 25804 + }, + { + "epoch": 0.25805, + "grad_norm": 1.1040619611740112, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 25805 + }, + { + "epoch": 0.25806, + "grad_norm": 1.0114960670471191, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 25806 + }, + { + "epoch": 0.25807, + "grad_norm": 0.9679039716720581, + "learning_rate": 0.003, + "loss": 4.032, + "step": 25807 + }, + { + "epoch": 0.25808, + "grad_norm": 0.9250187873840332, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 25808 + }, + { + "epoch": 0.25809, + "grad_norm": 0.7921593189239502, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 25809 + }, + { + "epoch": 0.2581, + "grad_norm": 0.7995185852050781, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 25810 + }, + { + "epoch": 0.25811, + "grad_norm": 0.8480748534202576, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 25811 + }, + { + "epoch": 0.25812, + "grad_norm": 0.824526846408844, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 25812 + }, + { + "epoch": 0.25813, + "grad_norm": 0.7234083414077759, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 25813 + }, + { + "epoch": 0.25814, + "grad_norm": 0.7057042121887207, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 25814 + }, + { + "epoch": 0.25815, + "grad_norm": 0.7693114280700684, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 25815 + }, + { + "epoch": 0.25816, + "grad_norm": 0.9493828415870667, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 25816 + }, + { + "epoch": 0.25817, + "grad_norm": 1.163997769355774, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 25817 + }, + { + "epoch": 0.25818, + "grad_norm": 0.891489565372467, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 25818 + }, + { + "epoch": 0.25819, + "grad_norm": 0.9550074934959412, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 25819 + }, + { + "epoch": 0.2582, + "grad_norm": 1.179107427597046, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 25820 + }, + { + "epoch": 0.25821, + "grad_norm": 1.1820309162139893, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 25821 + }, + { + "epoch": 0.25822, + "grad_norm": 0.8474862575531006, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 25822 + }, + { + "epoch": 0.25823, + "grad_norm": 0.7020347118377686, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 25823 + }, + { + "epoch": 0.25824, + "grad_norm": 0.613872230052948, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 25824 + }, + { + "epoch": 0.25825, + "grad_norm": 0.6908591389656067, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 25825 + }, + { + "epoch": 0.25826, + "grad_norm": 0.7043891549110413, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 25826 + }, + { + "epoch": 0.25827, + "grad_norm": 0.8030878305435181, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 25827 + }, + { + "epoch": 0.25828, + "grad_norm": 0.9838484525680542, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 25828 + }, + { + "epoch": 0.25829, + "grad_norm": 1.1212913990020752, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 25829 + }, + { + "epoch": 0.2583, + "grad_norm": 0.7046962380409241, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 25830 + }, + { + "epoch": 0.25831, + "grad_norm": 0.5883097052574158, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 25831 + }, + { + "epoch": 0.25832, + "grad_norm": 0.6974595785140991, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 25832 + }, + { + "epoch": 0.25833, + "grad_norm": 0.7001376748085022, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 25833 + }, + { + "epoch": 0.25834, + "grad_norm": 0.8415663242340088, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 25834 + }, + { + "epoch": 0.25835, + "grad_norm": 1.0844626426696777, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 25835 + }, + { + "epoch": 0.25836, + "grad_norm": 0.9797309041023254, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 25836 + }, + { + "epoch": 0.25837, + "grad_norm": 0.857633650302887, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 25837 + }, + { + "epoch": 0.25838, + "grad_norm": 0.8233782649040222, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 25838 + }, + { + "epoch": 0.25839, + "grad_norm": 0.9639068841934204, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 25839 + }, + { + "epoch": 0.2584, + "grad_norm": 1.1849944591522217, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 25840 + }, + { + "epoch": 0.25841, + "grad_norm": 0.7947565317153931, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 25841 + }, + { + "epoch": 0.25842, + "grad_norm": 0.7755822539329529, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 25842 + }, + { + "epoch": 0.25843, + "grad_norm": 0.8181538581848145, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 25843 + }, + { + "epoch": 0.25844, + "grad_norm": 0.8327009081840515, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 25844 + }, + { + "epoch": 0.25845, + "grad_norm": 0.8499844670295715, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 25845 + }, + { + "epoch": 0.25846, + "grad_norm": 0.8760747313499451, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 25846 + }, + { + "epoch": 0.25847, + "grad_norm": 0.9347440004348755, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 25847 + }, + { + "epoch": 0.25848, + "grad_norm": 1.1207454204559326, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 25848 + }, + { + "epoch": 0.25849, + "grad_norm": 0.8775489330291748, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 25849 + }, + { + "epoch": 0.2585, + "grad_norm": 0.8423935174942017, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 25850 + }, + { + "epoch": 0.25851, + "grad_norm": 0.8590739369392395, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 25851 + }, + { + "epoch": 0.25852, + "grad_norm": 0.6968228816986084, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 25852 + }, + { + "epoch": 0.25853, + "grad_norm": 0.6151673197746277, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 25853 + }, + { + "epoch": 0.25854, + "grad_norm": 0.6466443538665771, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 25854 + }, + { + "epoch": 0.25855, + "grad_norm": 0.7322021722793579, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 25855 + }, + { + "epoch": 0.25856, + "grad_norm": 0.671916127204895, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 25856 + }, + { + "epoch": 0.25857, + "grad_norm": 0.8052908778190613, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 25857 + }, + { + "epoch": 0.25858, + "grad_norm": 0.8748964667320251, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 25858 + }, + { + "epoch": 0.25859, + "grad_norm": 0.8244308829307556, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 25859 + }, + { + "epoch": 0.2586, + "grad_norm": 0.8058799505233765, + "learning_rate": 0.003, + "loss": 3.996, + "step": 25860 + }, + { + "epoch": 0.25861, + "grad_norm": 0.7465113997459412, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 25861 + }, + { + "epoch": 0.25862, + "grad_norm": 0.6154301166534424, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 25862 + }, + { + "epoch": 0.25863, + "grad_norm": 0.8742094039916992, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 25863 + }, + { + "epoch": 0.25864, + "grad_norm": 0.9450491666793823, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 25864 + }, + { + "epoch": 0.25865, + "grad_norm": 0.8916528224945068, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 25865 + }, + { + "epoch": 0.25866, + "grad_norm": 0.8411024212837219, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 25866 + }, + { + "epoch": 0.25867, + "grad_norm": 1.160959243774414, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 25867 + }, + { + "epoch": 0.25868, + "grad_norm": 1.1359484195709229, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 25868 + }, + { + "epoch": 0.25869, + "grad_norm": 1.0053073167800903, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 25869 + }, + { + "epoch": 0.2587, + "grad_norm": 1.062371015548706, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 25870 + }, + { + "epoch": 0.25871, + "grad_norm": 0.8989702463150024, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 25871 + }, + { + "epoch": 0.25872, + "grad_norm": 0.8367413282394409, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 25872 + }, + { + "epoch": 0.25873, + "grad_norm": 0.6667805910110474, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 25873 + }, + { + "epoch": 0.25874, + "grad_norm": 0.7142944931983948, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 25874 + }, + { + "epoch": 0.25875, + "grad_norm": 0.6907389760017395, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 25875 + }, + { + "epoch": 0.25876, + "grad_norm": 0.7605401277542114, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 25876 + }, + { + "epoch": 0.25877, + "grad_norm": 0.7984855771064758, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 25877 + }, + { + "epoch": 0.25878, + "grad_norm": 0.740105926990509, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 25878 + }, + { + "epoch": 0.25879, + "grad_norm": 0.8146708011627197, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 25879 + }, + { + "epoch": 0.2588, + "grad_norm": 0.9440900087356567, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 25880 + }, + { + "epoch": 0.25881, + "grad_norm": 1.2603927850723267, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 25881 + }, + { + "epoch": 0.25882, + "grad_norm": 0.6071096658706665, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 25882 + }, + { + "epoch": 0.25883, + "grad_norm": 0.7631679177284241, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 25883 + }, + { + "epoch": 0.25884, + "grad_norm": 1.067726969718933, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 25884 + }, + { + "epoch": 0.25885, + "grad_norm": 1.0687285661697388, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 25885 + }, + { + "epoch": 0.25886, + "grad_norm": 0.9491608142852783, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 25886 + }, + { + "epoch": 0.25887, + "grad_norm": 0.769534707069397, + "learning_rate": 0.003, + "loss": 4.014, + "step": 25887 + }, + { + "epoch": 0.25888, + "grad_norm": 0.5908053517341614, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 25888 + }, + { + "epoch": 0.25889, + "grad_norm": 0.7851896286010742, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 25889 + }, + { + "epoch": 0.2589, + "grad_norm": 0.9425569176673889, + "learning_rate": 0.003, + "loss": 3.984, + "step": 25890 + }, + { + "epoch": 0.25891, + "grad_norm": 1.019263744354248, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 25891 + }, + { + "epoch": 0.25892, + "grad_norm": 1.037838339805603, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 25892 + }, + { + "epoch": 0.25893, + "grad_norm": 1.0212059020996094, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 25893 + }, + { + "epoch": 0.25894, + "grad_norm": 0.8933007121086121, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 25894 + }, + { + "epoch": 0.25895, + "grad_norm": 0.8968614935874939, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 25895 + }, + { + "epoch": 0.25896, + "grad_norm": 0.9860278964042664, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 25896 + }, + { + "epoch": 0.25897, + "grad_norm": 0.9453736543655396, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 25897 + }, + { + "epoch": 0.25898, + "grad_norm": 0.9004547595977783, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 25898 + }, + { + "epoch": 0.25899, + "grad_norm": 0.8468442559242249, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 25899 + }, + { + "epoch": 0.259, + "grad_norm": 0.8706821203231812, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 25900 + }, + { + "epoch": 0.25901, + "grad_norm": 0.8651833534240723, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 25901 + }, + { + "epoch": 0.25902, + "grad_norm": 1.033781886100769, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 25902 + }, + { + "epoch": 0.25903, + "grad_norm": 1.1013420820236206, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 25903 + }, + { + "epoch": 0.25904, + "grad_norm": 0.9037925004959106, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 25904 + }, + { + "epoch": 0.25905, + "grad_norm": 0.8033300042152405, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 25905 + }, + { + "epoch": 0.25906, + "grad_norm": 0.9574422836303711, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 25906 + }, + { + "epoch": 0.25907, + "grad_norm": 1.000380039215088, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 25907 + }, + { + "epoch": 0.25908, + "grad_norm": 0.8852961659431458, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 25908 + }, + { + "epoch": 0.25909, + "grad_norm": 0.9282213449478149, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 25909 + }, + { + "epoch": 0.2591, + "grad_norm": 0.9120476245880127, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 25910 + }, + { + "epoch": 0.25911, + "grad_norm": 0.8363022804260254, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 25911 + }, + { + "epoch": 0.25912, + "grad_norm": 0.7790487408638, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 25912 + }, + { + "epoch": 0.25913, + "grad_norm": 0.9029703140258789, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 25913 + }, + { + "epoch": 0.25914, + "grad_norm": 0.9507086277008057, + "learning_rate": 0.003, + "loss": 3.981, + "step": 25914 + }, + { + "epoch": 0.25915, + "grad_norm": 1.043914794921875, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 25915 + }, + { + "epoch": 0.25916, + "grad_norm": 1.0656565427780151, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 25916 + }, + { + "epoch": 0.25917, + "grad_norm": 1.0790938138961792, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 25917 + }, + { + "epoch": 0.25918, + "grad_norm": 0.9749898910522461, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 25918 + }, + { + "epoch": 0.25919, + "grad_norm": 1.021247386932373, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 25919 + }, + { + "epoch": 0.2592, + "grad_norm": 0.9165574312210083, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 25920 + }, + { + "epoch": 0.25921, + "grad_norm": 0.9567174911499023, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 25921 + }, + { + "epoch": 0.25922, + "grad_norm": 1.0415518283843994, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 25922 + }, + { + "epoch": 0.25923, + "grad_norm": 0.9157006740570068, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 25923 + }, + { + "epoch": 0.25924, + "grad_norm": 0.7760864496231079, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 25924 + }, + { + "epoch": 0.25925, + "grad_norm": 0.7018623948097229, + "learning_rate": 0.003, + "loss": 3.978, + "step": 25925 + }, + { + "epoch": 0.25926, + "grad_norm": 0.705711305141449, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 25926 + }, + { + "epoch": 0.25927, + "grad_norm": 0.7335965037345886, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 25927 + }, + { + "epoch": 0.25928, + "grad_norm": 0.8530840873718262, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 25928 + }, + { + "epoch": 0.25929, + "grad_norm": 0.7701621055603027, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 25929 + }, + { + "epoch": 0.2593, + "grad_norm": 0.6791889667510986, + "learning_rate": 0.003, + "loss": 3.979, + "step": 25930 + }, + { + "epoch": 0.25931, + "grad_norm": 0.5936295986175537, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 25931 + }, + { + "epoch": 0.25932, + "grad_norm": 0.6546452045440674, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 25932 + }, + { + "epoch": 0.25933, + "grad_norm": 0.5575660467147827, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 25933 + }, + { + "epoch": 0.25934, + "grad_norm": 0.8412147164344788, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 25934 + }, + { + "epoch": 0.25935, + "grad_norm": 1.0489661693572998, + "learning_rate": 0.003, + "loss": 3.957, + "step": 25935 + }, + { + "epoch": 0.25936, + "grad_norm": 1.012406349182129, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 25936 + }, + { + "epoch": 0.25937, + "grad_norm": 1.2233247756958008, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 25937 + }, + { + "epoch": 0.25938, + "grad_norm": 0.6380748152732849, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 25938 + }, + { + "epoch": 0.25939, + "grad_norm": 0.6194359660148621, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 25939 + }, + { + "epoch": 0.2594, + "grad_norm": 0.8673077821731567, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 25940 + }, + { + "epoch": 0.25941, + "grad_norm": 0.8745825886726379, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 25941 + }, + { + "epoch": 0.25942, + "grad_norm": 0.8428331017494202, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 25942 + }, + { + "epoch": 0.25943, + "grad_norm": 0.7740764021873474, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 25943 + }, + { + "epoch": 0.25944, + "grad_norm": 0.684913694858551, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 25944 + }, + { + "epoch": 0.25945, + "grad_norm": 0.8072667717933655, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 25945 + }, + { + "epoch": 0.25946, + "grad_norm": 0.9668864011764526, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 25946 + }, + { + "epoch": 0.25947, + "grad_norm": 1.0614107847213745, + "learning_rate": 0.003, + "loss": 4.023, + "step": 25947 + }, + { + "epoch": 0.25948, + "grad_norm": 0.9031354188919067, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 25948 + }, + { + "epoch": 0.25949, + "grad_norm": 1.0145504474639893, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 25949 + }, + { + "epoch": 0.2595, + "grad_norm": 1.1248812675476074, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 25950 + }, + { + "epoch": 0.25951, + "grad_norm": 1.0499422550201416, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 25951 + }, + { + "epoch": 0.25952, + "grad_norm": 1.1418365240097046, + "learning_rate": 0.003, + "loss": 3.993, + "step": 25952 + }, + { + "epoch": 0.25953, + "grad_norm": 0.9452617168426514, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 25953 + }, + { + "epoch": 0.25954, + "grad_norm": 1.0036970376968384, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 25954 + }, + { + "epoch": 0.25955, + "grad_norm": 0.9354735016822815, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 25955 + }, + { + "epoch": 0.25956, + "grad_norm": 0.9437223076820374, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 25956 + }, + { + "epoch": 0.25957, + "grad_norm": 0.9209778308868408, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 25957 + }, + { + "epoch": 0.25958, + "grad_norm": 0.9755724668502808, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 25958 + }, + { + "epoch": 0.25959, + "grad_norm": 1.1002229452133179, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 25959 + }, + { + "epoch": 0.2596, + "grad_norm": 1.0856074094772339, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 25960 + }, + { + "epoch": 0.25961, + "grad_norm": 1.1520411968231201, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 25961 + }, + { + "epoch": 0.25962, + "grad_norm": 1.018645167350769, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 25962 + }, + { + "epoch": 0.25963, + "grad_norm": 1.0676405429840088, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 25963 + }, + { + "epoch": 0.25964, + "grad_norm": 0.9500738382339478, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 25964 + }, + { + "epoch": 0.25965, + "grad_norm": 0.7877424955368042, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 25965 + }, + { + "epoch": 0.25966, + "grad_norm": 0.7607520222663879, + "learning_rate": 0.003, + "loss": 3.998, + "step": 25966 + }, + { + "epoch": 0.25967, + "grad_norm": 0.779480516910553, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 25967 + }, + { + "epoch": 0.25968, + "grad_norm": 0.7897436618804932, + "learning_rate": 0.003, + "loss": 4.001, + "step": 25968 + }, + { + "epoch": 0.25969, + "grad_norm": 0.6808949112892151, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 25969 + }, + { + "epoch": 0.2597, + "grad_norm": 0.5582664608955383, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 25970 + }, + { + "epoch": 0.25971, + "grad_norm": 0.5594429969787598, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 25971 + }, + { + "epoch": 0.25972, + "grad_norm": 0.5572176575660706, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 25972 + }, + { + "epoch": 0.25973, + "grad_norm": 0.6118471622467041, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 25973 + }, + { + "epoch": 0.25974, + "grad_norm": 0.6598048806190491, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 25974 + }, + { + "epoch": 0.25975, + "grad_norm": 0.8110781311988831, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 25975 + }, + { + "epoch": 0.25976, + "grad_norm": 1.0262752771377563, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 25976 + }, + { + "epoch": 0.25977, + "grad_norm": 1.0079931020736694, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 25977 + }, + { + "epoch": 0.25978, + "grad_norm": 0.8735467791557312, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 25978 + }, + { + "epoch": 0.25979, + "grad_norm": 0.7437271475791931, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 25979 + }, + { + "epoch": 0.2598, + "grad_norm": 0.6837152242660522, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 25980 + }, + { + "epoch": 0.25981, + "grad_norm": 0.7171628475189209, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 25981 + }, + { + "epoch": 0.25982, + "grad_norm": 0.6967099905014038, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 25982 + }, + { + "epoch": 0.25983, + "grad_norm": 0.6324313879013062, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 25983 + }, + { + "epoch": 0.25984, + "grad_norm": 0.569660484790802, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 25984 + }, + { + "epoch": 0.25985, + "grad_norm": 0.6193564534187317, + "learning_rate": 0.003, + "loss": 3.946, + "step": 25985 + }, + { + "epoch": 0.25986, + "grad_norm": 0.653251588344574, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 25986 + }, + { + "epoch": 0.25987, + "grad_norm": 0.7069907188415527, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 25987 + }, + { + "epoch": 0.25988, + "grad_norm": 0.9087714552879333, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 25988 + }, + { + "epoch": 0.25989, + "grad_norm": 1.189262866973877, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 25989 + }, + { + "epoch": 0.2599, + "grad_norm": 0.961342990398407, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 25990 + }, + { + "epoch": 0.25991, + "grad_norm": 0.931216835975647, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 25991 + }, + { + "epoch": 0.25992, + "grad_norm": 0.8992535471916199, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 25992 + }, + { + "epoch": 0.25993, + "grad_norm": 0.8450477123260498, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 25993 + }, + { + "epoch": 0.25994, + "grad_norm": 0.7955902218818665, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 25994 + }, + { + "epoch": 0.25995, + "grad_norm": 0.850810170173645, + "learning_rate": 0.003, + "loss": 3.997, + "step": 25995 + }, + { + "epoch": 0.25996, + "grad_norm": 1.0162897109985352, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 25996 + }, + { + "epoch": 0.25997, + "grad_norm": 0.9954431056976318, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 25997 + }, + { + "epoch": 0.25998, + "grad_norm": 0.992469072341919, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 25998 + }, + { + "epoch": 0.25999, + "grad_norm": 0.9947081208229065, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 25999 + }, + { + "epoch": 0.26, + "grad_norm": 1.0166175365447998, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 26000 + }, + { + "epoch": 0.26001, + "grad_norm": 1.0827784538269043, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 26001 + }, + { + "epoch": 0.26002, + "grad_norm": 0.8862667679786682, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 26002 + }, + { + "epoch": 0.26003, + "grad_norm": 1.0098093748092651, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 26003 + }, + { + "epoch": 0.26004, + "grad_norm": 1.0928059816360474, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 26004 + }, + { + "epoch": 0.26005, + "grad_norm": 0.8219820857048035, + "learning_rate": 0.003, + "loss": 4.033, + "step": 26005 + }, + { + "epoch": 0.26006, + "grad_norm": 0.8297351002693176, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 26006 + }, + { + "epoch": 0.26007, + "grad_norm": 0.7845010757446289, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 26007 + }, + { + "epoch": 0.26008, + "grad_norm": 0.7932245135307312, + "learning_rate": 0.003, + "loss": 3.98, + "step": 26008 + }, + { + "epoch": 0.26009, + "grad_norm": 0.9111624956130981, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 26009 + }, + { + "epoch": 0.2601, + "grad_norm": 1.1172484159469604, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 26010 + }, + { + "epoch": 0.26011, + "grad_norm": 1.1286591291427612, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 26011 + }, + { + "epoch": 0.26012, + "grad_norm": 0.7879828810691833, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 26012 + }, + { + "epoch": 0.26013, + "grad_norm": 0.6245054602622986, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 26013 + }, + { + "epoch": 0.26014, + "grad_norm": 0.5778301954269409, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 26014 + }, + { + "epoch": 0.26015, + "grad_norm": 0.6178229451179504, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 26015 + }, + { + "epoch": 0.26016, + "grad_norm": 0.7216571569442749, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 26016 + }, + { + "epoch": 0.26017, + "grad_norm": 0.8518179655075073, + "learning_rate": 0.003, + "loss": 4.005, + "step": 26017 + }, + { + "epoch": 0.26018, + "grad_norm": 0.9163786768913269, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 26018 + }, + { + "epoch": 0.26019, + "grad_norm": 0.8660449981689453, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 26019 + }, + { + "epoch": 0.2602, + "grad_norm": 0.8376045227050781, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 26020 + }, + { + "epoch": 0.26021, + "grad_norm": 0.9022884368896484, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 26021 + }, + { + "epoch": 0.26022, + "grad_norm": 0.761573076248169, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 26022 + }, + { + "epoch": 0.26023, + "grad_norm": 0.7358976006507874, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 26023 + }, + { + "epoch": 0.26024, + "grad_norm": 0.8731980919837952, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 26024 + }, + { + "epoch": 0.26025, + "grad_norm": 0.9420176148414612, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 26025 + }, + { + "epoch": 0.26026, + "grad_norm": 1.0600532293319702, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 26026 + }, + { + "epoch": 0.26027, + "grad_norm": 1.1774088144302368, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 26027 + }, + { + "epoch": 0.26028, + "grad_norm": 0.8098871111869812, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 26028 + }, + { + "epoch": 0.26029, + "grad_norm": 0.6798940896987915, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 26029 + }, + { + "epoch": 0.2603, + "grad_norm": 0.7792561054229736, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 26030 + }, + { + "epoch": 0.26031, + "grad_norm": 0.9714436531066895, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 26031 + }, + { + "epoch": 0.26032, + "grad_norm": 0.884747326374054, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 26032 + }, + { + "epoch": 0.26033, + "grad_norm": 0.752316415309906, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 26033 + }, + { + "epoch": 0.26034, + "grad_norm": 0.7179027199745178, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 26034 + }, + { + "epoch": 0.26035, + "grad_norm": 0.6494106650352478, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 26035 + }, + { + "epoch": 0.26036, + "grad_norm": 0.5415992140769958, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 26036 + }, + { + "epoch": 0.26037, + "grad_norm": 0.5088874697685242, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 26037 + }, + { + "epoch": 0.26038, + "grad_norm": 0.6097182035446167, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 26038 + }, + { + "epoch": 0.26039, + "grad_norm": 0.7408965826034546, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 26039 + }, + { + "epoch": 0.2604, + "grad_norm": 1.0743622779846191, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 26040 + }, + { + "epoch": 0.26041, + "grad_norm": 1.0952346324920654, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 26041 + }, + { + "epoch": 0.26042, + "grad_norm": 0.9796099662780762, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 26042 + }, + { + "epoch": 0.26043, + "grad_norm": 1.0913171768188477, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 26043 + }, + { + "epoch": 0.26044, + "grad_norm": 1.0052175521850586, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 26044 + }, + { + "epoch": 0.26045, + "grad_norm": 1.1103712320327759, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 26045 + }, + { + "epoch": 0.26046, + "grad_norm": 0.8797158002853394, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 26046 + }, + { + "epoch": 0.26047, + "grad_norm": 0.7369373440742493, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 26047 + }, + { + "epoch": 0.26048, + "grad_norm": 0.6627705097198486, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 26048 + }, + { + "epoch": 0.26049, + "grad_norm": 0.7341480851173401, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 26049 + }, + { + "epoch": 0.2605, + "grad_norm": 0.8183398246765137, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 26050 + }, + { + "epoch": 0.26051, + "grad_norm": 0.7322660088539124, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 26051 + }, + { + "epoch": 0.26052, + "grad_norm": 0.7197584509849548, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 26052 + }, + { + "epoch": 0.26053, + "grad_norm": 0.7401795983314514, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 26053 + }, + { + "epoch": 0.26054, + "grad_norm": 0.7497231364250183, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 26054 + }, + { + "epoch": 0.26055, + "grad_norm": 0.8824513554573059, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 26055 + }, + { + "epoch": 0.26056, + "grad_norm": 1.0194402933120728, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 26056 + }, + { + "epoch": 0.26057, + "grad_norm": 1.0978212356567383, + "learning_rate": 0.003, + "loss": 3.991, + "step": 26057 + }, + { + "epoch": 0.26058, + "grad_norm": 1.007453203201294, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 26058 + }, + { + "epoch": 0.26059, + "grad_norm": 1.1754463911056519, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 26059 + }, + { + "epoch": 0.2606, + "grad_norm": 0.7121020555496216, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 26060 + }, + { + "epoch": 0.26061, + "grad_norm": 0.6577557921409607, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 26061 + }, + { + "epoch": 0.26062, + "grad_norm": 0.8202584981918335, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 26062 + }, + { + "epoch": 0.26063, + "grad_norm": 0.8864991664886475, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 26063 + }, + { + "epoch": 0.26064, + "grad_norm": 0.9445503354072571, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 26064 + }, + { + "epoch": 0.26065, + "grad_norm": 0.8596328496932983, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 26065 + }, + { + "epoch": 0.26066, + "grad_norm": 0.7207450866699219, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 26066 + }, + { + "epoch": 0.26067, + "grad_norm": 0.8190909028053284, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 26067 + }, + { + "epoch": 0.26068, + "grad_norm": 0.8733481764793396, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 26068 + }, + { + "epoch": 0.26069, + "grad_norm": 0.8340277671813965, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 26069 + }, + { + "epoch": 0.2607, + "grad_norm": 0.9255756735801697, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 26070 + }, + { + "epoch": 0.26071, + "grad_norm": 1.0153249502182007, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 26071 + }, + { + "epoch": 0.26072, + "grad_norm": 1.082266092300415, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 26072 + }, + { + "epoch": 0.26073, + "grad_norm": 0.8965131044387817, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 26073 + }, + { + "epoch": 0.26074, + "grad_norm": 0.7847994565963745, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 26074 + }, + { + "epoch": 0.26075, + "grad_norm": 0.7952308654785156, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 26075 + }, + { + "epoch": 0.26076, + "grad_norm": 0.6831323504447937, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 26076 + }, + { + "epoch": 0.26077, + "grad_norm": 0.6588191986083984, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 26077 + }, + { + "epoch": 0.26078, + "grad_norm": 0.8325164318084717, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 26078 + }, + { + "epoch": 0.26079, + "grad_norm": 0.9884130954742432, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 26079 + }, + { + "epoch": 0.2608, + "grad_norm": 1.0942987203598022, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 26080 + }, + { + "epoch": 0.26081, + "grad_norm": 1.1309094429016113, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 26081 + }, + { + "epoch": 0.26082, + "grad_norm": 1.2589694261550903, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 26082 + }, + { + "epoch": 0.26083, + "grad_norm": 0.6888445615768433, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 26083 + }, + { + "epoch": 0.26084, + "grad_norm": 0.6705524921417236, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 26084 + }, + { + "epoch": 0.26085, + "grad_norm": 0.6790623068809509, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 26085 + }, + { + "epoch": 0.26086, + "grad_norm": 0.6883323192596436, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 26086 + }, + { + "epoch": 0.26087, + "grad_norm": 0.8122868537902832, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 26087 + }, + { + "epoch": 0.26088, + "grad_norm": 0.9884844422340393, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 26088 + }, + { + "epoch": 0.26089, + "grad_norm": 1.0888664722442627, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 26089 + }, + { + "epoch": 0.2609, + "grad_norm": 0.8891667723655701, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 26090 + }, + { + "epoch": 0.26091, + "grad_norm": 0.7988774180412292, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 26091 + }, + { + "epoch": 0.26092, + "grad_norm": 0.775905430316925, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 26092 + }, + { + "epoch": 0.26093, + "grad_norm": 0.923873782157898, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 26093 + }, + { + "epoch": 0.26094, + "grad_norm": 0.9252065420150757, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 26094 + }, + { + "epoch": 0.26095, + "grad_norm": 0.8135464191436768, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 26095 + }, + { + "epoch": 0.26096, + "grad_norm": 0.9270109534263611, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 26096 + }, + { + "epoch": 0.26097, + "grad_norm": 1.0725722312927246, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 26097 + }, + { + "epoch": 0.26098, + "grad_norm": 0.8016641139984131, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 26098 + }, + { + "epoch": 0.26099, + "grad_norm": 0.6054804921150208, + "learning_rate": 0.003, + "loss": 3.996, + "step": 26099 + }, + { + "epoch": 0.261, + "grad_norm": 0.7556873559951782, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 26100 + }, + { + "epoch": 0.26101, + "grad_norm": 0.7187185883522034, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 26101 + }, + { + "epoch": 0.26102, + "grad_norm": 0.7237546443939209, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 26102 + }, + { + "epoch": 0.26103, + "grad_norm": 0.8489375710487366, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 26103 + }, + { + "epoch": 0.26104, + "grad_norm": 0.9379097819328308, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 26104 + }, + { + "epoch": 0.26105, + "grad_norm": 0.9334791898727417, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 26105 + }, + { + "epoch": 0.26106, + "grad_norm": 0.8662063479423523, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 26106 + }, + { + "epoch": 0.26107, + "grad_norm": 0.903740644454956, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 26107 + }, + { + "epoch": 0.26108, + "grad_norm": 0.9036640524864197, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 26108 + }, + { + "epoch": 0.26109, + "grad_norm": 0.9180763363838196, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 26109 + }, + { + "epoch": 0.2611, + "grad_norm": 1.2080146074295044, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 26110 + }, + { + "epoch": 0.26111, + "grad_norm": 0.7170922160148621, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 26111 + }, + { + "epoch": 0.26112, + "grad_norm": 0.750449538230896, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 26112 + }, + { + "epoch": 0.26113, + "grad_norm": 1.0889475345611572, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 26113 + }, + { + "epoch": 0.26114, + "grad_norm": 1.2774569988250732, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 26114 + }, + { + "epoch": 0.26115, + "grad_norm": 0.6673142910003662, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 26115 + }, + { + "epoch": 0.26116, + "grad_norm": 0.7468152642250061, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 26116 + }, + { + "epoch": 0.26117, + "grad_norm": 0.7240762114524841, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 26117 + }, + { + "epoch": 0.26118, + "grad_norm": 0.6921622157096863, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 26118 + }, + { + "epoch": 0.26119, + "grad_norm": 0.6969214677810669, + "learning_rate": 0.003, + "loss": 3.986, + "step": 26119 + }, + { + "epoch": 0.2612, + "grad_norm": 0.6466007828712463, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 26120 + }, + { + "epoch": 0.26121, + "grad_norm": 0.6213445663452148, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 26121 + }, + { + "epoch": 0.26122, + "grad_norm": 0.627733588218689, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 26122 + }, + { + "epoch": 0.26123, + "grad_norm": 0.6111557483673096, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 26123 + }, + { + "epoch": 0.26124, + "grad_norm": 0.8652463555335999, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 26124 + }, + { + "epoch": 0.26125, + "grad_norm": 1.2109947204589844, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 26125 + }, + { + "epoch": 0.26126, + "grad_norm": 1.0685865879058838, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 26126 + }, + { + "epoch": 0.26127, + "grad_norm": 1.0140823125839233, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 26127 + }, + { + "epoch": 0.26128, + "grad_norm": 0.9446596503257751, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 26128 + }, + { + "epoch": 0.26129, + "grad_norm": 0.8014139533042908, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 26129 + }, + { + "epoch": 0.2613, + "grad_norm": 0.8241828680038452, + "learning_rate": 0.003, + "loss": 4.002, + "step": 26130 + }, + { + "epoch": 0.26131, + "grad_norm": 0.8253116011619568, + "learning_rate": 0.003, + "loss": 3.986, + "step": 26131 + }, + { + "epoch": 0.26132, + "grad_norm": 0.8885040879249573, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 26132 + }, + { + "epoch": 0.26133, + "grad_norm": 0.939513623714447, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 26133 + }, + { + "epoch": 0.26134, + "grad_norm": 1.0416638851165771, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 26134 + }, + { + "epoch": 0.26135, + "grad_norm": 1.1034966707229614, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 26135 + }, + { + "epoch": 0.26136, + "grad_norm": 0.9799518585205078, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 26136 + }, + { + "epoch": 0.26137, + "grad_norm": 0.9487122893333435, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 26137 + }, + { + "epoch": 0.26138, + "grad_norm": 0.9069750905036926, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 26138 + }, + { + "epoch": 0.26139, + "grad_norm": 0.8710862398147583, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 26139 + }, + { + "epoch": 0.2614, + "grad_norm": 0.740614116191864, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 26140 + }, + { + "epoch": 0.26141, + "grad_norm": 0.8403662443161011, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 26141 + }, + { + "epoch": 0.26142, + "grad_norm": 0.9634315371513367, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 26142 + }, + { + "epoch": 0.26143, + "grad_norm": 1.047188639640808, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 26143 + }, + { + "epoch": 0.26144, + "grad_norm": 1.0012141466140747, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 26144 + }, + { + "epoch": 0.26145, + "grad_norm": 1.0742955207824707, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 26145 + }, + { + "epoch": 0.26146, + "grad_norm": 0.8760760426521301, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 26146 + }, + { + "epoch": 0.26147, + "grad_norm": 0.8868107795715332, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 26147 + }, + { + "epoch": 0.26148, + "grad_norm": 0.948391854763031, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 26148 + }, + { + "epoch": 0.26149, + "grad_norm": 0.9610210657119751, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 26149 + }, + { + "epoch": 0.2615, + "grad_norm": 1.165300965309143, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 26150 + }, + { + "epoch": 0.26151, + "grad_norm": 1.0720632076263428, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 26151 + }, + { + "epoch": 0.26152, + "grad_norm": 0.9005213975906372, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 26152 + }, + { + "epoch": 0.26153, + "grad_norm": 0.8290284276008606, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 26153 + }, + { + "epoch": 0.26154, + "grad_norm": 0.7439013123512268, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 26154 + }, + { + "epoch": 0.26155, + "grad_norm": 0.6381121277809143, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 26155 + }, + { + "epoch": 0.26156, + "grad_norm": 0.5811452269554138, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 26156 + }, + { + "epoch": 0.26157, + "grad_norm": 0.5715869069099426, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 26157 + }, + { + "epoch": 0.26158, + "grad_norm": 0.5928052067756653, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 26158 + }, + { + "epoch": 0.26159, + "grad_norm": 0.7776232957839966, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 26159 + }, + { + "epoch": 0.2616, + "grad_norm": 0.8656290173530579, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 26160 + }, + { + "epoch": 0.26161, + "grad_norm": 1.075168490409851, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 26161 + }, + { + "epoch": 0.26162, + "grad_norm": 1.1601680517196655, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 26162 + }, + { + "epoch": 0.26163, + "grad_norm": 0.7719733715057373, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 26163 + }, + { + "epoch": 0.26164, + "grad_norm": 0.6396301984786987, + "learning_rate": 0.003, + "loss": 3.968, + "step": 26164 + }, + { + "epoch": 0.26165, + "grad_norm": 0.6713854074478149, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 26165 + }, + { + "epoch": 0.26166, + "grad_norm": 0.7922201752662659, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 26166 + }, + { + "epoch": 0.26167, + "grad_norm": 0.8175195455551147, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 26167 + }, + { + "epoch": 0.26168, + "grad_norm": 0.9341580867767334, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 26168 + }, + { + "epoch": 0.26169, + "grad_norm": 0.8397361636161804, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 26169 + }, + { + "epoch": 0.2617, + "grad_norm": 0.6917333602905273, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 26170 + }, + { + "epoch": 0.26171, + "grad_norm": 0.7000739574432373, + "learning_rate": 0.003, + "loss": 3.975, + "step": 26171 + }, + { + "epoch": 0.26172, + "grad_norm": 0.7310563921928406, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 26172 + }, + { + "epoch": 0.26173, + "grad_norm": 0.9578790664672852, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 26173 + }, + { + "epoch": 0.26174, + "grad_norm": 1.1926627159118652, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 26174 + }, + { + "epoch": 0.26175, + "grad_norm": 0.6961535811424255, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 26175 + }, + { + "epoch": 0.26176, + "grad_norm": 0.7322376370429993, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 26176 + }, + { + "epoch": 0.26177, + "grad_norm": 0.772594690322876, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 26177 + }, + { + "epoch": 0.26178, + "grad_norm": 0.8888876438140869, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 26178 + }, + { + "epoch": 0.26179, + "grad_norm": 0.9135903716087341, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 26179 + }, + { + "epoch": 0.2618, + "grad_norm": 0.918860137462616, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 26180 + }, + { + "epoch": 0.26181, + "grad_norm": 1.0523300170898438, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 26181 + }, + { + "epoch": 0.26182, + "grad_norm": 1.141065239906311, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 26182 + }, + { + "epoch": 0.26183, + "grad_norm": 0.8255327343940735, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 26183 + }, + { + "epoch": 0.26184, + "grad_norm": 0.7999727725982666, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 26184 + }, + { + "epoch": 0.26185, + "grad_norm": 0.7504163384437561, + "learning_rate": 0.003, + "loss": 3.971, + "step": 26185 + }, + { + "epoch": 0.26186, + "grad_norm": 0.8033907413482666, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 26186 + }, + { + "epoch": 0.26187, + "grad_norm": 0.939363956451416, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 26187 + }, + { + "epoch": 0.26188, + "grad_norm": 0.9290593266487122, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 26188 + }, + { + "epoch": 0.26189, + "grad_norm": 0.9589370489120483, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 26189 + }, + { + "epoch": 0.2619, + "grad_norm": 1.1407946348190308, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 26190 + }, + { + "epoch": 0.26191, + "grad_norm": 0.9547918438911438, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 26191 + }, + { + "epoch": 0.26192, + "grad_norm": 0.8718344569206238, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 26192 + }, + { + "epoch": 0.26193, + "grad_norm": 0.8275954723358154, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 26193 + }, + { + "epoch": 0.26194, + "grad_norm": 0.8296892642974854, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 26194 + }, + { + "epoch": 0.26195, + "grad_norm": 0.8237044215202332, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 26195 + }, + { + "epoch": 0.26196, + "grad_norm": 0.8844059109687805, + "learning_rate": 0.003, + "loss": 3.982, + "step": 26196 + }, + { + "epoch": 0.26197, + "grad_norm": 1.0002353191375732, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 26197 + }, + { + "epoch": 0.26198, + "grad_norm": 0.9898791909217834, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 26198 + }, + { + "epoch": 0.26199, + "grad_norm": 0.9617847204208374, + "learning_rate": 0.003, + "loss": 4.033, + "step": 26199 + }, + { + "epoch": 0.262, + "grad_norm": 0.9881977438926697, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 26200 + }, + { + "epoch": 0.26201, + "grad_norm": 0.9661552906036377, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 26201 + }, + { + "epoch": 0.26202, + "grad_norm": 0.8673192262649536, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 26202 + }, + { + "epoch": 0.26203, + "grad_norm": 0.7687901854515076, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 26203 + }, + { + "epoch": 0.26204, + "grad_norm": 0.9056602716445923, + "learning_rate": 0.003, + "loss": 3.986, + "step": 26204 + }, + { + "epoch": 0.26205, + "grad_norm": 0.977712869644165, + "learning_rate": 0.003, + "loss": 4.014, + "step": 26205 + }, + { + "epoch": 0.26206, + "grad_norm": 0.9531207084655762, + "learning_rate": 0.003, + "loss": 3.97, + "step": 26206 + }, + { + "epoch": 0.26207, + "grad_norm": 0.9362176656723022, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 26207 + }, + { + "epoch": 0.26208, + "grad_norm": 0.994010865688324, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 26208 + }, + { + "epoch": 0.26209, + "grad_norm": 0.9880914092063904, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 26209 + }, + { + "epoch": 0.2621, + "grad_norm": 1.0496249198913574, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 26210 + }, + { + "epoch": 0.26211, + "grad_norm": 0.9843340516090393, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 26211 + }, + { + "epoch": 0.26212, + "grad_norm": 0.8727530241012573, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 26212 + }, + { + "epoch": 0.26213, + "grad_norm": 0.769472062587738, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 26213 + }, + { + "epoch": 0.26214, + "grad_norm": 0.8488693237304688, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 26214 + }, + { + "epoch": 0.26215, + "grad_norm": 1.0322341918945312, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 26215 + }, + { + "epoch": 0.26216, + "grad_norm": 1.031699776649475, + "learning_rate": 0.003, + "loss": 4.003, + "step": 26216 + }, + { + "epoch": 0.26217, + "grad_norm": 0.9240838885307312, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 26217 + }, + { + "epoch": 0.26218, + "grad_norm": 0.6283883452415466, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 26218 + }, + { + "epoch": 0.26219, + "grad_norm": 0.6544593572616577, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 26219 + }, + { + "epoch": 0.2622, + "grad_norm": 0.6957463026046753, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 26220 + }, + { + "epoch": 0.26221, + "grad_norm": 0.6686781048774719, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 26221 + }, + { + "epoch": 0.26222, + "grad_norm": 0.8353521227836609, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 26222 + }, + { + "epoch": 0.26223, + "grad_norm": 0.9412662982940674, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 26223 + }, + { + "epoch": 0.26224, + "grad_norm": 1.0198768377304077, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 26224 + }, + { + "epoch": 0.26225, + "grad_norm": 1.0233430862426758, + "learning_rate": 0.003, + "loss": 3.995, + "step": 26225 + }, + { + "epoch": 0.26226, + "grad_norm": 0.963280439376831, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 26226 + }, + { + "epoch": 0.26227, + "grad_norm": 0.9346410036087036, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 26227 + }, + { + "epoch": 0.26228, + "grad_norm": 0.9205958247184753, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 26228 + }, + { + "epoch": 0.26229, + "grad_norm": 0.8753637075424194, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 26229 + }, + { + "epoch": 0.2623, + "grad_norm": 0.8011388778686523, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 26230 + }, + { + "epoch": 0.26231, + "grad_norm": 0.7333263158798218, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 26231 + }, + { + "epoch": 0.26232, + "grad_norm": 0.8188323974609375, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 26232 + }, + { + "epoch": 0.26233, + "grad_norm": 1.064719796180725, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 26233 + }, + { + "epoch": 0.26234, + "grad_norm": 0.9831905961036682, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 26234 + }, + { + "epoch": 0.26235, + "grad_norm": 0.9711369872093201, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 26235 + }, + { + "epoch": 0.26236, + "grad_norm": 0.8881285786628723, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 26236 + }, + { + "epoch": 0.26237, + "grad_norm": 0.8305803537368774, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 26237 + }, + { + "epoch": 0.26238, + "grad_norm": 0.9598031640052795, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 26238 + }, + { + "epoch": 0.26239, + "grad_norm": 1.1944481134414673, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 26239 + }, + { + "epoch": 0.2624, + "grad_norm": 0.7163622975349426, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 26240 + }, + { + "epoch": 0.26241, + "grad_norm": 0.7165594696998596, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 26241 + }, + { + "epoch": 0.26242, + "grad_norm": 0.7519130706787109, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 26242 + }, + { + "epoch": 0.26243, + "grad_norm": 0.7892477512359619, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 26243 + }, + { + "epoch": 0.26244, + "grad_norm": 0.740693211555481, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 26244 + }, + { + "epoch": 0.26245, + "grad_norm": 0.6595543026924133, + "learning_rate": 0.003, + "loss": 3.986, + "step": 26245 + }, + { + "epoch": 0.26246, + "grad_norm": 0.7668536901473999, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 26246 + }, + { + "epoch": 0.26247, + "grad_norm": 0.9667211174964905, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 26247 + }, + { + "epoch": 0.26248, + "grad_norm": 1.1343917846679688, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 26248 + }, + { + "epoch": 0.26249, + "grad_norm": 0.9643491506576538, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 26249 + }, + { + "epoch": 0.2625, + "grad_norm": 0.9300940036773682, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 26250 + }, + { + "epoch": 0.26251, + "grad_norm": 0.8746671676635742, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 26251 + }, + { + "epoch": 0.26252, + "grad_norm": 0.8755052089691162, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 26252 + }, + { + "epoch": 0.26253, + "grad_norm": 0.8815786242485046, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 26253 + }, + { + "epoch": 0.26254, + "grad_norm": 0.948013961315155, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 26254 + }, + { + "epoch": 0.26255, + "grad_norm": 0.8346024751663208, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 26255 + }, + { + "epoch": 0.26256, + "grad_norm": 0.7882248759269714, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 26256 + }, + { + "epoch": 0.26257, + "grad_norm": 0.7358688116073608, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 26257 + }, + { + "epoch": 0.26258, + "grad_norm": 0.843254804611206, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 26258 + }, + { + "epoch": 0.26259, + "grad_norm": 0.7325842976570129, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 26259 + }, + { + "epoch": 0.2626, + "grad_norm": 0.5899859070777893, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 26260 + }, + { + "epoch": 0.26261, + "grad_norm": 0.6103318929672241, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 26261 + }, + { + "epoch": 0.26262, + "grad_norm": 0.762093186378479, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 26262 + }, + { + "epoch": 0.26263, + "grad_norm": 0.8497004508972168, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 26263 + }, + { + "epoch": 0.26264, + "grad_norm": 0.7882711291313171, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 26264 + }, + { + "epoch": 0.26265, + "grad_norm": 0.8978094458580017, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 26265 + }, + { + "epoch": 0.26266, + "grad_norm": 1.1767860651016235, + "learning_rate": 0.003, + "loss": 3.979, + "step": 26266 + }, + { + "epoch": 0.26267, + "grad_norm": 0.9363145232200623, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 26267 + }, + { + "epoch": 0.26268, + "grad_norm": 0.8490022420883179, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 26268 + }, + { + "epoch": 0.26269, + "grad_norm": 0.927634596824646, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 26269 + }, + { + "epoch": 0.2627, + "grad_norm": 1.1152746677398682, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 26270 + }, + { + "epoch": 0.26271, + "grad_norm": 1.051201343536377, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 26271 + }, + { + "epoch": 0.26272, + "grad_norm": 1.0176236629486084, + "learning_rate": 0.003, + "loss": 3.992, + "step": 26272 + }, + { + "epoch": 0.26273, + "grad_norm": 0.8677682876586914, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 26273 + }, + { + "epoch": 0.26274, + "grad_norm": 0.793242335319519, + "learning_rate": 0.003, + "loss": 3.958, + "step": 26274 + }, + { + "epoch": 0.26275, + "grad_norm": 0.7103435397148132, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 26275 + }, + { + "epoch": 0.26276, + "grad_norm": 0.7646177411079407, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 26276 + }, + { + "epoch": 0.26277, + "grad_norm": 1.0295369625091553, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 26277 + }, + { + "epoch": 0.26278, + "grad_norm": 1.1059845685958862, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 26278 + }, + { + "epoch": 0.26279, + "grad_norm": 0.8214289546012878, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 26279 + }, + { + "epoch": 0.2628, + "grad_norm": 0.6922842860221863, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 26280 + }, + { + "epoch": 0.26281, + "grad_norm": 0.6616455316543579, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 26281 + }, + { + "epoch": 0.26282, + "grad_norm": 0.7616168856620789, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 26282 + }, + { + "epoch": 0.26283, + "grad_norm": 0.8684755563735962, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 26283 + }, + { + "epoch": 0.26284, + "grad_norm": 0.9921523928642273, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 26284 + }, + { + "epoch": 0.26285, + "grad_norm": 0.9399673342704773, + "learning_rate": 0.003, + "loss": 3.976, + "step": 26285 + }, + { + "epoch": 0.26286, + "grad_norm": 1.2578797340393066, + "learning_rate": 0.003, + "loss": 4.009, + "step": 26286 + }, + { + "epoch": 0.26287, + "grad_norm": 0.9016884565353394, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 26287 + }, + { + "epoch": 0.26288, + "grad_norm": 0.759493887424469, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 26288 + }, + { + "epoch": 0.26289, + "grad_norm": 0.6765230298042297, + "learning_rate": 0.003, + "loss": 3.992, + "step": 26289 + }, + { + "epoch": 0.2629, + "grad_norm": 0.656478226184845, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 26290 + }, + { + "epoch": 0.26291, + "grad_norm": 0.7744309306144714, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 26291 + }, + { + "epoch": 0.26292, + "grad_norm": 0.8732783198356628, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 26292 + }, + { + "epoch": 0.26293, + "grad_norm": 1.0379987955093384, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 26293 + }, + { + "epoch": 0.26294, + "grad_norm": 1.2727183103561401, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 26294 + }, + { + "epoch": 0.26295, + "grad_norm": 0.7694161534309387, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 26295 + }, + { + "epoch": 0.26296, + "grad_norm": 0.6588641405105591, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 26296 + }, + { + "epoch": 0.26297, + "grad_norm": 0.5414159893989563, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 26297 + }, + { + "epoch": 0.26298, + "grad_norm": 0.610934317111969, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 26298 + }, + { + "epoch": 0.26299, + "grad_norm": 0.6057886481285095, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 26299 + }, + { + "epoch": 0.263, + "grad_norm": 0.6516247987747192, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 26300 + }, + { + "epoch": 0.26301, + "grad_norm": 0.721208393573761, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 26301 + }, + { + "epoch": 0.26302, + "grad_norm": 0.6706159710884094, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 26302 + }, + { + "epoch": 0.26303, + "grad_norm": 0.6943889260292053, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 26303 + }, + { + "epoch": 0.26304, + "grad_norm": 0.9945859313011169, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 26304 + }, + { + "epoch": 0.26305, + "grad_norm": 1.2872077226638794, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 26305 + }, + { + "epoch": 0.26306, + "grad_norm": 0.7545163035392761, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 26306 + }, + { + "epoch": 0.26307, + "grad_norm": 0.7252734899520874, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 26307 + }, + { + "epoch": 0.26308, + "grad_norm": 0.6983084678649902, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 26308 + }, + { + "epoch": 0.26309, + "grad_norm": 0.7367895245552063, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 26309 + }, + { + "epoch": 0.2631, + "grad_norm": 0.8705503940582275, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 26310 + }, + { + "epoch": 0.26311, + "grad_norm": 1.1918809413909912, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 26311 + }, + { + "epoch": 0.26312, + "grad_norm": 0.7633008360862732, + "learning_rate": 0.003, + "loss": 3.981, + "step": 26312 + }, + { + "epoch": 0.26313, + "grad_norm": 0.6117161512374878, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 26313 + }, + { + "epoch": 0.26314, + "grad_norm": 0.6238694787025452, + "learning_rate": 0.003, + "loss": 3.996, + "step": 26314 + }, + { + "epoch": 0.26315, + "grad_norm": 0.7115004658699036, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 26315 + }, + { + "epoch": 0.26316, + "grad_norm": 0.7816676497459412, + "learning_rate": 0.003, + "loss": 3.997, + "step": 26316 + }, + { + "epoch": 0.26317, + "grad_norm": 0.8141606450080872, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 26317 + }, + { + "epoch": 0.26318, + "grad_norm": 0.9022324085235596, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 26318 + }, + { + "epoch": 0.26319, + "grad_norm": 1.0549417734146118, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 26319 + }, + { + "epoch": 0.2632, + "grad_norm": 0.9886102080345154, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 26320 + }, + { + "epoch": 0.26321, + "grad_norm": 0.8798477649688721, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 26321 + }, + { + "epoch": 0.26322, + "grad_norm": 0.8114706873893738, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 26322 + }, + { + "epoch": 0.26323, + "grad_norm": 0.8313837647438049, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 26323 + }, + { + "epoch": 0.26324, + "grad_norm": 0.9767457246780396, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 26324 + }, + { + "epoch": 0.26325, + "grad_norm": 1.436343789100647, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 26325 + }, + { + "epoch": 0.26326, + "grad_norm": 0.9621006846427917, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 26326 + }, + { + "epoch": 0.26327, + "grad_norm": 1.0659526586532593, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 26327 + }, + { + "epoch": 0.26328, + "grad_norm": 0.9295737743377686, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 26328 + }, + { + "epoch": 0.26329, + "grad_norm": 0.8517880439758301, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 26329 + }, + { + "epoch": 0.2633, + "grad_norm": 0.8461743593215942, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 26330 + }, + { + "epoch": 0.26331, + "grad_norm": 0.8348371982574463, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 26331 + }, + { + "epoch": 0.26332, + "grad_norm": 0.8649199604988098, + "learning_rate": 0.003, + "loss": 4.009, + "step": 26332 + }, + { + "epoch": 0.26333, + "grad_norm": 0.9397104382514954, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 26333 + }, + { + "epoch": 0.26334, + "grad_norm": 0.9069020748138428, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 26334 + }, + { + "epoch": 0.26335, + "grad_norm": 0.8523068428039551, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 26335 + }, + { + "epoch": 0.26336, + "grad_norm": 0.7541300654411316, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 26336 + }, + { + "epoch": 0.26337, + "grad_norm": 0.8364758491516113, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 26337 + }, + { + "epoch": 0.26338, + "grad_norm": 0.8976424932479858, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 26338 + }, + { + "epoch": 0.26339, + "grad_norm": 1.018916130065918, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 26339 + }, + { + "epoch": 0.2634, + "grad_norm": 1.1499958038330078, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 26340 + }, + { + "epoch": 0.26341, + "grad_norm": 0.8625072836875916, + "learning_rate": 0.003, + "loss": 3.968, + "step": 26341 + }, + { + "epoch": 0.26342, + "grad_norm": 0.8117766380310059, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 26342 + }, + { + "epoch": 0.26343, + "grad_norm": 0.9497936964035034, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 26343 + }, + { + "epoch": 0.26344, + "grad_norm": 1.1073561906814575, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 26344 + }, + { + "epoch": 0.26345, + "grad_norm": 0.7805956602096558, + "learning_rate": 0.003, + "loss": 3.972, + "step": 26345 + }, + { + "epoch": 0.26346, + "grad_norm": 0.679230272769928, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 26346 + }, + { + "epoch": 0.26347, + "grad_norm": 0.7530626058578491, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 26347 + }, + { + "epoch": 0.26348, + "grad_norm": 0.9411347508430481, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 26348 + }, + { + "epoch": 0.26349, + "grad_norm": 1.1243294477462769, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 26349 + }, + { + "epoch": 0.2635, + "grad_norm": 0.9431162476539612, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 26350 + }, + { + "epoch": 0.26351, + "grad_norm": 0.9512896537780762, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 26351 + }, + { + "epoch": 0.26352, + "grad_norm": 1.0436831712722778, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 26352 + }, + { + "epoch": 0.26353, + "grad_norm": 1.0130467414855957, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 26353 + }, + { + "epoch": 0.26354, + "grad_norm": 0.9621747136116028, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 26354 + }, + { + "epoch": 0.26355, + "grad_norm": 0.8733596801757812, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 26355 + }, + { + "epoch": 0.26356, + "grad_norm": 0.9328529238700867, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 26356 + }, + { + "epoch": 0.26357, + "grad_norm": 0.8870083093643188, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 26357 + }, + { + "epoch": 0.26358, + "grad_norm": 0.8156624436378479, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 26358 + }, + { + "epoch": 0.26359, + "grad_norm": 0.6914650797843933, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 26359 + }, + { + "epoch": 0.2636, + "grad_norm": 0.6692876219749451, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 26360 + }, + { + "epoch": 0.26361, + "grad_norm": 0.6591487526893616, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 26361 + }, + { + "epoch": 0.26362, + "grad_norm": 0.6279633641242981, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 26362 + }, + { + "epoch": 0.26363, + "grad_norm": 0.735101044178009, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 26363 + }, + { + "epoch": 0.26364, + "grad_norm": 0.7852334976196289, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 26364 + }, + { + "epoch": 0.26365, + "grad_norm": 0.8564314842224121, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 26365 + }, + { + "epoch": 0.26366, + "grad_norm": 0.9732944369316101, + "learning_rate": 0.003, + "loss": 3.994, + "step": 26366 + }, + { + "epoch": 0.26367, + "grad_norm": 1.1483113765716553, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 26367 + }, + { + "epoch": 0.26368, + "grad_norm": 1.0346741676330566, + "learning_rate": 0.003, + "loss": 3.99, + "step": 26368 + }, + { + "epoch": 0.26369, + "grad_norm": 0.9876269102096558, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 26369 + }, + { + "epoch": 0.2637, + "grad_norm": 0.9756466150283813, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 26370 + }, + { + "epoch": 0.26371, + "grad_norm": 0.8384610414505005, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 26371 + }, + { + "epoch": 0.26372, + "grad_norm": 0.7655972242355347, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 26372 + }, + { + "epoch": 0.26373, + "grad_norm": 0.7886795997619629, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 26373 + }, + { + "epoch": 0.26374, + "grad_norm": 0.803052544593811, + "learning_rate": 0.003, + "loss": 3.9441, + "step": 26374 + }, + { + "epoch": 0.26375, + "grad_norm": 0.7946504354476929, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 26375 + }, + { + "epoch": 0.26376, + "grad_norm": 0.8543877005577087, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 26376 + }, + { + "epoch": 0.26377, + "grad_norm": 0.97053062915802, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 26377 + }, + { + "epoch": 0.26378, + "grad_norm": 1.0114985704421997, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 26378 + }, + { + "epoch": 0.26379, + "grad_norm": 0.8029046058654785, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 26379 + }, + { + "epoch": 0.2638, + "grad_norm": 0.7792451977729797, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 26380 + }, + { + "epoch": 0.26381, + "grad_norm": 0.839668869972229, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 26381 + }, + { + "epoch": 0.26382, + "grad_norm": 0.8915697336196899, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 26382 + }, + { + "epoch": 0.26383, + "grad_norm": 0.919077455997467, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 26383 + }, + { + "epoch": 0.26384, + "grad_norm": 0.9262064099311829, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 26384 + }, + { + "epoch": 0.26385, + "grad_norm": 0.8409391641616821, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 26385 + }, + { + "epoch": 0.26386, + "grad_norm": 0.8535761833190918, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 26386 + }, + { + "epoch": 0.26387, + "grad_norm": 0.8002924919128418, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 26387 + }, + { + "epoch": 0.26388, + "grad_norm": 0.7471354603767395, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 26388 + }, + { + "epoch": 0.26389, + "grad_norm": 0.6672201752662659, + "learning_rate": 0.003, + "loss": 3.99, + "step": 26389 + }, + { + "epoch": 0.2639, + "grad_norm": 0.6275830864906311, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 26390 + }, + { + "epoch": 0.26391, + "grad_norm": 0.6178247332572937, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 26391 + }, + { + "epoch": 0.26392, + "grad_norm": 0.647232174873352, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 26392 + }, + { + "epoch": 0.26393, + "grad_norm": 0.8555576801300049, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 26393 + }, + { + "epoch": 0.26394, + "grad_norm": 0.9734839200973511, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 26394 + }, + { + "epoch": 0.26395, + "grad_norm": 1.162287950515747, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 26395 + }, + { + "epoch": 0.26396, + "grad_norm": 1.0247159004211426, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 26396 + }, + { + "epoch": 0.26397, + "grad_norm": 0.9998897314071655, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 26397 + }, + { + "epoch": 0.26398, + "grad_norm": 0.9008488059043884, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 26398 + }, + { + "epoch": 0.26399, + "grad_norm": 1.001987099647522, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 26399 + }, + { + "epoch": 0.264, + "grad_norm": 1.1077755689620972, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 26400 + }, + { + "epoch": 0.26401, + "grad_norm": 0.9769701957702637, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 26401 + }, + { + "epoch": 0.26402, + "grad_norm": 0.9223307371139526, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 26402 + }, + { + "epoch": 0.26403, + "grad_norm": 0.8778688311576843, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 26403 + }, + { + "epoch": 0.26404, + "grad_norm": 0.9068320989608765, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 26404 + }, + { + "epoch": 0.26405, + "grad_norm": 0.867405891418457, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 26405 + }, + { + "epoch": 0.26406, + "grad_norm": 1.0178885459899902, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 26406 + }, + { + "epoch": 0.26407, + "grad_norm": 0.8494535684585571, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 26407 + }, + { + "epoch": 0.26408, + "grad_norm": 0.8869815468788147, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 26408 + }, + { + "epoch": 0.26409, + "grad_norm": 0.9095650911331177, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 26409 + }, + { + "epoch": 0.2641, + "grad_norm": 1.0663098096847534, + "learning_rate": 0.003, + "loss": 3.993, + "step": 26410 + }, + { + "epoch": 0.26411, + "grad_norm": 1.209734320640564, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 26411 + }, + { + "epoch": 0.26412, + "grad_norm": 0.7847728729248047, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 26412 + }, + { + "epoch": 0.26413, + "grad_norm": 0.6661299467086792, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 26413 + }, + { + "epoch": 0.26414, + "grad_norm": 0.9087114930152893, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 26414 + }, + { + "epoch": 0.26415, + "grad_norm": 1.0607949495315552, + "learning_rate": 0.003, + "loss": 4.0, + "step": 26415 + }, + { + "epoch": 0.26416, + "grad_norm": 0.867065966129303, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 26416 + }, + { + "epoch": 0.26417, + "grad_norm": 0.6344146132469177, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 26417 + }, + { + "epoch": 0.26418, + "grad_norm": 0.596205472946167, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 26418 + }, + { + "epoch": 0.26419, + "grad_norm": 0.5498960018157959, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 26419 + }, + { + "epoch": 0.2642, + "grad_norm": 0.5119738578796387, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 26420 + }, + { + "epoch": 0.26421, + "grad_norm": 0.6017084717750549, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 26421 + }, + { + "epoch": 0.26422, + "grad_norm": 0.7745648622512817, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 26422 + }, + { + "epoch": 0.26423, + "grad_norm": 0.9962769150733948, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 26423 + }, + { + "epoch": 0.26424, + "grad_norm": 1.1512432098388672, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 26424 + }, + { + "epoch": 0.26425, + "grad_norm": 0.8662477731704712, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 26425 + }, + { + "epoch": 0.26426, + "grad_norm": 0.8654990196228027, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 26426 + }, + { + "epoch": 0.26427, + "grad_norm": 0.8867961764335632, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 26427 + }, + { + "epoch": 0.26428, + "grad_norm": 0.9585190415382385, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 26428 + }, + { + "epoch": 0.26429, + "grad_norm": 0.9314472675323486, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 26429 + }, + { + "epoch": 0.2643, + "grad_norm": 1.109299659729004, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 26430 + }, + { + "epoch": 0.26431, + "grad_norm": 0.9780958890914917, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 26431 + }, + { + "epoch": 0.26432, + "grad_norm": 0.9495893716812134, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 26432 + }, + { + "epoch": 0.26433, + "grad_norm": 0.9403294324874878, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 26433 + }, + { + "epoch": 0.26434, + "grad_norm": 0.9594364762306213, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 26434 + }, + { + "epoch": 0.26435, + "grad_norm": 1.0262436866760254, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 26435 + }, + { + "epoch": 0.26436, + "grad_norm": 0.9119535684585571, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 26436 + }, + { + "epoch": 0.26437, + "grad_norm": 0.7700141668319702, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 26437 + }, + { + "epoch": 0.26438, + "grad_norm": 0.7231053113937378, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 26438 + }, + { + "epoch": 0.26439, + "grad_norm": 0.6933212280273438, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 26439 + }, + { + "epoch": 0.2644, + "grad_norm": 0.8134849667549133, + "learning_rate": 0.003, + "loss": 3.979, + "step": 26440 + }, + { + "epoch": 0.26441, + "grad_norm": 0.9607251882553101, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 26441 + }, + { + "epoch": 0.26442, + "grad_norm": 0.8323318362236023, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 26442 + }, + { + "epoch": 0.26443, + "grad_norm": 0.6596994400024414, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 26443 + }, + { + "epoch": 0.26444, + "grad_norm": 0.6827681064605713, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 26444 + }, + { + "epoch": 0.26445, + "grad_norm": 0.8596805930137634, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 26445 + }, + { + "epoch": 0.26446, + "grad_norm": 1.2196558713912964, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 26446 + }, + { + "epoch": 0.26447, + "grad_norm": 0.7904372215270996, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 26447 + }, + { + "epoch": 0.26448, + "grad_norm": 0.8257365822792053, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 26448 + }, + { + "epoch": 0.26449, + "grad_norm": 1.0550010204315186, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 26449 + }, + { + "epoch": 0.2645, + "grad_norm": 1.0479947328567505, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 26450 + }, + { + "epoch": 0.26451, + "grad_norm": 0.8491488695144653, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 26451 + }, + { + "epoch": 0.26452, + "grad_norm": 0.7498260736465454, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 26452 + }, + { + "epoch": 0.26453, + "grad_norm": 0.8202844262123108, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 26453 + }, + { + "epoch": 0.26454, + "grad_norm": 0.9297610521316528, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 26454 + }, + { + "epoch": 0.26455, + "grad_norm": 0.8806290030479431, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 26455 + }, + { + "epoch": 0.26456, + "grad_norm": 0.7885338664054871, + "learning_rate": 0.003, + "loss": 3.993, + "step": 26456 + }, + { + "epoch": 0.26457, + "grad_norm": 0.7856360077857971, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 26457 + }, + { + "epoch": 0.26458, + "grad_norm": 0.7050808072090149, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 26458 + }, + { + "epoch": 0.26459, + "grad_norm": 0.6673949360847473, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 26459 + }, + { + "epoch": 0.2646, + "grad_norm": 0.6345787644386292, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 26460 + }, + { + "epoch": 0.26461, + "grad_norm": 0.6821427941322327, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 26461 + }, + { + "epoch": 0.26462, + "grad_norm": 0.6267688274383545, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 26462 + }, + { + "epoch": 0.26463, + "grad_norm": 0.6140087246894836, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 26463 + }, + { + "epoch": 0.26464, + "grad_norm": 0.660267174243927, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 26464 + }, + { + "epoch": 0.26465, + "grad_norm": 0.8273265957832336, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 26465 + }, + { + "epoch": 0.26466, + "grad_norm": 1.1865296363830566, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 26466 + }, + { + "epoch": 0.26467, + "grad_norm": 1.0199133157730103, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 26467 + }, + { + "epoch": 0.26468, + "grad_norm": 1.0120984315872192, + "learning_rate": 0.003, + "loss": 3.992, + "step": 26468 + }, + { + "epoch": 0.26469, + "grad_norm": 1.0639883279800415, + "learning_rate": 0.003, + "loss": 3.963, + "step": 26469 + }, + { + "epoch": 0.2647, + "grad_norm": 0.9432227611541748, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 26470 + }, + { + "epoch": 0.26471, + "grad_norm": 0.8450754284858704, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 26471 + }, + { + "epoch": 0.26472, + "grad_norm": 0.6919029951095581, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 26472 + }, + { + "epoch": 0.26473, + "grad_norm": 0.6320092678070068, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 26473 + }, + { + "epoch": 0.26474, + "grad_norm": 0.7342188358306885, + "learning_rate": 0.003, + "loss": 4.008, + "step": 26474 + }, + { + "epoch": 0.26475, + "grad_norm": 0.8626300692558289, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 26475 + }, + { + "epoch": 0.26476, + "grad_norm": 1.067351222038269, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 26476 + }, + { + "epoch": 0.26477, + "grad_norm": 1.1984059810638428, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 26477 + }, + { + "epoch": 0.26478, + "grad_norm": 0.7183151841163635, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 26478 + }, + { + "epoch": 0.26479, + "grad_norm": 0.6163790822029114, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 26479 + }, + { + "epoch": 0.2648, + "grad_norm": 0.5707510709762573, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 26480 + }, + { + "epoch": 0.26481, + "grad_norm": 0.6848487854003906, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 26481 + }, + { + "epoch": 0.26482, + "grad_norm": 0.696398138999939, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 26482 + }, + { + "epoch": 0.26483, + "grad_norm": 0.7499039173126221, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 26483 + }, + { + "epoch": 0.26484, + "grad_norm": 0.8067950010299683, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 26484 + }, + { + "epoch": 0.26485, + "grad_norm": 0.8977285623550415, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 26485 + }, + { + "epoch": 0.26486, + "grad_norm": 1.0539058446884155, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 26486 + }, + { + "epoch": 0.26487, + "grad_norm": 1.0996514558792114, + "learning_rate": 0.003, + "loss": 3.973, + "step": 26487 + }, + { + "epoch": 0.26488, + "grad_norm": 1.0835779905319214, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 26488 + }, + { + "epoch": 0.26489, + "grad_norm": 0.8817256093025208, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 26489 + }, + { + "epoch": 0.2649, + "grad_norm": 0.7925683259963989, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 26490 + }, + { + "epoch": 0.26491, + "grad_norm": 0.8073363304138184, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 26491 + }, + { + "epoch": 0.26492, + "grad_norm": 0.7876495122909546, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 26492 + }, + { + "epoch": 0.26493, + "grad_norm": 0.6958380937576294, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 26493 + }, + { + "epoch": 0.26494, + "grad_norm": 0.6519895792007446, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 26494 + }, + { + "epoch": 0.26495, + "grad_norm": 0.6350019574165344, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 26495 + }, + { + "epoch": 0.26496, + "grad_norm": 0.6607838273048401, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 26496 + }, + { + "epoch": 0.26497, + "grad_norm": 0.8638601303100586, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 26497 + }, + { + "epoch": 0.26498, + "grad_norm": 1.0543031692504883, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 26498 + }, + { + "epoch": 0.26499, + "grad_norm": 1.087007999420166, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 26499 + }, + { + "epoch": 0.265, + "grad_norm": 0.8985573053359985, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 26500 + }, + { + "epoch": 0.26501, + "grad_norm": 0.9086453318595886, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 26501 + }, + { + "epoch": 0.26502, + "grad_norm": 0.9008393287658691, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 26502 + }, + { + "epoch": 0.26503, + "grad_norm": 0.828372597694397, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 26503 + }, + { + "epoch": 0.26504, + "grad_norm": 0.7871405482292175, + "learning_rate": 0.003, + "loss": 3.958, + "step": 26504 + }, + { + "epoch": 0.26505, + "grad_norm": 0.849120020866394, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 26505 + }, + { + "epoch": 0.26506, + "grad_norm": 1.0368517637252808, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 26506 + }, + { + "epoch": 0.26507, + "grad_norm": 1.088584303855896, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 26507 + }, + { + "epoch": 0.26508, + "grad_norm": 0.9591417908668518, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 26508 + }, + { + "epoch": 0.26509, + "grad_norm": 0.8841323852539062, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 26509 + }, + { + "epoch": 0.2651, + "grad_norm": 0.8521095514297485, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 26510 + }, + { + "epoch": 0.26511, + "grad_norm": 1.017820954322815, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 26511 + }, + { + "epoch": 0.26512, + "grad_norm": 0.921698272228241, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 26512 + }, + { + "epoch": 0.26513, + "grad_norm": 0.932417094707489, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 26513 + }, + { + "epoch": 0.26514, + "grad_norm": 1.114182710647583, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 26514 + }, + { + "epoch": 0.26515, + "grad_norm": 1.247523307800293, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 26515 + }, + { + "epoch": 0.26516, + "grad_norm": 0.7800676822662354, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 26516 + }, + { + "epoch": 0.26517, + "grad_norm": 0.7460652589797974, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 26517 + }, + { + "epoch": 0.26518, + "grad_norm": 0.7832653522491455, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 26518 + }, + { + "epoch": 0.26519, + "grad_norm": 0.8060523867607117, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 26519 + }, + { + "epoch": 0.2652, + "grad_norm": 0.9557890892028809, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 26520 + }, + { + "epoch": 0.26521, + "grad_norm": 0.9906416535377502, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 26521 + }, + { + "epoch": 0.26522, + "grad_norm": 1.0174890756607056, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 26522 + }, + { + "epoch": 0.26523, + "grad_norm": 1.1568467617034912, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 26523 + }, + { + "epoch": 0.26524, + "grad_norm": 1.0548182725906372, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 26524 + }, + { + "epoch": 0.26525, + "grad_norm": 1.0439348220825195, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 26525 + }, + { + "epoch": 0.26526, + "grad_norm": 0.8862600326538086, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 26526 + }, + { + "epoch": 0.26527, + "grad_norm": 0.7524644732475281, + "learning_rate": 0.003, + "loss": 4.001, + "step": 26527 + }, + { + "epoch": 0.26528, + "grad_norm": 0.7543812394142151, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 26528 + }, + { + "epoch": 0.26529, + "grad_norm": 0.8100801706314087, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 26529 + }, + { + "epoch": 0.2653, + "grad_norm": 0.7848318815231323, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 26530 + }, + { + "epoch": 0.26531, + "grad_norm": 0.8375314474105835, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 26531 + }, + { + "epoch": 0.26532, + "grad_norm": 0.8764406442642212, + "learning_rate": 0.003, + "loss": 3.979, + "step": 26532 + }, + { + "epoch": 0.26533, + "grad_norm": 0.8510498404502869, + "learning_rate": 0.003, + "loss": 4.008, + "step": 26533 + }, + { + "epoch": 0.26534, + "grad_norm": 0.9211528301239014, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 26534 + }, + { + "epoch": 0.26535, + "grad_norm": 0.867523729801178, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 26535 + }, + { + "epoch": 0.26536, + "grad_norm": 0.9513092041015625, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 26536 + }, + { + "epoch": 0.26537, + "grad_norm": 1.133223533630371, + "learning_rate": 0.003, + "loss": 4.002, + "step": 26537 + }, + { + "epoch": 0.26538, + "grad_norm": 1.0575989484786987, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 26538 + }, + { + "epoch": 0.26539, + "grad_norm": 0.8717256784439087, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 26539 + }, + { + "epoch": 0.2654, + "grad_norm": 0.8113469481468201, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 26540 + }, + { + "epoch": 0.26541, + "grad_norm": 0.8158908486366272, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 26541 + }, + { + "epoch": 0.26542, + "grad_norm": 0.6956957578659058, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 26542 + }, + { + "epoch": 0.26543, + "grad_norm": 0.5493845343589783, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 26543 + }, + { + "epoch": 0.26544, + "grad_norm": 0.5506685376167297, + "learning_rate": 0.003, + "loss": 4.012, + "step": 26544 + }, + { + "epoch": 0.26545, + "grad_norm": 0.5926285982131958, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 26545 + }, + { + "epoch": 0.26546, + "grad_norm": 0.5904402732849121, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 26546 + }, + { + "epoch": 0.26547, + "grad_norm": 0.5075915455818176, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 26547 + }, + { + "epoch": 0.26548, + "grad_norm": 0.5210120677947998, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 26548 + }, + { + "epoch": 0.26549, + "grad_norm": 0.5043071508407593, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 26549 + }, + { + "epoch": 0.2655, + "grad_norm": 0.6163516044616699, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 26550 + }, + { + "epoch": 0.26551, + "grad_norm": 0.8241298198699951, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 26551 + }, + { + "epoch": 0.26552, + "grad_norm": 1.1715075969696045, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 26552 + }, + { + "epoch": 0.26553, + "grad_norm": 1.0114527940750122, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 26553 + }, + { + "epoch": 0.26554, + "grad_norm": 0.9092875719070435, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 26554 + }, + { + "epoch": 0.26555, + "grad_norm": 0.80528724193573, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 26555 + }, + { + "epoch": 0.26556, + "grad_norm": 0.7883594632148743, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 26556 + }, + { + "epoch": 0.26557, + "grad_norm": 0.7474246025085449, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 26557 + }, + { + "epoch": 0.26558, + "grad_norm": 0.8490984439849854, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 26558 + }, + { + "epoch": 0.26559, + "grad_norm": 0.8439042568206787, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 26559 + }, + { + "epoch": 0.2656, + "grad_norm": 0.8373095393180847, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 26560 + }, + { + "epoch": 0.26561, + "grad_norm": 0.8999047875404358, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 26561 + }, + { + "epoch": 0.26562, + "grad_norm": 0.9899620413780212, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 26562 + }, + { + "epoch": 0.26563, + "grad_norm": 1.108747959136963, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 26563 + }, + { + "epoch": 0.26564, + "grad_norm": 0.9010823369026184, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 26564 + }, + { + "epoch": 0.26565, + "grad_norm": 0.9883604645729065, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 26565 + }, + { + "epoch": 0.26566, + "grad_norm": 1.325946569442749, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 26566 + }, + { + "epoch": 0.26567, + "grad_norm": 0.9813773036003113, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 26567 + }, + { + "epoch": 0.26568, + "grad_norm": 0.9627945423126221, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 26568 + }, + { + "epoch": 0.26569, + "grad_norm": 0.944141149520874, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 26569 + }, + { + "epoch": 0.2657, + "grad_norm": 0.9212344884872437, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 26570 + }, + { + "epoch": 0.26571, + "grad_norm": 0.9159365892410278, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 26571 + }, + { + "epoch": 0.26572, + "grad_norm": 0.8671747446060181, + "learning_rate": 0.003, + "loss": 3.977, + "step": 26572 + }, + { + "epoch": 0.26573, + "grad_norm": 0.9012168645858765, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 26573 + }, + { + "epoch": 0.26574, + "grad_norm": 0.9177530407905579, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 26574 + }, + { + "epoch": 0.26575, + "grad_norm": 0.9231523871421814, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 26575 + }, + { + "epoch": 0.26576, + "grad_norm": 0.9149707555770874, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 26576 + }, + { + "epoch": 0.26577, + "grad_norm": 0.9310858845710754, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 26577 + }, + { + "epoch": 0.26578, + "grad_norm": 0.9424949288368225, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 26578 + }, + { + "epoch": 0.26579, + "grad_norm": 0.8869303464889526, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 26579 + }, + { + "epoch": 0.2658, + "grad_norm": 0.8013001680374146, + "learning_rate": 0.003, + "loss": 3.997, + "step": 26580 + }, + { + "epoch": 0.26581, + "grad_norm": 0.7137444615364075, + "learning_rate": 0.003, + "loss": 4.012, + "step": 26581 + }, + { + "epoch": 0.26582, + "grad_norm": 0.6990264058113098, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 26582 + }, + { + "epoch": 0.26583, + "grad_norm": 0.7623013257980347, + "learning_rate": 0.003, + "loss": 4.0, + "step": 26583 + }, + { + "epoch": 0.26584, + "grad_norm": 0.7454833984375, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 26584 + }, + { + "epoch": 0.26585, + "grad_norm": 0.600673258304596, + "learning_rate": 0.003, + "loss": 4.005, + "step": 26585 + }, + { + "epoch": 0.26586, + "grad_norm": 0.6771314144134521, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 26586 + }, + { + "epoch": 0.26587, + "grad_norm": 0.7412838935852051, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 26587 + }, + { + "epoch": 0.26588, + "grad_norm": 0.8380226492881775, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 26588 + }, + { + "epoch": 0.26589, + "grad_norm": 1.2564139366149902, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 26589 + }, + { + "epoch": 0.2659, + "grad_norm": 1.2194620370864868, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 26590 + }, + { + "epoch": 0.26591, + "grad_norm": 0.6755321025848389, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 26591 + }, + { + "epoch": 0.26592, + "grad_norm": 0.6561030745506287, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 26592 + }, + { + "epoch": 0.26593, + "grad_norm": 0.8157505393028259, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 26593 + }, + { + "epoch": 0.26594, + "grad_norm": 1.0473004579544067, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 26594 + }, + { + "epoch": 0.26595, + "grad_norm": 1.1667290925979614, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 26595 + }, + { + "epoch": 0.26596, + "grad_norm": 0.6928238868713379, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 26596 + }, + { + "epoch": 0.26597, + "grad_norm": 0.7913288474082947, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 26597 + }, + { + "epoch": 0.26598, + "grad_norm": 0.8181057572364807, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 26598 + }, + { + "epoch": 0.26599, + "grad_norm": 0.7558661699295044, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 26599 + }, + { + "epoch": 0.266, + "grad_norm": 0.8597836494445801, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 26600 + }, + { + "epoch": 0.26601, + "grad_norm": 0.9852092862129211, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 26601 + }, + { + "epoch": 0.26602, + "grad_norm": 1.0522921085357666, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 26602 + }, + { + "epoch": 0.26603, + "grad_norm": 0.8292137980461121, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 26603 + }, + { + "epoch": 0.26604, + "grad_norm": 0.8421301245689392, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 26604 + }, + { + "epoch": 0.26605, + "grad_norm": 0.8772748708724976, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 26605 + }, + { + "epoch": 0.26606, + "grad_norm": 0.8846122026443481, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 26606 + }, + { + "epoch": 0.26607, + "grad_norm": 0.9800306558609009, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 26607 + }, + { + "epoch": 0.26608, + "grad_norm": 1.0793317556381226, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 26608 + }, + { + "epoch": 0.26609, + "grad_norm": 0.9044927358627319, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 26609 + }, + { + "epoch": 0.2661, + "grad_norm": 0.789547324180603, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 26610 + }, + { + "epoch": 0.26611, + "grad_norm": 0.7369686961174011, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 26611 + }, + { + "epoch": 0.26612, + "grad_norm": 0.7097822427749634, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 26612 + }, + { + "epoch": 0.26613, + "grad_norm": 1.0335359573364258, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 26613 + }, + { + "epoch": 0.26614, + "grad_norm": 1.1728456020355225, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 26614 + }, + { + "epoch": 0.26615, + "grad_norm": 0.7514881491661072, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 26615 + }, + { + "epoch": 0.26616, + "grad_norm": 0.6575021147727966, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 26616 + }, + { + "epoch": 0.26617, + "grad_norm": 0.6885526180267334, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 26617 + }, + { + "epoch": 0.26618, + "grad_norm": 0.7441917657852173, + "learning_rate": 0.003, + "loss": 3.999, + "step": 26618 + }, + { + "epoch": 0.26619, + "grad_norm": 0.8479630351066589, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 26619 + }, + { + "epoch": 0.2662, + "grad_norm": 1.0762484073638916, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 26620 + }, + { + "epoch": 0.26621, + "grad_norm": 1.1570357084274292, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 26621 + }, + { + "epoch": 0.26622, + "grad_norm": 0.7672175168991089, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 26622 + }, + { + "epoch": 0.26623, + "grad_norm": 0.8116321563720703, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 26623 + }, + { + "epoch": 0.26624, + "grad_norm": 0.8070601224899292, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 26624 + }, + { + "epoch": 0.26625, + "grad_norm": 0.8558129668235779, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 26625 + }, + { + "epoch": 0.26626, + "grad_norm": 0.9192191362380981, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 26626 + }, + { + "epoch": 0.26627, + "grad_norm": 0.8810770511627197, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 26627 + }, + { + "epoch": 0.26628, + "grad_norm": 0.8165067434310913, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 26628 + }, + { + "epoch": 0.26629, + "grad_norm": 0.8137277364730835, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 26629 + }, + { + "epoch": 0.2663, + "grad_norm": 0.864626407623291, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 26630 + }, + { + "epoch": 0.26631, + "grad_norm": 1.010443091392517, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 26631 + }, + { + "epoch": 0.26632, + "grad_norm": 1.250076413154602, + "learning_rate": 0.003, + "loss": 4.04, + "step": 26632 + }, + { + "epoch": 0.26633, + "grad_norm": 0.7683765292167664, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 26633 + }, + { + "epoch": 0.26634, + "grad_norm": 0.697792112827301, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 26634 + }, + { + "epoch": 0.26635, + "grad_norm": 0.701888918876648, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 26635 + }, + { + "epoch": 0.26636, + "grad_norm": 0.6172318458557129, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 26636 + }, + { + "epoch": 0.26637, + "grad_norm": 0.6364462375640869, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 26637 + }, + { + "epoch": 0.26638, + "grad_norm": 0.6378108263015747, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 26638 + }, + { + "epoch": 0.26639, + "grad_norm": 0.7824150323867798, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 26639 + }, + { + "epoch": 0.2664, + "grad_norm": 1.016945719718933, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 26640 + }, + { + "epoch": 0.26641, + "grad_norm": 1.164945125579834, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 26641 + }, + { + "epoch": 0.26642, + "grad_norm": 0.8106678128242493, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 26642 + }, + { + "epoch": 0.26643, + "grad_norm": 0.74677574634552, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 26643 + }, + { + "epoch": 0.26644, + "grad_norm": 0.7602241039276123, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 26644 + }, + { + "epoch": 0.26645, + "grad_norm": 0.7173739671707153, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 26645 + }, + { + "epoch": 0.26646, + "grad_norm": 0.849429190158844, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 26646 + }, + { + "epoch": 0.26647, + "grad_norm": 1.0224213600158691, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 26647 + }, + { + "epoch": 0.26648, + "grad_norm": 1.104501724243164, + "learning_rate": 0.003, + "loss": 4.017, + "step": 26648 + }, + { + "epoch": 0.26649, + "grad_norm": 1.0253686904907227, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 26649 + }, + { + "epoch": 0.2665, + "grad_norm": 0.9986797571182251, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 26650 + }, + { + "epoch": 0.26651, + "grad_norm": 1.0513263940811157, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 26651 + }, + { + "epoch": 0.26652, + "grad_norm": 0.780060350894928, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 26652 + }, + { + "epoch": 0.26653, + "grad_norm": 0.7767507433891296, + "learning_rate": 0.003, + "loss": 3.987, + "step": 26653 + }, + { + "epoch": 0.26654, + "grad_norm": 0.9481044411659241, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 26654 + }, + { + "epoch": 0.26655, + "grad_norm": 1.2071799039840698, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 26655 + }, + { + "epoch": 0.26656, + "grad_norm": 0.9685212969779968, + "learning_rate": 0.003, + "loss": 3.998, + "step": 26656 + }, + { + "epoch": 0.26657, + "grad_norm": 0.8923317790031433, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 26657 + }, + { + "epoch": 0.26658, + "grad_norm": 0.8353278040885925, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 26658 + }, + { + "epoch": 0.26659, + "grad_norm": 0.8796067237854004, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 26659 + }, + { + "epoch": 0.2666, + "grad_norm": 0.9195642471313477, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 26660 + }, + { + "epoch": 0.26661, + "grad_norm": 1.1639740467071533, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 26661 + }, + { + "epoch": 0.26662, + "grad_norm": 0.8347406387329102, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 26662 + }, + { + "epoch": 0.26663, + "grad_norm": 0.7930341362953186, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 26663 + }, + { + "epoch": 0.26664, + "grad_norm": 0.7322008013725281, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 26664 + }, + { + "epoch": 0.26665, + "grad_norm": 0.7529243230819702, + "learning_rate": 0.003, + "loss": 3.991, + "step": 26665 + }, + { + "epoch": 0.26666, + "grad_norm": 0.7689900994300842, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 26666 + }, + { + "epoch": 0.26667, + "grad_norm": 0.8936759233474731, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 26667 + }, + { + "epoch": 0.26668, + "grad_norm": 0.9129641056060791, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 26668 + }, + { + "epoch": 0.26669, + "grad_norm": 1.0326377153396606, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 26669 + }, + { + "epoch": 0.2667, + "grad_norm": 1.0939067602157593, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 26670 + }, + { + "epoch": 0.26671, + "grad_norm": 0.9808085560798645, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 26671 + }, + { + "epoch": 0.26672, + "grad_norm": 1.0181254148483276, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 26672 + }, + { + "epoch": 0.26673, + "grad_norm": 0.9154828190803528, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 26673 + }, + { + "epoch": 0.26674, + "grad_norm": 0.8811129927635193, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 26674 + }, + { + "epoch": 0.26675, + "grad_norm": 0.8225769996643066, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 26675 + }, + { + "epoch": 0.26676, + "grad_norm": 0.7715761065483093, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 26676 + }, + { + "epoch": 0.26677, + "grad_norm": 0.7704134583473206, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 26677 + }, + { + "epoch": 0.26678, + "grad_norm": 0.8493550419807434, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 26678 + }, + { + "epoch": 0.26679, + "grad_norm": 0.8250501751899719, + "learning_rate": 0.003, + "loss": 4.006, + "step": 26679 + }, + { + "epoch": 0.2668, + "grad_norm": 0.7438123226165771, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 26680 + }, + { + "epoch": 0.26681, + "grad_norm": 0.6603222489356995, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 26681 + }, + { + "epoch": 0.26682, + "grad_norm": 0.5865319967269897, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 26682 + }, + { + "epoch": 0.26683, + "grad_norm": 0.7140552997589111, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 26683 + }, + { + "epoch": 0.26684, + "grad_norm": 0.7702814936637878, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 26684 + }, + { + "epoch": 0.26685, + "grad_norm": 0.6827938556671143, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 26685 + }, + { + "epoch": 0.26686, + "grad_norm": 0.5899706482887268, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 26686 + }, + { + "epoch": 0.26687, + "grad_norm": 0.9270033240318298, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 26687 + }, + { + "epoch": 0.26688, + "grad_norm": 1.2600271701812744, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 26688 + }, + { + "epoch": 0.26689, + "grad_norm": 1.0068048238754272, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 26689 + }, + { + "epoch": 0.2669, + "grad_norm": 0.9965534210205078, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 26690 + }, + { + "epoch": 0.26691, + "grad_norm": 0.8586801886558533, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 26691 + }, + { + "epoch": 0.26692, + "grad_norm": 0.7489359378814697, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 26692 + }, + { + "epoch": 0.26693, + "grad_norm": 0.8505041599273682, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 26693 + }, + { + "epoch": 0.26694, + "grad_norm": 0.894472062587738, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 26694 + }, + { + "epoch": 0.26695, + "grad_norm": 0.9180401563644409, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 26695 + }, + { + "epoch": 0.26696, + "grad_norm": 0.8566429018974304, + "learning_rate": 0.003, + "loss": 4.007, + "step": 26696 + }, + { + "epoch": 0.26697, + "grad_norm": 0.776520848274231, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 26697 + }, + { + "epoch": 0.26698, + "grad_norm": 0.6564178466796875, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 26698 + }, + { + "epoch": 0.26699, + "grad_norm": 0.6125232577323914, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 26699 + }, + { + "epoch": 0.267, + "grad_norm": 0.605782151222229, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 26700 + }, + { + "epoch": 0.26701, + "grad_norm": 0.6502355933189392, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 26701 + }, + { + "epoch": 0.26702, + "grad_norm": 0.7998926043510437, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 26702 + }, + { + "epoch": 0.26703, + "grad_norm": 0.9836147427558899, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 26703 + }, + { + "epoch": 0.26704, + "grad_norm": 1.2195185422897339, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 26704 + }, + { + "epoch": 0.26705, + "grad_norm": 0.9445772171020508, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 26705 + }, + { + "epoch": 0.26706, + "grad_norm": 0.9498836994171143, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 26706 + }, + { + "epoch": 0.26707, + "grad_norm": 1.0772334337234497, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 26707 + }, + { + "epoch": 0.26708, + "grad_norm": 1.0719894170761108, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 26708 + }, + { + "epoch": 0.26709, + "grad_norm": 0.8797013163566589, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 26709 + }, + { + "epoch": 0.2671, + "grad_norm": 0.9144425392150879, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 26710 + }, + { + "epoch": 0.26711, + "grad_norm": 0.8874399662017822, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 26711 + }, + { + "epoch": 0.26712, + "grad_norm": 0.9173582196235657, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 26712 + }, + { + "epoch": 0.26713, + "grad_norm": 1.1814895868301392, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 26713 + }, + { + "epoch": 0.26714, + "grad_norm": 1.1180397272109985, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 26714 + }, + { + "epoch": 0.26715, + "grad_norm": 0.8776730298995972, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 26715 + }, + { + "epoch": 0.26716, + "grad_norm": 0.6648668646812439, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 26716 + }, + { + "epoch": 0.26717, + "grad_norm": 0.6418883204460144, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 26717 + }, + { + "epoch": 0.26718, + "grad_norm": 0.6972147226333618, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 26718 + }, + { + "epoch": 0.26719, + "grad_norm": 0.7217696905136108, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 26719 + }, + { + "epoch": 0.2672, + "grad_norm": 0.6641680002212524, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 26720 + }, + { + "epoch": 0.26721, + "grad_norm": 0.5991376638412476, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 26721 + }, + { + "epoch": 0.26722, + "grad_norm": 0.6972905993461609, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 26722 + }, + { + "epoch": 0.26723, + "grad_norm": 0.8910608291625977, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 26723 + }, + { + "epoch": 0.26724, + "grad_norm": 1.0783928632736206, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 26724 + }, + { + "epoch": 0.26725, + "grad_norm": 1.0828315019607544, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 26725 + }, + { + "epoch": 0.26726, + "grad_norm": 0.9382603764533997, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 26726 + }, + { + "epoch": 0.26727, + "grad_norm": 0.868273138999939, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 26727 + }, + { + "epoch": 0.26728, + "grad_norm": 0.8580745458602905, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 26728 + }, + { + "epoch": 0.26729, + "grad_norm": 0.9479647278785706, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 26729 + }, + { + "epoch": 0.2673, + "grad_norm": 0.8537881970405579, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 26730 + }, + { + "epoch": 0.26731, + "grad_norm": 0.8383710980415344, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 26731 + }, + { + "epoch": 0.26732, + "grad_norm": 0.8494365811347961, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 26732 + }, + { + "epoch": 0.26733, + "grad_norm": 0.8609455823898315, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 26733 + }, + { + "epoch": 0.26734, + "grad_norm": 0.8556392192840576, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 26734 + }, + { + "epoch": 0.26735, + "grad_norm": 0.8714920282363892, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 26735 + }, + { + "epoch": 0.26736, + "grad_norm": 0.8974188566207886, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 26736 + }, + { + "epoch": 0.26737, + "grad_norm": 0.7693655490875244, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 26737 + }, + { + "epoch": 0.26738, + "grad_norm": 0.7117119431495667, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 26738 + }, + { + "epoch": 0.26739, + "grad_norm": 0.8055451512336731, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 26739 + }, + { + "epoch": 0.2674, + "grad_norm": 1.1087113618850708, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 26740 + }, + { + "epoch": 0.26741, + "grad_norm": 1.1799360513687134, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 26741 + }, + { + "epoch": 0.26742, + "grad_norm": 0.8879600763320923, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 26742 + }, + { + "epoch": 0.26743, + "grad_norm": 0.875536322593689, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 26743 + }, + { + "epoch": 0.26744, + "grad_norm": 0.9812246561050415, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 26744 + }, + { + "epoch": 0.26745, + "grad_norm": 0.9170215129852295, + "learning_rate": 0.003, + "loss": 3.966, + "step": 26745 + }, + { + "epoch": 0.26746, + "grad_norm": 0.858008623123169, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 26746 + }, + { + "epoch": 0.26747, + "grad_norm": 0.806045651435852, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 26747 + }, + { + "epoch": 0.26748, + "grad_norm": 0.8406345844268799, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 26748 + }, + { + "epoch": 0.26749, + "grad_norm": 1.031674861907959, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 26749 + }, + { + "epoch": 0.2675, + "grad_norm": 1.2192113399505615, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 26750 + }, + { + "epoch": 0.26751, + "grad_norm": 0.8070982694625854, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 26751 + }, + { + "epoch": 0.26752, + "grad_norm": 0.6329705119132996, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 26752 + }, + { + "epoch": 0.26753, + "grad_norm": 0.7715604305267334, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 26753 + }, + { + "epoch": 0.26754, + "grad_norm": 0.8061716556549072, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 26754 + }, + { + "epoch": 0.26755, + "grad_norm": 0.7728832960128784, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 26755 + }, + { + "epoch": 0.26756, + "grad_norm": 0.7718921303749084, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 26756 + }, + { + "epoch": 0.26757, + "grad_norm": 0.7994676232337952, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 26757 + }, + { + "epoch": 0.26758, + "grad_norm": 0.7494291067123413, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 26758 + }, + { + "epoch": 0.26759, + "grad_norm": 0.8563376665115356, + "learning_rate": 0.003, + "loss": 4.012, + "step": 26759 + }, + { + "epoch": 0.2676, + "grad_norm": 1.0573689937591553, + "learning_rate": 0.003, + "loss": 3.994, + "step": 26760 + }, + { + "epoch": 0.26761, + "grad_norm": 1.0011420249938965, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 26761 + }, + { + "epoch": 0.26762, + "grad_norm": 0.9281361103057861, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 26762 + }, + { + "epoch": 0.26763, + "grad_norm": 0.9442180395126343, + "learning_rate": 0.003, + "loss": 3.99, + "step": 26763 + }, + { + "epoch": 0.26764, + "grad_norm": 1.0118286609649658, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 26764 + }, + { + "epoch": 0.26765, + "grad_norm": 0.9458659887313843, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 26765 + }, + { + "epoch": 0.26766, + "grad_norm": 0.9904401302337646, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 26766 + }, + { + "epoch": 0.26767, + "grad_norm": 0.9920666217803955, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 26767 + }, + { + "epoch": 0.26768, + "grad_norm": 1.0138250589370728, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 26768 + }, + { + "epoch": 0.26769, + "grad_norm": 1.0258861780166626, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 26769 + }, + { + "epoch": 0.2677, + "grad_norm": 0.9270117878913879, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 26770 + }, + { + "epoch": 0.26771, + "grad_norm": 0.8244357109069824, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 26771 + }, + { + "epoch": 0.26772, + "grad_norm": 0.8674041628837585, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 26772 + }, + { + "epoch": 0.26773, + "grad_norm": 0.940833330154419, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 26773 + }, + { + "epoch": 0.26774, + "grad_norm": 0.9858083724975586, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 26774 + }, + { + "epoch": 0.26775, + "grad_norm": 1.0467134714126587, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 26775 + }, + { + "epoch": 0.26776, + "grad_norm": 1.1193077564239502, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 26776 + }, + { + "epoch": 0.26777, + "grad_norm": 1.018762469291687, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 26777 + }, + { + "epoch": 0.26778, + "grad_norm": 1.131069540977478, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 26778 + }, + { + "epoch": 0.26779, + "grad_norm": 0.8830070495605469, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 26779 + }, + { + "epoch": 0.2678, + "grad_norm": 0.809949517250061, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 26780 + }, + { + "epoch": 0.26781, + "grad_norm": 0.6683459281921387, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 26781 + }, + { + "epoch": 0.26782, + "grad_norm": 0.6698920130729675, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 26782 + }, + { + "epoch": 0.26783, + "grad_norm": 0.7504865527153015, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 26783 + }, + { + "epoch": 0.26784, + "grad_norm": 0.8909552693367004, + "learning_rate": 0.003, + "loss": 3.976, + "step": 26784 + }, + { + "epoch": 0.26785, + "grad_norm": 1.0628223419189453, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 26785 + }, + { + "epoch": 0.26786, + "grad_norm": 1.051287055015564, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 26786 + }, + { + "epoch": 0.26787, + "grad_norm": 1.1323392391204834, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 26787 + }, + { + "epoch": 0.26788, + "grad_norm": 0.8481760621070862, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 26788 + }, + { + "epoch": 0.26789, + "grad_norm": 0.7028201222419739, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 26789 + }, + { + "epoch": 0.2679, + "grad_norm": 0.7769451141357422, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 26790 + }, + { + "epoch": 0.26791, + "grad_norm": 0.8812412619590759, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 26791 + }, + { + "epoch": 0.26792, + "grad_norm": 1.0501583814620972, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 26792 + }, + { + "epoch": 0.26793, + "grad_norm": 1.0075722932815552, + "learning_rate": 0.003, + "loss": 3.981, + "step": 26793 + }, + { + "epoch": 0.26794, + "grad_norm": 0.9740227460861206, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 26794 + }, + { + "epoch": 0.26795, + "grad_norm": 0.9859083294868469, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 26795 + }, + { + "epoch": 0.26796, + "grad_norm": 1.0084415674209595, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 26796 + }, + { + "epoch": 0.26797, + "grad_norm": 0.8562464118003845, + "learning_rate": 0.003, + "loss": 3.999, + "step": 26797 + }, + { + "epoch": 0.26798, + "grad_norm": 0.7321401834487915, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 26798 + }, + { + "epoch": 0.26799, + "grad_norm": 0.6699439287185669, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 26799 + }, + { + "epoch": 0.268, + "grad_norm": 0.6800193786621094, + "learning_rate": 0.003, + "loss": 4.003, + "step": 26800 + }, + { + "epoch": 0.26801, + "grad_norm": 0.6331163048744202, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 26801 + }, + { + "epoch": 0.26802, + "grad_norm": 0.6259260177612305, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 26802 + }, + { + "epoch": 0.26803, + "grad_norm": 0.6989444494247437, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 26803 + }, + { + "epoch": 0.26804, + "grad_norm": 0.6643627882003784, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 26804 + }, + { + "epoch": 0.26805, + "grad_norm": 0.606626570224762, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 26805 + }, + { + "epoch": 0.26806, + "grad_norm": 0.7345000505447388, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 26806 + }, + { + "epoch": 0.26807, + "grad_norm": 0.9186859726905823, + "learning_rate": 0.003, + "loss": 3.975, + "step": 26807 + }, + { + "epoch": 0.26808, + "grad_norm": 0.9697563052177429, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 26808 + }, + { + "epoch": 0.26809, + "grad_norm": 1.0474400520324707, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 26809 + }, + { + "epoch": 0.2681, + "grad_norm": 1.240935206413269, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 26810 + }, + { + "epoch": 0.26811, + "grad_norm": 0.7855817079544067, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 26811 + }, + { + "epoch": 0.26812, + "grad_norm": 0.6030044555664062, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 26812 + }, + { + "epoch": 0.26813, + "grad_norm": 0.7496215105056763, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 26813 + }, + { + "epoch": 0.26814, + "grad_norm": 1.0056443214416504, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 26814 + }, + { + "epoch": 0.26815, + "grad_norm": 0.998940110206604, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 26815 + }, + { + "epoch": 0.26816, + "grad_norm": 0.9617518186569214, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 26816 + }, + { + "epoch": 0.26817, + "grad_norm": 0.8608899712562561, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 26817 + }, + { + "epoch": 0.26818, + "grad_norm": 0.7278012037277222, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 26818 + }, + { + "epoch": 0.26819, + "grad_norm": 0.885563313961029, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 26819 + }, + { + "epoch": 0.2682, + "grad_norm": 1.0107579231262207, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 26820 + }, + { + "epoch": 0.26821, + "grad_norm": 1.1783747673034668, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 26821 + }, + { + "epoch": 0.26822, + "grad_norm": 0.6321266889572144, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 26822 + }, + { + "epoch": 0.26823, + "grad_norm": 0.7427105903625488, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 26823 + }, + { + "epoch": 0.26824, + "grad_norm": 0.9048545956611633, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 26824 + }, + { + "epoch": 0.26825, + "grad_norm": 0.8423759937286377, + "learning_rate": 0.003, + "loss": 4.005, + "step": 26825 + }, + { + "epoch": 0.26826, + "grad_norm": 0.8534355759620667, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 26826 + }, + { + "epoch": 0.26827, + "grad_norm": 0.8399826884269714, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 26827 + }, + { + "epoch": 0.26828, + "grad_norm": 0.7448134422302246, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 26828 + }, + { + "epoch": 0.26829, + "grad_norm": 0.8032739162445068, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 26829 + }, + { + "epoch": 0.2683, + "grad_norm": 0.8899061679840088, + "learning_rate": 0.003, + "loss": 3.98, + "step": 26830 + }, + { + "epoch": 0.26831, + "grad_norm": 0.9580153226852417, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 26831 + }, + { + "epoch": 0.26832, + "grad_norm": 0.9839617013931274, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 26832 + }, + { + "epoch": 0.26833, + "grad_norm": 0.879736065864563, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 26833 + }, + { + "epoch": 0.26834, + "grad_norm": 0.761864185333252, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 26834 + }, + { + "epoch": 0.26835, + "grad_norm": 0.7169963121414185, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 26835 + }, + { + "epoch": 0.26836, + "grad_norm": 0.7252517342567444, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 26836 + }, + { + "epoch": 0.26837, + "grad_norm": 0.8372198939323425, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 26837 + }, + { + "epoch": 0.26838, + "grad_norm": 1.0471982955932617, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 26838 + }, + { + "epoch": 0.26839, + "grad_norm": 1.123673677444458, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 26839 + }, + { + "epoch": 0.2684, + "grad_norm": 0.99821537733078, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 26840 + }, + { + "epoch": 0.26841, + "grad_norm": 1.0857452154159546, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 26841 + }, + { + "epoch": 0.26842, + "grad_norm": 1.0824800729751587, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 26842 + }, + { + "epoch": 0.26843, + "grad_norm": 0.8289909958839417, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 26843 + }, + { + "epoch": 0.26844, + "grad_norm": 0.7467186450958252, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 26844 + }, + { + "epoch": 0.26845, + "grad_norm": 0.7229576706886292, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 26845 + }, + { + "epoch": 0.26846, + "grad_norm": 0.7063528895378113, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 26846 + }, + { + "epoch": 0.26847, + "grad_norm": 0.6613720655441284, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 26847 + }, + { + "epoch": 0.26848, + "grad_norm": 0.6699565052986145, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 26848 + }, + { + "epoch": 0.26849, + "grad_norm": 0.7581509351730347, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 26849 + }, + { + "epoch": 0.2685, + "grad_norm": 1.0681291818618774, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 26850 + }, + { + "epoch": 0.26851, + "grad_norm": 1.5306317806243896, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 26851 + }, + { + "epoch": 0.26852, + "grad_norm": 0.5006439685821533, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 26852 + }, + { + "epoch": 0.26853, + "grad_norm": 1.1453466415405273, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 26853 + }, + { + "epoch": 0.26854, + "grad_norm": 1.083205223083496, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 26854 + }, + { + "epoch": 0.26855, + "grad_norm": 0.7569419145584106, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 26855 + }, + { + "epoch": 0.26856, + "grad_norm": 0.6247507333755493, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 26856 + }, + { + "epoch": 0.26857, + "grad_norm": 0.544352114200592, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 26857 + }, + { + "epoch": 0.26858, + "grad_norm": 0.5798980593681335, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 26858 + }, + { + "epoch": 0.26859, + "grad_norm": 0.6364675164222717, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 26859 + }, + { + "epoch": 0.2686, + "grad_norm": 0.6833467483520508, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 26860 + }, + { + "epoch": 0.26861, + "grad_norm": 0.7785399556159973, + "learning_rate": 0.003, + "loss": 3.963, + "step": 26861 + }, + { + "epoch": 0.26862, + "grad_norm": 0.8291352987289429, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 26862 + }, + { + "epoch": 0.26863, + "grad_norm": 0.8500729203224182, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 26863 + }, + { + "epoch": 0.26864, + "grad_norm": 0.894162654876709, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 26864 + }, + { + "epoch": 0.26865, + "grad_norm": 1.182952880859375, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 26865 + }, + { + "epoch": 0.26866, + "grad_norm": 1.0060524940490723, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 26866 + }, + { + "epoch": 0.26867, + "grad_norm": 0.8175178170204163, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 26867 + }, + { + "epoch": 0.26868, + "grad_norm": 0.8039507865905762, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 26868 + }, + { + "epoch": 0.26869, + "grad_norm": 0.8884865641593933, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 26869 + }, + { + "epoch": 0.2687, + "grad_norm": 0.9426500201225281, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 26870 + }, + { + "epoch": 0.26871, + "grad_norm": 0.9491475224494934, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 26871 + }, + { + "epoch": 0.26872, + "grad_norm": 0.9801953434944153, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 26872 + }, + { + "epoch": 0.26873, + "grad_norm": 0.9668394923210144, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 26873 + }, + { + "epoch": 0.26874, + "grad_norm": 0.7871923446655273, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 26874 + }, + { + "epoch": 0.26875, + "grad_norm": 0.7934064269065857, + "learning_rate": 0.003, + "loss": 3.985, + "step": 26875 + }, + { + "epoch": 0.26876, + "grad_norm": 0.8205674290657043, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 26876 + }, + { + "epoch": 0.26877, + "grad_norm": 0.8672219514846802, + "learning_rate": 0.003, + "loss": 4.005, + "step": 26877 + }, + { + "epoch": 0.26878, + "grad_norm": 0.8086318373680115, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 26878 + }, + { + "epoch": 0.26879, + "grad_norm": 0.8868533968925476, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 26879 + }, + { + "epoch": 0.2688, + "grad_norm": 0.9758858680725098, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 26880 + }, + { + "epoch": 0.26881, + "grad_norm": 0.9611081480979919, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 26881 + }, + { + "epoch": 0.26882, + "grad_norm": 1.086177110671997, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 26882 + }, + { + "epoch": 0.26883, + "grad_norm": 1.191046118736267, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 26883 + }, + { + "epoch": 0.26884, + "grad_norm": 0.8313100934028625, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 26884 + }, + { + "epoch": 0.26885, + "grad_norm": 0.7404287457466125, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 26885 + }, + { + "epoch": 0.26886, + "grad_norm": 0.6126376986503601, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 26886 + }, + { + "epoch": 0.26887, + "grad_norm": 0.6593831181526184, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 26887 + }, + { + "epoch": 0.26888, + "grad_norm": 0.7050580978393555, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 26888 + }, + { + "epoch": 0.26889, + "grad_norm": 0.8946563005447388, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 26889 + }, + { + "epoch": 0.2689, + "grad_norm": 1.1619454622268677, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 26890 + }, + { + "epoch": 0.26891, + "grad_norm": 0.9593244194984436, + "learning_rate": 0.003, + "loss": 3.995, + "step": 26891 + }, + { + "epoch": 0.26892, + "grad_norm": 1.0426104068756104, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 26892 + }, + { + "epoch": 0.26893, + "grad_norm": 1.0251026153564453, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 26893 + }, + { + "epoch": 0.26894, + "grad_norm": 0.7339516282081604, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 26894 + }, + { + "epoch": 0.26895, + "grad_norm": 0.6703333854675293, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 26895 + }, + { + "epoch": 0.26896, + "grad_norm": 0.6508586406707764, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 26896 + }, + { + "epoch": 0.26897, + "grad_norm": 0.6395254731178284, + "learning_rate": 0.003, + "loss": 4.0, + "step": 26897 + }, + { + "epoch": 0.26898, + "grad_norm": 0.789559543132782, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 26898 + }, + { + "epoch": 0.26899, + "grad_norm": 0.8113242387771606, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 26899 + }, + { + "epoch": 0.269, + "grad_norm": 0.8799034953117371, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 26900 + }, + { + "epoch": 0.26901, + "grad_norm": 1.0058872699737549, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 26901 + }, + { + "epoch": 0.26902, + "grad_norm": 0.9480020999908447, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 26902 + }, + { + "epoch": 0.26903, + "grad_norm": 1.2013781070709229, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 26903 + }, + { + "epoch": 0.26904, + "grad_norm": 0.9146543145179749, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 26904 + }, + { + "epoch": 0.26905, + "grad_norm": 1.086599588394165, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 26905 + }, + { + "epoch": 0.26906, + "grad_norm": 1.0874896049499512, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 26906 + }, + { + "epoch": 0.26907, + "grad_norm": 0.9202030301094055, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 26907 + }, + { + "epoch": 0.26908, + "grad_norm": 0.8318597078323364, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 26908 + }, + { + "epoch": 0.26909, + "grad_norm": 0.7929989099502563, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 26909 + }, + { + "epoch": 0.2691, + "grad_norm": 0.7963500618934631, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 26910 + }, + { + "epoch": 0.26911, + "grad_norm": 0.6995571255683899, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 26911 + }, + { + "epoch": 0.26912, + "grad_norm": 0.5849372148513794, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 26912 + }, + { + "epoch": 0.26913, + "grad_norm": 0.5557720065116882, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 26913 + }, + { + "epoch": 0.26914, + "grad_norm": 0.5734115839004517, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 26914 + }, + { + "epoch": 0.26915, + "grad_norm": 0.6619975566864014, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 26915 + }, + { + "epoch": 0.26916, + "grad_norm": 0.6793997883796692, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 26916 + }, + { + "epoch": 0.26917, + "grad_norm": 0.6831039786338806, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 26917 + }, + { + "epoch": 0.26918, + "grad_norm": 0.8539472222328186, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 26918 + }, + { + "epoch": 0.26919, + "grad_norm": 1.3271632194519043, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 26919 + }, + { + "epoch": 0.2692, + "grad_norm": 0.9335992336273193, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 26920 + }, + { + "epoch": 0.26921, + "grad_norm": 0.6899843215942383, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 26921 + }, + { + "epoch": 0.26922, + "grad_norm": 0.5932106971740723, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 26922 + }, + { + "epoch": 0.26923, + "grad_norm": 0.7297183871269226, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 26923 + }, + { + "epoch": 0.26924, + "grad_norm": 0.7606067657470703, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 26924 + }, + { + "epoch": 0.26925, + "grad_norm": 0.7986975908279419, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 26925 + }, + { + "epoch": 0.26926, + "grad_norm": 0.9611266851425171, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 26926 + }, + { + "epoch": 0.26927, + "grad_norm": 1.0915277004241943, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 26927 + }, + { + "epoch": 0.26928, + "grad_norm": 0.8452636003494263, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 26928 + }, + { + "epoch": 0.26929, + "grad_norm": 0.6491087675094604, + "learning_rate": 0.003, + "loss": 3.998, + "step": 26929 + }, + { + "epoch": 0.2693, + "grad_norm": 0.706055760383606, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 26930 + }, + { + "epoch": 0.26931, + "grad_norm": 0.7248590588569641, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 26931 + }, + { + "epoch": 0.26932, + "grad_norm": 0.7218074798583984, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 26932 + }, + { + "epoch": 0.26933, + "grad_norm": 0.7340793609619141, + "learning_rate": 0.003, + "loss": 3.967, + "step": 26933 + }, + { + "epoch": 0.26934, + "grad_norm": 0.9130221009254456, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 26934 + }, + { + "epoch": 0.26935, + "grad_norm": 1.2785820960998535, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 26935 + }, + { + "epoch": 0.26936, + "grad_norm": 0.956989586353302, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 26936 + }, + { + "epoch": 0.26937, + "grad_norm": 0.8088443875312805, + "learning_rate": 0.003, + "loss": 3.988, + "step": 26937 + }, + { + "epoch": 0.26938, + "grad_norm": 0.7912686467170715, + "learning_rate": 0.003, + "loss": 3.976, + "step": 26938 + }, + { + "epoch": 0.26939, + "grad_norm": 0.7353087067604065, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 26939 + }, + { + "epoch": 0.2694, + "grad_norm": 0.7239277362823486, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 26940 + }, + { + "epoch": 0.26941, + "grad_norm": 0.9445987939834595, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 26941 + }, + { + "epoch": 0.26942, + "grad_norm": 1.0276241302490234, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 26942 + }, + { + "epoch": 0.26943, + "grad_norm": 0.9777191281318665, + "learning_rate": 0.003, + "loss": 3.989, + "step": 26943 + }, + { + "epoch": 0.26944, + "grad_norm": 0.9267406463623047, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 26944 + }, + { + "epoch": 0.26945, + "grad_norm": 0.8610537052154541, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 26945 + }, + { + "epoch": 0.26946, + "grad_norm": 0.8419165015220642, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 26946 + }, + { + "epoch": 0.26947, + "grad_norm": 0.904140055179596, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 26947 + }, + { + "epoch": 0.26948, + "grad_norm": 0.9626393914222717, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 26948 + }, + { + "epoch": 0.26949, + "grad_norm": 1.0379447937011719, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 26949 + }, + { + "epoch": 0.2695, + "grad_norm": 0.8856220841407776, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 26950 + }, + { + "epoch": 0.26951, + "grad_norm": 0.848025918006897, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 26951 + }, + { + "epoch": 0.26952, + "grad_norm": 0.9671321511268616, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 26952 + }, + { + "epoch": 0.26953, + "grad_norm": 0.9879338145256042, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 26953 + }, + { + "epoch": 0.26954, + "grad_norm": 1.036647081375122, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 26954 + }, + { + "epoch": 0.26955, + "grad_norm": 1.0067524909973145, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 26955 + }, + { + "epoch": 0.26956, + "grad_norm": 0.9981648921966553, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 26956 + }, + { + "epoch": 0.26957, + "grad_norm": 0.9263343214988708, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 26957 + }, + { + "epoch": 0.26958, + "grad_norm": 0.9223698377609253, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 26958 + }, + { + "epoch": 0.26959, + "grad_norm": 1.1469440460205078, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 26959 + }, + { + "epoch": 0.2696, + "grad_norm": 1.2300925254821777, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 26960 + }, + { + "epoch": 0.26961, + "grad_norm": 0.7712873816490173, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 26961 + }, + { + "epoch": 0.26962, + "grad_norm": 0.7557539939880371, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 26962 + }, + { + "epoch": 0.26963, + "grad_norm": 0.736118733882904, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 26963 + }, + { + "epoch": 0.26964, + "grad_norm": 0.7034887075424194, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 26964 + }, + { + "epoch": 0.26965, + "grad_norm": 0.7543181777000427, + "learning_rate": 0.003, + "loss": 3.987, + "step": 26965 + }, + { + "epoch": 0.26966, + "grad_norm": 0.7381757497787476, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 26966 + }, + { + "epoch": 0.26967, + "grad_norm": 0.8954094052314758, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 26967 + }, + { + "epoch": 0.26968, + "grad_norm": 0.9873430728912354, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 26968 + }, + { + "epoch": 0.26969, + "grad_norm": 1.23573637008667, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 26969 + }, + { + "epoch": 0.2697, + "grad_norm": 0.7933611273765564, + "learning_rate": 0.003, + "loss": 3.981, + "step": 26970 + }, + { + "epoch": 0.26971, + "grad_norm": 0.7668277025222778, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 26971 + }, + { + "epoch": 0.26972, + "grad_norm": 0.725156307220459, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 26972 + }, + { + "epoch": 0.26973, + "grad_norm": 0.7790424227714539, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 26973 + }, + { + "epoch": 0.26974, + "grad_norm": 0.8758139610290527, + "learning_rate": 0.003, + "loss": 3.983, + "step": 26974 + }, + { + "epoch": 0.26975, + "grad_norm": 1.0123920440673828, + "learning_rate": 0.003, + "loss": 3.978, + "step": 26975 + }, + { + "epoch": 0.26976, + "grad_norm": 1.1889832019805908, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 26976 + }, + { + "epoch": 0.26977, + "grad_norm": 0.7493115067481995, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 26977 + }, + { + "epoch": 0.26978, + "grad_norm": 0.6523833870887756, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 26978 + }, + { + "epoch": 0.26979, + "grad_norm": 0.6388462781906128, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 26979 + }, + { + "epoch": 0.2698, + "grad_norm": 0.710640013217926, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 26980 + }, + { + "epoch": 0.26981, + "grad_norm": 0.7177200317382812, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 26981 + }, + { + "epoch": 0.26982, + "grad_norm": 0.7387577891349792, + "learning_rate": 0.003, + "loss": 4.007, + "step": 26982 + }, + { + "epoch": 0.26983, + "grad_norm": 0.7730430960655212, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 26983 + }, + { + "epoch": 0.26984, + "grad_norm": 0.6620581150054932, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 26984 + }, + { + "epoch": 0.26985, + "grad_norm": 0.6303138136863708, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 26985 + }, + { + "epoch": 0.26986, + "grad_norm": 0.6939755082130432, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 26986 + }, + { + "epoch": 0.26987, + "grad_norm": 0.8214789628982544, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 26987 + }, + { + "epoch": 0.26988, + "grad_norm": 1.014513373374939, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 26988 + }, + { + "epoch": 0.26989, + "grad_norm": 1.1641255617141724, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 26989 + }, + { + "epoch": 0.2699, + "grad_norm": 0.9508329033851624, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 26990 + }, + { + "epoch": 0.26991, + "grad_norm": 0.9357451796531677, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 26991 + }, + { + "epoch": 0.26992, + "grad_norm": 0.9069871306419373, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 26992 + }, + { + "epoch": 0.26993, + "grad_norm": 0.7724646925926208, + "learning_rate": 0.003, + "loss": 3.989, + "step": 26993 + }, + { + "epoch": 0.26994, + "grad_norm": 0.6176658868789673, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 26994 + }, + { + "epoch": 0.26995, + "grad_norm": 0.5527603626251221, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 26995 + }, + { + "epoch": 0.26996, + "grad_norm": 0.5936351418495178, + "learning_rate": 0.003, + "loss": 3.975, + "step": 26996 + }, + { + "epoch": 0.26997, + "grad_norm": 0.6627873182296753, + "learning_rate": 0.003, + "loss": 3.979, + "step": 26997 + }, + { + "epoch": 0.26998, + "grad_norm": 0.7797447443008423, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 26998 + }, + { + "epoch": 0.26999, + "grad_norm": 0.8486611247062683, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 26999 + }, + { + "epoch": 0.27, + "grad_norm": 0.8138474822044373, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 27000 + }, + { + "epoch": 0.27001, + "grad_norm": 0.7168936133384705, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 27001 + }, + { + "epoch": 0.27002, + "grad_norm": 0.611851692199707, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 27002 + }, + { + "epoch": 0.27003, + "grad_norm": 0.570483922958374, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 27003 + }, + { + "epoch": 0.27004, + "grad_norm": 0.5827093124389648, + "learning_rate": 0.003, + "loss": 3.9474, + "step": 27004 + }, + { + "epoch": 0.27005, + "grad_norm": 0.605319619178772, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 27005 + }, + { + "epoch": 0.27006, + "grad_norm": 0.7284801602363586, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 27006 + }, + { + "epoch": 0.27007, + "grad_norm": 0.919279158115387, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 27007 + }, + { + "epoch": 0.27008, + "grad_norm": 1.0125707387924194, + "learning_rate": 0.003, + "loss": 4.011, + "step": 27008 + }, + { + "epoch": 0.27009, + "grad_norm": 1.1423592567443848, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 27009 + }, + { + "epoch": 0.2701, + "grad_norm": 0.9353553056716919, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 27010 + }, + { + "epoch": 0.27011, + "grad_norm": 1.015319585800171, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 27011 + }, + { + "epoch": 0.27012, + "grad_norm": 1.080467700958252, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 27012 + }, + { + "epoch": 0.27013, + "grad_norm": 0.970727264881134, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 27013 + }, + { + "epoch": 0.27014, + "grad_norm": 1.0511382818222046, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 27014 + }, + { + "epoch": 0.27015, + "grad_norm": 1.2037031650543213, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 27015 + }, + { + "epoch": 0.27016, + "grad_norm": 0.9713382124900818, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 27016 + }, + { + "epoch": 0.27017, + "grad_norm": 0.9327167272567749, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 27017 + }, + { + "epoch": 0.27018, + "grad_norm": 1.0313044786453247, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 27018 + }, + { + "epoch": 0.27019, + "grad_norm": 0.992304801940918, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 27019 + }, + { + "epoch": 0.2702, + "grad_norm": 0.8205800652503967, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 27020 + }, + { + "epoch": 0.27021, + "grad_norm": 0.8233579397201538, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 27021 + }, + { + "epoch": 0.27022, + "grad_norm": 0.7800913453102112, + "learning_rate": 0.003, + "loss": 3.97, + "step": 27022 + }, + { + "epoch": 0.27023, + "grad_norm": 0.7496599555015564, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 27023 + }, + { + "epoch": 0.27024, + "grad_norm": 0.7044620513916016, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 27024 + }, + { + "epoch": 0.27025, + "grad_norm": 0.6980661749839783, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 27025 + }, + { + "epoch": 0.27026, + "grad_norm": 0.9149898290634155, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 27026 + }, + { + "epoch": 0.27027, + "grad_norm": 1.2622218132019043, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 27027 + }, + { + "epoch": 0.27028, + "grad_norm": 0.8074148893356323, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 27028 + }, + { + "epoch": 0.27029, + "grad_norm": 0.695311963558197, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 27029 + }, + { + "epoch": 0.2703, + "grad_norm": 0.7994462847709656, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 27030 + }, + { + "epoch": 0.27031, + "grad_norm": 0.9340416193008423, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 27031 + }, + { + "epoch": 0.27032, + "grad_norm": 1.1298831701278687, + "learning_rate": 0.003, + "loss": 3.996, + "step": 27032 + }, + { + "epoch": 0.27033, + "grad_norm": 1.0094680786132812, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 27033 + }, + { + "epoch": 0.27034, + "grad_norm": 1.123949646949768, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 27034 + }, + { + "epoch": 0.27035, + "grad_norm": 0.7015727758407593, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 27035 + }, + { + "epoch": 0.27036, + "grad_norm": 0.7106392979621887, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 27036 + }, + { + "epoch": 0.27037, + "grad_norm": 0.6487913727760315, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 27037 + }, + { + "epoch": 0.27038, + "grad_norm": 0.7397325038909912, + "learning_rate": 0.003, + "loss": 3.983, + "step": 27038 + }, + { + "epoch": 0.27039, + "grad_norm": 0.972796618938446, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 27039 + }, + { + "epoch": 0.2704, + "grad_norm": 1.213052749633789, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 27040 + }, + { + "epoch": 0.27041, + "grad_norm": 0.6418881416320801, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 27041 + }, + { + "epoch": 0.27042, + "grad_norm": 0.764031708240509, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 27042 + }, + { + "epoch": 0.27043, + "grad_norm": 0.7654557228088379, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 27043 + }, + { + "epoch": 0.27044, + "grad_norm": 0.8657233119010925, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 27044 + }, + { + "epoch": 0.27045, + "grad_norm": 0.8698302507400513, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 27045 + }, + { + "epoch": 0.27046, + "grad_norm": 0.8358195424079895, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 27046 + }, + { + "epoch": 0.27047, + "grad_norm": 0.8536239862442017, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 27047 + }, + { + "epoch": 0.27048, + "grad_norm": 0.8509815335273743, + "learning_rate": 0.003, + "loss": 3.94, + "step": 27048 + }, + { + "epoch": 0.27049, + "grad_norm": 1.0559519529342651, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 27049 + }, + { + "epoch": 0.2705, + "grad_norm": 0.9866148233413696, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 27050 + }, + { + "epoch": 0.27051, + "grad_norm": 0.8380328416824341, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 27051 + }, + { + "epoch": 0.27052, + "grad_norm": 0.7213466167449951, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 27052 + }, + { + "epoch": 0.27053, + "grad_norm": 0.74525386095047, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 27053 + }, + { + "epoch": 0.27054, + "grad_norm": 0.7096536755561829, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 27054 + }, + { + "epoch": 0.27055, + "grad_norm": 0.9088784456253052, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 27055 + }, + { + "epoch": 0.27056, + "grad_norm": 1.1989259719848633, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 27056 + }, + { + "epoch": 0.27057, + "grad_norm": 0.8783380389213562, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 27057 + }, + { + "epoch": 0.27058, + "grad_norm": 0.8699001669883728, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 27058 + }, + { + "epoch": 0.27059, + "grad_norm": 0.8570318818092346, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 27059 + }, + { + "epoch": 0.2706, + "grad_norm": 0.983121395111084, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 27060 + }, + { + "epoch": 0.27061, + "grad_norm": 1.0251221656799316, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 27061 + }, + { + "epoch": 0.27062, + "grad_norm": 1.039657711982727, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 27062 + }, + { + "epoch": 0.27063, + "grad_norm": 0.9121429920196533, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 27063 + }, + { + "epoch": 0.27064, + "grad_norm": 0.9213479161262512, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 27064 + }, + { + "epoch": 0.27065, + "grad_norm": 0.9955871105194092, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 27065 + }, + { + "epoch": 0.27066, + "grad_norm": 0.8172794580459595, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 27066 + }, + { + "epoch": 0.27067, + "grad_norm": 0.7169832587242126, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 27067 + }, + { + "epoch": 0.27068, + "grad_norm": 0.7305949926376343, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 27068 + }, + { + "epoch": 0.27069, + "grad_norm": 0.7135331034660339, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 27069 + }, + { + "epoch": 0.2707, + "grad_norm": 0.7542229294776917, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 27070 + }, + { + "epoch": 0.27071, + "grad_norm": 0.8077606558799744, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 27071 + }, + { + "epoch": 0.27072, + "grad_norm": 0.8102840781211853, + "learning_rate": 0.003, + "loss": 3.977, + "step": 27072 + }, + { + "epoch": 0.27073, + "grad_norm": 0.9839593172073364, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 27073 + }, + { + "epoch": 0.27074, + "grad_norm": 1.0182253122329712, + "learning_rate": 0.003, + "loss": 3.99, + "step": 27074 + }, + { + "epoch": 0.27075, + "grad_norm": 1.2457844018936157, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 27075 + }, + { + "epoch": 0.27076, + "grad_norm": 0.7913490533828735, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 27076 + }, + { + "epoch": 0.27077, + "grad_norm": 0.6804322600364685, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 27077 + }, + { + "epoch": 0.27078, + "grad_norm": 0.7987557649612427, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 27078 + }, + { + "epoch": 0.27079, + "grad_norm": 0.8166942596435547, + "learning_rate": 0.003, + "loss": 4.019, + "step": 27079 + }, + { + "epoch": 0.2708, + "grad_norm": 0.8048427700996399, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 27080 + }, + { + "epoch": 0.27081, + "grad_norm": 0.7861166596412659, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 27081 + }, + { + "epoch": 0.27082, + "grad_norm": 0.8630853891372681, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 27082 + }, + { + "epoch": 0.27083, + "grad_norm": 0.92915940284729, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 27083 + }, + { + "epoch": 0.27084, + "grad_norm": 1.2619038820266724, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 27084 + }, + { + "epoch": 0.27085, + "grad_norm": 0.9356670379638672, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 27085 + }, + { + "epoch": 0.27086, + "grad_norm": 0.9232414960861206, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 27086 + }, + { + "epoch": 0.27087, + "grad_norm": 0.8015287518501282, + "learning_rate": 0.003, + "loss": 3.993, + "step": 27087 + }, + { + "epoch": 0.27088, + "grad_norm": 0.9218838214874268, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 27088 + }, + { + "epoch": 0.27089, + "grad_norm": 0.8363850116729736, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 27089 + }, + { + "epoch": 0.2709, + "grad_norm": 0.7275590300559998, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 27090 + }, + { + "epoch": 0.27091, + "grad_norm": 0.6608415246009827, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 27091 + }, + { + "epoch": 0.27092, + "grad_norm": 0.653886079788208, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 27092 + }, + { + "epoch": 0.27093, + "grad_norm": 0.6315980553627014, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 27093 + }, + { + "epoch": 0.27094, + "grad_norm": 0.6968366503715515, + "learning_rate": 0.003, + "loss": 3.97, + "step": 27094 + }, + { + "epoch": 0.27095, + "grad_norm": 0.8643034100532532, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 27095 + }, + { + "epoch": 0.27096, + "grad_norm": 1.0968397855758667, + "learning_rate": 0.003, + "loss": 3.988, + "step": 27096 + }, + { + "epoch": 0.27097, + "grad_norm": 1.2314077615737915, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 27097 + }, + { + "epoch": 0.27098, + "grad_norm": 0.889805018901825, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 27098 + }, + { + "epoch": 0.27099, + "grad_norm": 0.8302732110023499, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 27099 + }, + { + "epoch": 0.271, + "grad_norm": 0.6643983721733093, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 27100 + }, + { + "epoch": 0.27101, + "grad_norm": 0.6238534450531006, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 27101 + }, + { + "epoch": 0.27102, + "grad_norm": 0.7224483489990234, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 27102 + }, + { + "epoch": 0.27103, + "grad_norm": 0.7820181846618652, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 27103 + }, + { + "epoch": 0.27104, + "grad_norm": 0.8623741269111633, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 27104 + }, + { + "epoch": 0.27105, + "grad_norm": 1.1277861595153809, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 27105 + }, + { + "epoch": 0.27106, + "grad_norm": 1.118759036064148, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 27106 + }, + { + "epoch": 0.27107, + "grad_norm": 0.774941086769104, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 27107 + }, + { + "epoch": 0.27108, + "grad_norm": 0.782593309879303, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 27108 + }, + { + "epoch": 0.27109, + "grad_norm": 0.8423106074333191, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 27109 + }, + { + "epoch": 0.2711, + "grad_norm": 0.8127303719520569, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 27110 + }, + { + "epoch": 0.27111, + "grad_norm": 1.0027467012405396, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 27111 + }, + { + "epoch": 0.27112, + "grad_norm": 1.0094614028930664, + "learning_rate": 0.003, + "loss": 3.975, + "step": 27112 + }, + { + "epoch": 0.27113, + "grad_norm": 0.7967077493667603, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 27113 + }, + { + "epoch": 0.27114, + "grad_norm": 0.699492335319519, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 27114 + }, + { + "epoch": 0.27115, + "grad_norm": 0.7075315713882446, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 27115 + }, + { + "epoch": 0.27116, + "grad_norm": 0.6354879140853882, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 27116 + }, + { + "epoch": 0.27117, + "grad_norm": 0.5803404450416565, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 27117 + }, + { + "epoch": 0.27118, + "grad_norm": 0.6277279257774353, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 27118 + }, + { + "epoch": 0.27119, + "grad_norm": 0.7396233677864075, + "learning_rate": 0.003, + "loss": 3.99, + "step": 27119 + }, + { + "epoch": 0.2712, + "grad_norm": 0.947019636631012, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 27120 + }, + { + "epoch": 0.27121, + "grad_norm": 1.16558837890625, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 27121 + }, + { + "epoch": 0.27122, + "grad_norm": 0.8131913542747498, + "learning_rate": 0.003, + "loss": 3.976, + "step": 27122 + }, + { + "epoch": 0.27123, + "grad_norm": 0.9423155784606934, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 27123 + }, + { + "epoch": 0.27124, + "grad_norm": 1.0810471773147583, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 27124 + }, + { + "epoch": 0.27125, + "grad_norm": 1.2719948291778564, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 27125 + }, + { + "epoch": 0.27126, + "grad_norm": 0.7827710509300232, + "learning_rate": 0.003, + "loss": 3.976, + "step": 27126 + }, + { + "epoch": 0.27127, + "grad_norm": 0.8469380736351013, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 27127 + }, + { + "epoch": 0.27128, + "grad_norm": 0.8559473156929016, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 27128 + }, + { + "epoch": 0.27129, + "grad_norm": 0.893719494342804, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 27129 + }, + { + "epoch": 0.2713, + "grad_norm": 0.9191077351570129, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 27130 + }, + { + "epoch": 0.27131, + "grad_norm": 0.9912793040275574, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 27131 + }, + { + "epoch": 0.27132, + "grad_norm": 1.1225228309631348, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 27132 + }, + { + "epoch": 0.27133, + "grad_norm": 0.8143816590309143, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 27133 + }, + { + "epoch": 0.27134, + "grad_norm": 0.739786684513092, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 27134 + }, + { + "epoch": 0.27135, + "grad_norm": 0.7543268203735352, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 27135 + }, + { + "epoch": 0.27136, + "grad_norm": 1.0323151350021362, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 27136 + }, + { + "epoch": 0.27137, + "grad_norm": 1.3026156425476074, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 27137 + }, + { + "epoch": 0.27138, + "grad_norm": 0.7119749188423157, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 27138 + }, + { + "epoch": 0.27139, + "grad_norm": 0.6979197859764099, + "learning_rate": 0.003, + "loss": 3.998, + "step": 27139 + }, + { + "epoch": 0.2714, + "grad_norm": 0.7298131585121155, + "learning_rate": 0.003, + "loss": 3.985, + "step": 27140 + }, + { + "epoch": 0.27141, + "grad_norm": 0.7905148267745972, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 27141 + }, + { + "epoch": 0.27142, + "grad_norm": 0.7504250407218933, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 27142 + }, + { + "epoch": 0.27143, + "grad_norm": 0.6078124046325684, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 27143 + }, + { + "epoch": 0.27144, + "grad_norm": 0.7183386087417603, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 27144 + }, + { + "epoch": 0.27145, + "grad_norm": 0.7868930101394653, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 27145 + }, + { + "epoch": 0.27146, + "grad_norm": 0.803274929523468, + "learning_rate": 0.003, + "loss": 4.008, + "step": 27146 + }, + { + "epoch": 0.27147, + "grad_norm": 0.8626024723052979, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 27147 + }, + { + "epoch": 0.27148, + "grad_norm": 0.8712923526763916, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 27148 + }, + { + "epoch": 0.27149, + "grad_norm": 0.9050530195236206, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 27149 + }, + { + "epoch": 0.2715, + "grad_norm": 0.7697364091873169, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 27150 + }, + { + "epoch": 0.27151, + "grad_norm": 0.7703530192375183, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 27151 + }, + { + "epoch": 0.27152, + "grad_norm": 0.7330682873725891, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 27152 + }, + { + "epoch": 0.27153, + "grad_norm": 0.8106885552406311, + "learning_rate": 0.003, + "loss": 3.99, + "step": 27153 + }, + { + "epoch": 0.27154, + "grad_norm": 0.9453380107879639, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 27154 + }, + { + "epoch": 0.27155, + "grad_norm": 1.0304653644561768, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 27155 + }, + { + "epoch": 0.27156, + "grad_norm": 1.1518059968948364, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 27156 + }, + { + "epoch": 0.27157, + "grad_norm": 1.0118807554244995, + "learning_rate": 0.003, + "loss": 4.023, + "step": 27157 + }, + { + "epoch": 0.27158, + "grad_norm": 1.1427245140075684, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 27158 + }, + { + "epoch": 0.27159, + "grad_norm": 1.061833143234253, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 27159 + }, + { + "epoch": 0.2716, + "grad_norm": 1.0246587991714478, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 27160 + }, + { + "epoch": 0.27161, + "grad_norm": 0.9639211893081665, + "learning_rate": 0.003, + "loss": 3.969, + "step": 27161 + }, + { + "epoch": 0.27162, + "grad_norm": 0.843187689781189, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 27162 + }, + { + "epoch": 0.27163, + "grad_norm": 0.8422991037368774, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 27163 + }, + { + "epoch": 0.27164, + "grad_norm": 0.841580867767334, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 27164 + }, + { + "epoch": 0.27165, + "grad_norm": 0.7627331018447876, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 27165 + }, + { + "epoch": 0.27166, + "grad_norm": 0.9395751953125, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 27166 + }, + { + "epoch": 0.27167, + "grad_norm": 0.9231085777282715, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 27167 + }, + { + "epoch": 0.27168, + "grad_norm": 0.9610833525657654, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 27168 + }, + { + "epoch": 0.27169, + "grad_norm": 1.2825160026550293, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 27169 + }, + { + "epoch": 0.2717, + "grad_norm": 0.8354895114898682, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 27170 + }, + { + "epoch": 0.27171, + "grad_norm": 0.7563830614089966, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 27171 + }, + { + "epoch": 0.27172, + "grad_norm": 0.8750113844871521, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 27172 + }, + { + "epoch": 0.27173, + "grad_norm": 0.997057318687439, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 27173 + }, + { + "epoch": 0.27174, + "grad_norm": 0.9230407476425171, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 27174 + }, + { + "epoch": 0.27175, + "grad_norm": 0.8273273706436157, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 27175 + }, + { + "epoch": 0.27176, + "grad_norm": 0.8211009502410889, + "learning_rate": 0.003, + "loss": 3.9417, + "step": 27176 + }, + { + "epoch": 0.27177, + "grad_norm": 0.9607328176498413, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 27177 + }, + { + "epoch": 0.27178, + "grad_norm": 1.4490911960601807, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 27178 + }, + { + "epoch": 0.27179, + "grad_norm": 0.7204159498214722, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 27179 + }, + { + "epoch": 0.2718, + "grad_norm": 0.6921504139900208, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 27180 + }, + { + "epoch": 0.27181, + "grad_norm": 0.7559910416603088, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 27181 + }, + { + "epoch": 0.27182, + "grad_norm": 0.9365822672843933, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 27182 + }, + { + "epoch": 0.27183, + "grad_norm": 1.000426173210144, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 27183 + }, + { + "epoch": 0.27184, + "grad_norm": 1.0509922504425049, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 27184 + }, + { + "epoch": 0.27185, + "grad_norm": 1.0948946475982666, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 27185 + }, + { + "epoch": 0.27186, + "grad_norm": 1.1486098766326904, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 27186 + }, + { + "epoch": 0.27187, + "grad_norm": 0.8881915211677551, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 27187 + }, + { + "epoch": 0.27188, + "grad_norm": 0.9096392393112183, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 27188 + }, + { + "epoch": 0.27189, + "grad_norm": 0.8356248140335083, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 27189 + }, + { + "epoch": 0.2719, + "grad_norm": 0.8184696435928345, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 27190 + }, + { + "epoch": 0.27191, + "grad_norm": 0.7764760255813599, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 27191 + }, + { + "epoch": 0.27192, + "grad_norm": 0.7106375694274902, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 27192 + }, + { + "epoch": 0.27193, + "grad_norm": 0.678717851638794, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 27193 + }, + { + "epoch": 0.27194, + "grad_norm": 0.7185769081115723, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 27194 + }, + { + "epoch": 0.27195, + "grad_norm": 0.7798089385032654, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 27195 + }, + { + "epoch": 0.27196, + "grad_norm": 0.8697138428688049, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 27196 + }, + { + "epoch": 0.27197, + "grad_norm": 0.869999885559082, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 27197 + }, + { + "epoch": 0.27198, + "grad_norm": 0.9579412937164307, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 27198 + }, + { + "epoch": 0.27199, + "grad_norm": 0.9346488118171692, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 27199 + }, + { + "epoch": 0.272, + "grad_norm": 0.935490608215332, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 27200 + }, + { + "epoch": 0.27201, + "grad_norm": 1.0314116477966309, + "learning_rate": 0.003, + "loss": 3.994, + "step": 27201 + }, + { + "epoch": 0.27202, + "grad_norm": 1.078041672706604, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 27202 + }, + { + "epoch": 0.27203, + "grad_norm": 0.9542617797851562, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 27203 + }, + { + "epoch": 0.27204, + "grad_norm": 0.9121935367584229, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 27204 + }, + { + "epoch": 0.27205, + "grad_norm": 0.8171749711036682, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 27205 + }, + { + "epoch": 0.27206, + "grad_norm": 0.5972840785980225, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 27206 + }, + { + "epoch": 0.27207, + "grad_norm": 0.5274034142494202, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 27207 + }, + { + "epoch": 0.27208, + "grad_norm": 0.5602524876594543, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 27208 + }, + { + "epoch": 0.27209, + "grad_norm": 0.6150363683700562, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 27209 + }, + { + "epoch": 0.2721, + "grad_norm": 0.6244784593582153, + "learning_rate": 0.003, + "loss": 3.97, + "step": 27210 + }, + { + "epoch": 0.27211, + "grad_norm": 0.6069692969322205, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 27211 + }, + { + "epoch": 0.27212, + "grad_norm": 0.6651700735092163, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 27212 + }, + { + "epoch": 0.27213, + "grad_norm": 0.6566125750541687, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 27213 + }, + { + "epoch": 0.27214, + "grad_norm": 0.5520392656326294, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 27214 + }, + { + "epoch": 0.27215, + "grad_norm": 0.49440258741378784, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 27215 + }, + { + "epoch": 0.27216, + "grad_norm": 0.606573760509491, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 27216 + }, + { + "epoch": 0.27217, + "grad_norm": 0.6308159828186035, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 27217 + }, + { + "epoch": 0.27218, + "grad_norm": 0.693194568157196, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 27218 + }, + { + "epoch": 0.27219, + "grad_norm": 0.7690804600715637, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 27219 + }, + { + "epoch": 0.2722, + "grad_norm": 0.8495365977287292, + "learning_rate": 0.003, + "loss": 3.969, + "step": 27220 + }, + { + "epoch": 0.27221, + "grad_norm": 0.9529397487640381, + "learning_rate": 0.003, + "loss": 4.02, + "step": 27221 + }, + { + "epoch": 0.27222, + "grad_norm": 1.0452810525894165, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 27222 + }, + { + "epoch": 0.27223, + "grad_norm": 0.9911584854125977, + "learning_rate": 0.003, + "loss": 3.999, + "step": 27223 + }, + { + "epoch": 0.27224, + "grad_norm": 1.4090977907180786, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 27224 + }, + { + "epoch": 0.27225, + "grad_norm": 1.0390053987503052, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 27225 + }, + { + "epoch": 0.27226, + "grad_norm": 1.1865270137786865, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 27226 + }, + { + "epoch": 0.27227, + "grad_norm": 1.0211189985275269, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 27227 + }, + { + "epoch": 0.27228, + "grad_norm": 1.0255944728851318, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 27228 + }, + { + "epoch": 0.27229, + "grad_norm": 0.9929599761962891, + "learning_rate": 0.003, + "loss": 4.007, + "step": 27229 + }, + { + "epoch": 0.2723, + "grad_norm": 1.0446662902832031, + "learning_rate": 0.003, + "loss": 4.016, + "step": 27230 + }, + { + "epoch": 0.27231, + "grad_norm": 0.964654266834259, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 27231 + }, + { + "epoch": 0.27232, + "grad_norm": 1.1118906736373901, + "learning_rate": 0.003, + "loss": 4.007, + "step": 27232 + }, + { + "epoch": 0.27233, + "grad_norm": 0.7566955089569092, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 27233 + }, + { + "epoch": 0.27234, + "grad_norm": 0.7525182962417603, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 27234 + }, + { + "epoch": 0.27235, + "grad_norm": 0.7999179363250732, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 27235 + }, + { + "epoch": 0.27236, + "grad_norm": 0.9057989120483398, + "learning_rate": 0.003, + "loss": 4.002, + "step": 27236 + }, + { + "epoch": 0.27237, + "grad_norm": 1.0185986757278442, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 27237 + }, + { + "epoch": 0.27238, + "grad_norm": 1.2569376230239868, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 27238 + }, + { + "epoch": 0.27239, + "grad_norm": 0.7446125149726868, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 27239 + }, + { + "epoch": 0.2724, + "grad_norm": 0.6141840815544128, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 27240 + }, + { + "epoch": 0.27241, + "grad_norm": 0.6022250056266785, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 27241 + }, + { + "epoch": 0.27242, + "grad_norm": 0.6781219840049744, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 27242 + }, + { + "epoch": 0.27243, + "grad_norm": 0.7307631969451904, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 27243 + }, + { + "epoch": 0.27244, + "grad_norm": 0.8763177394866943, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 27244 + }, + { + "epoch": 0.27245, + "grad_norm": 1.042679786682129, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 27245 + }, + { + "epoch": 0.27246, + "grad_norm": 0.9895907640457153, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 27246 + }, + { + "epoch": 0.27247, + "grad_norm": 1.0544564723968506, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 27247 + }, + { + "epoch": 0.27248, + "grad_norm": 0.921110987663269, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 27248 + }, + { + "epoch": 0.27249, + "grad_norm": 0.8244520425796509, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 27249 + }, + { + "epoch": 0.2725, + "grad_norm": 0.9618173837661743, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 27250 + }, + { + "epoch": 0.27251, + "grad_norm": 1.1916728019714355, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 27251 + }, + { + "epoch": 0.27252, + "grad_norm": 0.7322824001312256, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 27252 + }, + { + "epoch": 0.27253, + "grad_norm": 0.761198103427887, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 27253 + }, + { + "epoch": 0.27254, + "grad_norm": 0.8882818818092346, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 27254 + }, + { + "epoch": 0.27255, + "grad_norm": 0.9485483169555664, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 27255 + }, + { + "epoch": 0.27256, + "grad_norm": 0.969332754611969, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 27256 + }, + { + "epoch": 0.27257, + "grad_norm": 1.1952522993087769, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 27257 + }, + { + "epoch": 0.27258, + "grad_norm": 0.8981733918190002, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 27258 + }, + { + "epoch": 0.27259, + "grad_norm": 1.0290237665176392, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 27259 + }, + { + "epoch": 0.2726, + "grad_norm": 1.265031099319458, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 27260 + }, + { + "epoch": 0.27261, + "grad_norm": 0.8211906552314758, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 27261 + }, + { + "epoch": 0.27262, + "grad_norm": 0.6734006404876709, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 27262 + }, + { + "epoch": 0.27263, + "grad_norm": 0.7765024304389954, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 27263 + }, + { + "epoch": 0.27264, + "grad_norm": 0.8262984156608582, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 27264 + }, + { + "epoch": 0.27265, + "grad_norm": 0.8394256830215454, + "learning_rate": 0.003, + "loss": 4.002, + "step": 27265 + }, + { + "epoch": 0.27266, + "grad_norm": 0.8419214487075806, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 27266 + }, + { + "epoch": 0.27267, + "grad_norm": 0.8167539238929749, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 27267 + }, + { + "epoch": 0.27268, + "grad_norm": 0.6872770190238953, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 27268 + }, + { + "epoch": 0.27269, + "grad_norm": 0.690771758556366, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 27269 + }, + { + "epoch": 0.2727, + "grad_norm": 0.6885172724723816, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 27270 + }, + { + "epoch": 0.27271, + "grad_norm": 0.6401276588439941, + "learning_rate": 0.003, + "loss": 4.003, + "step": 27271 + }, + { + "epoch": 0.27272, + "grad_norm": 0.6128309965133667, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 27272 + }, + { + "epoch": 0.27273, + "grad_norm": 0.5943838357925415, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 27273 + }, + { + "epoch": 0.27274, + "grad_norm": 0.6126666069030762, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 27274 + }, + { + "epoch": 0.27275, + "grad_norm": 0.667704164981842, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 27275 + }, + { + "epoch": 0.27276, + "grad_norm": 0.7444532513618469, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 27276 + }, + { + "epoch": 0.27277, + "grad_norm": 0.8858018517494202, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 27277 + }, + { + "epoch": 0.27278, + "grad_norm": 0.9171189069747925, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 27278 + }, + { + "epoch": 0.27279, + "grad_norm": 0.8386469483375549, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 27279 + }, + { + "epoch": 0.2728, + "grad_norm": 0.8351879715919495, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 27280 + }, + { + "epoch": 0.27281, + "grad_norm": 0.7272552847862244, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 27281 + }, + { + "epoch": 0.27282, + "grad_norm": 0.8957657217979431, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 27282 + }, + { + "epoch": 0.27283, + "grad_norm": 1.2471429109573364, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 27283 + }, + { + "epoch": 0.27284, + "grad_norm": 1.2508060932159424, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 27284 + }, + { + "epoch": 0.27285, + "grad_norm": 0.7700511813163757, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 27285 + }, + { + "epoch": 0.27286, + "grad_norm": 0.7314867377281189, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 27286 + }, + { + "epoch": 0.27287, + "grad_norm": 0.6534942388534546, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 27287 + }, + { + "epoch": 0.27288, + "grad_norm": 0.6133785843849182, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 27288 + }, + { + "epoch": 0.27289, + "grad_norm": 0.6281283497810364, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 27289 + }, + { + "epoch": 0.2729, + "grad_norm": 0.6758475303649902, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 27290 + }, + { + "epoch": 0.27291, + "grad_norm": 0.7814037799835205, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 27291 + }, + { + "epoch": 0.27292, + "grad_norm": 0.8717766404151917, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 27292 + }, + { + "epoch": 0.27293, + "grad_norm": 1.2573468685150146, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 27293 + }, + { + "epoch": 0.27294, + "grad_norm": 1.0176265239715576, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 27294 + }, + { + "epoch": 0.27295, + "grad_norm": 0.8197279572486877, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 27295 + }, + { + "epoch": 0.27296, + "grad_norm": 0.8305360674858093, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 27296 + }, + { + "epoch": 0.27297, + "grad_norm": 0.9737041592597961, + "learning_rate": 0.003, + "loss": 3.977, + "step": 27297 + }, + { + "epoch": 0.27298, + "grad_norm": 0.9749774932861328, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 27298 + }, + { + "epoch": 0.27299, + "grad_norm": 0.9580719470977783, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 27299 + }, + { + "epoch": 0.273, + "grad_norm": 0.9398443698883057, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 27300 + }, + { + "epoch": 0.27301, + "grad_norm": 1.0283048152923584, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 27301 + }, + { + "epoch": 0.27302, + "grad_norm": 1.1554996967315674, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 27302 + }, + { + "epoch": 0.27303, + "grad_norm": 0.8888179063796997, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 27303 + }, + { + "epoch": 0.27304, + "grad_norm": 0.9627481698989868, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 27304 + }, + { + "epoch": 0.27305, + "grad_norm": 0.9836884140968323, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 27305 + }, + { + "epoch": 0.27306, + "grad_norm": 1.1546335220336914, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 27306 + }, + { + "epoch": 0.27307, + "grad_norm": 1.0716745853424072, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 27307 + }, + { + "epoch": 0.27308, + "grad_norm": 1.2177176475524902, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 27308 + }, + { + "epoch": 0.27309, + "grad_norm": 0.8941708207130432, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 27309 + }, + { + "epoch": 0.2731, + "grad_norm": 0.7694114446640015, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 27310 + }, + { + "epoch": 0.27311, + "grad_norm": 0.7914281487464905, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 27311 + }, + { + "epoch": 0.27312, + "grad_norm": 0.7556042075157166, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 27312 + }, + { + "epoch": 0.27313, + "grad_norm": 0.7213252782821655, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 27313 + }, + { + "epoch": 0.27314, + "grad_norm": 0.8217189908027649, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 27314 + }, + { + "epoch": 0.27315, + "grad_norm": 0.7581725716590881, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 27315 + }, + { + "epoch": 0.27316, + "grad_norm": 0.7063874006271362, + "learning_rate": 0.003, + "loss": 3.9451, + "step": 27316 + }, + { + "epoch": 0.27317, + "grad_norm": 0.8897295594215393, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 27317 + }, + { + "epoch": 0.27318, + "grad_norm": 1.173671007156372, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 27318 + }, + { + "epoch": 0.27319, + "grad_norm": 1.2417079210281372, + "learning_rate": 0.003, + "loss": 3.981, + "step": 27319 + }, + { + "epoch": 0.2732, + "grad_norm": 0.6676192283630371, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 27320 + }, + { + "epoch": 0.27321, + "grad_norm": 0.7630194425582886, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 27321 + }, + { + "epoch": 0.27322, + "grad_norm": 0.8606192469596863, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 27322 + }, + { + "epoch": 0.27323, + "grad_norm": 0.7657390236854553, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 27323 + }, + { + "epoch": 0.27324, + "grad_norm": 0.7882050275802612, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 27324 + }, + { + "epoch": 0.27325, + "grad_norm": 0.9199216365814209, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 27325 + }, + { + "epoch": 0.27326, + "grad_norm": 0.965779185295105, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 27326 + }, + { + "epoch": 0.27327, + "grad_norm": 0.7427788376808167, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 27327 + }, + { + "epoch": 0.27328, + "grad_norm": 0.6253379583358765, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 27328 + }, + { + "epoch": 0.27329, + "grad_norm": 0.5912899374961853, + "learning_rate": 0.003, + "loss": 4.0, + "step": 27329 + }, + { + "epoch": 0.2733, + "grad_norm": 0.5637574195861816, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 27330 + }, + { + "epoch": 0.27331, + "grad_norm": 0.5472348928451538, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 27331 + }, + { + "epoch": 0.27332, + "grad_norm": 0.6405960321426392, + "learning_rate": 0.003, + "loss": 3.982, + "step": 27332 + }, + { + "epoch": 0.27333, + "grad_norm": 0.7980006337165833, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 27333 + }, + { + "epoch": 0.27334, + "grad_norm": 1.013510823249817, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 27334 + }, + { + "epoch": 0.27335, + "grad_norm": 1.0679645538330078, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 27335 + }, + { + "epoch": 0.27336, + "grad_norm": 1.063496470451355, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 27336 + }, + { + "epoch": 0.27337, + "grad_norm": 0.8360173106193542, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 27337 + }, + { + "epoch": 0.27338, + "grad_norm": 0.581982433795929, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 27338 + }, + { + "epoch": 0.27339, + "grad_norm": 0.5968117117881775, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 27339 + }, + { + "epoch": 0.2734, + "grad_norm": 0.6353313326835632, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 27340 + }, + { + "epoch": 0.27341, + "grad_norm": 0.7435879707336426, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 27341 + }, + { + "epoch": 0.27342, + "grad_norm": 0.8228645324707031, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 27342 + }, + { + "epoch": 0.27343, + "grad_norm": 0.9968677759170532, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 27343 + }, + { + "epoch": 0.27344, + "grad_norm": 1.188807725906372, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 27344 + }, + { + "epoch": 0.27345, + "grad_norm": 0.7267497777938843, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 27345 + }, + { + "epoch": 0.27346, + "grad_norm": 0.7930703163146973, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 27346 + }, + { + "epoch": 0.27347, + "grad_norm": 0.847551167011261, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 27347 + }, + { + "epoch": 0.27348, + "grad_norm": 0.9229040145874023, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 27348 + }, + { + "epoch": 0.27349, + "grad_norm": 0.9789646863937378, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 27349 + }, + { + "epoch": 0.2735, + "grad_norm": 0.9304605722427368, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 27350 + }, + { + "epoch": 0.27351, + "grad_norm": 0.8507494330406189, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 27351 + }, + { + "epoch": 0.27352, + "grad_norm": 0.8558912873268127, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 27352 + }, + { + "epoch": 0.27353, + "grad_norm": 0.8083445429801941, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 27353 + }, + { + "epoch": 0.27354, + "grad_norm": 0.9377551674842834, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 27354 + }, + { + "epoch": 0.27355, + "grad_norm": 1.0137124061584473, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 27355 + }, + { + "epoch": 0.27356, + "grad_norm": 1.0788735151290894, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 27356 + }, + { + "epoch": 0.27357, + "grad_norm": 0.8949525356292725, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 27357 + }, + { + "epoch": 0.27358, + "grad_norm": 0.894335925579071, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 27358 + }, + { + "epoch": 0.27359, + "grad_norm": 0.94383704662323, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 27359 + }, + { + "epoch": 0.2736, + "grad_norm": 1.0513166189193726, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 27360 + }, + { + "epoch": 0.27361, + "grad_norm": 0.9563454389572144, + "learning_rate": 0.003, + "loss": 3.989, + "step": 27361 + }, + { + "epoch": 0.27362, + "grad_norm": 0.8201004862785339, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 27362 + }, + { + "epoch": 0.27363, + "grad_norm": 0.6713995337486267, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 27363 + }, + { + "epoch": 0.27364, + "grad_norm": 0.7297184467315674, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 27364 + }, + { + "epoch": 0.27365, + "grad_norm": 0.7061269879341125, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 27365 + }, + { + "epoch": 0.27366, + "grad_norm": 0.6055716872215271, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 27366 + }, + { + "epoch": 0.27367, + "grad_norm": 0.6170346736907959, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 27367 + }, + { + "epoch": 0.27368, + "grad_norm": 0.6307644844055176, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 27368 + }, + { + "epoch": 0.27369, + "grad_norm": 0.581829845905304, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 27369 + }, + { + "epoch": 0.2737, + "grad_norm": 0.557957649230957, + "learning_rate": 0.003, + "loss": 3.987, + "step": 27370 + }, + { + "epoch": 0.27371, + "grad_norm": 0.7993025183677673, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 27371 + }, + { + "epoch": 0.27372, + "grad_norm": 1.2655781507492065, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 27372 + }, + { + "epoch": 0.27373, + "grad_norm": 0.9484772682189941, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 27373 + }, + { + "epoch": 0.27374, + "grad_norm": 0.929388165473938, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 27374 + }, + { + "epoch": 0.27375, + "grad_norm": 1.0165555477142334, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 27375 + }, + { + "epoch": 0.27376, + "grad_norm": 1.0672074556350708, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 27376 + }, + { + "epoch": 0.27377, + "grad_norm": 0.9315049052238464, + "learning_rate": 0.003, + "loss": 4.0, + "step": 27377 + }, + { + "epoch": 0.27378, + "grad_norm": 0.9467721581459045, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 27378 + }, + { + "epoch": 0.27379, + "grad_norm": 0.8678153157234192, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 27379 + }, + { + "epoch": 0.2738, + "grad_norm": 0.9870476722717285, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 27380 + }, + { + "epoch": 0.27381, + "grad_norm": 1.0339574813842773, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 27381 + }, + { + "epoch": 0.27382, + "grad_norm": 1.0721737146377563, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 27382 + }, + { + "epoch": 0.27383, + "grad_norm": 0.9055377244949341, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 27383 + }, + { + "epoch": 0.27384, + "grad_norm": 0.7617312073707581, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 27384 + }, + { + "epoch": 0.27385, + "grad_norm": 0.7064080834388733, + "learning_rate": 0.003, + "loss": 3.9443, + "step": 27385 + }, + { + "epoch": 0.27386, + "grad_norm": 0.7582418918609619, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 27386 + }, + { + "epoch": 0.27387, + "grad_norm": 0.7680725455284119, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 27387 + }, + { + "epoch": 0.27388, + "grad_norm": 0.9470184445381165, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 27388 + }, + { + "epoch": 0.27389, + "grad_norm": 1.0031965970993042, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 27389 + }, + { + "epoch": 0.2739, + "grad_norm": 1.0622776746749878, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 27390 + }, + { + "epoch": 0.27391, + "grad_norm": 0.9663096070289612, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 27391 + }, + { + "epoch": 0.27392, + "grad_norm": 0.9677345752716064, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 27392 + }, + { + "epoch": 0.27393, + "grad_norm": 0.9702807664871216, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 27393 + }, + { + "epoch": 0.27394, + "grad_norm": 0.9231035709381104, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 27394 + }, + { + "epoch": 0.27395, + "grad_norm": 0.8324052095413208, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 27395 + }, + { + "epoch": 0.27396, + "grad_norm": 0.7304432392120361, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 27396 + }, + { + "epoch": 0.27397, + "grad_norm": 0.6926811933517456, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 27397 + }, + { + "epoch": 0.27398, + "grad_norm": 0.6700541377067566, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 27398 + }, + { + "epoch": 0.27399, + "grad_norm": 0.7024574279785156, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 27399 + }, + { + "epoch": 0.274, + "grad_norm": 0.9483926296234131, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 27400 + }, + { + "epoch": 0.27401, + "grad_norm": 1.3934913873672485, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 27401 + }, + { + "epoch": 0.27402, + "grad_norm": 0.7396705746650696, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 27402 + }, + { + "epoch": 0.27403, + "grad_norm": 0.6489509344100952, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 27403 + }, + { + "epoch": 0.27404, + "grad_norm": 0.8406927585601807, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 27404 + }, + { + "epoch": 0.27405, + "grad_norm": 1.101593017578125, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 27405 + }, + { + "epoch": 0.27406, + "grad_norm": 1.1046550273895264, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 27406 + }, + { + "epoch": 0.27407, + "grad_norm": 0.7189905047416687, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 27407 + }, + { + "epoch": 0.27408, + "grad_norm": 0.5775378346443176, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 27408 + }, + { + "epoch": 0.27409, + "grad_norm": 0.6669825315475464, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 27409 + }, + { + "epoch": 0.2741, + "grad_norm": 0.6955987811088562, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 27410 + }, + { + "epoch": 0.27411, + "grad_norm": 0.8756113648414612, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 27411 + }, + { + "epoch": 0.27412, + "grad_norm": 0.8369191884994507, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 27412 + }, + { + "epoch": 0.27413, + "grad_norm": 0.8644777536392212, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 27413 + }, + { + "epoch": 0.27414, + "grad_norm": 0.8973546028137207, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 27414 + }, + { + "epoch": 0.27415, + "grad_norm": 0.890508234500885, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 27415 + }, + { + "epoch": 0.27416, + "grad_norm": 0.9468024373054504, + "learning_rate": 0.003, + "loss": 4.012, + "step": 27416 + }, + { + "epoch": 0.27417, + "grad_norm": 1.055660367012024, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 27417 + }, + { + "epoch": 0.27418, + "grad_norm": 1.08524751663208, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 27418 + }, + { + "epoch": 0.27419, + "grad_norm": 0.847043514251709, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 27419 + }, + { + "epoch": 0.2742, + "grad_norm": 0.9440910816192627, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 27420 + }, + { + "epoch": 0.27421, + "grad_norm": 0.9510536193847656, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 27421 + }, + { + "epoch": 0.27422, + "grad_norm": 0.9584581851959229, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 27422 + }, + { + "epoch": 0.27423, + "grad_norm": 0.9429883360862732, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 27423 + }, + { + "epoch": 0.27424, + "grad_norm": 0.9682146310806274, + "learning_rate": 0.003, + "loss": 4.025, + "step": 27424 + }, + { + "epoch": 0.27425, + "grad_norm": 0.9778080582618713, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 27425 + }, + { + "epoch": 0.27426, + "grad_norm": 1.0793708562850952, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 27426 + }, + { + "epoch": 0.27427, + "grad_norm": 1.0525439977645874, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 27427 + }, + { + "epoch": 0.27428, + "grad_norm": 0.9899699687957764, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 27428 + }, + { + "epoch": 0.27429, + "grad_norm": 1.19954514503479, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 27429 + }, + { + "epoch": 0.2743, + "grad_norm": 1.017873764038086, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 27430 + }, + { + "epoch": 0.27431, + "grad_norm": 0.9253607392311096, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 27431 + }, + { + "epoch": 0.27432, + "grad_norm": 0.7603287696838379, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 27432 + }, + { + "epoch": 0.27433, + "grad_norm": 0.8039819002151489, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 27433 + }, + { + "epoch": 0.27434, + "grad_norm": 0.7720279097557068, + "learning_rate": 0.003, + "loss": 3.964, + "step": 27434 + }, + { + "epoch": 0.27435, + "grad_norm": 0.7914043664932251, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 27435 + }, + { + "epoch": 0.27436, + "grad_norm": 0.9943209886550903, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 27436 + }, + { + "epoch": 0.27437, + "grad_norm": 1.2014057636260986, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 27437 + }, + { + "epoch": 0.27438, + "grad_norm": 0.8046541810035706, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 27438 + }, + { + "epoch": 0.27439, + "grad_norm": 0.6859484314918518, + "learning_rate": 0.003, + "loss": 4.014, + "step": 27439 + }, + { + "epoch": 0.2744, + "grad_norm": 0.6971654891967773, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 27440 + }, + { + "epoch": 0.27441, + "grad_norm": 0.7635741233825684, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 27441 + }, + { + "epoch": 0.27442, + "grad_norm": 0.8075980544090271, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 27442 + }, + { + "epoch": 0.27443, + "grad_norm": 0.8907822966575623, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 27443 + }, + { + "epoch": 0.27444, + "grad_norm": 1.0140950679779053, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 27444 + }, + { + "epoch": 0.27445, + "grad_norm": 0.8976731300354004, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 27445 + }, + { + "epoch": 0.27446, + "grad_norm": 0.7725752592086792, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 27446 + }, + { + "epoch": 0.27447, + "grad_norm": 0.8473606705665588, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 27447 + }, + { + "epoch": 0.27448, + "grad_norm": 0.9964342713356018, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 27448 + }, + { + "epoch": 0.27449, + "grad_norm": 1.065108060836792, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 27449 + }, + { + "epoch": 0.2745, + "grad_norm": 1.007391333580017, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 27450 + }, + { + "epoch": 0.27451, + "grad_norm": 1.1021496057510376, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 27451 + }, + { + "epoch": 0.27452, + "grad_norm": 1.0233157873153687, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 27452 + }, + { + "epoch": 0.27453, + "grad_norm": 0.9199975728988647, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 27453 + }, + { + "epoch": 0.27454, + "grad_norm": 1.0091447830200195, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 27454 + }, + { + "epoch": 0.27455, + "grad_norm": 0.9866530299186707, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 27455 + }, + { + "epoch": 0.27456, + "grad_norm": 0.7936580181121826, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 27456 + }, + { + "epoch": 0.27457, + "grad_norm": 0.6754306554794312, + "learning_rate": 0.003, + "loss": 3.98, + "step": 27457 + }, + { + "epoch": 0.27458, + "grad_norm": 0.5775511860847473, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 27458 + }, + { + "epoch": 0.27459, + "grad_norm": 0.479630708694458, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 27459 + }, + { + "epoch": 0.2746, + "grad_norm": 0.4192330241203308, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 27460 + }, + { + "epoch": 0.27461, + "grad_norm": 0.4248686730861664, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 27461 + }, + { + "epoch": 0.27462, + "grad_norm": 0.4587532877922058, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 27462 + }, + { + "epoch": 0.27463, + "grad_norm": 0.5318459272384644, + "learning_rate": 0.003, + "loss": 3.987, + "step": 27463 + }, + { + "epoch": 0.27464, + "grad_norm": 0.6105327606201172, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 27464 + }, + { + "epoch": 0.27465, + "grad_norm": 0.8750140070915222, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 27465 + }, + { + "epoch": 0.27466, + "grad_norm": 1.2729243040084839, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 27466 + }, + { + "epoch": 0.27467, + "grad_norm": 0.7628155946731567, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 27467 + }, + { + "epoch": 0.27468, + "grad_norm": 0.5491991639137268, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 27468 + }, + { + "epoch": 0.27469, + "grad_norm": 0.6777762174606323, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 27469 + }, + { + "epoch": 0.2747, + "grad_norm": 0.826667845249176, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 27470 + }, + { + "epoch": 0.27471, + "grad_norm": 0.9650562405586243, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 27471 + }, + { + "epoch": 0.27472, + "grad_norm": 1.17313551902771, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 27472 + }, + { + "epoch": 0.27473, + "grad_norm": 0.8615050315856934, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 27473 + }, + { + "epoch": 0.27474, + "grad_norm": 0.8405173420906067, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 27474 + }, + { + "epoch": 0.27475, + "grad_norm": 0.9134665131568909, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 27475 + }, + { + "epoch": 0.27476, + "grad_norm": 1.0892829895019531, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 27476 + }, + { + "epoch": 0.27477, + "grad_norm": 0.8567143082618713, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 27477 + }, + { + "epoch": 0.27478, + "grad_norm": 0.815953254699707, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 27478 + }, + { + "epoch": 0.27479, + "grad_norm": 0.8717374801635742, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 27479 + }, + { + "epoch": 0.2748, + "grad_norm": 1.102596640586853, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 27480 + }, + { + "epoch": 0.27481, + "grad_norm": 1.0612961053848267, + "learning_rate": 0.003, + "loss": 4.001, + "step": 27481 + }, + { + "epoch": 0.27482, + "grad_norm": 1.156240701675415, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 27482 + }, + { + "epoch": 0.27483, + "grad_norm": 0.9280985593795776, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 27483 + }, + { + "epoch": 0.27484, + "grad_norm": 0.766474723815918, + "learning_rate": 0.003, + "loss": 3.986, + "step": 27484 + }, + { + "epoch": 0.27485, + "grad_norm": 0.7318002581596375, + "learning_rate": 0.003, + "loss": 3.995, + "step": 27485 + }, + { + "epoch": 0.27486, + "grad_norm": 0.8468157649040222, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 27486 + }, + { + "epoch": 0.27487, + "grad_norm": 0.9415206909179688, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 27487 + }, + { + "epoch": 0.27488, + "grad_norm": 1.020017147064209, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 27488 + }, + { + "epoch": 0.27489, + "grad_norm": 0.9005305171012878, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 27489 + }, + { + "epoch": 0.2749, + "grad_norm": 0.9366637468338013, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 27490 + }, + { + "epoch": 0.27491, + "grad_norm": 0.8568800091743469, + "learning_rate": 0.003, + "loss": 3.99, + "step": 27491 + }, + { + "epoch": 0.27492, + "grad_norm": 0.6716027855873108, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 27492 + }, + { + "epoch": 0.27493, + "grad_norm": 0.6491971015930176, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 27493 + }, + { + "epoch": 0.27494, + "grad_norm": 0.813419759273529, + "learning_rate": 0.003, + "loss": 3.995, + "step": 27494 + }, + { + "epoch": 0.27495, + "grad_norm": 1.0293679237365723, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 27495 + }, + { + "epoch": 0.27496, + "grad_norm": 1.350870966911316, + "learning_rate": 0.003, + "loss": 3.987, + "step": 27496 + }, + { + "epoch": 0.27497, + "grad_norm": 0.5310719609260559, + "learning_rate": 0.003, + "loss": 3.973, + "step": 27497 + }, + { + "epoch": 0.27498, + "grad_norm": 0.7857516407966614, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 27498 + }, + { + "epoch": 0.27499, + "grad_norm": 1.296915888786316, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 27499 + }, + { + "epoch": 0.275, + "grad_norm": 0.4156470000743866, + "learning_rate": 0.003, + "loss": 3.971, + "step": 27500 + }, + { + "epoch": 0.27501, + "grad_norm": 0.8435004949569702, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 27501 + }, + { + "epoch": 0.27502, + "grad_norm": 1.1184828281402588, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 27502 + }, + { + "epoch": 0.27503, + "grad_norm": 0.8004366755485535, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 27503 + }, + { + "epoch": 0.27504, + "grad_norm": 0.812540590763092, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 27504 + }, + { + "epoch": 0.27505, + "grad_norm": 1.0017023086547852, + "learning_rate": 0.003, + "loss": 4.019, + "step": 27505 + }, + { + "epoch": 0.27506, + "grad_norm": 1.0073845386505127, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 27506 + }, + { + "epoch": 0.27507, + "grad_norm": 0.9725349545478821, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 27507 + }, + { + "epoch": 0.27508, + "grad_norm": 0.9632261991500854, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 27508 + }, + { + "epoch": 0.27509, + "grad_norm": 0.8313358426094055, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 27509 + }, + { + "epoch": 0.2751, + "grad_norm": 0.678312361240387, + "learning_rate": 0.003, + "loss": 3.989, + "step": 27510 + }, + { + "epoch": 0.27511, + "grad_norm": 0.8360702991485596, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 27511 + }, + { + "epoch": 0.27512, + "grad_norm": 1.0218812227249146, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 27512 + }, + { + "epoch": 0.27513, + "grad_norm": 1.1618809700012207, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 27513 + }, + { + "epoch": 0.27514, + "grad_norm": 0.6880694627761841, + "learning_rate": 0.003, + "loss": 3.985, + "step": 27514 + }, + { + "epoch": 0.27515, + "grad_norm": 0.6590391993522644, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 27515 + }, + { + "epoch": 0.27516, + "grad_norm": 0.7101160883903503, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 27516 + }, + { + "epoch": 0.27517, + "grad_norm": 0.6804440021514893, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 27517 + }, + { + "epoch": 0.27518, + "grad_norm": 0.8066008687019348, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 27518 + }, + { + "epoch": 0.27519, + "grad_norm": 1.0922895669937134, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 27519 + }, + { + "epoch": 0.2752, + "grad_norm": 1.0006365776062012, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 27520 + }, + { + "epoch": 0.27521, + "grad_norm": 0.7713634371757507, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 27521 + }, + { + "epoch": 0.27522, + "grad_norm": 0.7892709374427795, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 27522 + }, + { + "epoch": 0.27523, + "grad_norm": 1.0661677122116089, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 27523 + }, + { + "epoch": 0.27524, + "grad_norm": 1.0689290761947632, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 27524 + }, + { + "epoch": 0.27525, + "grad_norm": 0.9537630081176758, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 27525 + }, + { + "epoch": 0.27526, + "grad_norm": 1.0265555381774902, + "learning_rate": 0.003, + "loss": 3.983, + "step": 27526 + }, + { + "epoch": 0.27527, + "grad_norm": 0.9586579203605652, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 27527 + }, + { + "epoch": 0.27528, + "grad_norm": 0.9130793809890747, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 27528 + }, + { + "epoch": 0.27529, + "grad_norm": 0.8579205274581909, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 27529 + }, + { + "epoch": 0.2753, + "grad_norm": 0.7251337170600891, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 27530 + }, + { + "epoch": 0.27531, + "grad_norm": 0.7411168217658997, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 27531 + }, + { + "epoch": 0.27532, + "grad_norm": 0.7535727620124817, + "learning_rate": 0.003, + "loss": 4.001, + "step": 27532 + }, + { + "epoch": 0.27533, + "grad_norm": 0.6408496499061584, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 27533 + }, + { + "epoch": 0.27534, + "grad_norm": 0.6368839740753174, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 27534 + }, + { + "epoch": 0.27535, + "grad_norm": 0.6420371532440186, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 27535 + }, + { + "epoch": 0.27536, + "grad_norm": 0.6309686899185181, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 27536 + }, + { + "epoch": 0.27537, + "grad_norm": 0.6537293195724487, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 27537 + }, + { + "epoch": 0.27538, + "grad_norm": 0.7411226630210876, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 27538 + }, + { + "epoch": 0.27539, + "grad_norm": 0.9489669799804688, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 27539 + }, + { + "epoch": 0.2754, + "grad_norm": 1.1660888195037842, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 27540 + }, + { + "epoch": 0.27541, + "grad_norm": 0.9532082676887512, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 27541 + }, + { + "epoch": 0.27542, + "grad_norm": 0.8071833848953247, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 27542 + }, + { + "epoch": 0.27543, + "grad_norm": 0.5678730607032776, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 27543 + }, + { + "epoch": 0.27544, + "grad_norm": 0.5539964437484741, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 27544 + }, + { + "epoch": 0.27545, + "grad_norm": 0.557013213634491, + "learning_rate": 0.003, + "loss": 3.9509, + "step": 27545 + }, + { + "epoch": 0.27546, + "grad_norm": 0.629891037940979, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 27546 + }, + { + "epoch": 0.27547, + "grad_norm": 0.6205410957336426, + "learning_rate": 0.003, + "loss": 3.989, + "step": 27547 + }, + { + "epoch": 0.27548, + "grad_norm": 0.5941054821014404, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 27548 + }, + { + "epoch": 0.27549, + "grad_norm": 0.6259490251541138, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 27549 + }, + { + "epoch": 0.2755, + "grad_norm": 0.7338967323303223, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 27550 + }, + { + "epoch": 0.27551, + "grad_norm": 0.9306878447532654, + "learning_rate": 0.003, + "loss": 3.993, + "step": 27551 + }, + { + "epoch": 0.27552, + "grad_norm": 1.0630313158035278, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 27552 + }, + { + "epoch": 0.27553, + "grad_norm": 0.943733274936676, + "learning_rate": 0.003, + "loss": 3.958, + "step": 27553 + }, + { + "epoch": 0.27554, + "grad_norm": 0.9695666432380676, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 27554 + }, + { + "epoch": 0.27555, + "grad_norm": 0.7960732579231262, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 27555 + }, + { + "epoch": 0.27556, + "grad_norm": 0.7781399488449097, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 27556 + }, + { + "epoch": 0.27557, + "grad_norm": 0.96524578332901, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 27557 + }, + { + "epoch": 0.27558, + "grad_norm": 1.2231404781341553, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 27558 + }, + { + "epoch": 0.27559, + "grad_norm": 0.8243578672409058, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 27559 + }, + { + "epoch": 0.2756, + "grad_norm": 0.8287889361381531, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 27560 + }, + { + "epoch": 0.27561, + "grad_norm": 0.8953477144241333, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 27561 + }, + { + "epoch": 0.27562, + "grad_norm": 1.019101858139038, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 27562 + }, + { + "epoch": 0.27563, + "grad_norm": 1.08418607711792, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 27563 + }, + { + "epoch": 0.27564, + "grad_norm": 0.8863173723220825, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 27564 + }, + { + "epoch": 0.27565, + "grad_norm": 0.9965258240699768, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 27565 + }, + { + "epoch": 0.27566, + "grad_norm": 1.0025477409362793, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 27566 + }, + { + "epoch": 0.27567, + "grad_norm": 1.036256194114685, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 27567 + }, + { + "epoch": 0.27568, + "grad_norm": 1.0217703580856323, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 27568 + }, + { + "epoch": 0.27569, + "grad_norm": 0.9422799348831177, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 27569 + }, + { + "epoch": 0.2757, + "grad_norm": 0.9570853114128113, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 27570 + }, + { + "epoch": 0.27571, + "grad_norm": 0.975795567035675, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 27571 + }, + { + "epoch": 0.27572, + "grad_norm": 1.0673190355300903, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 27572 + }, + { + "epoch": 0.27573, + "grad_norm": 1.325131893157959, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 27573 + }, + { + "epoch": 0.27574, + "grad_norm": 0.5839540362358093, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 27574 + }, + { + "epoch": 0.27575, + "grad_norm": 0.6633597016334534, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 27575 + }, + { + "epoch": 0.27576, + "grad_norm": 0.6938775181770325, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 27576 + }, + { + "epoch": 0.27577, + "grad_norm": 0.7029494047164917, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 27577 + }, + { + "epoch": 0.27578, + "grad_norm": 1.0904148817062378, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 27578 + }, + { + "epoch": 0.27579, + "grad_norm": 1.2696795463562012, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 27579 + }, + { + "epoch": 0.2758, + "grad_norm": 0.7552760243415833, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 27580 + }, + { + "epoch": 0.27581, + "grad_norm": 0.7514081001281738, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 27581 + }, + { + "epoch": 0.27582, + "grad_norm": 0.7512006163597107, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 27582 + }, + { + "epoch": 0.27583, + "grad_norm": 0.7773392200469971, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 27583 + }, + { + "epoch": 0.27584, + "grad_norm": 0.7710407972335815, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 27584 + }, + { + "epoch": 0.27585, + "grad_norm": 0.81489497423172, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 27585 + }, + { + "epoch": 0.27586, + "grad_norm": 0.8341657519340515, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 27586 + }, + { + "epoch": 0.27587, + "grad_norm": 0.7888920903205872, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 27587 + }, + { + "epoch": 0.27588, + "grad_norm": 0.7754246592521667, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 27588 + }, + { + "epoch": 0.27589, + "grad_norm": 0.7535818815231323, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 27589 + }, + { + "epoch": 0.2759, + "grad_norm": 0.7481028437614441, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 27590 + }, + { + "epoch": 0.27591, + "grad_norm": 0.7951812148094177, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 27591 + }, + { + "epoch": 0.27592, + "grad_norm": 1.012080430984497, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 27592 + }, + { + "epoch": 0.27593, + "grad_norm": 1.3410909175872803, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 27593 + }, + { + "epoch": 0.27594, + "grad_norm": 0.7224690318107605, + "learning_rate": 0.003, + "loss": 4.008, + "step": 27594 + }, + { + "epoch": 0.27595, + "grad_norm": 0.6289100646972656, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 27595 + }, + { + "epoch": 0.27596, + "grad_norm": 0.63004070520401, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 27596 + }, + { + "epoch": 0.27597, + "grad_norm": 0.569611668586731, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 27597 + }, + { + "epoch": 0.27598, + "grad_norm": 0.5850928425788879, + "learning_rate": 0.003, + "loss": 3.973, + "step": 27598 + }, + { + "epoch": 0.27599, + "grad_norm": 0.6180763244628906, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 27599 + }, + { + "epoch": 0.276, + "grad_norm": 0.6322371363639832, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 27600 + }, + { + "epoch": 0.27601, + "grad_norm": 0.6504416465759277, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 27601 + }, + { + "epoch": 0.27602, + "grad_norm": 0.6719006896018982, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 27602 + }, + { + "epoch": 0.27603, + "grad_norm": 0.8494760990142822, + "learning_rate": 0.003, + "loss": 3.9439, + "step": 27603 + }, + { + "epoch": 0.27604, + "grad_norm": 1.2337288856506348, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 27604 + }, + { + "epoch": 0.27605, + "grad_norm": 1.1627254486083984, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 27605 + }, + { + "epoch": 0.27606, + "grad_norm": 0.7290728688240051, + "learning_rate": 0.003, + "loss": 3.975, + "step": 27606 + }, + { + "epoch": 0.27607, + "grad_norm": 0.7046745419502258, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 27607 + }, + { + "epoch": 0.27608, + "grad_norm": 0.7319915890693665, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 27608 + }, + { + "epoch": 0.27609, + "grad_norm": 0.8766193985939026, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 27609 + }, + { + "epoch": 0.2761, + "grad_norm": 1.0066797733306885, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 27610 + }, + { + "epoch": 0.27611, + "grad_norm": 0.9772179126739502, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 27611 + }, + { + "epoch": 0.27612, + "grad_norm": 1.0188242197036743, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 27612 + }, + { + "epoch": 0.27613, + "grad_norm": 1.1321760416030884, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 27613 + }, + { + "epoch": 0.27614, + "grad_norm": 0.9182440042495728, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 27614 + }, + { + "epoch": 0.27615, + "grad_norm": 0.9125707149505615, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 27615 + }, + { + "epoch": 0.27616, + "grad_norm": 1.0261332988739014, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 27616 + }, + { + "epoch": 0.27617, + "grad_norm": 1.0960075855255127, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 27617 + }, + { + "epoch": 0.27618, + "grad_norm": 1.0078213214874268, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 27618 + }, + { + "epoch": 0.27619, + "grad_norm": 0.9920313954353333, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 27619 + }, + { + "epoch": 0.2762, + "grad_norm": 0.9627675414085388, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 27620 + }, + { + "epoch": 0.27621, + "grad_norm": 0.878381073474884, + "learning_rate": 0.003, + "loss": 4.008, + "step": 27621 + }, + { + "epoch": 0.27622, + "grad_norm": 0.961880624294281, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 27622 + }, + { + "epoch": 0.27623, + "grad_norm": 1.1784189939498901, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 27623 + }, + { + "epoch": 0.27624, + "grad_norm": 1.1749062538146973, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 27624 + }, + { + "epoch": 0.27625, + "grad_norm": 0.8761844635009766, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 27625 + }, + { + "epoch": 0.27626, + "grad_norm": 0.8164211511611938, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 27626 + }, + { + "epoch": 0.27627, + "grad_norm": 0.9072662591934204, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 27627 + }, + { + "epoch": 0.27628, + "grad_norm": 0.9282047748565674, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 27628 + }, + { + "epoch": 0.27629, + "grad_norm": 0.9141059517860413, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 27629 + }, + { + "epoch": 0.2763, + "grad_norm": 1.0321937799453735, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 27630 + }, + { + "epoch": 0.27631, + "grad_norm": 1.0813441276550293, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 27631 + }, + { + "epoch": 0.27632, + "grad_norm": 0.769786536693573, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 27632 + }, + { + "epoch": 0.27633, + "grad_norm": 0.6619625091552734, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 27633 + }, + { + "epoch": 0.27634, + "grad_norm": 0.7238293290138245, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 27634 + }, + { + "epoch": 0.27635, + "grad_norm": 0.9702540040016174, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 27635 + }, + { + "epoch": 0.27636, + "grad_norm": 1.0177339315414429, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 27636 + }, + { + "epoch": 0.27637, + "grad_norm": 0.8585287928581238, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 27637 + }, + { + "epoch": 0.27638, + "grad_norm": 0.8455610275268555, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 27638 + }, + { + "epoch": 0.27639, + "grad_norm": 0.8833653926849365, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 27639 + }, + { + "epoch": 0.2764, + "grad_norm": 0.7517426609992981, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 27640 + }, + { + "epoch": 0.27641, + "grad_norm": 0.7192456722259521, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 27641 + }, + { + "epoch": 0.27642, + "grad_norm": 0.7157459855079651, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 27642 + }, + { + "epoch": 0.27643, + "grad_norm": 0.780162513256073, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 27643 + }, + { + "epoch": 0.27644, + "grad_norm": 1.056547999382019, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 27644 + }, + { + "epoch": 0.27645, + "grad_norm": 1.0472464561462402, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 27645 + }, + { + "epoch": 0.27646, + "grad_norm": 0.8640562891960144, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 27646 + }, + { + "epoch": 0.27647, + "grad_norm": 0.7283574342727661, + "learning_rate": 0.003, + "loss": 3.985, + "step": 27647 + }, + { + "epoch": 0.27648, + "grad_norm": 0.7836434841156006, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 27648 + }, + { + "epoch": 0.27649, + "grad_norm": 0.7346542477607727, + "learning_rate": 0.003, + "loss": 3.968, + "step": 27649 + }, + { + "epoch": 0.2765, + "grad_norm": 0.6129751801490784, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 27650 + }, + { + "epoch": 0.27651, + "grad_norm": 0.5396965146064758, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 27651 + }, + { + "epoch": 0.27652, + "grad_norm": 0.607039749622345, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 27652 + }, + { + "epoch": 0.27653, + "grad_norm": 0.7185578346252441, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 27653 + }, + { + "epoch": 0.27654, + "grad_norm": 0.6915170550346375, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 27654 + }, + { + "epoch": 0.27655, + "grad_norm": 0.7017138600349426, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 27655 + }, + { + "epoch": 0.27656, + "grad_norm": 0.8866867423057556, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 27656 + }, + { + "epoch": 0.27657, + "grad_norm": 1.4228912591934204, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 27657 + }, + { + "epoch": 0.27658, + "grad_norm": 0.6795519590377808, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 27658 + }, + { + "epoch": 0.27659, + "grad_norm": 0.6810477375984192, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 27659 + }, + { + "epoch": 0.2766, + "grad_norm": 0.7013441324234009, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 27660 + }, + { + "epoch": 0.27661, + "grad_norm": 0.7482807040214539, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 27661 + }, + { + "epoch": 0.27662, + "grad_norm": 0.6525097489356995, + "learning_rate": 0.003, + "loss": 3.989, + "step": 27662 + }, + { + "epoch": 0.27663, + "grad_norm": 0.6516523957252502, + "learning_rate": 0.003, + "loss": 3.992, + "step": 27663 + }, + { + "epoch": 0.27664, + "grad_norm": 0.626291036605835, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 27664 + }, + { + "epoch": 0.27665, + "grad_norm": 0.5703355669975281, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 27665 + }, + { + "epoch": 0.27666, + "grad_norm": 0.6864889860153198, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 27666 + }, + { + "epoch": 0.27667, + "grad_norm": 0.7095913887023926, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 27667 + }, + { + "epoch": 0.27668, + "grad_norm": 0.788063108921051, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 27668 + }, + { + "epoch": 0.27669, + "grad_norm": 0.8611177206039429, + "learning_rate": 0.003, + "loss": 3.9466, + "step": 27669 + }, + { + "epoch": 0.2767, + "grad_norm": 1.1890418529510498, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 27670 + }, + { + "epoch": 0.27671, + "grad_norm": 1.2339247465133667, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 27671 + }, + { + "epoch": 0.27672, + "grad_norm": 0.7009509801864624, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 27672 + }, + { + "epoch": 0.27673, + "grad_norm": 0.6841902732849121, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 27673 + }, + { + "epoch": 0.27674, + "grad_norm": 0.7050684690475464, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 27674 + }, + { + "epoch": 0.27675, + "grad_norm": 0.7763504981994629, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 27675 + }, + { + "epoch": 0.27676, + "grad_norm": 1.0081148147583008, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 27676 + }, + { + "epoch": 0.27677, + "grad_norm": 1.2700501680374146, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 27677 + }, + { + "epoch": 0.27678, + "grad_norm": 0.8274831175804138, + "learning_rate": 0.003, + "loss": 3.991, + "step": 27678 + }, + { + "epoch": 0.27679, + "grad_norm": 0.9295189380645752, + "learning_rate": 0.003, + "loss": 3.978, + "step": 27679 + }, + { + "epoch": 0.2768, + "grad_norm": 0.8896917700767517, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 27680 + }, + { + "epoch": 0.27681, + "grad_norm": 0.8219906091690063, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 27681 + }, + { + "epoch": 0.27682, + "grad_norm": 0.773920476436615, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 27682 + }, + { + "epoch": 0.27683, + "grad_norm": 0.7282680869102478, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 27683 + }, + { + "epoch": 0.27684, + "grad_norm": 0.7062919735908508, + "learning_rate": 0.003, + "loss": 3.957, + "step": 27684 + }, + { + "epoch": 0.27685, + "grad_norm": 0.7712883353233337, + "learning_rate": 0.003, + "loss": 3.965, + "step": 27685 + }, + { + "epoch": 0.27686, + "grad_norm": 0.9149674773216248, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 27686 + }, + { + "epoch": 0.27687, + "grad_norm": 1.0577096939086914, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 27687 + }, + { + "epoch": 0.27688, + "grad_norm": 1.0462485551834106, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 27688 + }, + { + "epoch": 0.27689, + "grad_norm": 1.0027856826782227, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 27689 + }, + { + "epoch": 0.2769, + "grad_norm": 1.1406887769699097, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 27690 + }, + { + "epoch": 0.27691, + "grad_norm": 1.0405666828155518, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 27691 + }, + { + "epoch": 0.27692, + "grad_norm": 0.9702868461608887, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 27692 + }, + { + "epoch": 0.27693, + "grad_norm": 0.9196146130561829, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 27693 + }, + { + "epoch": 0.27694, + "grad_norm": 0.8421724438667297, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 27694 + }, + { + "epoch": 0.27695, + "grad_norm": 0.9130584001541138, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 27695 + }, + { + "epoch": 0.27696, + "grad_norm": 0.8568956851959229, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 27696 + }, + { + "epoch": 0.27697, + "grad_norm": 0.9214305281639099, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 27697 + }, + { + "epoch": 0.27698, + "grad_norm": 1.2216482162475586, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 27698 + }, + { + "epoch": 0.27699, + "grad_norm": 0.9366539716720581, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 27699 + }, + { + "epoch": 0.277, + "grad_norm": 0.9639722108840942, + "learning_rate": 0.003, + "loss": 4.009, + "step": 27700 + }, + { + "epoch": 0.27701, + "grad_norm": 0.979809045791626, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 27701 + }, + { + "epoch": 0.27702, + "grad_norm": 0.8172654509544373, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 27702 + }, + { + "epoch": 0.27703, + "grad_norm": 0.6757631897926331, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 27703 + }, + { + "epoch": 0.27704, + "grad_norm": 0.6882668733596802, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 27704 + }, + { + "epoch": 0.27705, + "grad_norm": 0.6857265830039978, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 27705 + }, + { + "epoch": 0.27706, + "grad_norm": 0.6467714309692383, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 27706 + }, + { + "epoch": 0.27707, + "grad_norm": 0.6279401779174805, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 27707 + }, + { + "epoch": 0.27708, + "grad_norm": 0.694170355796814, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 27708 + }, + { + "epoch": 0.27709, + "grad_norm": 0.7964757680892944, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 27709 + }, + { + "epoch": 0.2771, + "grad_norm": 1.177504539489746, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 27710 + }, + { + "epoch": 0.27711, + "grad_norm": 1.192767858505249, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 27711 + }, + { + "epoch": 0.27712, + "grad_norm": 0.8171151876449585, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 27712 + }, + { + "epoch": 0.27713, + "grad_norm": 0.7543219327926636, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 27713 + }, + { + "epoch": 0.27714, + "grad_norm": 0.6892006993293762, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 27714 + }, + { + "epoch": 0.27715, + "grad_norm": 0.7056968808174133, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 27715 + }, + { + "epoch": 0.27716, + "grad_norm": 0.6919156908988953, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 27716 + }, + { + "epoch": 0.27717, + "grad_norm": 0.8895740509033203, + "learning_rate": 0.003, + "loss": 4.003, + "step": 27717 + }, + { + "epoch": 0.27718, + "grad_norm": 1.1315759420394897, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 27718 + }, + { + "epoch": 0.27719, + "grad_norm": 0.8864564895629883, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 27719 + }, + { + "epoch": 0.2772, + "grad_norm": 0.6320936679840088, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 27720 + }, + { + "epoch": 0.27721, + "grad_norm": 0.610953152179718, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 27721 + }, + { + "epoch": 0.27722, + "grad_norm": 0.8592227697372437, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 27722 + }, + { + "epoch": 0.27723, + "grad_norm": 1.0675663948059082, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 27723 + }, + { + "epoch": 0.27724, + "grad_norm": 1.3175380229949951, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 27724 + }, + { + "epoch": 0.27725, + "grad_norm": 0.8346483707427979, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 27725 + }, + { + "epoch": 0.27726, + "grad_norm": 0.920992374420166, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 27726 + }, + { + "epoch": 0.27727, + "grad_norm": 0.9628113508224487, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 27727 + }, + { + "epoch": 0.27728, + "grad_norm": 1.0104995965957642, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 27728 + }, + { + "epoch": 0.27729, + "grad_norm": 1.0632432699203491, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 27729 + }, + { + "epoch": 0.2773, + "grad_norm": 0.8651211857795715, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 27730 + }, + { + "epoch": 0.27731, + "grad_norm": 0.7183332443237305, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 27731 + }, + { + "epoch": 0.27732, + "grad_norm": 0.8873608112335205, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 27732 + }, + { + "epoch": 0.27733, + "grad_norm": 0.8606005311012268, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 27733 + }, + { + "epoch": 0.27734, + "grad_norm": 0.8071790337562561, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 27734 + }, + { + "epoch": 0.27735, + "grad_norm": 0.9066339135169983, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 27735 + }, + { + "epoch": 0.27736, + "grad_norm": 0.8769851326942444, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 27736 + }, + { + "epoch": 0.27737, + "grad_norm": 0.726504921913147, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 27737 + }, + { + "epoch": 0.27738, + "grad_norm": 0.8206346035003662, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 27738 + }, + { + "epoch": 0.27739, + "grad_norm": 1.0534707307815552, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 27739 + }, + { + "epoch": 0.2774, + "grad_norm": 0.9052572250366211, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 27740 + }, + { + "epoch": 0.27741, + "grad_norm": 0.8630173802375793, + "learning_rate": 0.003, + "loss": 3.988, + "step": 27741 + }, + { + "epoch": 0.27742, + "grad_norm": 0.9660598039627075, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 27742 + }, + { + "epoch": 0.27743, + "grad_norm": 0.9441372156143188, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 27743 + }, + { + "epoch": 0.27744, + "grad_norm": 0.9781750440597534, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 27744 + }, + { + "epoch": 0.27745, + "grad_norm": 0.9728083610534668, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 27745 + }, + { + "epoch": 0.27746, + "grad_norm": 0.8032357692718506, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 27746 + }, + { + "epoch": 0.27747, + "grad_norm": 0.8013231754302979, + "learning_rate": 0.003, + "loss": 3.975, + "step": 27747 + }, + { + "epoch": 0.27748, + "grad_norm": 1.0289177894592285, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 27748 + }, + { + "epoch": 0.27749, + "grad_norm": 1.104108214378357, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 27749 + }, + { + "epoch": 0.2775, + "grad_norm": 1.0047310590744019, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 27750 + }, + { + "epoch": 0.27751, + "grad_norm": 0.927300751209259, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 27751 + }, + { + "epoch": 0.27752, + "grad_norm": 0.8779742121696472, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 27752 + }, + { + "epoch": 0.27753, + "grad_norm": 0.9722281694412231, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 27753 + }, + { + "epoch": 0.27754, + "grad_norm": 0.8552652597427368, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 27754 + }, + { + "epoch": 0.27755, + "grad_norm": 1.0086098909378052, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 27755 + }, + { + "epoch": 0.27756, + "grad_norm": 1.3072072267532349, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 27756 + }, + { + "epoch": 0.27757, + "grad_norm": 0.749038577079773, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 27757 + }, + { + "epoch": 0.27758, + "grad_norm": 0.7614638209342957, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 27758 + }, + { + "epoch": 0.27759, + "grad_norm": 0.7812074422836304, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 27759 + }, + { + "epoch": 0.2776, + "grad_norm": 0.8772993087768555, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 27760 + }, + { + "epoch": 0.27761, + "grad_norm": 0.8677672147750854, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 27761 + }, + { + "epoch": 0.27762, + "grad_norm": 0.7295716404914856, + "learning_rate": 0.003, + "loss": 4.008, + "step": 27762 + }, + { + "epoch": 0.27763, + "grad_norm": 0.6481612920761108, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 27763 + }, + { + "epoch": 0.27764, + "grad_norm": 0.6667928099632263, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 27764 + }, + { + "epoch": 0.27765, + "grad_norm": 1.0837918519973755, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 27765 + }, + { + "epoch": 0.27766, + "grad_norm": 1.2024157047271729, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 27766 + }, + { + "epoch": 0.27767, + "grad_norm": 0.7869893312454224, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 27767 + }, + { + "epoch": 0.27768, + "grad_norm": 0.661437451839447, + "learning_rate": 0.003, + "loss": 3.983, + "step": 27768 + }, + { + "epoch": 0.27769, + "grad_norm": 0.5328320860862732, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 27769 + }, + { + "epoch": 0.2777, + "grad_norm": 0.5301913022994995, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 27770 + }, + { + "epoch": 0.27771, + "grad_norm": 0.5643831491470337, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 27771 + }, + { + "epoch": 0.27772, + "grad_norm": 0.6149654388427734, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 27772 + }, + { + "epoch": 0.27773, + "grad_norm": 0.6744791865348816, + "learning_rate": 0.003, + "loss": 3.982, + "step": 27773 + }, + { + "epoch": 0.27774, + "grad_norm": 0.7040554881095886, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 27774 + }, + { + "epoch": 0.27775, + "grad_norm": 0.7363758683204651, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 27775 + }, + { + "epoch": 0.27776, + "grad_norm": 0.9637890458106995, + "learning_rate": 0.003, + "loss": 3.995, + "step": 27776 + }, + { + "epoch": 0.27777, + "grad_norm": 1.2297592163085938, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 27777 + }, + { + "epoch": 0.27778, + "grad_norm": 0.7385910153388977, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 27778 + }, + { + "epoch": 0.27779, + "grad_norm": 0.7131724953651428, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 27779 + }, + { + "epoch": 0.2778, + "grad_norm": 0.6743952035903931, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 27780 + }, + { + "epoch": 0.27781, + "grad_norm": 0.6336796879768372, + "learning_rate": 0.003, + "loss": 3.9427, + "step": 27781 + }, + { + "epoch": 0.27782, + "grad_norm": 0.7124606966972351, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 27782 + }, + { + "epoch": 0.27783, + "grad_norm": 0.8658545017242432, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 27783 + }, + { + "epoch": 0.27784, + "grad_norm": 0.9689455628395081, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 27784 + }, + { + "epoch": 0.27785, + "grad_norm": 1.0661171674728394, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 27785 + }, + { + "epoch": 0.27786, + "grad_norm": 0.9615309238433838, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 27786 + }, + { + "epoch": 0.27787, + "grad_norm": 0.916214108467102, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 27787 + }, + { + "epoch": 0.27788, + "grad_norm": 0.7734590768814087, + "learning_rate": 0.003, + "loss": 4.005, + "step": 27788 + }, + { + "epoch": 0.27789, + "grad_norm": 0.8341209888458252, + "learning_rate": 0.003, + "loss": 3.983, + "step": 27789 + }, + { + "epoch": 0.2779, + "grad_norm": 0.9605803489685059, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 27790 + }, + { + "epoch": 0.27791, + "grad_norm": 1.5301934480667114, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 27791 + }, + { + "epoch": 0.27792, + "grad_norm": 0.6900938749313354, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 27792 + }, + { + "epoch": 0.27793, + "grad_norm": 0.7278600335121155, + "learning_rate": 0.003, + "loss": 3.999, + "step": 27793 + }, + { + "epoch": 0.27794, + "grad_norm": 0.8395312428474426, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 27794 + }, + { + "epoch": 0.27795, + "grad_norm": 1.159361481666565, + "learning_rate": 0.003, + "loss": 3.9401, + "step": 27795 + }, + { + "epoch": 0.27796, + "grad_norm": 1.2808589935302734, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 27796 + }, + { + "epoch": 0.27797, + "grad_norm": 0.7417958974838257, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 27797 + }, + { + "epoch": 0.27798, + "grad_norm": 0.8626911044120789, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 27798 + }, + { + "epoch": 0.27799, + "grad_norm": 1.1022182703018188, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 27799 + }, + { + "epoch": 0.278, + "grad_norm": 1.1757081747055054, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 27800 + }, + { + "epoch": 0.27801, + "grad_norm": 0.8726761341094971, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 27801 + }, + { + "epoch": 0.27802, + "grad_norm": 0.8222735524177551, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 27802 + }, + { + "epoch": 0.27803, + "grad_norm": 0.8794581294059753, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 27803 + }, + { + "epoch": 0.27804, + "grad_norm": 0.8326951265335083, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 27804 + }, + { + "epoch": 0.27805, + "grad_norm": 0.7858664989471436, + "learning_rate": 0.003, + "loss": 3.994, + "step": 27805 + }, + { + "epoch": 0.27806, + "grad_norm": 0.751673698425293, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 27806 + }, + { + "epoch": 0.27807, + "grad_norm": 0.7059588432312012, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 27807 + }, + { + "epoch": 0.27808, + "grad_norm": 0.8100309371948242, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 27808 + }, + { + "epoch": 0.27809, + "grad_norm": 1.019904613494873, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 27809 + }, + { + "epoch": 0.2781, + "grad_norm": 1.0804487466812134, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 27810 + }, + { + "epoch": 0.27811, + "grad_norm": 0.9376332759857178, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 27811 + }, + { + "epoch": 0.27812, + "grad_norm": 1.0027427673339844, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 27812 + }, + { + "epoch": 0.27813, + "grad_norm": 1.0320101976394653, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 27813 + }, + { + "epoch": 0.27814, + "grad_norm": 0.8571596741676331, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 27814 + }, + { + "epoch": 0.27815, + "grad_norm": 0.7778292894363403, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 27815 + }, + { + "epoch": 0.27816, + "grad_norm": 0.9632325172424316, + "learning_rate": 0.003, + "loss": 3.981, + "step": 27816 + }, + { + "epoch": 0.27817, + "grad_norm": 0.9855098128318787, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 27817 + }, + { + "epoch": 0.27818, + "grad_norm": 1.0761218070983887, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 27818 + }, + { + "epoch": 0.27819, + "grad_norm": 1.006109595298767, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 27819 + }, + { + "epoch": 0.2782, + "grad_norm": 0.9139106273651123, + "learning_rate": 0.003, + "loss": 4.013, + "step": 27820 + }, + { + "epoch": 0.27821, + "grad_norm": 0.8483889102935791, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 27821 + }, + { + "epoch": 0.27822, + "grad_norm": 0.6912669539451599, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 27822 + }, + { + "epoch": 0.27823, + "grad_norm": 0.8110162615776062, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 27823 + }, + { + "epoch": 0.27824, + "grad_norm": 0.8279449939727783, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 27824 + }, + { + "epoch": 0.27825, + "grad_norm": 0.8023001551628113, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 27825 + }, + { + "epoch": 0.27826, + "grad_norm": 0.7356520295143127, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 27826 + }, + { + "epoch": 0.27827, + "grad_norm": 0.7163177132606506, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 27827 + }, + { + "epoch": 0.27828, + "grad_norm": 0.8987792730331421, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 27828 + }, + { + "epoch": 0.27829, + "grad_norm": 0.9271143674850464, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 27829 + }, + { + "epoch": 0.2783, + "grad_norm": 0.9693663120269775, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 27830 + }, + { + "epoch": 0.27831, + "grad_norm": 0.8926043510437012, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 27831 + }, + { + "epoch": 0.27832, + "grad_norm": 0.9148818254470825, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 27832 + }, + { + "epoch": 0.27833, + "grad_norm": 1.1280072927474976, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 27833 + }, + { + "epoch": 0.27834, + "grad_norm": 0.9514440894126892, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 27834 + }, + { + "epoch": 0.27835, + "grad_norm": 0.7403717041015625, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 27835 + }, + { + "epoch": 0.27836, + "grad_norm": 0.696217954158783, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 27836 + }, + { + "epoch": 0.27837, + "grad_norm": 0.7104209065437317, + "learning_rate": 0.003, + "loss": 4.005, + "step": 27837 + }, + { + "epoch": 0.27838, + "grad_norm": 0.7648135423660278, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 27838 + }, + { + "epoch": 0.27839, + "grad_norm": 1.0590239763259888, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 27839 + }, + { + "epoch": 0.2784, + "grad_norm": 1.1478782892227173, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 27840 + }, + { + "epoch": 0.27841, + "grad_norm": 0.8546966910362244, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 27841 + }, + { + "epoch": 0.27842, + "grad_norm": 0.855537474155426, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 27842 + }, + { + "epoch": 0.27843, + "grad_norm": 0.9221584796905518, + "learning_rate": 0.003, + "loss": 3.987, + "step": 27843 + }, + { + "epoch": 0.27844, + "grad_norm": 0.9528251886367798, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 27844 + }, + { + "epoch": 0.27845, + "grad_norm": 1.0208154916763306, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 27845 + }, + { + "epoch": 0.27846, + "grad_norm": 1.2354847192764282, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 27846 + }, + { + "epoch": 0.27847, + "grad_norm": 0.8137288093566895, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 27847 + }, + { + "epoch": 0.27848, + "grad_norm": 0.6588727831840515, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 27848 + }, + { + "epoch": 0.27849, + "grad_norm": 0.6079830527305603, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 27849 + }, + { + "epoch": 0.2785, + "grad_norm": 0.618229329586029, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 27850 + }, + { + "epoch": 0.27851, + "grad_norm": 0.71535325050354, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 27851 + }, + { + "epoch": 0.27852, + "grad_norm": 0.9079141020774841, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 27852 + }, + { + "epoch": 0.27853, + "grad_norm": 1.1960772275924683, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 27853 + }, + { + "epoch": 0.27854, + "grad_norm": 0.894469678401947, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 27854 + }, + { + "epoch": 0.27855, + "grad_norm": 0.6717695593833923, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 27855 + }, + { + "epoch": 0.27856, + "grad_norm": 0.6776416301727295, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 27856 + }, + { + "epoch": 0.27857, + "grad_norm": 0.6130914092063904, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 27857 + }, + { + "epoch": 0.27858, + "grad_norm": 0.5296789407730103, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 27858 + }, + { + "epoch": 0.27859, + "grad_norm": 0.5796956419944763, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 27859 + }, + { + "epoch": 0.2786, + "grad_norm": 0.5380564332008362, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 27860 + }, + { + "epoch": 0.27861, + "grad_norm": 0.4699243903160095, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 27861 + }, + { + "epoch": 0.27862, + "grad_norm": 0.5363632440567017, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 27862 + }, + { + "epoch": 0.27863, + "grad_norm": 0.6234567165374756, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 27863 + }, + { + "epoch": 0.27864, + "grad_norm": 0.673678994178772, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 27864 + }, + { + "epoch": 0.27865, + "grad_norm": 0.7822198867797852, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 27865 + }, + { + "epoch": 0.27866, + "grad_norm": 1.0108805894851685, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 27866 + }, + { + "epoch": 0.27867, + "grad_norm": 1.1334481239318848, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 27867 + }, + { + "epoch": 0.27868, + "grad_norm": 1.0018328428268433, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 27868 + }, + { + "epoch": 0.27869, + "grad_norm": 1.2883968353271484, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 27869 + }, + { + "epoch": 0.2787, + "grad_norm": 0.8515567183494568, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 27870 + }, + { + "epoch": 0.27871, + "grad_norm": 0.7704970836639404, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 27871 + }, + { + "epoch": 0.27872, + "grad_norm": 0.7333813309669495, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 27872 + }, + { + "epoch": 0.27873, + "grad_norm": 0.7341665029525757, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 27873 + }, + { + "epoch": 0.27874, + "grad_norm": 0.764582633972168, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 27874 + }, + { + "epoch": 0.27875, + "grad_norm": 0.9898257851600647, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 27875 + }, + { + "epoch": 0.27876, + "grad_norm": 1.2061021327972412, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 27876 + }, + { + "epoch": 0.27877, + "grad_norm": 0.8607355952262878, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 27877 + }, + { + "epoch": 0.27878, + "grad_norm": 0.9202953577041626, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 27878 + }, + { + "epoch": 0.27879, + "grad_norm": 1.0093989372253418, + "learning_rate": 0.003, + "loss": 3.953, + "step": 27879 + }, + { + "epoch": 0.2788, + "grad_norm": 1.066453218460083, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 27880 + }, + { + "epoch": 0.27881, + "grad_norm": 0.8863624930381775, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 27881 + }, + { + "epoch": 0.27882, + "grad_norm": 0.8597854375839233, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 27882 + }, + { + "epoch": 0.27883, + "grad_norm": 0.9050495028495789, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 27883 + }, + { + "epoch": 0.27884, + "grad_norm": 0.8738573789596558, + "learning_rate": 0.003, + "loss": 3.993, + "step": 27884 + }, + { + "epoch": 0.27885, + "grad_norm": 0.7890745997428894, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 27885 + }, + { + "epoch": 0.27886, + "grad_norm": 0.8743499517440796, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 27886 + }, + { + "epoch": 0.27887, + "grad_norm": 0.7630235552787781, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 27887 + }, + { + "epoch": 0.27888, + "grad_norm": 0.7463727593421936, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 27888 + }, + { + "epoch": 0.27889, + "grad_norm": 0.8171845078468323, + "learning_rate": 0.003, + "loss": 3.974, + "step": 27889 + }, + { + "epoch": 0.2789, + "grad_norm": 1.0348668098449707, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 27890 + }, + { + "epoch": 0.27891, + "grad_norm": 1.4427764415740967, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 27891 + }, + { + "epoch": 0.27892, + "grad_norm": 0.6015771627426147, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 27892 + }, + { + "epoch": 0.27893, + "grad_norm": 0.8810895085334778, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 27893 + }, + { + "epoch": 0.27894, + "grad_norm": 1.2527889013290405, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 27894 + }, + { + "epoch": 0.27895, + "grad_norm": 0.6628817915916443, + "learning_rate": 0.003, + "loss": 3.99, + "step": 27895 + }, + { + "epoch": 0.27896, + "grad_norm": 0.741752564907074, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 27896 + }, + { + "epoch": 0.27897, + "grad_norm": 0.7943568229675293, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 27897 + }, + { + "epoch": 0.27898, + "grad_norm": 0.8744359016418457, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 27898 + }, + { + "epoch": 0.27899, + "grad_norm": 0.9680164456367493, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 27899 + }, + { + "epoch": 0.279, + "grad_norm": 0.9025626182556152, + "learning_rate": 0.003, + "loss": 3.989, + "step": 27900 + }, + { + "epoch": 0.27901, + "grad_norm": 0.8622552156448364, + "learning_rate": 0.003, + "loss": 3.9295, + "step": 27901 + }, + { + "epoch": 0.27902, + "grad_norm": 0.8894708752632141, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 27902 + }, + { + "epoch": 0.27903, + "grad_norm": 0.8909054398536682, + "learning_rate": 0.003, + "loss": 3.986, + "step": 27903 + }, + { + "epoch": 0.27904, + "grad_norm": 0.8117542266845703, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 27904 + }, + { + "epoch": 0.27905, + "grad_norm": 0.7497193813323975, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 27905 + }, + { + "epoch": 0.27906, + "grad_norm": 0.7483964562416077, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 27906 + }, + { + "epoch": 0.27907, + "grad_norm": 0.7953978776931763, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 27907 + }, + { + "epoch": 0.27908, + "grad_norm": 0.854555070400238, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 27908 + }, + { + "epoch": 0.27909, + "grad_norm": 0.9705546498298645, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 27909 + }, + { + "epoch": 0.2791, + "grad_norm": 1.0302709341049194, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 27910 + }, + { + "epoch": 0.27911, + "grad_norm": 0.981521725654602, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 27911 + }, + { + "epoch": 0.27912, + "grad_norm": 1.069940209388733, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 27912 + }, + { + "epoch": 0.27913, + "grad_norm": 0.9629427790641785, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 27913 + }, + { + "epoch": 0.27914, + "grad_norm": 0.7876320481300354, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 27914 + }, + { + "epoch": 0.27915, + "grad_norm": 0.830550491809845, + "learning_rate": 0.003, + "loss": 3.986, + "step": 27915 + }, + { + "epoch": 0.27916, + "grad_norm": 0.768650472164154, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 27916 + }, + { + "epoch": 0.27917, + "grad_norm": 0.8358531594276428, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 27917 + }, + { + "epoch": 0.27918, + "grad_norm": 0.8154186010360718, + "learning_rate": 0.003, + "loss": 3.995, + "step": 27918 + }, + { + "epoch": 0.27919, + "grad_norm": 0.9205735325813293, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 27919 + }, + { + "epoch": 0.2792, + "grad_norm": 0.9606949687004089, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 27920 + }, + { + "epoch": 0.27921, + "grad_norm": 1.2259432077407837, + "learning_rate": 0.003, + "loss": 4.007, + "step": 27921 + }, + { + "epoch": 0.27922, + "grad_norm": 1.0013662576675415, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 27922 + }, + { + "epoch": 0.27923, + "grad_norm": 1.157562017440796, + "learning_rate": 0.003, + "loss": 3.989, + "step": 27923 + }, + { + "epoch": 0.27924, + "grad_norm": 0.9500899314880371, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 27924 + }, + { + "epoch": 0.27925, + "grad_norm": 0.8681535124778748, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 27925 + }, + { + "epoch": 0.27926, + "grad_norm": 0.7753390669822693, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 27926 + }, + { + "epoch": 0.27927, + "grad_norm": 0.6933079957962036, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 27927 + }, + { + "epoch": 0.27928, + "grad_norm": 0.6713147163391113, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 27928 + }, + { + "epoch": 0.27929, + "grad_norm": 0.8364907503128052, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 27929 + }, + { + "epoch": 0.2793, + "grad_norm": 1.03541100025177, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 27930 + }, + { + "epoch": 0.27931, + "grad_norm": 0.9594789743423462, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 27931 + }, + { + "epoch": 0.27932, + "grad_norm": 0.9041743874549866, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 27932 + }, + { + "epoch": 0.27933, + "grad_norm": 0.8665250539779663, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 27933 + }, + { + "epoch": 0.27934, + "grad_norm": 0.7668823003768921, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 27934 + }, + { + "epoch": 0.27935, + "grad_norm": 0.6909937858581543, + "learning_rate": 0.003, + "loss": 3.975, + "step": 27935 + }, + { + "epoch": 0.27936, + "grad_norm": 0.6823228597640991, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 27936 + }, + { + "epoch": 0.27937, + "grad_norm": 0.6584793329238892, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 27937 + }, + { + "epoch": 0.27938, + "grad_norm": 0.9105526804924011, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 27938 + }, + { + "epoch": 0.27939, + "grad_norm": 1.2705765962600708, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 27939 + }, + { + "epoch": 0.2794, + "grad_norm": 0.700515866279602, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 27940 + }, + { + "epoch": 0.27941, + "grad_norm": 0.5821597576141357, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 27941 + }, + { + "epoch": 0.27942, + "grad_norm": 0.6693868041038513, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 27942 + }, + { + "epoch": 0.27943, + "grad_norm": 0.7024862766265869, + "learning_rate": 0.003, + "loss": 3.969, + "step": 27943 + }, + { + "epoch": 0.27944, + "grad_norm": 0.7196487784385681, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 27944 + }, + { + "epoch": 0.27945, + "grad_norm": 0.7410542964935303, + "learning_rate": 0.003, + "loss": 3.987, + "step": 27945 + }, + { + "epoch": 0.27946, + "grad_norm": 0.9183028340339661, + "learning_rate": 0.003, + "loss": 3.978, + "step": 27946 + }, + { + "epoch": 0.27947, + "grad_norm": 1.0296945571899414, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 27947 + }, + { + "epoch": 0.27948, + "grad_norm": 1.0017321109771729, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 27948 + }, + { + "epoch": 0.27949, + "grad_norm": 0.7936185002326965, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 27949 + }, + { + "epoch": 0.2795, + "grad_norm": 0.7335847020149231, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 27950 + }, + { + "epoch": 0.27951, + "grad_norm": 0.9070531725883484, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 27951 + }, + { + "epoch": 0.27952, + "grad_norm": 1.1568922996520996, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 27952 + }, + { + "epoch": 0.27953, + "grad_norm": 1.1040410995483398, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 27953 + }, + { + "epoch": 0.27954, + "grad_norm": 0.876828670501709, + "learning_rate": 0.003, + "loss": 3.99, + "step": 27954 + }, + { + "epoch": 0.27955, + "grad_norm": 0.8576979041099548, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 27955 + }, + { + "epoch": 0.27956, + "grad_norm": 0.8433840274810791, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 27956 + }, + { + "epoch": 0.27957, + "grad_norm": 0.7889150977134705, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 27957 + }, + { + "epoch": 0.27958, + "grad_norm": 0.8527053594589233, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 27958 + }, + { + "epoch": 0.27959, + "grad_norm": 0.9317420125007629, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 27959 + }, + { + "epoch": 0.2796, + "grad_norm": 0.9960830211639404, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 27960 + }, + { + "epoch": 0.27961, + "grad_norm": 1.0491758584976196, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 27961 + }, + { + "epoch": 0.27962, + "grad_norm": 1.1054775714874268, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 27962 + }, + { + "epoch": 0.27963, + "grad_norm": 1.0609893798828125, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 27963 + }, + { + "epoch": 0.27964, + "grad_norm": 0.9247941374778748, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 27964 + }, + { + "epoch": 0.27965, + "grad_norm": 0.8862802386283875, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 27965 + }, + { + "epoch": 0.27966, + "grad_norm": 0.9390245079994202, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 27966 + }, + { + "epoch": 0.27967, + "grad_norm": 0.9765099287033081, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 27967 + }, + { + "epoch": 0.27968, + "grad_norm": 1.1837857961654663, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 27968 + }, + { + "epoch": 0.27969, + "grad_norm": 1.127743124961853, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 27969 + }, + { + "epoch": 0.2797, + "grad_norm": 0.9191488027572632, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 27970 + }, + { + "epoch": 0.27971, + "grad_norm": 0.9105640053749084, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 27971 + }, + { + "epoch": 0.27972, + "grad_norm": 1.0358827114105225, + "learning_rate": 0.003, + "loss": 4.011, + "step": 27972 + }, + { + "epoch": 0.27973, + "grad_norm": 1.056662678718567, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 27973 + }, + { + "epoch": 0.27974, + "grad_norm": 0.8376032710075378, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 27974 + }, + { + "epoch": 0.27975, + "grad_norm": 0.5619061589241028, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 27975 + }, + { + "epoch": 0.27976, + "grad_norm": 0.4898180067539215, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 27976 + }, + { + "epoch": 0.27977, + "grad_norm": 0.5737332701683044, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 27977 + }, + { + "epoch": 0.27978, + "grad_norm": 0.6704110503196716, + "learning_rate": 0.003, + "loss": 4.001, + "step": 27978 + }, + { + "epoch": 0.27979, + "grad_norm": 0.6934046149253845, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 27979 + }, + { + "epoch": 0.2798, + "grad_norm": 0.6398918628692627, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 27980 + }, + { + "epoch": 0.27981, + "grad_norm": 0.7006879448890686, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 27981 + }, + { + "epoch": 0.27982, + "grad_norm": 0.9369761943817139, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 27982 + }, + { + "epoch": 0.27983, + "grad_norm": 1.04574716091156, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 27983 + }, + { + "epoch": 0.27984, + "grad_norm": 0.9275797009468079, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 27984 + }, + { + "epoch": 0.27985, + "grad_norm": 0.8649306297302246, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 27985 + }, + { + "epoch": 0.27986, + "grad_norm": 0.8022871017456055, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 27986 + }, + { + "epoch": 0.27987, + "grad_norm": 0.9036130309104919, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 27987 + }, + { + "epoch": 0.27988, + "grad_norm": 1.0106843709945679, + "learning_rate": 0.003, + "loss": 3.997, + "step": 27988 + }, + { + "epoch": 0.27989, + "grad_norm": 1.0240379571914673, + "learning_rate": 0.003, + "loss": 3.983, + "step": 27989 + }, + { + "epoch": 0.2799, + "grad_norm": 1.0996307134628296, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 27990 + }, + { + "epoch": 0.27991, + "grad_norm": 1.0508149862289429, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 27991 + }, + { + "epoch": 0.27992, + "grad_norm": 0.7803552746772766, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 27992 + }, + { + "epoch": 0.27993, + "grad_norm": 0.768828272819519, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 27993 + }, + { + "epoch": 0.27994, + "grad_norm": 0.6478749513626099, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 27994 + }, + { + "epoch": 0.27995, + "grad_norm": 0.6465196013450623, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 27995 + }, + { + "epoch": 0.27996, + "grad_norm": 0.672635555267334, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 27996 + }, + { + "epoch": 0.27997, + "grad_norm": 0.8837477564811707, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 27997 + }, + { + "epoch": 0.27998, + "grad_norm": 1.1262245178222656, + "learning_rate": 0.003, + "loss": 3.989, + "step": 27998 + }, + { + "epoch": 0.27999, + "grad_norm": 1.1547943353652954, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 27999 + }, + { + "epoch": 0.28, + "grad_norm": 0.7800897359848022, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 28000 + }, + { + "epoch": 0.28001, + "grad_norm": 0.6189377307891846, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 28001 + }, + { + "epoch": 0.28002, + "grad_norm": 0.7867535948753357, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 28002 + }, + { + "epoch": 0.28003, + "grad_norm": 0.9303097724914551, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 28003 + }, + { + "epoch": 0.28004, + "grad_norm": 1.0655561685562134, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 28004 + }, + { + "epoch": 0.28005, + "grad_norm": 1.076712965965271, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 28005 + }, + { + "epoch": 0.28006, + "grad_norm": 0.8724036812782288, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 28006 + }, + { + "epoch": 0.28007, + "grad_norm": 0.6995612978935242, + "learning_rate": 0.003, + "loss": 4.005, + "step": 28007 + }, + { + "epoch": 0.28008, + "grad_norm": 0.7123009562492371, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 28008 + }, + { + "epoch": 0.28009, + "grad_norm": 0.705467700958252, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 28009 + }, + { + "epoch": 0.2801, + "grad_norm": 0.7540966272354126, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 28010 + }, + { + "epoch": 0.28011, + "grad_norm": 0.8740915060043335, + "learning_rate": 0.003, + "loss": 4.002, + "step": 28011 + }, + { + "epoch": 0.28012, + "grad_norm": 1.1475307941436768, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 28012 + }, + { + "epoch": 0.28013, + "grad_norm": 1.0170767307281494, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 28013 + }, + { + "epoch": 0.28014, + "grad_norm": 0.919968843460083, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 28014 + }, + { + "epoch": 0.28015, + "grad_norm": 0.898516833782196, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 28015 + }, + { + "epoch": 0.28016, + "grad_norm": 0.7948172092437744, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 28016 + }, + { + "epoch": 0.28017, + "grad_norm": 0.6527708172798157, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 28017 + }, + { + "epoch": 0.28018, + "grad_norm": 0.6162542700767517, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 28018 + }, + { + "epoch": 0.28019, + "grad_norm": 0.6477906703948975, + "learning_rate": 0.003, + "loss": 3.972, + "step": 28019 + }, + { + "epoch": 0.2802, + "grad_norm": 0.6946806311607361, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 28020 + }, + { + "epoch": 0.28021, + "grad_norm": 0.7504633069038391, + "learning_rate": 0.003, + "loss": 4.002, + "step": 28021 + }, + { + "epoch": 0.28022, + "grad_norm": 0.886394739151001, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 28022 + }, + { + "epoch": 0.28023, + "grad_norm": 0.9345799684524536, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 28023 + }, + { + "epoch": 0.28024, + "grad_norm": 0.8718189001083374, + "learning_rate": 0.003, + "loss": 3.971, + "step": 28024 + }, + { + "epoch": 0.28025, + "grad_norm": 1.0935503244400024, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 28025 + }, + { + "epoch": 0.28026, + "grad_norm": 0.8928748965263367, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 28026 + }, + { + "epoch": 0.28027, + "grad_norm": 0.8563785552978516, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 28027 + }, + { + "epoch": 0.28028, + "grad_norm": 0.8536560535430908, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 28028 + }, + { + "epoch": 0.28029, + "grad_norm": 0.9550771713256836, + "learning_rate": 0.003, + "loss": 3.983, + "step": 28029 + }, + { + "epoch": 0.2803, + "grad_norm": 1.0775690078735352, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 28030 + }, + { + "epoch": 0.28031, + "grad_norm": 1.0150153636932373, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 28031 + }, + { + "epoch": 0.28032, + "grad_norm": 0.9562099575996399, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 28032 + }, + { + "epoch": 0.28033, + "grad_norm": 0.9163222312927246, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 28033 + }, + { + "epoch": 0.28034, + "grad_norm": 0.7881798148155212, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 28034 + }, + { + "epoch": 0.28035, + "grad_norm": 0.8334857225418091, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 28035 + }, + { + "epoch": 0.28036, + "grad_norm": 0.910777747631073, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 28036 + }, + { + "epoch": 0.28037, + "grad_norm": 0.9865319132804871, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 28037 + }, + { + "epoch": 0.28038, + "grad_norm": 0.9942948818206787, + "learning_rate": 0.003, + "loss": 3.988, + "step": 28038 + }, + { + "epoch": 0.28039, + "grad_norm": 1.050466775894165, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 28039 + }, + { + "epoch": 0.2804, + "grad_norm": 0.9826627373695374, + "learning_rate": 0.003, + "loss": 3.992, + "step": 28040 + }, + { + "epoch": 0.28041, + "grad_norm": 0.8827066421508789, + "learning_rate": 0.003, + "loss": 3.998, + "step": 28041 + }, + { + "epoch": 0.28042, + "grad_norm": 0.9923157691955566, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 28042 + }, + { + "epoch": 0.28043, + "grad_norm": 1.1366883516311646, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 28043 + }, + { + "epoch": 0.28044, + "grad_norm": 0.8214313983917236, + "learning_rate": 0.003, + "loss": 4.009, + "step": 28044 + }, + { + "epoch": 0.28045, + "grad_norm": 0.8683178424835205, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 28045 + }, + { + "epoch": 0.28046, + "grad_norm": 1.0448417663574219, + "learning_rate": 0.003, + "loss": 4.031, + "step": 28046 + }, + { + "epoch": 0.28047, + "grad_norm": 1.0810054540634155, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 28047 + }, + { + "epoch": 0.28048, + "grad_norm": 1.0368294715881348, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 28048 + }, + { + "epoch": 0.28049, + "grad_norm": 0.9151131510734558, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 28049 + }, + { + "epoch": 0.2805, + "grad_norm": 0.8094285130500793, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 28050 + }, + { + "epoch": 0.28051, + "grad_norm": 0.7310207486152649, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 28051 + }, + { + "epoch": 0.28052, + "grad_norm": 0.7244188189506531, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 28052 + }, + { + "epoch": 0.28053, + "grad_norm": 0.9477925896644592, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 28053 + }, + { + "epoch": 0.28054, + "grad_norm": 1.0533430576324463, + "learning_rate": 0.003, + "loss": 3.975, + "step": 28054 + }, + { + "epoch": 0.28055, + "grad_norm": 0.8127114772796631, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 28055 + }, + { + "epoch": 0.28056, + "grad_norm": 0.8363418579101562, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 28056 + }, + { + "epoch": 0.28057, + "grad_norm": 0.9006823897361755, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 28057 + }, + { + "epoch": 0.28058, + "grad_norm": 1.2030056715011597, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 28058 + }, + { + "epoch": 0.28059, + "grad_norm": 0.7985123991966248, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 28059 + }, + { + "epoch": 0.2806, + "grad_norm": 0.6712624430656433, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 28060 + }, + { + "epoch": 0.28061, + "grad_norm": 0.7235991358757019, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 28061 + }, + { + "epoch": 0.28062, + "grad_norm": 0.6027777194976807, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 28062 + }, + { + "epoch": 0.28063, + "grad_norm": 0.5721040964126587, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 28063 + }, + { + "epoch": 0.28064, + "grad_norm": 0.5519540905952454, + "learning_rate": 0.003, + "loss": 3.962, + "step": 28064 + }, + { + "epoch": 0.28065, + "grad_norm": 0.7200703024864197, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 28065 + }, + { + "epoch": 0.28066, + "grad_norm": 0.9409999847412109, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 28066 + }, + { + "epoch": 0.28067, + "grad_norm": 1.203487753868103, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 28067 + }, + { + "epoch": 0.28068, + "grad_norm": 0.6718849539756775, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 28068 + }, + { + "epoch": 0.28069, + "grad_norm": 0.5697625875473022, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 28069 + }, + { + "epoch": 0.2807, + "grad_norm": 0.6122552752494812, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 28070 + }, + { + "epoch": 0.28071, + "grad_norm": 0.7346797585487366, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 28071 + }, + { + "epoch": 0.28072, + "grad_norm": 0.9388381242752075, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 28072 + }, + { + "epoch": 0.28073, + "grad_norm": 1.0284267663955688, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 28073 + }, + { + "epoch": 0.28074, + "grad_norm": 0.9556589722633362, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 28074 + }, + { + "epoch": 0.28075, + "grad_norm": 0.9155338406562805, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 28075 + }, + { + "epoch": 0.28076, + "grad_norm": 0.8468285202980042, + "learning_rate": 0.003, + "loss": 3.964, + "step": 28076 + }, + { + "epoch": 0.28077, + "grad_norm": 0.7861206531524658, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 28077 + }, + { + "epoch": 0.28078, + "grad_norm": 0.7558118104934692, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 28078 + }, + { + "epoch": 0.28079, + "grad_norm": 0.8134294152259827, + "learning_rate": 0.003, + "loss": 3.962, + "step": 28079 + }, + { + "epoch": 0.2808, + "grad_norm": 0.8861095905303955, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 28080 + }, + { + "epoch": 0.28081, + "grad_norm": 1.1685034036636353, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 28081 + }, + { + "epoch": 0.28082, + "grad_norm": 0.9859872460365295, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 28082 + }, + { + "epoch": 0.28083, + "grad_norm": 0.853253185749054, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 28083 + }, + { + "epoch": 0.28084, + "grad_norm": 0.8980066776275635, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 28084 + }, + { + "epoch": 0.28085, + "grad_norm": 1.1378260850906372, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 28085 + }, + { + "epoch": 0.28086, + "grad_norm": 0.7693279385566711, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 28086 + }, + { + "epoch": 0.28087, + "grad_norm": 0.7046682834625244, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 28087 + }, + { + "epoch": 0.28088, + "grad_norm": 0.7984728813171387, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 28088 + }, + { + "epoch": 0.28089, + "grad_norm": 0.7653061151504517, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 28089 + }, + { + "epoch": 0.2809, + "grad_norm": 0.8639857172966003, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 28090 + }, + { + "epoch": 0.28091, + "grad_norm": 1.2494438886642456, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 28091 + }, + { + "epoch": 0.28092, + "grad_norm": 0.8435816168785095, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 28092 + }, + { + "epoch": 0.28093, + "grad_norm": 0.8402063846588135, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 28093 + }, + { + "epoch": 0.28094, + "grad_norm": 0.7691982984542847, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 28094 + }, + { + "epoch": 0.28095, + "grad_norm": 0.6840458512306213, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 28095 + }, + { + "epoch": 0.28096, + "grad_norm": 0.8768066763877869, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 28096 + }, + { + "epoch": 0.28097, + "grad_norm": 1.069383978843689, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 28097 + }, + { + "epoch": 0.28098, + "grad_norm": 0.9912194609642029, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 28098 + }, + { + "epoch": 0.28099, + "grad_norm": 0.9437240958213806, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 28099 + }, + { + "epoch": 0.281, + "grad_norm": 0.9491842985153198, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 28100 + }, + { + "epoch": 0.28101, + "grad_norm": 1.1180216073989868, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 28101 + }, + { + "epoch": 0.28102, + "grad_norm": 0.9096631407737732, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 28102 + }, + { + "epoch": 0.28103, + "grad_norm": 0.8980797529220581, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 28103 + }, + { + "epoch": 0.28104, + "grad_norm": 1.001442790031433, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 28104 + }, + { + "epoch": 0.28105, + "grad_norm": 1.1718544960021973, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 28105 + }, + { + "epoch": 0.28106, + "grad_norm": 0.8204549551010132, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 28106 + }, + { + "epoch": 0.28107, + "grad_norm": 0.8064346313476562, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 28107 + }, + { + "epoch": 0.28108, + "grad_norm": 0.7647159099578857, + "learning_rate": 0.003, + "loss": 4.004, + "step": 28108 + }, + { + "epoch": 0.28109, + "grad_norm": 0.6047796010971069, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 28109 + }, + { + "epoch": 0.2811, + "grad_norm": 0.6091798543930054, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 28110 + }, + { + "epoch": 0.28111, + "grad_norm": 0.7285879850387573, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 28111 + }, + { + "epoch": 0.28112, + "grad_norm": 0.9307032823562622, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 28112 + }, + { + "epoch": 0.28113, + "grad_norm": 1.0864630937576294, + "learning_rate": 0.003, + "loss": 4.001, + "step": 28113 + }, + { + "epoch": 0.28114, + "grad_norm": 0.8894389271736145, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 28114 + }, + { + "epoch": 0.28115, + "grad_norm": 0.720097005367279, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 28115 + }, + { + "epoch": 0.28116, + "grad_norm": 0.6336705088615417, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 28116 + }, + { + "epoch": 0.28117, + "grad_norm": 0.6234661936759949, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 28117 + }, + { + "epoch": 0.28118, + "grad_norm": 0.6438164114952087, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 28118 + }, + { + "epoch": 0.28119, + "grad_norm": 0.6752090454101562, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 28119 + }, + { + "epoch": 0.2812, + "grad_norm": 0.7544713616371155, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 28120 + }, + { + "epoch": 0.28121, + "grad_norm": 1.0361852645874023, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 28121 + }, + { + "epoch": 0.28122, + "grad_norm": 1.0524405241012573, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 28122 + }, + { + "epoch": 0.28123, + "grad_norm": 0.6551626920700073, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 28123 + }, + { + "epoch": 0.28124, + "grad_norm": 0.5907923579216003, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 28124 + }, + { + "epoch": 0.28125, + "grad_norm": 0.7255517840385437, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 28125 + }, + { + "epoch": 0.28126, + "grad_norm": 0.870631217956543, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 28126 + }, + { + "epoch": 0.28127, + "grad_norm": 1.0482186079025269, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 28127 + }, + { + "epoch": 0.28128, + "grad_norm": 0.963068425655365, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 28128 + }, + { + "epoch": 0.28129, + "grad_norm": 0.8812925219535828, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 28129 + }, + { + "epoch": 0.2813, + "grad_norm": 1.0045311450958252, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 28130 + }, + { + "epoch": 0.28131, + "grad_norm": 1.0961953401565552, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 28131 + }, + { + "epoch": 0.28132, + "grad_norm": 0.9322014451026917, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 28132 + }, + { + "epoch": 0.28133, + "grad_norm": 1.131105899810791, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 28133 + }, + { + "epoch": 0.28134, + "grad_norm": 0.9428444504737854, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 28134 + }, + { + "epoch": 0.28135, + "grad_norm": 0.9698842167854309, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 28135 + }, + { + "epoch": 0.28136, + "grad_norm": 1.053078532218933, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 28136 + }, + { + "epoch": 0.28137, + "grad_norm": 0.96568363904953, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 28137 + }, + { + "epoch": 0.28138, + "grad_norm": 0.8671183586120605, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 28138 + }, + { + "epoch": 0.28139, + "grad_norm": 0.7763950824737549, + "learning_rate": 0.003, + "loss": 4.001, + "step": 28139 + }, + { + "epoch": 0.2814, + "grad_norm": 0.8313345909118652, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 28140 + }, + { + "epoch": 0.28141, + "grad_norm": 0.9500892758369446, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 28141 + }, + { + "epoch": 0.28142, + "grad_norm": 0.8887403011322021, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 28142 + }, + { + "epoch": 0.28143, + "grad_norm": 0.9415918588638306, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 28143 + }, + { + "epoch": 0.28144, + "grad_norm": 1.1329251527786255, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 28144 + }, + { + "epoch": 0.28145, + "grad_norm": 0.8397986888885498, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 28145 + }, + { + "epoch": 0.28146, + "grad_norm": 0.7886302471160889, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 28146 + }, + { + "epoch": 0.28147, + "grad_norm": 0.706418514251709, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 28147 + }, + { + "epoch": 0.28148, + "grad_norm": 0.7218039631843567, + "learning_rate": 0.003, + "loss": 3.966, + "step": 28148 + }, + { + "epoch": 0.28149, + "grad_norm": 0.7978153228759766, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 28149 + }, + { + "epoch": 0.2815, + "grad_norm": 0.9879612922668457, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 28150 + }, + { + "epoch": 0.28151, + "grad_norm": 1.1444724798202515, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 28151 + }, + { + "epoch": 0.28152, + "grad_norm": 0.8184658885002136, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 28152 + }, + { + "epoch": 0.28153, + "grad_norm": 0.7681998014450073, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 28153 + }, + { + "epoch": 0.28154, + "grad_norm": 0.9024993777275085, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 28154 + }, + { + "epoch": 0.28155, + "grad_norm": 1.0558745861053467, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 28155 + }, + { + "epoch": 0.28156, + "grad_norm": 1.013386607170105, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 28156 + }, + { + "epoch": 0.28157, + "grad_norm": 1.1688734292984009, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 28157 + }, + { + "epoch": 0.28158, + "grad_norm": 0.8869865536689758, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 28158 + }, + { + "epoch": 0.28159, + "grad_norm": 0.7299702763557434, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 28159 + }, + { + "epoch": 0.2816, + "grad_norm": 0.8300586342811584, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 28160 + }, + { + "epoch": 0.28161, + "grad_norm": 0.8836418986320496, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 28161 + }, + { + "epoch": 0.28162, + "grad_norm": 0.8306666016578674, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 28162 + }, + { + "epoch": 0.28163, + "grad_norm": 0.8514958024024963, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 28163 + }, + { + "epoch": 0.28164, + "grad_norm": 0.8538970947265625, + "learning_rate": 0.003, + "loss": 3.996, + "step": 28164 + }, + { + "epoch": 0.28165, + "grad_norm": 0.8820816874504089, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 28165 + }, + { + "epoch": 0.28166, + "grad_norm": 0.7938548922538757, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 28166 + }, + { + "epoch": 0.28167, + "grad_norm": 0.8435572385787964, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 28167 + }, + { + "epoch": 0.28168, + "grad_norm": 0.8815374970436096, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 28168 + }, + { + "epoch": 0.28169, + "grad_norm": 1.0968661308288574, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 28169 + }, + { + "epoch": 0.2817, + "grad_norm": 1.0195279121398926, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 28170 + }, + { + "epoch": 0.28171, + "grad_norm": 0.9155490398406982, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 28171 + }, + { + "epoch": 0.28172, + "grad_norm": 0.8858137130737305, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 28172 + }, + { + "epoch": 0.28173, + "grad_norm": 0.8097586035728455, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 28173 + }, + { + "epoch": 0.28174, + "grad_norm": 0.8612500429153442, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 28174 + }, + { + "epoch": 0.28175, + "grad_norm": 0.6913039684295654, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 28175 + }, + { + "epoch": 0.28176, + "grad_norm": 0.6258280873298645, + "learning_rate": 0.003, + "loss": 3.968, + "step": 28176 + }, + { + "epoch": 0.28177, + "grad_norm": 0.6701483130455017, + "learning_rate": 0.003, + "loss": 3.99, + "step": 28177 + }, + { + "epoch": 0.28178, + "grad_norm": 0.6290279626846313, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 28178 + }, + { + "epoch": 0.28179, + "grad_norm": 0.5758747458457947, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 28179 + }, + { + "epoch": 0.2818, + "grad_norm": 0.7551039457321167, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 28180 + }, + { + "epoch": 0.28181, + "grad_norm": 0.8342494368553162, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 28181 + }, + { + "epoch": 0.28182, + "grad_norm": 0.9032070636749268, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 28182 + }, + { + "epoch": 0.28183, + "grad_norm": 1.072236180305481, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 28183 + }, + { + "epoch": 0.28184, + "grad_norm": 0.9327178597450256, + "learning_rate": 0.003, + "loss": 3.9254, + "step": 28184 + }, + { + "epoch": 0.28185, + "grad_norm": 0.7676780819892883, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 28185 + }, + { + "epoch": 0.28186, + "grad_norm": 0.8014476895332336, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 28186 + }, + { + "epoch": 0.28187, + "grad_norm": 0.8759790062904358, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 28187 + }, + { + "epoch": 0.28188, + "grad_norm": 0.9651340842247009, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 28188 + }, + { + "epoch": 0.28189, + "grad_norm": 1.199195146560669, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 28189 + }, + { + "epoch": 0.2819, + "grad_norm": 0.8995957374572754, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 28190 + }, + { + "epoch": 0.28191, + "grad_norm": 0.9609403610229492, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 28191 + }, + { + "epoch": 0.28192, + "grad_norm": 1.2720884084701538, + "learning_rate": 0.003, + "loss": 4.024, + "step": 28192 + }, + { + "epoch": 0.28193, + "grad_norm": 0.8241503238677979, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 28193 + }, + { + "epoch": 0.28194, + "grad_norm": 0.7152815461158752, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 28194 + }, + { + "epoch": 0.28195, + "grad_norm": 0.7517530918121338, + "learning_rate": 0.003, + "loss": 3.987, + "step": 28195 + }, + { + "epoch": 0.28196, + "grad_norm": 0.8125039935112, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 28196 + }, + { + "epoch": 0.28197, + "grad_norm": 0.7432599663734436, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 28197 + }, + { + "epoch": 0.28198, + "grad_norm": 0.7398964166641235, + "learning_rate": 0.003, + "loss": 4.02, + "step": 28198 + }, + { + "epoch": 0.28199, + "grad_norm": 0.7316308617591858, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 28199 + }, + { + "epoch": 0.282, + "grad_norm": 0.7118121385574341, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 28200 + }, + { + "epoch": 0.28201, + "grad_norm": 0.710850715637207, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 28201 + }, + { + "epoch": 0.28202, + "grad_norm": 0.7224267721176147, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 28202 + }, + { + "epoch": 0.28203, + "grad_norm": 0.6967412233352661, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 28203 + }, + { + "epoch": 0.28204, + "grad_norm": 0.8825206160545349, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 28204 + }, + { + "epoch": 0.28205, + "grad_norm": 1.2199665307998657, + "learning_rate": 0.003, + "loss": 4.006, + "step": 28205 + }, + { + "epoch": 0.28206, + "grad_norm": 0.8089765906333923, + "learning_rate": 0.003, + "loss": 3.994, + "step": 28206 + }, + { + "epoch": 0.28207, + "grad_norm": 0.6690025329589844, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 28207 + }, + { + "epoch": 0.28208, + "grad_norm": 0.6555092334747314, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 28208 + }, + { + "epoch": 0.28209, + "grad_norm": 0.6694302558898926, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 28209 + }, + { + "epoch": 0.2821, + "grad_norm": 0.8655672073364258, + "learning_rate": 0.003, + "loss": 3.943, + "step": 28210 + }, + { + "epoch": 0.28211, + "grad_norm": 1.1464319229125977, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 28211 + }, + { + "epoch": 0.28212, + "grad_norm": 0.8974980711936951, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 28212 + }, + { + "epoch": 0.28213, + "grad_norm": 0.8577998280525208, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 28213 + }, + { + "epoch": 0.28214, + "grad_norm": 0.8548339009284973, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 28214 + }, + { + "epoch": 0.28215, + "grad_norm": 0.6758325695991516, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 28215 + }, + { + "epoch": 0.28216, + "grad_norm": 0.6070768237113953, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 28216 + }, + { + "epoch": 0.28217, + "grad_norm": 0.7416322827339172, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 28217 + }, + { + "epoch": 0.28218, + "grad_norm": 0.9519782662391663, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 28218 + }, + { + "epoch": 0.28219, + "grad_norm": 0.9918175339698792, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 28219 + }, + { + "epoch": 0.2822, + "grad_norm": 0.9144274592399597, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 28220 + }, + { + "epoch": 0.28221, + "grad_norm": 0.9143868684768677, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 28221 + }, + { + "epoch": 0.28222, + "grad_norm": 1.0406198501586914, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 28222 + }, + { + "epoch": 0.28223, + "grad_norm": 1.076452612876892, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 28223 + }, + { + "epoch": 0.28224, + "grad_norm": 0.9073577523231506, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 28224 + }, + { + "epoch": 0.28225, + "grad_norm": 0.8666258454322815, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 28225 + }, + { + "epoch": 0.28226, + "grad_norm": 0.8291043043136597, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 28226 + }, + { + "epoch": 0.28227, + "grad_norm": 0.7271919250488281, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 28227 + }, + { + "epoch": 0.28228, + "grad_norm": 0.7047246694564819, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 28228 + }, + { + "epoch": 0.28229, + "grad_norm": 0.7411260008811951, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 28229 + }, + { + "epoch": 0.2823, + "grad_norm": 0.8109962940216064, + "learning_rate": 0.003, + "loss": 3.962, + "step": 28230 + }, + { + "epoch": 0.28231, + "grad_norm": 0.7907114624977112, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 28231 + }, + { + "epoch": 0.28232, + "grad_norm": 0.8694118857383728, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 28232 + }, + { + "epoch": 0.28233, + "grad_norm": 0.9229956865310669, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 28233 + }, + { + "epoch": 0.28234, + "grad_norm": 1.0789731740951538, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 28234 + }, + { + "epoch": 0.28235, + "grad_norm": 1.0362281799316406, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 28235 + }, + { + "epoch": 0.28236, + "grad_norm": 1.0868817567825317, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 28236 + }, + { + "epoch": 0.28237, + "grad_norm": 1.0108400583267212, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 28237 + }, + { + "epoch": 0.28238, + "grad_norm": 0.9263377785682678, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 28238 + }, + { + "epoch": 0.28239, + "grad_norm": 0.7835015058517456, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 28239 + }, + { + "epoch": 0.2824, + "grad_norm": 0.8188238739967346, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 28240 + }, + { + "epoch": 0.28241, + "grad_norm": 0.8736681938171387, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 28241 + }, + { + "epoch": 0.28242, + "grad_norm": 1.0038799047470093, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 28242 + }, + { + "epoch": 0.28243, + "grad_norm": 1.0271044969558716, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 28243 + }, + { + "epoch": 0.28244, + "grad_norm": 0.9738487005233765, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 28244 + }, + { + "epoch": 0.28245, + "grad_norm": 1.0952926874160767, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 28245 + }, + { + "epoch": 0.28246, + "grad_norm": 0.8878053426742554, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 28246 + }, + { + "epoch": 0.28247, + "grad_norm": 0.7013972997665405, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 28247 + }, + { + "epoch": 0.28248, + "grad_norm": 0.7192615866661072, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 28248 + }, + { + "epoch": 0.28249, + "grad_norm": 0.7681589722633362, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 28249 + }, + { + "epoch": 0.2825, + "grad_norm": 0.8447195291519165, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 28250 + }, + { + "epoch": 0.28251, + "grad_norm": 0.95870441198349, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 28251 + }, + { + "epoch": 0.28252, + "grad_norm": 1.0330144166946411, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 28252 + }, + { + "epoch": 0.28253, + "grad_norm": 0.8654082417488098, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 28253 + }, + { + "epoch": 0.28254, + "grad_norm": 0.9333270192146301, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 28254 + }, + { + "epoch": 0.28255, + "grad_norm": 0.9377995133399963, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 28255 + }, + { + "epoch": 0.28256, + "grad_norm": 0.8972247838973999, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 28256 + }, + { + "epoch": 0.28257, + "grad_norm": 1.1581801176071167, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 28257 + }, + { + "epoch": 0.28258, + "grad_norm": 0.867113471031189, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 28258 + }, + { + "epoch": 0.28259, + "grad_norm": 0.7649432420730591, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 28259 + }, + { + "epoch": 0.2826, + "grad_norm": 0.751146137714386, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 28260 + }, + { + "epoch": 0.28261, + "grad_norm": 0.7327833771705627, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 28261 + }, + { + "epoch": 0.28262, + "grad_norm": 0.8004399538040161, + "learning_rate": 0.003, + "loss": 3.9442, + "step": 28262 + }, + { + "epoch": 0.28263, + "grad_norm": 0.8902596235275269, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 28263 + }, + { + "epoch": 0.28264, + "grad_norm": 0.8758144378662109, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 28264 + }, + { + "epoch": 0.28265, + "grad_norm": 0.8940191268920898, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 28265 + }, + { + "epoch": 0.28266, + "grad_norm": 0.9842545390129089, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 28266 + }, + { + "epoch": 0.28267, + "grad_norm": 1.184718370437622, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 28267 + }, + { + "epoch": 0.28268, + "grad_norm": 0.7736249566078186, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 28268 + }, + { + "epoch": 0.28269, + "grad_norm": 0.672602117061615, + "learning_rate": 0.003, + "loss": 3.982, + "step": 28269 + }, + { + "epoch": 0.2827, + "grad_norm": 0.7421377301216125, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 28270 + }, + { + "epoch": 0.28271, + "grad_norm": 1.1061298847198486, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 28271 + }, + { + "epoch": 0.28272, + "grad_norm": 1.0099159479141235, + "learning_rate": 0.003, + "loss": 3.953, + "step": 28272 + }, + { + "epoch": 0.28273, + "grad_norm": 0.9220610857009888, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 28273 + }, + { + "epoch": 0.28274, + "grad_norm": 0.9376311302185059, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 28274 + }, + { + "epoch": 0.28275, + "grad_norm": 0.945244312286377, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 28275 + }, + { + "epoch": 0.28276, + "grad_norm": 0.788723349571228, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 28276 + }, + { + "epoch": 0.28277, + "grad_norm": 0.7244881987571716, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 28277 + }, + { + "epoch": 0.28278, + "grad_norm": 0.7880487442016602, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 28278 + }, + { + "epoch": 0.28279, + "grad_norm": 0.7756094932556152, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 28279 + }, + { + "epoch": 0.2828, + "grad_norm": 0.7567785382270813, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 28280 + }, + { + "epoch": 0.28281, + "grad_norm": 0.7093895077705383, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 28281 + }, + { + "epoch": 0.28282, + "grad_norm": 0.6929912567138672, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 28282 + }, + { + "epoch": 0.28283, + "grad_norm": 0.7265345454216003, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 28283 + }, + { + "epoch": 0.28284, + "grad_norm": 0.6563029289245605, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 28284 + }, + { + "epoch": 0.28285, + "grad_norm": 0.7335759401321411, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 28285 + }, + { + "epoch": 0.28286, + "grad_norm": 0.9198670983314514, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 28286 + }, + { + "epoch": 0.28287, + "grad_norm": 1.410281777381897, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 28287 + }, + { + "epoch": 0.28288, + "grad_norm": 0.5311173796653748, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 28288 + }, + { + "epoch": 0.28289, + "grad_norm": 0.8892162442207336, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 28289 + }, + { + "epoch": 0.2829, + "grad_norm": 1.2206262350082397, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 28290 + }, + { + "epoch": 0.28291, + "grad_norm": 0.7961776256561279, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 28291 + }, + { + "epoch": 0.28292, + "grad_norm": 0.6026272177696228, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 28292 + }, + { + "epoch": 0.28293, + "grad_norm": 0.648118257522583, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 28293 + }, + { + "epoch": 0.28294, + "grad_norm": 0.6796482801437378, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 28294 + }, + { + "epoch": 0.28295, + "grad_norm": 0.6342206001281738, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 28295 + }, + { + "epoch": 0.28296, + "grad_norm": 0.5825303792953491, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 28296 + }, + { + "epoch": 0.28297, + "grad_norm": 0.5659521222114563, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 28297 + }, + { + "epoch": 0.28298, + "grad_norm": 0.6755831837654114, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 28298 + }, + { + "epoch": 0.28299, + "grad_norm": 0.9651461839675903, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 28299 + }, + { + "epoch": 0.283, + "grad_norm": 1.2552289962768555, + "learning_rate": 0.003, + "loss": 3.9468, + "step": 28300 + }, + { + "epoch": 0.28301, + "grad_norm": 0.9193450808525085, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 28301 + }, + { + "epoch": 0.28302, + "grad_norm": 0.9583123326301575, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 28302 + }, + { + "epoch": 0.28303, + "grad_norm": 0.9180430769920349, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 28303 + }, + { + "epoch": 0.28304, + "grad_norm": 1.1599667072296143, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 28304 + }, + { + "epoch": 0.28305, + "grad_norm": 0.9351750612258911, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 28305 + }, + { + "epoch": 0.28306, + "grad_norm": 0.7567969560623169, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 28306 + }, + { + "epoch": 0.28307, + "grad_norm": 0.7039915919303894, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 28307 + }, + { + "epoch": 0.28308, + "grad_norm": 0.7531316876411438, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 28308 + }, + { + "epoch": 0.28309, + "grad_norm": 0.8406116366386414, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 28309 + }, + { + "epoch": 0.2831, + "grad_norm": 0.9203659892082214, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 28310 + }, + { + "epoch": 0.28311, + "grad_norm": 1.1296354532241821, + "learning_rate": 0.003, + "loss": 3.979, + "step": 28311 + }, + { + "epoch": 0.28312, + "grad_norm": 1.048190712928772, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 28312 + }, + { + "epoch": 0.28313, + "grad_norm": 1.0188981294631958, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 28313 + }, + { + "epoch": 0.28314, + "grad_norm": 0.9437536597251892, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 28314 + }, + { + "epoch": 0.28315, + "grad_norm": 0.9790985584259033, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 28315 + }, + { + "epoch": 0.28316, + "grad_norm": 1.0365021228790283, + "learning_rate": 0.003, + "loss": 3.988, + "step": 28316 + }, + { + "epoch": 0.28317, + "grad_norm": 1.2152554988861084, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 28317 + }, + { + "epoch": 0.28318, + "grad_norm": 0.9137653708457947, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 28318 + }, + { + "epoch": 0.28319, + "grad_norm": 0.9851117134094238, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 28319 + }, + { + "epoch": 0.2832, + "grad_norm": 0.9505988955497742, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 28320 + }, + { + "epoch": 0.28321, + "grad_norm": 1.004146695137024, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 28321 + }, + { + "epoch": 0.28322, + "grad_norm": 1.0989820957183838, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 28322 + }, + { + "epoch": 0.28323, + "grad_norm": 0.8711421489715576, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 28323 + }, + { + "epoch": 0.28324, + "grad_norm": 0.8665750026702881, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 28324 + }, + { + "epoch": 0.28325, + "grad_norm": 0.9028405547142029, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 28325 + }, + { + "epoch": 0.28326, + "grad_norm": 1.0594485998153687, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 28326 + }, + { + "epoch": 0.28327, + "grad_norm": 1.0426563024520874, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 28327 + }, + { + "epoch": 0.28328, + "grad_norm": 0.7066168785095215, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 28328 + }, + { + "epoch": 0.28329, + "grad_norm": 0.6255882382392883, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 28329 + }, + { + "epoch": 0.2833, + "grad_norm": 0.8477040529251099, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 28330 + }, + { + "epoch": 0.28331, + "grad_norm": 0.8305228352546692, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 28331 + }, + { + "epoch": 0.28332, + "grad_norm": 0.8315953612327576, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 28332 + }, + { + "epoch": 0.28333, + "grad_norm": 0.8528433442115784, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 28333 + }, + { + "epoch": 0.28334, + "grad_norm": 0.7634260654449463, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 28334 + }, + { + "epoch": 0.28335, + "grad_norm": 0.682229220867157, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 28335 + }, + { + "epoch": 0.28336, + "grad_norm": 0.6495056748390198, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 28336 + }, + { + "epoch": 0.28337, + "grad_norm": 0.9332974553108215, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 28337 + }, + { + "epoch": 0.28338, + "grad_norm": 1.1316347122192383, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 28338 + }, + { + "epoch": 0.28339, + "grad_norm": 0.8880043625831604, + "learning_rate": 0.003, + "loss": 3.992, + "step": 28339 + }, + { + "epoch": 0.2834, + "grad_norm": 0.8643243312835693, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 28340 + }, + { + "epoch": 0.28341, + "grad_norm": 0.8617450594902039, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 28341 + }, + { + "epoch": 0.28342, + "grad_norm": 0.8886629939079285, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 28342 + }, + { + "epoch": 0.28343, + "grad_norm": 1.0440126657485962, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 28343 + }, + { + "epoch": 0.28344, + "grad_norm": 1.1481295824050903, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 28344 + }, + { + "epoch": 0.28345, + "grad_norm": 0.860749363899231, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 28345 + }, + { + "epoch": 0.28346, + "grad_norm": 0.6513493657112122, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 28346 + }, + { + "epoch": 0.28347, + "grad_norm": 0.9491482377052307, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 28347 + }, + { + "epoch": 0.28348, + "grad_norm": 1.03617525100708, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 28348 + }, + { + "epoch": 0.28349, + "grad_norm": 0.9842594861984253, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 28349 + }, + { + "epoch": 0.2835, + "grad_norm": 0.8462135195732117, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 28350 + }, + { + "epoch": 0.28351, + "grad_norm": 0.6516234278678894, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 28351 + }, + { + "epoch": 0.28352, + "grad_norm": 0.7368994951248169, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 28352 + }, + { + "epoch": 0.28353, + "grad_norm": 0.7708935141563416, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 28353 + }, + { + "epoch": 0.28354, + "grad_norm": 0.7989370822906494, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 28354 + }, + { + "epoch": 0.28355, + "grad_norm": 1.0919520854949951, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 28355 + }, + { + "epoch": 0.28356, + "grad_norm": 1.0968377590179443, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 28356 + }, + { + "epoch": 0.28357, + "grad_norm": 0.7845099568367004, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 28357 + }, + { + "epoch": 0.28358, + "grad_norm": 0.6964684128761292, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 28358 + }, + { + "epoch": 0.28359, + "grad_norm": 0.7284111380577087, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 28359 + }, + { + "epoch": 0.2836, + "grad_norm": 0.8928865790367126, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 28360 + }, + { + "epoch": 0.28361, + "grad_norm": 1.031795859336853, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 28361 + }, + { + "epoch": 0.28362, + "grad_norm": 0.944645881652832, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 28362 + }, + { + "epoch": 0.28363, + "grad_norm": 0.7839560508728027, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 28363 + }, + { + "epoch": 0.28364, + "grad_norm": 0.6103917956352234, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 28364 + }, + { + "epoch": 0.28365, + "grad_norm": 0.5713149905204773, + "learning_rate": 0.003, + "loss": 3.9368, + "step": 28365 + }, + { + "epoch": 0.28366, + "grad_norm": 0.6628426313400269, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 28366 + }, + { + "epoch": 0.28367, + "grad_norm": 0.6853655576705933, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 28367 + }, + { + "epoch": 0.28368, + "grad_norm": 0.7193369269371033, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 28368 + }, + { + "epoch": 0.28369, + "grad_norm": 0.7826032042503357, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 28369 + }, + { + "epoch": 0.2837, + "grad_norm": 0.7871755361557007, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 28370 + }, + { + "epoch": 0.28371, + "grad_norm": 0.7989972829818726, + "learning_rate": 0.003, + "loss": 3.9464, + "step": 28371 + }, + { + "epoch": 0.28372, + "grad_norm": 0.7901728749275208, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 28372 + }, + { + "epoch": 0.28373, + "grad_norm": 0.8818800449371338, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 28373 + }, + { + "epoch": 0.28374, + "grad_norm": 0.9527623057365417, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 28374 + }, + { + "epoch": 0.28375, + "grad_norm": 0.9083271622657776, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 28375 + }, + { + "epoch": 0.28376, + "grad_norm": 0.9846487641334534, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 28376 + }, + { + "epoch": 0.28377, + "grad_norm": 1.1770119667053223, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 28377 + }, + { + "epoch": 0.28378, + "grad_norm": 1.0570073127746582, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 28378 + }, + { + "epoch": 0.28379, + "grad_norm": 0.9042402505874634, + "learning_rate": 0.003, + "loss": 4.012, + "step": 28379 + }, + { + "epoch": 0.2838, + "grad_norm": 0.8664714694023132, + "learning_rate": 0.003, + "loss": 3.977, + "step": 28380 + }, + { + "epoch": 0.28381, + "grad_norm": 0.7914630770683289, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 28381 + }, + { + "epoch": 0.28382, + "grad_norm": 0.7762675881385803, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 28382 + }, + { + "epoch": 0.28383, + "grad_norm": 0.7529070973396301, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 28383 + }, + { + "epoch": 0.28384, + "grad_norm": 0.7753962874412537, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 28384 + }, + { + "epoch": 0.28385, + "grad_norm": 0.7736270427703857, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 28385 + }, + { + "epoch": 0.28386, + "grad_norm": 0.7044236660003662, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 28386 + }, + { + "epoch": 0.28387, + "grad_norm": 0.7249996066093445, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 28387 + }, + { + "epoch": 0.28388, + "grad_norm": 0.9839141964912415, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 28388 + }, + { + "epoch": 0.28389, + "grad_norm": 1.155472993850708, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 28389 + }, + { + "epoch": 0.2839, + "grad_norm": 0.877102255821228, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 28390 + }, + { + "epoch": 0.28391, + "grad_norm": 0.8855808973312378, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 28391 + }, + { + "epoch": 0.28392, + "grad_norm": 0.9862276911735535, + "learning_rate": 0.003, + "loss": 3.979, + "step": 28392 + }, + { + "epoch": 0.28393, + "grad_norm": 1.0914745330810547, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 28393 + }, + { + "epoch": 0.28394, + "grad_norm": 1.0252339839935303, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 28394 + }, + { + "epoch": 0.28395, + "grad_norm": 1.0536857843399048, + "learning_rate": 0.003, + "loss": 4.036, + "step": 28395 + }, + { + "epoch": 0.28396, + "grad_norm": 1.2006999254226685, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 28396 + }, + { + "epoch": 0.28397, + "grad_norm": 0.7658678293228149, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 28397 + }, + { + "epoch": 0.28398, + "grad_norm": 0.6544502377510071, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 28398 + }, + { + "epoch": 0.28399, + "grad_norm": 0.6363487839698792, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 28399 + }, + { + "epoch": 0.284, + "grad_norm": 0.6817333102226257, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 28400 + }, + { + "epoch": 0.28401, + "grad_norm": 0.8085818290710449, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 28401 + }, + { + "epoch": 0.28402, + "grad_norm": 0.97398442029953, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 28402 + }, + { + "epoch": 0.28403, + "grad_norm": 1.071284532546997, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 28403 + }, + { + "epoch": 0.28404, + "grad_norm": 1.0433417558670044, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 28404 + }, + { + "epoch": 0.28405, + "grad_norm": 0.9383113384246826, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 28405 + }, + { + "epoch": 0.28406, + "grad_norm": 0.9312393069267273, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 28406 + }, + { + "epoch": 0.28407, + "grad_norm": 0.9633320569992065, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 28407 + }, + { + "epoch": 0.28408, + "grad_norm": 1.0363105535507202, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 28408 + }, + { + "epoch": 0.28409, + "grad_norm": 0.9378328919410706, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 28409 + }, + { + "epoch": 0.2841, + "grad_norm": 0.8796053528785706, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 28410 + }, + { + "epoch": 0.28411, + "grad_norm": 0.8459492325782776, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 28411 + }, + { + "epoch": 0.28412, + "grad_norm": 0.8717344999313354, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 28412 + }, + { + "epoch": 0.28413, + "grad_norm": 0.7837488651275635, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 28413 + }, + { + "epoch": 0.28414, + "grad_norm": 0.8188886642456055, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 28414 + }, + { + "epoch": 0.28415, + "grad_norm": 0.8667516708374023, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 28415 + }, + { + "epoch": 0.28416, + "grad_norm": 0.9738795161247253, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 28416 + }, + { + "epoch": 0.28417, + "grad_norm": 1.0533355474472046, + "learning_rate": 0.003, + "loss": 3.974, + "step": 28417 + }, + { + "epoch": 0.28418, + "grad_norm": 0.8608596324920654, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 28418 + }, + { + "epoch": 0.28419, + "grad_norm": 0.9177553653717041, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 28419 + }, + { + "epoch": 0.2842, + "grad_norm": 1.0594284534454346, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 28420 + }, + { + "epoch": 0.28421, + "grad_norm": 1.0282220840454102, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 28421 + }, + { + "epoch": 0.28422, + "grad_norm": 0.949529230594635, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 28422 + }, + { + "epoch": 0.28423, + "grad_norm": 0.9822243452072144, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 28423 + }, + { + "epoch": 0.28424, + "grad_norm": 1.1076371669769287, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 28424 + }, + { + "epoch": 0.28425, + "grad_norm": 0.8326815366744995, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 28425 + }, + { + "epoch": 0.28426, + "grad_norm": 0.7476851344108582, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 28426 + }, + { + "epoch": 0.28427, + "grad_norm": 0.8394558429718018, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 28427 + }, + { + "epoch": 0.28428, + "grad_norm": 0.7517449259757996, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 28428 + }, + { + "epoch": 0.28429, + "grad_norm": 0.6769333481788635, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 28429 + }, + { + "epoch": 0.2843, + "grad_norm": 0.6571972966194153, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 28430 + }, + { + "epoch": 0.28431, + "grad_norm": 0.6934214234352112, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 28431 + }, + { + "epoch": 0.28432, + "grad_norm": 0.619805097579956, + "learning_rate": 0.003, + "loss": 3.955, + "step": 28432 + }, + { + "epoch": 0.28433, + "grad_norm": 0.6180689334869385, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 28433 + }, + { + "epoch": 0.28434, + "grad_norm": 0.6035190224647522, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 28434 + }, + { + "epoch": 0.28435, + "grad_norm": 0.6248013377189636, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 28435 + }, + { + "epoch": 0.28436, + "grad_norm": 0.6510832905769348, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 28436 + }, + { + "epoch": 0.28437, + "grad_norm": 0.6860239505767822, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 28437 + }, + { + "epoch": 0.28438, + "grad_norm": 0.6545128226280212, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 28438 + }, + { + "epoch": 0.28439, + "grad_norm": 0.6205263137817383, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 28439 + }, + { + "epoch": 0.2844, + "grad_norm": 0.7747625708580017, + "learning_rate": 0.003, + "loss": 3.968, + "step": 28440 + }, + { + "epoch": 0.28441, + "grad_norm": 1.1499238014221191, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 28441 + }, + { + "epoch": 0.28442, + "grad_norm": 1.247251033782959, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 28442 + }, + { + "epoch": 0.28443, + "grad_norm": 0.8900379538536072, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 28443 + }, + { + "epoch": 0.28444, + "grad_norm": 0.9597698450088501, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 28444 + }, + { + "epoch": 0.28445, + "grad_norm": 1.1726428270339966, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 28445 + }, + { + "epoch": 0.28446, + "grad_norm": 0.8535525798797607, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 28446 + }, + { + "epoch": 0.28447, + "grad_norm": 0.759624183177948, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 28447 + }, + { + "epoch": 0.28448, + "grad_norm": 0.7417036294937134, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 28448 + }, + { + "epoch": 0.28449, + "grad_norm": 0.8676761984825134, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 28449 + }, + { + "epoch": 0.2845, + "grad_norm": 0.9352858066558838, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 28450 + }, + { + "epoch": 0.28451, + "grad_norm": 1.1156691312789917, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 28451 + }, + { + "epoch": 0.28452, + "grad_norm": 1.0329469442367554, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 28452 + }, + { + "epoch": 0.28453, + "grad_norm": 0.8795492053031921, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 28453 + }, + { + "epoch": 0.28454, + "grad_norm": 0.8820512890815735, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 28454 + }, + { + "epoch": 0.28455, + "grad_norm": 0.9537266492843628, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 28455 + }, + { + "epoch": 0.28456, + "grad_norm": 0.9331933259963989, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 28456 + }, + { + "epoch": 0.28457, + "grad_norm": 0.8545373678207397, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 28457 + }, + { + "epoch": 0.28458, + "grad_norm": 0.7797852754592896, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 28458 + }, + { + "epoch": 0.28459, + "grad_norm": 0.7614102363586426, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 28459 + }, + { + "epoch": 0.2846, + "grad_norm": 0.7373015284538269, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 28460 + }, + { + "epoch": 0.28461, + "grad_norm": 0.8318608999252319, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 28461 + }, + { + "epoch": 0.28462, + "grad_norm": 1.0507146120071411, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 28462 + }, + { + "epoch": 0.28463, + "grad_norm": 1.176414132118225, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 28463 + }, + { + "epoch": 0.28464, + "grad_norm": 0.8835601806640625, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 28464 + }, + { + "epoch": 0.28465, + "grad_norm": 0.8986618518829346, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 28465 + }, + { + "epoch": 0.28466, + "grad_norm": 0.9616013169288635, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 28466 + }, + { + "epoch": 0.28467, + "grad_norm": 0.9902765154838562, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 28467 + }, + { + "epoch": 0.28468, + "grad_norm": 0.8869798183441162, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 28468 + }, + { + "epoch": 0.28469, + "grad_norm": 0.7587646842002869, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 28469 + }, + { + "epoch": 0.2847, + "grad_norm": 0.8901192545890808, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 28470 + }, + { + "epoch": 0.28471, + "grad_norm": 0.9013084769248962, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 28471 + }, + { + "epoch": 0.28472, + "grad_norm": 0.8857449293136597, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 28472 + }, + { + "epoch": 0.28473, + "grad_norm": 0.8305806517601013, + "learning_rate": 0.003, + "loss": 3.993, + "step": 28473 + }, + { + "epoch": 0.28474, + "grad_norm": 0.7371850609779358, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 28474 + }, + { + "epoch": 0.28475, + "grad_norm": 0.7753798365592957, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 28475 + }, + { + "epoch": 0.28476, + "grad_norm": 0.8286325335502625, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 28476 + }, + { + "epoch": 0.28477, + "grad_norm": 0.8798861503601074, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 28477 + }, + { + "epoch": 0.28478, + "grad_norm": 1.0372954607009888, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 28478 + }, + { + "epoch": 0.28479, + "grad_norm": 1.0896039009094238, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 28479 + }, + { + "epoch": 0.2848, + "grad_norm": 1.0095199346542358, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 28480 + }, + { + "epoch": 0.28481, + "grad_norm": 0.856417179107666, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 28481 + }, + { + "epoch": 0.28482, + "grad_norm": 0.8709612488746643, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 28482 + }, + { + "epoch": 0.28483, + "grad_norm": 0.9233640432357788, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 28483 + }, + { + "epoch": 0.28484, + "grad_norm": 1.0113693475723267, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 28484 + }, + { + "epoch": 0.28485, + "grad_norm": 0.9888996481895447, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 28485 + }, + { + "epoch": 0.28486, + "grad_norm": 0.9253531098365784, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 28486 + }, + { + "epoch": 0.28487, + "grad_norm": 0.9223387837409973, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 28487 + }, + { + "epoch": 0.28488, + "grad_norm": 1.0234888792037964, + "learning_rate": 0.003, + "loss": 3.98, + "step": 28488 + }, + { + "epoch": 0.28489, + "grad_norm": 0.9903175234794617, + "learning_rate": 0.003, + "loss": 3.991, + "step": 28489 + }, + { + "epoch": 0.2849, + "grad_norm": 0.8739856481552124, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 28490 + }, + { + "epoch": 0.28491, + "grad_norm": 0.9221113920211792, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 28491 + }, + { + "epoch": 0.28492, + "grad_norm": 1.1447842121124268, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 28492 + }, + { + "epoch": 0.28493, + "grad_norm": 1.2279328107833862, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 28493 + }, + { + "epoch": 0.28494, + "grad_norm": 0.9469074010848999, + "learning_rate": 0.003, + "loss": 4.023, + "step": 28494 + }, + { + "epoch": 0.28495, + "grad_norm": 0.9383217096328735, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 28495 + }, + { + "epoch": 0.28496, + "grad_norm": 0.7887699007987976, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 28496 + }, + { + "epoch": 0.28497, + "grad_norm": 0.7602590322494507, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 28497 + }, + { + "epoch": 0.28498, + "grad_norm": 0.6186055541038513, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 28498 + }, + { + "epoch": 0.28499, + "grad_norm": 0.6941701173782349, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 28499 + }, + { + "epoch": 0.285, + "grad_norm": 0.7202759981155396, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 28500 + }, + { + "epoch": 0.28501, + "grad_norm": 0.9308323264122009, + "learning_rate": 0.003, + "loss": 3.997, + "step": 28501 + }, + { + "epoch": 0.28502, + "grad_norm": 1.2977548837661743, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 28502 + }, + { + "epoch": 0.28503, + "grad_norm": 1.1781874895095825, + "learning_rate": 0.003, + "loss": 3.982, + "step": 28503 + }, + { + "epoch": 0.28504, + "grad_norm": 0.781490683555603, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 28504 + }, + { + "epoch": 0.28505, + "grad_norm": 0.7123475670814514, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 28505 + }, + { + "epoch": 0.28506, + "grad_norm": 0.8734947443008423, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 28506 + }, + { + "epoch": 0.28507, + "grad_norm": 0.9815630316734314, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 28507 + }, + { + "epoch": 0.28508, + "grad_norm": 1.0145596265792847, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 28508 + }, + { + "epoch": 0.28509, + "grad_norm": 0.8833717703819275, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 28509 + }, + { + "epoch": 0.2851, + "grad_norm": 0.805304229259491, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 28510 + }, + { + "epoch": 0.28511, + "grad_norm": 0.6793612837791443, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 28511 + }, + { + "epoch": 0.28512, + "grad_norm": 0.5940540432929993, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 28512 + }, + { + "epoch": 0.28513, + "grad_norm": 0.6180674433708191, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 28513 + }, + { + "epoch": 0.28514, + "grad_norm": 0.5954532027244568, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 28514 + }, + { + "epoch": 0.28515, + "grad_norm": 0.587099015712738, + "learning_rate": 0.003, + "loss": 3.998, + "step": 28515 + }, + { + "epoch": 0.28516, + "grad_norm": 0.654094398021698, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 28516 + }, + { + "epoch": 0.28517, + "grad_norm": 0.6329269409179688, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 28517 + }, + { + "epoch": 0.28518, + "grad_norm": 0.649957001209259, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 28518 + }, + { + "epoch": 0.28519, + "grad_norm": 0.7775930762290955, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 28519 + }, + { + "epoch": 0.2852, + "grad_norm": 1.0449607372283936, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 28520 + }, + { + "epoch": 0.28521, + "grad_norm": 1.1237751245498657, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 28521 + }, + { + "epoch": 0.28522, + "grad_norm": 0.9188483357429504, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 28522 + }, + { + "epoch": 0.28523, + "grad_norm": 0.7888297438621521, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 28523 + }, + { + "epoch": 0.28524, + "grad_norm": 0.748757004737854, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 28524 + }, + { + "epoch": 0.28525, + "grad_norm": 0.7855581641197205, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 28525 + }, + { + "epoch": 0.28526, + "grad_norm": 0.8806647658348083, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 28526 + }, + { + "epoch": 0.28527, + "grad_norm": 1.0807563066482544, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 28527 + }, + { + "epoch": 0.28528, + "grad_norm": 1.0141240358352661, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 28528 + }, + { + "epoch": 0.28529, + "grad_norm": 1.0105332136154175, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 28529 + }, + { + "epoch": 0.2853, + "grad_norm": 1.0035057067871094, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 28530 + }, + { + "epoch": 0.28531, + "grad_norm": 0.9926250576972961, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 28531 + }, + { + "epoch": 0.28532, + "grad_norm": 0.8710039854049683, + "learning_rate": 0.003, + "loss": 3.959, + "step": 28532 + }, + { + "epoch": 0.28533, + "grad_norm": 0.8991121649742126, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 28533 + }, + { + "epoch": 0.28534, + "grad_norm": 0.7294273376464844, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 28534 + }, + { + "epoch": 0.28535, + "grad_norm": 0.6911311745643616, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 28535 + }, + { + "epoch": 0.28536, + "grad_norm": 0.7752890586853027, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 28536 + }, + { + "epoch": 0.28537, + "grad_norm": 0.8517906069755554, + "learning_rate": 0.003, + "loss": 3.987, + "step": 28537 + }, + { + "epoch": 0.28538, + "grad_norm": 0.7416520714759827, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 28538 + }, + { + "epoch": 0.28539, + "grad_norm": 0.7435680031776428, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 28539 + }, + { + "epoch": 0.2854, + "grad_norm": 0.8514683246612549, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 28540 + }, + { + "epoch": 0.28541, + "grad_norm": 0.8195257186889648, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 28541 + }, + { + "epoch": 0.28542, + "grad_norm": 0.8660439848899841, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 28542 + }, + { + "epoch": 0.28543, + "grad_norm": 1.1997615098953247, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 28543 + }, + { + "epoch": 0.28544, + "grad_norm": 1.1134164333343506, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 28544 + }, + { + "epoch": 0.28545, + "grad_norm": 0.8458535075187683, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 28545 + }, + { + "epoch": 0.28546, + "grad_norm": 0.7921898365020752, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 28546 + }, + { + "epoch": 0.28547, + "grad_norm": 0.8299499154090881, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 28547 + }, + { + "epoch": 0.28548, + "grad_norm": 0.913594126701355, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 28548 + }, + { + "epoch": 0.28549, + "grad_norm": 0.8628702163696289, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 28549 + }, + { + "epoch": 0.2855, + "grad_norm": 0.8673170804977417, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 28550 + }, + { + "epoch": 0.28551, + "grad_norm": 0.9247093200683594, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 28551 + }, + { + "epoch": 0.28552, + "grad_norm": 1.1119500398635864, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 28552 + }, + { + "epoch": 0.28553, + "grad_norm": 1.0460302829742432, + "learning_rate": 0.003, + "loss": 4.014, + "step": 28553 + }, + { + "epoch": 0.28554, + "grad_norm": 0.8953739404678345, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 28554 + }, + { + "epoch": 0.28555, + "grad_norm": 0.7610073089599609, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 28555 + }, + { + "epoch": 0.28556, + "grad_norm": 0.7207679748535156, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 28556 + }, + { + "epoch": 0.28557, + "grad_norm": 0.7916410565376282, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 28557 + }, + { + "epoch": 0.28558, + "grad_norm": 0.7734299302101135, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 28558 + }, + { + "epoch": 0.28559, + "grad_norm": 0.8925231099128723, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 28559 + }, + { + "epoch": 0.2856, + "grad_norm": 0.9033007025718689, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 28560 + }, + { + "epoch": 0.28561, + "grad_norm": 0.7876929044723511, + "learning_rate": 0.003, + "loss": 4.012, + "step": 28561 + }, + { + "epoch": 0.28562, + "grad_norm": 0.8726317882537842, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 28562 + }, + { + "epoch": 0.28563, + "grad_norm": 0.7989873886108398, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 28563 + }, + { + "epoch": 0.28564, + "grad_norm": 0.7313238978385925, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 28564 + }, + { + "epoch": 0.28565, + "grad_norm": 0.7708633542060852, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 28565 + }, + { + "epoch": 0.28566, + "grad_norm": 0.8703032732009888, + "learning_rate": 0.003, + "loss": 3.965, + "step": 28566 + }, + { + "epoch": 0.28567, + "grad_norm": 1.1094977855682373, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 28567 + }, + { + "epoch": 0.28568, + "grad_norm": 1.2129648923873901, + "learning_rate": 0.003, + "loss": 3.997, + "step": 28568 + }, + { + "epoch": 0.28569, + "grad_norm": 0.8852020502090454, + "learning_rate": 0.003, + "loss": 3.9423, + "step": 28569 + }, + { + "epoch": 0.2857, + "grad_norm": 0.7711713910102844, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 28570 + }, + { + "epoch": 0.28571, + "grad_norm": 0.7680934071540833, + "learning_rate": 0.003, + "loss": 4.005, + "step": 28571 + }, + { + "epoch": 0.28572, + "grad_norm": 0.7733069658279419, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 28572 + }, + { + "epoch": 0.28573, + "grad_norm": 0.678919792175293, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 28573 + }, + { + "epoch": 0.28574, + "grad_norm": 0.798120379447937, + "learning_rate": 0.003, + "loss": 3.991, + "step": 28574 + }, + { + "epoch": 0.28575, + "grad_norm": 0.8650692105293274, + "learning_rate": 0.003, + "loss": 3.954, + "step": 28575 + }, + { + "epoch": 0.28576, + "grad_norm": 0.9906229376792908, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 28576 + }, + { + "epoch": 0.28577, + "grad_norm": 1.1647663116455078, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 28577 + }, + { + "epoch": 0.28578, + "grad_norm": 0.9284600019454956, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 28578 + }, + { + "epoch": 0.28579, + "grad_norm": 0.8483290672302246, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 28579 + }, + { + "epoch": 0.2858, + "grad_norm": 0.7020196318626404, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 28580 + }, + { + "epoch": 0.28581, + "grad_norm": 0.6444172859191895, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 28581 + }, + { + "epoch": 0.28582, + "grad_norm": 0.703364372253418, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 28582 + }, + { + "epoch": 0.28583, + "grad_norm": 0.9739793539047241, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 28583 + }, + { + "epoch": 0.28584, + "grad_norm": 1.2476285696029663, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 28584 + }, + { + "epoch": 0.28585, + "grad_norm": 1.00211763381958, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 28585 + }, + { + "epoch": 0.28586, + "grad_norm": 0.9480902552604675, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 28586 + }, + { + "epoch": 0.28587, + "grad_norm": 0.8672526478767395, + "learning_rate": 0.003, + "loss": 3.989, + "step": 28587 + }, + { + "epoch": 0.28588, + "grad_norm": 0.8214858174324036, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 28588 + }, + { + "epoch": 0.28589, + "grad_norm": 0.7679252624511719, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 28589 + }, + { + "epoch": 0.2859, + "grad_norm": 0.7817029356956482, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 28590 + }, + { + "epoch": 0.28591, + "grad_norm": 0.7559792995452881, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 28591 + }, + { + "epoch": 0.28592, + "grad_norm": 0.7231417894363403, + "learning_rate": 0.003, + "loss": 3.9471, + "step": 28592 + }, + { + "epoch": 0.28593, + "grad_norm": 0.7934483289718628, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 28593 + }, + { + "epoch": 0.28594, + "grad_norm": 0.7770290970802307, + "learning_rate": 0.003, + "loss": 4.013, + "step": 28594 + }, + { + "epoch": 0.28595, + "grad_norm": 0.7620813250541687, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 28595 + }, + { + "epoch": 0.28596, + "grad_norm": 0.7633997201919556, + "learning_rate": 0.003, + "loss": 3.983, + "step": 28596 + }, + { + "epoch": 0.28597, + "grad_norm": 0.800887405872345, + "learning_rate": 0.003, + "loss": 3.986, + "step": 28597 + }, + { + "epoch": 0.28598, + "grad_norm": 0.9247352480888367, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 28598 + }, + { + "epoch": 0.28599, + "grad_norm": 1.0488711595535278, + "learning_rate": 0.003, + "loss": 3.968, + "step": 28599 + }, + { + "epoch": 0.286, + "grad_norm": 1.0912647247314453, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 28600 + }, + { + "epoch": 0.28601, + "grad_norm": 1.108940601348877, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 28601 + }, + { + "epoch": 0.28602, + "grad_norm": 0.9006870985031128, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 28602 + }, + { + "epoch": 0.28603, + "grad_norm": 0.9184116125106812, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 28603 + }, + { + "epoch": 0.28604, + "grad_norm": 0.9313085675239563, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 28604 + }, + { + "epoch": 0.28605, + "grad_norm": 0.8099063634872437, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 28605 + }, + { + "epoch": 0.28606, + "grad_norm": 1.0149097442626953, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 28606 + }, + { + "epoch": 0.28607, + "grad_norm": 1.181838035583496, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 28607 + }, + { + "epoch": 0.28608, + "grad_norm": 0.7284578680992126, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 28608 + }, + { + "epoch": 0.28609, + "grad_norm": 0.7293453812599182, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 28609 + }, + { + "epoch": 0.2861, + "grad_norm": 0.8059666752815247, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 28610 + }, + { + "epoch": 0.28611, + "grad_norm": 0.9836856126785278, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 28611 + }, + { + "epoch": 0.28612, + "grad_norm": 1.1391695737838745, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 28612 + }, + { + "epoch": 0.28613, + "grad_norm": 0.812428891658783, + "learning_rate": 0.003, + "loss": 3.991, + "step": 28613 + }, + { + "epoch": 0.28614, + "grad_norm": 0.7770878672599792, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 28614 + }, + { + "epoch": 0.28615, + "grad_norm": 0.7148449420928955, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 28615 + }, + { + "epoch": 0.28616, + "grad_norm": 0.7180427312850952, + "learning_rate": 0.003, + "loss": 3.9265, + "step": 28616 + }, + { + "epoch": 0.28617, + "grad_norm": 0.8864667415618896, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 28617 + }, + { + "epoch": 0.28618, + "grad_norm": 1.158481240272522, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 28618 + }, + { + "epoch": 0.28619, + "grad_norm": 0.9192246794700623, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 28619 + }, + { + "epoch": 0.2862, + "grad_norm": 0.762881338596344, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 28620 + }, + { + "epoch": 0.28621, + "grad_norm": 0.6315200924873352, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 28621 + }, + { + "epoch": 0.28622, + "grad_norm": 0.5817621350288391, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 28622 + }, + { + "epoch": 0.28623, + "grad_norm": 0.554206907749176, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 28623 + }, + { + "epoch": 0.28624, + "grad_norm": 0.6631183624267578, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 28624 + }, + { + "epoch": 0.28625, + "grad_norm": 0.6795054078102112, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 28625 + }, + { + "epoch": 0.28626, + "grad_norm": 0.6431945562362671, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 28626 + }, + { + "epoch": 0.28627, + "grad_norm": 0.7162263989448547, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 28627 + }, + { + "epoch": 0.28628, + "grad_norm": 0.9073095321655273, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 28628 + }, + { + "epoch": 0.28629, + "grad_norm": 1.114111304283142, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 28629 + }, + { + "epoch": 0.2863, + "grad_norm": 0.950124979019165, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 28630 + }, + { + "epoch": 0.28631, + "grad_norm": 0.8651631474494934, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 28631 + }, + { + "epoch": 0.28632, + "grad_norm": 0.8380487561225891, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 28632 + }, + { + "epoch": 0.28633, + "grad_norm": 0.8556308746337891, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 28633 + }, + { + "epoch": 0.28634, + "grad_norm": 0.835232138633728, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 28634 + }, + { + "epoch": 0.28635, + "grad_norm": 0.7206199765205383, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 28635 + }, + { + "epoch": 0.28636, + "grad_norm": 0.7000617384910583, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 28636 + }, + { + "epoch": 0.28637, + "grad_norm": 0.771648108959198, + "learning_rate": 0.003, + "loss": 3.988, + "step": 28637 + }, + { + "epoch": 0.28638, + "grad_norm": 1.1081701517105103, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 28638 + }, + { + "epoch": 0.28639, + "grad_norm": 1.209431529045105, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 28639 + }, + { + "epoch": 0.2864, + "grad_norm": 1.031816840171814, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 28640 + }, + { + "epoch": 0.28641, + "grad_norm": 0.9457855820655823, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 28641 + }, + { + "epoch": 0.28642, + "grad_norm": 1.0481157302856445, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 28642 + }, + { + "epoch": 0.28643, + "grad_norm": 1.1583733558654785, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 28643 + }, + { + "epoch": 0.28644, + "grad_norm": 0.6455049514770508, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 28644 + }, + { + "epoch": 0.28645, + "grad_norm": 0.7751355767250061, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 28645 + }, + { + "epoch": 0.28646, + "grad_norm": 0.9021131992340088, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 28646 + }, + { + "epoch": 0.28647, + "grad_norm": 0.903308093547821, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 28647 + }, + { + "epoch": 0.28648, + "grad_norm": 0.8665527105331421, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 28648 + }, + { + "epoch": 0.28649, + "grad_norm": 0.8465485572814941, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 28649 + }, + { + "epoch": 0.2865, + "grad_norm": 0.8353832364082336, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 28650 + }, + { + "epoch": 0.28651, + "grad_norm": 0.9166907668113708, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 28651 + }, + { + "epoch": 0.28652, + "grad_norm": 1.0536826848983765, + "learning_rate": 0.003, + "loss": 3.98, + "step": 28652 + }, + { + "epoch": 0.28653, + "grad_norm": 1.0385136604309082, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 28653 + }, + { + "epoch": 0.28654, + "grad_norm": 0.9335598349571228, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 28654 + }, + { + "epoch": 0.28655, + "grad_norm": 0.9789895415306091, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 28655 + }, + { + "epoch": 0.28656, + "grad_norm": 0.9875578284263611, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 28656 + }, + { + "epoch": 0.28657, + "grad_norm": 0.9852379560470581, + "learning_rate": 0.003, + "loss": 3.98, + "step": 28657 + }, + { + "epoch": 0.28658, + "grad_norm": 1.0274102687835693, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 28658 + }, + { + "epoch": 0.28659, + "grad_norm": 0.9544501304626465, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 28659 + }, + { + "epoch": 0.2866, + "grad_norm": 0.9119637608528137, + "learning_rate": 0.003, + "loss": 3.9464, + "step": 28660 + }, + { + "epoch": 0.28661, + "grad_norm": 1.0930522680282593, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 28661 + }, + { + "epoch": 0.28662, + "grad_norm": 0.725115954875946, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 28662 + }, + { + "epoch": 0.28663, + "grad_norm": 0.7503170967102051, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 28663 + }, + { + "epoch": 0.28664, + "grad_norm": 0.9767497181892395, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 28664 + }, + { + "epoch": 0.28665, + "grad_norm": 0.9723098278045654, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 28665 + }, + { + "epoch": 0.28666, + "grad_norm": 0.8569937348365784, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 28666 + }, + { + "epoch": 0.28667, + "grad_norm": 0.8491440415382385, + "learning_rate": 0.003, + "loss": 3.967, + "step": 28667 + }, + { + "epoch": 0.28668, + "grad_norm": 0.8112216591835022, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 28668 + }, + { + "epoch": 0.28669, + "grad_norm": 0.9795170426368713, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 28669 + }, + { + "epoch": 0.2867, + "grad_norm": 1.3003654479980469, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 28670 + }, + { + "epoch": 0.28671, + "grad_norm": 0.6938055753707886, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 28671 + }, + { + "epoch": 0.28672, + "grad_norm": 0.6239778995513916, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 28672 + }, + { + "epoch": 0.28673, + "grad_norm": 0.5993738174438477, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 28673 + }, + { + "epoch": 0.28674, + "grad_norm": 0.6026144027709961, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 28674 + }, + { + "epoch": 0.28675, + "grad_norm": 0.691068708896637, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 28675 + }, + { + "epoch": 0.28676, + "grad_norm": 0.719160258769989, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 28676 + }, + { + "epoch": 0.28677, + "grad_norm": 0.7264537215232849, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 28677 + }, + { + "epoch": 0.28678, + "grad_norm": 0.7164241671562195, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 28678 + }, + { + "epoch": 0.28679, + "grad_norm": 0.8721011281013489, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 28679 + }, + { + "epoch": 0.2868, + "grad_norm": 1.0602717399597168, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 28680 + }, + { + "epoch": 0.28681, + "grad_norm": 1.112969994544983, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 28681 + }, + { + "epoch": 0.28682, + "grad_norm": 0.8002830743789673, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 28682 + }, + { + "epoch": 0.28683, + "grad_norm": 0.776226818561554, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 28683 + }, + { + "epoch": 0.28684, + "grad_norm": 1.0315361022949219, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 28684 + }, + { + "epoch": 0.28685, + "grad_norm": 1.0091657638549805, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 28685 + }, + { + "epoch": 0.28686, + "grad_norm": 0.9872680902481079, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 28686 + }, + { + "epoch": 0.28687, + "grad_norm": 0.9890653491020203, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 28687 + }, + { + "epoch": 0.28688, + "grad_norm": 1.0870004892349243, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 28688 + }, + { + "epoch": 0.28689, + "grad_norm": 0.7783733606338501, + "learning_rate": 0.003, + "loss": 4.005, + "step": 28689 + }, + { + "epoch": 0.2869, + "grad_norm": 0.6864676475524902, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 28690 + }, + { + "epoch": 0.28691, + "grad_norm": 0.7421690821647644, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 28691 + }, + { + "epoch": 0.28692, + "grad_norm": 0.7174811363220215, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 28692 + }, + { + "epoch": 0.28693, + "grad_norm": 0.7736871242523193, + "learning_rate": 0.003, + "loss": 3.982, + "step": 28693 + }, + { + "epoch": 0.28694, + "grad_norm": 0.8353952169418335, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 28694 + }, + { + "epoch": 0.28695, + "grad_norm": 0.8264212012290955, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 28695 + }, + { + "epoch": 0.28696, + "grad_norm": 0.6980035305023193, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 28696 + }, + { + "epoch": 0.28697, + "grad_norm": 0.7503317594528198, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 28697 + }, + { + "epoch": 0.28698, + "grad_norm": 0.7482432723045349, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 28698 + }, + { + "epoch": 0.28699, + "grad_norm": 0.8161530494689941, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 28699 + }, + { + "epoch": 0.287, + "grad_norm": 0.9810968041419983, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 28700 + }, + { + "epoch": 0.28701, + "grad_norm": 1.1739548444747925, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 28701 + }, + { + "epoch": 0.28702, + "grad_norm": 0.840529203414917, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 28702 + }, + { + "epoch": 0.28703, + "grad_norm": 0.7665707468986511, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 28703 + }, + { + "epoch": 0.28704, + "grad_norm": 0.7602238655090332, + "learning_rate": 0.003, + "loss": 3.987, + "step": 28704 + }, + { + "epoch": 0.28705, + "grad_norm": 0.8292978405952454, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 28705 + }, + { + "epoch": 0.28706, + "grad_norm": 0.8637842535972595, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 28706 + }, + { + "epoch": 0.28707, + "grad_norm": 0.8341313600540161, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 28707 + }, + { + "epoch": 0.28708, + "grad_norm": 0.793847382068634, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 28708 + }, + { + "epoch": 0.28709, + "grad_norm": 0.7601509690284729, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 28709 + }, + { + "epoch": 0.2871, + "grad_norm": 0.7589887380599976, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 28710 + }, + { + "epoch": 0.28711, + "grad_norm": 0.9487878084182739, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 28711 + }, + { + "epoch": 0.28712, + "grad_norm": 1.0400398969650269, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 28712 + }, + { + "epoch": 0.28713, + "grad_norm": 1.155285120010376, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 28713 + }, + { + "epoch": 0.28714, + "grad_norm": 0.9179924726486206, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 28714 + }, + { + "epoch": 0.28715, + "grad_norm": 1.0273882150650024, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 28715 + }, + { + "epoch": 0.28716, + "grad_norm": 0.9156093597412109, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 28716 + }, + { + "epoch": 0.28717, + "grad_norm": 0.9216423630714417, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 28717 + }, + { + "epoch": 0.28718, + "grad_norm": 1.0004853010177612, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 28718 + }, + { + "epoch": 0.28719, + "grad_norm": 1.0586398839950562, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 28719 + }, + { + "epoch": 0.2872, + "grad_norm": 1.0535048246383667, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 28720 + }, + { + "epoch": 0.28721, + "grad_norm": 1.054696798324585, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 28721 + }, + { + "epoch": 0.28722, + "grad_norm": 1.0283198356628418, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 28722 + }, + { + "epoch": 0.28723, + "grad_norm": 1.108062505722046, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 28723 + }, + { + "epoch": 0.28724, + "grad_norm": 0.914276659488678, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 28724 + }, + { + "epoch": 0.28725, + "grad_norm": 0.7993897199630737, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 28725 + }, + { + "epoch": 0.28726, + "grad_norm": 0.8370265364646912, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 28726 + }, + { + "epoch": 0.28727, + "grad_norm": 0.8771136999130249, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 28727 + }, + { + "epoch": 0.28728, + "grad_norm": 0.8703499436378479, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 28728 + }, + { + "epoch": 0.28729, + "grad_norm": 0.8442876935005188, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 28729 + }, + { + "epoch": 0.2873, + "grad_norm": 0.748560905456543, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 28730 + }, + { + "epoch": 0.28731, + "grad_norm": 0.6645280718803406, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 28731 + }, + { + "epoch": 0.28732, + "grad_norm": 0.6842381954193115, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 28732 + }, + { + "epoch": 0.28733, + "grad_norm": 0.7485685348510742, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 28733 + }, + { + "epoch": 0.28734, + "grad_norm": 0.7008538842201233, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 28734 + }, + { + "epoch": 0.28735, + "grad_norm": 0.6356719136238098, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 28735 + }, + { + "epoch": 0.28736, + "grad_norm": 0.5529755353927612, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 28736 + }, + { + "epoch": 0.28737, + "grad_norm": 0.4947183132171631, + "learning_rate": 0.003, + "loss": 4.015, + "step": 28737 + }, + { + "epoch": 0.28738, + "grad_norm": 0.5649681091308594, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 28738 + }, + { + "epoch": 0.28739, + "grad_norm": 0.4802338182926178, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 28739 + }, + { + "epoch": 0.2874, + "grad_norm": 0.5246257185935974, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 28740 + }, + { + "epoch": 0.28741, + "grad_norm": 0.5724809765815735, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 28741 + }, + { + "epoch": 0.28742, + "grad_norm": 0.6972962021827698, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 28742 + }, + { + "epoch": 0.28743, + "grad_norm": 1.1391441822052002, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 28743 + }, + { + "epoch": 0.28744, + "grad_norm": 1.3247994184494019, + "learning_rate": 0.003, + "loss": 3.9347, + "step": 28744 + }, + { + "epoch": 0.28745, + "grad_norm": 0.7246697545051575, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 28745 + }, + { + "epoch": 0.28746, + "grad_norm": 0.626922607421875, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 28746 + }, + { + "epoch": 0.28747, + "grad_norm": 0.7625699043273926, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 28747 + }, + { + "epoch": 0.28748, + "grad_norm": 1.0059596300125122, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 28748 + }, + { + "epoch": 0.28749, + "grad_norm": 0.9682197570800781, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 28749 + }, + { + "epoch": 0.2875, + "grad_norm": 0.9575952887535095, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 28750 + }, + { + "epoch": 0.28751, + "grad_norm": 0.9325644969940186, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 28751 + }, + { + "epoch": 0.28752, + "grad_norm": 0.8282651305198669, + "learning_rate": 0.003, + "loss": 3.98, + "step": 28752 + }, + { + "epoch": 0.28753, + "grad_norm": 0.8324861526489258, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 28753 + }, + { + "epoch": 0.28754, + "grad_norm": 0.8970338106155396, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 28754 + }, + { + "epoch": 0.28755, + "grad_norm": 1.0089665651321411, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 28755 + }, + { + "epoch": 0.28756, + "grad_norm": 1.2605860233306885, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 28756 + }, + { + "epoch": 0.28757, + "grad_norm": 0.7713136672973633, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 28757 + }, + { + "epoch": 0.28758, + "grad_norm": 0.7452896237373352, + "learning_rate": 0.003, + "loss": 3.987, + "step": 28758 + }, + { + "epoch": 0.28759, + "grad_norm": 0.8402251601219177, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 28759 + }, + { + "epoch": 0.2876, + "grad_norm": 0.9039424061775208, + "learning_rate": 0.003, + "loss": 3.987, + "step": 28760 + }, + { + "epoch": 0.28761, + "grad_norm": 0.9564940929412842, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 28761 + }, + { + "epoch": 0.28762, + "grad_norm": 0.9862769842147827, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 28762 + }, + { + "epoch": 0.28763, + "grad_norm": 1.1875437498092651, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 28763 + }, + { + "epoch": 0.28764, + "grad_norm": 0.9459707736968994, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 28764 + }, + { + "epoch": 0.28765, + "grad_norm": 0.8635523319244385, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 28765 + }, + { + "epoch": 0.28766, + "grad_norm": 0.783240020275116, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 28766 + }, + { + "epoch": 0.28767, + "grad_norm": 0.7378678321838379, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 28767 + }, + { + "epoch": 0.28768, + "grad_norm": 1.01416015625, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 28768 + }, + { + "epoch": 0.28769, + "grad_norm": 1.1705304384231567, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 28769 + }, + { + "epoch": 0.2877, + "grad_norm": 1.2544589042663574, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 28770 + }, + { + "epoch": 0.28771, + "grad_norm": 0.8095514178276062, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 28771 + }, + { + "epoch": 0.28772, + "grad_norm": 0.8167713284492493, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 28772 + }, + { + "epoch": 0.28773, + "grad_norm": 1.0521198511123657, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 28773 + }, + { + "epoch": 0.28774, + "grad_norm": 1.104849934577942, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 28774 + }, + { + "epoch": 0.28775, + "grad_norm": 0.6875806450843811, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 28775 + }, + { + "epoch": 0.28776, + "grad_norm": 0.6683295965194702, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 28776 + }, + { + "epoch": 0.28777, + "grad_norm": 0.7330931425094604, + "learning_rate": 0.003, + "loss": 3.971, + "step": 28777 + }, + { + "epoch": 0.28778, + "grad_norm": 0.7450997829437256, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 28778 + }, + { + "epoch": 0.28779, + "grad_norm": 0.5767092108726501, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 28779 + }, + { + "epoch": 0.2878, + "grad_norm": 0.6418662667274475, + "learning_rate": 0.003, + "loss": 3.971, + "step": 28780 + }, + { + "epoch": 0.28781, + "grad_norm": 0.6646447777748108, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 28781 + }, + { + "epoch": 0.28782, + "grad_norm": 0.7545374631881714, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 28782 + }, + { + "epoch": 0.28783, + "grad_norm": 0.9670398831367493, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 28783 + }, + { + "epoch": 0.28784, + "grad_norm": 1.2868390083312988, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 28784 + }, + { + "epoch": 0.28785, + "grad_norm": 0.6737198829650879, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 28785 + }, + { + "epoch": 0.28786, + "grad_norm": 0.8518375754356384, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 28786 + }, + { + "epoch": 0.28787, + "grad_norm": 1.295261263847351, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 28787 + }, + { + "epoch": 0.28788, + "grad_norm": 0.865143895149231, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 28788 + }, + { + "epoch": 0.28789, + "grad_norm": 0.6813794374465942, + "learning_rate": 0.003, + "loss": 3.996, + "step": 28789 + }, + { + "epoch": 0.2879, + "grad_norm": 0.6431641578674316, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 28790 + }, + { + "epoch": 0.28791, + "grad_norm": 0.7180762887001038, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 28791 + }, + { + "epoch": 0.28792, + "grad_norm": 0.7663118243217468, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 28792 + }, + { + "epoch": 0.28793, + "grad_norm": 0.7012357115745544, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 28793 + }, + { + "epoch": 0.28794, + "grad_norm": 0.6633811593055725, + "learning_rate": 0.003, + "loss": 3.988, + "step": 28794 + }, + { + "epoch": 0.28795, + "grad_norm": 0.6335874199867249, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 28795 + }, + { + "epoch": 0.28796, + "grad_norm": 0.6676688194274902, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 28796 + }, + { + "epoch": 0.28797, + "grad_norm": 0.7705386877059937, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 28797 + }, + { + "epoch": 0.28798, + "grad_norm": 0.8808639645576477, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 28798 + }, + { + "epoch": 0.28799, + "grad_norm": 1.1526060104370117, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 28799 + }, + { + "epoch": 0.288, + "grad_norm": 0.8686129450798035, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 28800 + }, + { + "epoch": 0.28801, + "grad_norm": 0.7654739618301392, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 28801 + }, + { + "epoch": 0.28802, + "grad_norm": 0.6644590497016907, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 28802 + }, + { + "epoch": 0.28803, + "grad_norm": 0.685867428779602, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 28803 + }, + { + "epoch": 0.28804, + "grad_norm": 0.7641301155090332, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 28804 + }, + { + "epoch": 0.28805, + "grad_norm": 0.8641214370727539, + "learning_rate": 0.003, + "loss": 3.975, + "step": 28805 + }, + { + "epoch": 0.28806, + "grad_norm": 1.0392861366271973, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 28806 + }, + { + "epoch": 0.28807, + "grad_norm": 1.140297770500183, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 28807 + }, + { + "epoch": 0.28808, + "grad_norm": 0.7039082646369934, + "learning_rate": 0.003, + "loss": 3.97, + "step": 28808 + }, + { + "epoch": 0.28809, + "grad_norm": 0.8383872509002686, + "learning_rate": 0.003, + "loss": 3.942, + "step": 28809 + }, + { + "epoch": 0.2881, + "grad_norm": 0.9402018189430237, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 28810 + }, + { + "epoch": 0.28811, + "grad_norm": 0.9145536422729492, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 28811 + }, + { + "epoch": 0.28812, + "grad_norm": 0.9092174768447876, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 28812 + }, + { + "epoch": 0.28813, + "grad_norm": 0.9207307696342468, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 28813 + }, + { + "epoch": 0.28814, + "grad_norm": 1.0899847745895386, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 28814 + }, + { + "epoch": 0.28815, + "grad_norm": 0.9935137033462524, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 28815 + }, + { + "epoch": 0.28816, + "grad_norm": 1.1687514781951904, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 28816 + }, + { + "epoch": 0.28817, + "grad_norm": 1.0931270122528076, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 28817 + }, + { + "epoch": 0.28818, + "grad_norm": 0.8963140845298767, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 28818 + }, + { + "epoch": 0.28819, + "grad_norm": 0.7795963883399963, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 28819 + }, + { + "epoch": 0.2882, + "grad_norm": 0.7250871062278748, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 28820 + }, + { + "epoch": 0.28821, + "grad_norm": 0.7369036674499512, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 28821 + }, + { + "epoch": 0.28822, + "grad_norm": 0.9630380868911743, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 28822 + }, + { + "epoch": 0.28823, + "grad_norm": 1.0341534614562988, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 28823 + }, + { + "epoch": 0.28824, + "grad_norm": 0.8876447677612305, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 28824 + }, + { + "epoch": 0.28825, + "grad_norm": 0.8573097586631775, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 28825 + }, + { + "epoch": 0.28826, + "grad_norm": 0.9768937826156616, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 28826 + }, + { + "epoch": 0.28827, + "grad_norm": 1.0085773468017578, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 28827 + }, + { + "epoch": 0.28828, + "grad_norm": 1.2009880542755127, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 28828 + }, + { + "epoch": 0.28829, + "grad_norm": 0.8763651847839355, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 28829 + }, + { + "epoch": 0.2883, + "grad_norm": 1.001397728919983, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 28830 + }, + { + "epoch": 0.28831, + "grad_norm": 1.0398681163787842, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 28831 + }, + { + "epoch": 0.28832, + "grad_norm": 1.031926155090332, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 28832 + }, + { + "epoch": 0.28833, + "grad_norm": 1.0480366945266724, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 28833 + }, + { + "epoch": 0.28834, + "grad_norm": 1.0934337377548218, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 28834 + }, + { + "epoch": 0.28835, + "grad_norm": 0.9880228042602539, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 28835 + }, + { + "epoch": 0.28836, + "grad_norm": 0.8780912160873413, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 28836 + }, + { + "epoch": 0.28837, + "grad_norm": 0.7740278244018555, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 28837 + }, + { + "epoch": 0.28838, + "grad_norm": 0.7449983954429626, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 28838 + }, + { + "epoch": 0.28839, + "grad_norm": 0.8202010989189148, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 28839 + }, + { + "epoch": 0.2884, + "grad_norm": 0.9795507192611694, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 28840 + }, + { + "epoch": 0.28841, + "grad_norm": 1.2277157306671143, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 28841 + }, + { + "epoch": 0.28842, + "grad_norm": 1.0121188163757324, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 28842 + }, + { + "epoch": 0.28843, + "grad_norm": 1.0581437349319458, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 28843 + }, + { + "epoch": 0.28844, + "grad_norm": 0.9867231845855713, + "learning_rate": 0.003, + "loss": 3.994, + "step": 28844 + }, + { + "epoch": 0.28845, + "grad_norm": 0.7808648943901062, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 28845 + }, + { + "epoch": 0.28846, + "grad_norm": 0.5226108431816101, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 28846 + }, + { + "epoch": 0.28847, + "grad_norm": 0.47433316707611084, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 28847 + }, + { + "epoch": 0.28848, + "grad_norm": 0.5059252381324768, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 28848 + }, + { + "epoch": 0.28849, + "grad_norm": 0.648259699344635, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 28849 + }, + { + "epoch": 0.2885, + "grad_norm": 0.7547573447227478, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 28850 + }, + { + "epoch": 0.28851, + "grad_norm": 0.8742242455482483, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 28851 + }, + { + "epoch": 0.28852, + "grad_norm": 0.8710440993309021, + "learning_rate": 0.003, + "loss": 3.989, + "step": 28852 + }, + { + "epoch": 0.28853, + "grad_norm": 0.7826887369155884, + "learning_rate": 0.003, + "loss": 3.988, + "step": 28853 + }, + { + "epoch": 0.28854, + "grad_norm": 0.6785268187522888, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 28854 + }, + { + "epoch": 0.28855, + "grad_norm": 0.5506678819656372, + "learning_rate": 0.003, + "loss": 3.997, + "step": 28855 + }, + { + "epoch": 0.28856, + "grad_norm": 0.5927761793136597, + "learning_rate": 0.003, + "loss": 3.954, + "step": 28856 + }, + { + "epoch": 0.28857, + "grad_norm": 0.6698287129402161, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 28857 + }, + { + "epoch": 0.28858, + "grad_norm": 0.5635228157043457, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 28858 + }, + { + "epoch": 0.28859, + "grad_norm": 0.5124345421791077, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 28859 + }, + { + "epoch": 0.2886, + "grad_norm": 0.5932596921920776, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 28860 + }, + { + "epoch": 0.28861, + "grad_norm": 0.6337072253227234, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 28861 + }, + { + "epoch": 0.28862, + "grad_norm": 0.6417006850242615, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 28862 + }, + { + "epoch": 0.28863, + "grad_norm": 0.7269140481948853, + "learning_rate": 0.003, + "loss": 3.9493, + "step": 28863 + }, + { + "epoch": 0.28864, + "grad_norm": 1.0041913986206055, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 28864 + }, + { + "epoch": 0.28865, + "grad_norm": 1.5231961011886597, + "learning_rate": 0.003, + "loss": 3.978, + "step": 28865 + }, + { + "epoch": 0.28866, + "grad_norm": 0.7829987406730652, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 28866 + }, + { + "epoch": 0.28867, + "grad_norm": 0.7408373355865479, + "learning_rate": 0.003, + "loss": 3.991, + "step": 28867 + }, + { + "epoch": 0.28868, + "grad_norm": 0.7696337103843689, + "learning_rate": 0.003, + "loss": 3.973, + "step": 28868 + }, + { + "epoch": 0.28869, + "grad_norm": 0.8911351561546326, + "learning_rate": 0.003, + "loss": 3.977, + "step": 28869 + }, + { + "epoch": 0.2887, + "grad_norm": 1.005966305732727, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 28870 + }, + { + "epoch": 0.28871, + "grad_norm": 0.9288684129714966, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 28871 + }, + { + "epoch": 0.28872, + "grad_norm": 0.9614913463592529, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 28872 + }, + { + "epoch": 0.28873, + "grad_norm": 1.0734885931015015, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 28873 + }, + { + "epoch": 0.28874, + "grad_norm": 0.9335092306137085, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 28874 + }, + { + "epoch": 0.28875, + "grad_norm": 1.1709104776382446, + "learning_rate": 0.003, + "loss": 4.027, + "step": 28875 + }, + { + "epoch": 0.28876, + "grad_norm": 0.9036986231803894, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 28876 + }, + { + "epoch": 0.28877, + "grad_norm": 0.9687111973762512, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 28877 + }, + { + "epoch": 0.28878, + "grad_norm": 1.0800405740737915, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 28878 + }, + { + "epoch": 0.28879, + "grad_norm": 1.0443323850631714, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 28879 + }, + { + "epoch": 0.2888, + "grad_norm": 0.9807519912719727, + "learning_rate": 0.003, + "loss": 4.016, + "step": 28880 + }, + { + "epoch": 0.28881, + "grad_norm": 1.0353236198425293, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 28881 + }, + { + "epoch": 0.28882, + "grad_norm": 1.001981258392334, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 28882 + }, + { + "epoch": 0.28883, + "grad_norm": 0.9029473066329956, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 28883 + }, + { + "epoch": 0.28884, + "grad_norm": 0.9397034049034119, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 28884 + }, + { + "epoch": 0.28885, + "grad_norm": 1.1023056507110596, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 28885 + }, + { + "epoch": 0.28886, + "grad_norm": 0.9696000218391418, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 28886 + }, + { + "epoch": 0.28887, + "grad_norm": 1.0936033725738525, + "learning_rate": 0.003, + "loss": 3.981, + "step": 28887 + }, + { + "epoch": 0.28888, + "grad_norm": 0.9516791105270386, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 28888 + }, + { + "epoch": 0.28889, + "grad_norm": 0.8509979248046875, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 28889 + }, + { + "epoch": 0.2889, + "grad_norm": 0.707859218120575, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 28890 + }, + { + "epoch": 0.28891, + "grad_norm": 0.7699530720710754, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 28891 + }, + { + "epoch": 0.28892, + "grad_norm": 0.6849014163017273, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 28892 + }, + { + "epoch": 0.28893, + "grad_norm": 0.7187605500221252, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 28893 + }, + { + "epoch": 0.28894, + "grad_norm": 0.7328411340713501, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 28894 + }, + { + "epoch": 0.28895, + "grad_norm": 0.5712026953697205, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 28895 + }, + { + "epoch": 0.28896, + "grad_norm": 0.5375128984451294, + "learning_rate": 0.003, + "loss": 3.978, + "step": 28896 + }, + { + "epoch": 0.28897, + "grad_norm": 0.5727459192276001, + "learning_rate": 0.003, + "loss": 3.979, + "step": 28897 + }, + { + "epoch": 0.28898, + "grad_norm": 0.6933693885803223, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 28898 + }, + { + "epoch": 0.28899, + "grad_norm": 0.8718135356903076, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 28899 + }, + { + "epoch": 0.289, + "grad_norm": 1.0803806781768799, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 28900 + }, + { + "epoch": 0.28901, + "grad_norm": 1.0645544528961182, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 28901 + }, + { + "epoch": 0.28902, + "grad_norm": 1.0098421573638916, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 28902 + }, + { + "epoch": 0.28903, + "grad_norm": 0.97967928647995, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 28903 + }, + { + "epoch": 0.28904, + "grad_norm": 0.9887168407440186, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 28904 + }, + { + "epoch": 0.28905, + "grad_norm": 0.9269795417785645, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 28905 + }, + { + "epoch": 0.28906, + "grad_norm": 0.7184670567512512, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 28906 + }, + { + "epoch": 0.28907, + "grad_norm": 0.6991847157478333, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 28907 + }, + { + "epoch": 0.28908, + "grad_norm": 0.7141897082328796, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 28908 + }, + { + "epoch": 0.28909, + "grad_norm": 0.7307652831077576, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 28909 + }, + { + "epoch": 0.2891, + "grad_norm": 0.8507932424545288, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 28910 + }, + { + "epoch": 0.28911, + "grad_norm": 1.104758381843567, + "learning_rate": 0.003, + "loss": 3.992, + "step": 28911 + }, + { + "epoch": 0.28912, + "grad_norm": 1.3602882623672485, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 28912 + }, + { + "epoch": 0.28913, + "grad_norm": 0.6180644035339355, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 28913 + }, + { + "epoch": 0.28914, + "grad_norm": 0.6300908327102661, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 28914 + }, + { + "epoch": 0.28915, + "grad_norm": 0.6690584421157837, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 28915 + }, + { + "epoch": 0.28916, + "grad_norm": 0.7525841593742371, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 28916 + }, + { + "epoch": 0.28917, + "grad_norm": 0.9217771291732788, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 28917 + }, + { + "epoch": 0.28918, + "grad_norm": 1.023508071899414, + "learning_rate": 0.003, + "loss": 3.999, + "step": 28918 + }, + { + "epoch": 0.28919, + "grad_norm": 0.9008146524429321, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 28919 + }, + { + "epoch": 0.2892, + "grad_norm": 0.7702022790908813, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 28920 + }, + { + "epoch": 0.28921, + "grad_norm": 0.72856605052948, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 28921 + }, + { + "epoch": 0.28922, + "grad_norm": 0.8461525440216064, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 28922 + }, + { + "epoch": 0.28923, + "grad_norm": 1.0114396810531616, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 28923 + }, + { + "epoch": 0.28924, + "grad_norm": 1.1152881383895874, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 28924 + }, + { + "epoch": 0.28925, + "grad_norm": 0.7133868336677551, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 28925 + }, + { + "epoch": 0.28926, + "grad_norm": 0.8010155558586121, + "learning_rate": 0.003, + "loss": 3.9466, + "step": 28926 + }, + { + "epoch": 0.28927, + "grad_norm": 1.1089547872543335, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 28927 + }, + { + "epoch": 0.28928, + "grad_norm": 1.0063297748565674, + "learning_rate": 0.003, + "loss": 3.954, + "step": 28928 + }, + { + "epoch": 0.28929, + "grad_norm": 0.8608474135398865, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 28929 + }, + { + "epoch": 0.2893, + "grad_norm": 0.7458817958831787, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 28930 + }, + { + "epoch": 0.28931, + "grad_norm": 0.7365446090698242, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 28931 + }, + { + "epoch": 0.28932, + "grad_norm": 0.8817208409309387, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 28932 + }, + { + "epoch": 0.28933, + "grad_norm": 0.9952393174171448, + "learning_rate": 0.003, + "loss": 3.998, + "step": 28933 + }, + { + "epoch": 0.28934, + "grad_norm": 1.21809983253479, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 28934 + }, + { + "epoch": 0.28935, + "grad_norm": 1.011305809020996, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 28935 + }, + { + "epoch": 0.28936, + "grad_norm": 0.9867362976074219, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 28936 + }, + { + "epoch": 0.28937, + "grad_norm": 0.9053528308868408, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 28937 + }, + { + "epoch": 0.28938, + "grad_norm": 0.8969165086746216, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 28938 + }, + { + "epoch": 0.28939, + "grad_norm": 0.8951830267906189, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 28939 + }, + { + "epoch": 0.2894, + "grad_norm": 1.0581719875335693, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 28940 + }, + { + "epoch": 0.28941, + "grad_norm": 0.9814466834068298, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 28941 + }, + { + "epoch": 0.28942, + "grad_norm": 0.9632497429847717, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 28942 + }, + { + "epoch": 0.28943, + "grad_norm": 0.8814110159873962, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 28943 + }, + { + "epoch": 0.28944, + "grad_norm": 0.7981817722320557, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 28944 + }, + { + "epoch": 0.28945, + "grad_norm": 0.7879018783569336, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 28945 + }, + { + "epoch": 0.28946, + "grad_norm": 0.9945456981658936, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 28946 + }, + { + "epoch": 0.28947, + "grad_norm": 1.1087219715118408, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 28947 + }, + { + "epoch": 0.28948, + "grad_norm": 0.9770598411560059, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 28948 + }, + { + "epoch": 0.28949, + "grad_norm": 1.0288100242614746, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 28949 + }, + { + "epoch": 0.2895, + "grad_norm": 1.1807372570037842, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 28950 + }, + { + "epoch": 0.28951, + "grad_norm": 1.2259107828140259, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 28951 + }, + { + "epoch": 0.28952, + "grad_norm": 0.7064810395240784, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 28952 + }, + { + "epoch": 0.28953, + "grad_norm": 0.6787410378456116, + "learning_rate": 0.003, + "loss": 3.978, + "step": 28953 + }, + { + "epoch": 0.28954, + "grad_norm": 0.7541005611419678, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 28954 + }, + { + "epoch": 0.28955, + "grad_norm": 0.7184621691703796, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 28955 + }, + { + "epoch": 0.28956, + "grad_norm": 0.6632510423660278, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 28956 + }, + { + "epoch": 0.28957, + "grad_norm": 0.795120358467102, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 28957 + }, + { + "epoch": 0.28958, + "grad_norm": 0.8115760087966919, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 28958 + }, + { + "epoch": 0.28959, + "grad_norm": 0.8091943860054016, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 28959 + }, + { + "epoch": 0.2896, + "grad_norm": 0.6932988166809082, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 28960 + }, + { + "epoch": 0.28961, + "grad_norm": 0.7118127942085266, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 28961 + }, + { + "epoch": 0.28962, + "grad_norm": 0.8560382127761841, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 28962 + }, + { + "epoch": 0.28963, + "grad_norm": 1.2349399328231812, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 28963 + }, + { + "epoch": 0.28964, + "grad_norm": 0.8476227521896362, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 28964 + }, + { + "epoch": 0.28965, + "grad_norm": 0.7192935943603516, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 28965 + }, + { + "epoch": 0.28966, + "grad_norm": 0.8021114468574524, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 28966 + }, + { + "epoch": 0.28967, + "grad_norm": 0.9954956769943237, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 28967 + }, + { + "epoch": 0.28968, + "grad_norm": 1.0046252012252808, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 28968 + }, + { + "epoch": 0.28969, + "grad_norm": 0.9479559063911438, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 28969 + }, + { + "epoch": 0.2897, + "grad_norm": 0.855967104434967, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 28970 + }, + { + "epoch": 0.28971, + "grad_norm": 0.6479618549346924, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 28971 + }, + { + "epoch": 0.28972, + "grad_norm": 0.558454155921936, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 28972 + }, + { + "epoch": 0.28973, + "grad_norm": 0.5472530126571655, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 28973 + }, + { + "epoch": 0.28974, + "grad_norm": 0.5774360299110413, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 28974 + }, + { + "epoch": 0.28975, + "grad_norm": 0.7031108140945435, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 28975 + }, + { + "epoch": 0.28976, + "grad_norm": 0.8945627808570862, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 28976 + }, + { + "epoch": 0.28977, + "grad_norm": 1.0409646034240723, + "learning_rate": 0.003, + "loss": 4.004, + "step": 28977 + }, + { + "epoch": 0.28978, + "grad_norm": 1.0860941410064697, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 28978 + }, + { + "epoch": 0.28979, + "grad_norm": 0.8141595721244812, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 28979 + }, + { + "epoch": 0.2898, + "grad_norm": 0.6812304854393005, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 28980 + }, + { + "epoch": 0.28981, + "grad_norm": 0.6579672694206238, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 28981 + }, + { + "epoch": 0.28982, + "grad_norm": 0.6865857839584351, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 28982 + }, + { + "epoch": 0.28983, + "grad_norm": 0.7445271611213684, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 28983 + }, + { + "epoch": 0.28984, + "grad_norm": 0.6952838897705078, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 28984 + }, + { + "epoch": 0.28985, + "grad_norm": 0.6812704205513, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 28985 + }, + { + "epoch": 0.28986, + "grad_norm": 0.8002928495407104, + "learning_rate": 0.003, + "loss": 3.991, + "step": 28986 + }, + { + "epoch": 0.28987, + "grad_norm": 0.9271587133407593, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 28987 + }, + { + "epoch": 0.28988, + "grad_norm": 0.9575695395469666, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 28988 + }, + { + "epoch": 0.28989, + "grad_norm": 0.9721452593803406, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 28989 + }, + { + "epoch": 0.2899, + "grad_norm": 1.1662588119506836, + "learning_rate": 0.003, + "loss": 3.994, + "step": 28990 + }, + { + "epoch": 0.28991, + "grad_norm": 1.0475409030914307, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 28991 + }, + { + "epoch": 0.28992, + "grad_norm": 1.2892935276031494, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 28992 + }, + { + "epoch": 0.28993, + "grad_norm": 0.7578301429748535, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 28993 + }, + { + "epoch": 0.28994, + "grad_norm": 0.5981103181838989, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 28994 + }, + { + "epoch": 0.28995, + "grad_norm": 0.5970335602760315, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 28995 + }, + { + "epoch": 0.28996, + "grad_norm": 0.6347265243530273, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 28996 + }, + { + "epoch": 0.28997, + "grad_norm": 0.8277536630630493, + "learning_rate": 0.003, + "loss": 3.997, + "step": 28997 + }, + { + "epoch": 0.28998, + "grad_norm": 0.950858473777771, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 28998 + }, + { + "epoch": 0.28999, + "grad_norm": 0.9128184914588928, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 28999 + }, + { + "epoch": 0.29, + "grad_norm": 0.8858577609062195, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 29000 + }, + { + "epoch": 0.29001, + "grad_norm": 1.1265193223953247, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 29001 + }, + { + "epoch": 0.29002, + "grad_norm": 0.8128500580787659, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 29002 + }, + { + "epoch": 0.29003, + "grad_norm": 0.8144601583480835, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 29003 + }, + { + "epoch": 0.29004, + "grad_norm": 0.8815447092056274, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 29004 + }, + { + "epoch": 0.29005, + "grad_norm": 1.0002143383026123, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 29005 + }, + { + "epoch": 0.29006, + "grad_norm": 1.0262566804885864, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 29006 + }, + { + "epoch": 0.29007, + "grad_norm": 0.8476322293281555, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 29007 + }, + { + "epoch": 0.29008, + "grad_norm": 0.7470033168792725, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 29008 + }, + { + "epoch": 0.29009, + "grad_norm": 0.7919071316719055, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 29009 + }, + { + "epoch": 0.2901, + "grad_norm": 0.9192633628845215, + "learning_rate": 0.003, + "loss": 4.009, + "step": 29010 + }, + { + "epoch": 0.29011, + "grad_norm": 1.1116960048675537, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 29011 + }, + { + "epoch": 0.29012, + "grad_norm": 0.8382503390312195, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 29012 + }, + { + "epoch": 0.29013, + "grad_norm": 0.7571344375610352, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 29013 + }, + { + "epoch": 0.29014, + "grad_norm": 0.8691380620002747, + "learning_rate": 0.003, + "loss": 4.013, + "step": 29014 + }, + { + "epoch": 0.29015, + "grad_norm": 1.071904182434082, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 29015 + }, + { + "epoch": 0.29016, + "grad_norm": 1.2634536027908325, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 29016 + }, + { + "epoch": 0.29017, + "grad_norm": 0.6748138070106506, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 29017 + }, + { + "epoch": 0.29018, + "grad_norm": 0.5864481925964355, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 29018 + }, + { + "epoch": 0.29019, + "grad_norm": 0.8671297430992126, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 29019 + }, + { + "epoch": 0.2902, + "grad_norm": 1.0212777853012085, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 29020 + }, + { + "epoch": 0.29021, + "grad_norm": 0.8770200610160828, + "learning_rate": 0.003, + "loss": 3.959, + "step": 29021 + }, + { + "epoch": 0.29022, + "grad_norm": 0.7804937958717346, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 29022 + }, + { + "epoch": 0.29023, + "grad_norm": 0.7627916932106018, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 29023 + }, + { + "epoch": 0.29024, + "grad_norm": 0.7592844367027283, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 29024 + }, + { + "epoch": 0.29025, + "grad_norm": 0.7141545414924622, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 29025 + }, + { + "epoch": 0.29026, + "grad_norm": 0.7312930226325989, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 29026 + }, + { + "epoch": 0.29027, + "grad_norm": 0.7952088117599487, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 29027 + }, + { + "epoch": 0.29028, + "grad_norm": 0.7387848496437073, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 29028 + }, + { + "epoch": 0.29029, + "grad_norm": 0.7242494225502014, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 29029 + }, + { + "epoch": 0.2903, + "grad_norm": 0.7841684222221375, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 29030 + }, + { + "epoch": 0.29031, + "grad_norm": 0.8776065707206726, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 29031 + }, + { + "epoch": 0.29032, + "grad_norm": 0.8603924512863159, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 29032 + }, + { + "epoch": 0.29033, + "grad_norm": 0.8497645854949951, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 29033 + }, + { + "epoch": 0.29034, + "grad_norm": 1.0446467399597168, + "learning_rate": 0.003, + "loss": 4.003, + "step": 29034 + }, + { + "epoch": 0.29035, + "grad_norm": 1.226523756980896, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 29035 + }, + { + "epoch": 0.29036, + "grad_norm": 0.7440049648284912, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 29036 + }, + { + "epoch": 0.29037, + "grad_norm": 0.9464699625968933, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 29037 + }, + { + "epoch": 0.29038, + "grad_norm": 1.1137549877166748, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 29038 + }, + { + "epoch": 0.29039, + "grad_norm": 1.1873594522476196, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 29039 + }, + { + "epoch": 0.2904, + "grad_norm": 0.9341791272163391, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 29040 + }, + { + "epoch": 0.29041, + "grad_norm": 0.8500692844390869, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 29041 + }, + { + "epoch": 0.29042, + "grad_norm": 0.7202122211456299, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 29042 + }, + { + "epoch": 0.29043, + "grad_norm": 0.702830970287323, + "learning_rate": 0.003, + "loss": 3.98, + "step": 29043 + }, + { + "epoch": 0.29044, + "grad_norm": 0.7628594636917114, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 29044 + }, + { + "epoch": 0.29045, + "grad_norm": 0.7980680465698242, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 29045 + }, + { + "epoch": 0.29046, + "grad_norm": 0.7479470372200012, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 29046 + }, + { + "epoch": 0.29047, + "grad_norm": 0.7880944609642029, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 29047 + }, + { + "epoch": 0.29048, + "grad_norm": 0.8659824728965759, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 29048 + }, + { + "epoch": 0.29049, + "grad_norm": 0.877802848815918, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 29049 + }, + { + "epoch": 0.2905, + "grad_norm": 0.8870892524719238, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 29050 + }, + { + "epoch": 0.29051, + "grad_norm": 0.9068344831466675, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 29051 + }, + { + "epoch": 0.29052, + "grad_norm": 1.1183232069015503, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 29052 + }, + { + "epoch": 0.29053, + "grad_norm": 1.1106723546981812, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 29053 + }, + { + "epoch": 0.29054, + "grad_norm": 0.7954079508781433, + "learning_rate": 0.003, + "loss": 3.959, + "step": 29054 + }, + { + "epoch": 0.29055, + "grad_norm": 0.7728688716888428, + "learning_rate": 0.003, + "loss": 3.981, + "step": 29055 + }, + { + "epoch": 0.29056, + "grad_norm": 0.7000091671943665, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 29056 + }, + { + "epoch": 0.29057, + "grad_norm": 0.6387006044387817, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 29057 + }, + { + "epoch": 0.29058, + "grad_norm": 0.6317980885505676, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 29058 + }, + { + "epoch": 0.29059, + "grad_norm": 0.661799430847168, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 29059 + }, + { + "epoch": 0.2906, + "grad_norm": 0.6999861001968384, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 29060 + }, + { + "epoch": 0.29061, + "grad_norm": 0.9766762852668762, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 29061 + }, + { + "epoch": 0.29062, + "grad_norm": 1.3159908056259155, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 29062 + }, + { + "epoch": 0.29063, + "grad_norm": 0.8390430808067322, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 29063 + }, + { + "epoch": 0.29064, + "grad_norm": 0.7377676963806152, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 29064 + }, + { + "epoch": 0.29065, + "grad_norm": 0.6666255593299866, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 29065 + }, + { + "epoch": 0.29066, + "grad_norm": 0.6708187460899353, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 29066 + }, + { + "epoch": 0.29067, + "grad_norm": 0.7444360852241516, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 29067 + }, + { + "epoch": 0.29068, + "grad_norm": 0.656787097454071, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 29068 + }, + { + "epoch": 0.29069, + "grad_norm": 0.6715625524520874, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 29069 + }, + { + "epoch": 0.2907, + "grad_norm": 0.7290574908256531, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 29070 + }, + { + "epoch": 0.29071, + "grad_norm": 0.7984590530395508, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 29071 + }, + { + "epoch": 0.29072, + "grad_norm": 0.8420333862304688, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 29072 + }, + { + "epoch": 0.29073, + "grad_norm": 0.8428716659545898, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 29073 + }, + { + "epoch": 0.29074, + "grad_norm": 0.9218831658363342, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 29074 + }, + { + "epoch": 0.29075, + "grad_norm": 1.1427799463272095, + "learning_rate": 0.003, + "loss": 3.976, + "step": 29075 + }, + { + "epoch": 0.29076, + "grad_norm": 0.8824218511581421, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 29076 + }, + { + "epoch": 0.29077, + "grad_norm": 0.708073079586029, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 29077 + }, + { + "epoch": 0.29078, + "grad_norm": 0.6253885626792908, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 29078 + }, + { + "epoch": 0.29079, + "grad_norm": 0.6240352392196655, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 29079 + }, + { + "epoch": 0.2908, + "grad_norm": 0.6844454407691956, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 29080 + }, + { + "epoch": 0.29081, + "grad_norm": 1.0164313316345215, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 29081 + }, + { + "epoch": 0.29082, + "grad_norm": 1.2960569858551025, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 29082 + }, + { + "epoch": 0.29083, + "grad_norm": 0.710368812084198, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 29083 + }, + { + "epoch": 0.29084, + "grad_norm": 0.6953149437904358, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 29084 + }, + { + "epoch": 0.29085, + "grad_norm": 0.752515435218811, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 29085 + }, + { + "epoch": 0.29086, + "grad_norm": 0.8468724489212036, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 29086 + }, + { + "epoch": 0.29087, + "grad_norm": 1.1469382047653198, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 29087 + }, + { + "epoch": 0.29088, + "grad_norm": 0.8762564063072205, + "learning_rate": 0.003, + "loss": 3.978, + "step": 29088 + }, + { + "epoch": 0.29089, + "grad_norm": 0.8044001460075378, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 29089 + }, + { + "epoch": 0.2909, + "grad_norm": 0.8248804211616516, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 29090 + }, + { + "epoch": 0.29091, + "grad_norm": 0.8836286664009094, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 29091 + }, + { + "epoch": 0.29092, + "grad_norm": 0.9813172221183777, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 29092 + }, + { + "epoch": 0.29093, + "grad_norm": 1.0707111358642578, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 29093 + }, + { + "epoch": 0.29094, + "grad_norm": 1.1566649675369263, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 29094 + }, + { + "epoch": 0.29095, + "grad_norm": 0.8425024151802063, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 29095 + }, + { + "epoch": 0.29096, + "grad_norm": 0.8229311108589172, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 29096 + }, + { + "epoch": 0.29097, + "grad_norm": 1.0229823589324951, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 29097 + }, + { + "epoch": 0.29098, + "grad_norm": 1.048198938369751, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 29098 + }, + { + "epoch": 0.29099, + "grad_norm": 1.0543988943099976, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 29099 + }, + { + "epoch": 0.291, + "grad_norm": 0.9616487622261047, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 29100 + }, + { + "epoch": 0.29101, + "grad_norm": 1.229850172996521, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 29101 + }, + { + "epoch": 0.29102, + "grad_norm": 0.9451892971992493, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 29102 + }, + { + "epoch": 0.29103, + "grad_norm": 1.1262131929397583, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 29103 + }, + { + "epoch": 0.29104, + "grad_norm": 0.9493424892425537, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 29104 + }, + { + "epoch": 0.29105, + "grad_norm": 0.8151524066925049, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 29105 + }, + { + "epoch": 0.29106, + "grad_norm": 0.8535007834434509, + "learning_rate": 0.003, + "loss": 4.004, + "step": 29106 + }, + { + "epoch": 0.29107, + "grad_norm": 0.8018661141395569, + "learning_rate": 0.003, + "loss": 3.967, + "step": 29107 + }, + { + "epoch": 0.29108, + "grad_norm": 0.7357597947120667, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 29108 + }, + { + "epoch": 0.29109, + "grad_norm": 0.8225977420806885, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 29109 + }, + { + "epoch": 0.2911, + "grad_norm": 0.9635128974914551, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 29110 + }, + { + "epoch": 0.29111, + "grad_norm": 1.0726709365844727, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 29111 + }, + { + "epoch": 0.29112, + "grad_norm": 0.9254584908485413, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 29112 + }, + { + "epoch": 0.29113, + "grad_norm": 1.0285121202468872, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 29113 + }, + { + "epoch": 0.29114, + "grad_norm": 1.222135305404663, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 29114 + }, + { + "epoch": 0.29115, + "grad_norm": 0.8261186480522156, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 29115 + }, + { + "epoch": 0.29116, + "grad_norm": 0.7160526514053345, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 29116 + }, + { + "epoch": 0.29117, + "grad_norm": 0.684047281742096, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 29117 + }, + { + "epoch": 0.29118, + "grad_norm": 0.8303099274635315, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 29118 + }, + { + "epoch": 0.29119, + "grad_norm": 0.9164711236953735, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 29119 + }, + { + "epoch": 0.2912, + "grad_norm": 0.8639779686927795, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 29120 + }, + { + "epoch": 0.29121, + "grad_norm": 0.8591728806495667, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 29121 + }, + { + "epoch": 0.29122, + "grad_norm": 1.0446499586105347, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 29122 + }, + { + "epoch": 0.29123, + "grad_norm": 1.2550945281982422, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 29123 + }, + { + "epoch": 0.29124, + "grad_norm": 0.9949382543563843, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 29124 + }, + { + "epoch": 0.29125, + "grad_norm": 0.8178385496139526, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 29125 + }, + { + "epoch": 0.29126, + "grad_norm": 0.6640225648880005, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 29126 + }, + { + "epoch": 0.29127, + "grad_norm": 0.7248461246490479, + "learning_rate": 0.003, + "loss": 3.964, + "step": 29127 + }, + { + "epoch": 0.29128, + "grad_norm": 0.7678794860839844, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 29128 + }, + { + "epoch": 0.29129, + "grad_norm": 0.9037972092628479, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 29129 + }, + { + "epoch": 0.2913, + "grad_norm": 0.9878264665603638, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 29130 + }, + { + "epoch": 0.29131, + "grad_norm": 0.9598780870437622, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 29131 + }, + { + "epoch": 0.29132, + "grad_norm": 0.9007064700126648, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 29132 + }, + { + "epoch": 0.29133, + "grad_norm": 0.9206957817077637, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 29133 + }, + { + "epoch": 0.29134, + "grad_norm": 0.9214566349983215, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 29134 + }, + { + "epoch": 0.29135, + "grad_norm": 0.8530617952346802, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 29135 + }, + { + "epoch": 0.29136, + "grad_norm": 0.7517346143722534, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 29136 + }, + { + "epoch": 0.29137, + "grad_norm": 0.8036770224571228, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 29137 + }, + { + "epoch": 0.29138, + "grad_norm": 0.9288204312324524, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 29138 + }, + { + "epoch": 0.29139, + "grad_norm": 0.9850285649299622, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 29139 + }, + { + "epoch": 0.2914, + "grad_norm": 0.8792307376861572, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 29140 + }, + { + "epoch": 0.29141, + "grad_norm": 0.8612130880355835, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 29141 + }, + { + "epoch": 0.29142, + "grad_norm": 0.8895094394683838, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 29142 + }, + { + "epoch": 0.29143, + "grad_norm": 0.9036446213722229, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 29143 + }, + { + "epoch": 0.29144, + "grad_norm": 0.988093376159668, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 29144 + }, + { + "epoch": 0.29145, + "grad_norm": 1.2282615900039673, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 29145 + }, + { + "epoch": 0.29146, + "grad_norm": 0.7327998876571655, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 29146 + }, + { + "epoch": 0.29147, + "grad_norm": 0.6757382750511169, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 29147 + }, + { + "epoch": 0.29148, + "grad_norm": 0.8298256993293762, + "learning_rate": 0.003, + "loss": 4.021, + "step": 29148 + }, + { + "epoch": 0.29149, + "grad_norm": 0.8512889742851257, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 29149 + }, + { + "epoch": 0.2915, + "grad_norm": 0.9529120922088623, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 29150 + }, + { + "epoch": 0.29151, + "grad_norm": 1.1277652978897095, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 29151 + }, + { + "epoch": 0.29152, + "grad_norm": 1.2719060182571411, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 29152 + }, + { + "epoch": 0.29153, + "grad_norm": 0.6643340587615967, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 29153 + }, + { + "epoch": 0.29154, + "grad_norm": 0.8010669946670532, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 29154 + }, + { + "epoch": 0.29155, + "grad_norm": 0.7080842852592468, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 29155 + }, + { + "epoch": 0.29156, + "grad_norm": 0.84911048412323, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 29156 + }, + { + "epoch": 0.29157, + "grad_norm": 0.9367660880088806, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 29157 + }, + { + "epoch": 0.29158, + "grad_norm": 0.996724545955658, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 29158 + }, + { + "epoch": 0.29159, + "grad_norm": 0.9063332676887512, + "learning_rate": 0.003, + "loss": 3.994, + "step": 29159 + }, + { + "epoch": 0.2916, + "grad_norm": 0.791290819644928, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 29160 + }, + { + "epoch": 0.29161, + "grad_norm": 0.6401340365409851, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 29161 + }, + { + "epoch": 0.29162, + "grad_norm": 0.5129435658454895, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 29162 + }, + { + "epoch": 0.29163, + "grad_norm": 0.5428082346916199, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 29163 + }, + { + "epoch": 0.29164, + "grad_norm": 0.6088681817054749, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 29164 + }, + { + "epoch": 0.29165, + "grad_norm": 0.580108642578125, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 29165 + }, + { + "epoch": 0.29166, + "grad_norm": 0.7038347125053406, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 29166 + }, + { + "epoch": 0.29167, + "grad_norm": 0.8075817823410034, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 29167 + }, + { + "epoch": 0.29168, + "grad_norm": 1.0836613178253174, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 29168 + }, + { + "epoch": 0.29169, + "grad_norm": 1.0090515613555908, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 29169 + }, + { + "epoch": 0.2917, + "grad_norm": 1.0472471714019775, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 29170 + }, + { + "epoch": 0.29171, + "grad_norm": 0.9498915672302246, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 29171 + }, + { + "epoch": 0.29172, + "grad_norm": 0.9241468906402588, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 29172 + }, + { + "epoch": 0.29173, + "grad_norm": 0.8396495580673218, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 29173 + }, + { + "epoch": 0.29174, + "grad_norm": 0.8846224546432495, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 29174 + }, + { + "epoch": 0.29175, + "grad_norm": 0.9156776070594788, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 29175 + }, + { + "epoch": 0.29176, + "grad_norm": 0.931320309638977, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 29176 + }, + { + "epoch": 0.29177, + "grad_norm": 0.789471447467804, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 29177 + }, + { + "epoch": 0.29178, + "grad_norm": 0.629005491733551, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 29178 + }, + { + "epoch": 0.29179, + "grad_norm": 0.7003027200698853, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 29179 + }, + { + "epoch": 0.2918, + "grad_norm": 0.7474744915962219, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 29180 + }, + { + "epoch": 0.29181, + "grad_norm": 0.7580128908157349, + "learning_rate": 0.003, + "loss": 4.0, + "step": 29181 + }, + { + "epoch": 0.29182, + "grad_norm": 0.7144866585731506, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 29182 + }, + { + "epoch": 0.29183, + "grad_norm": 0.6723872423171997, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 29183 + }, + { + "epoch": 0.29184, + "grad_norm": 0.6724082231521606, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 29184 + }, + { + "epoch": 0.29185, + "grad_norm": 0.6809003949165344, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 29185 + }, + { + "epoch": 0.29186, + "grad_norm": 0.6850377917289734, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 29186 + }, + { + "epoch": 0.29187, + "grad_norm": 0.6929375529289246, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 29187 + }, + { + "epoch": 0.29188, + "grad_norm": 0.8198863863945007, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 29188 + }, + { + "epoch": 0.29189, + "grad_norm": 1.0433882474899292, + "learning_rate": 0.003, + "loss": 4.008, + "step": 29189 + }, + { + "epoch": 0.2919, + "grad_norm": 1.1796213388442993, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 29190 + }, + { + "epoch": 0.29191, + "grad_norm": 0.860930860042572, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 29191 + }, + { + "epoch": 0.29192, + "grad_norm": 0.8166892528533936, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 29192 + }, + { + "epoch": 0.29193, + "grad_norm": 1.0184515714645386, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 29193 + }, + { + "epoch": 0.29194, + "grad_norm": 1.0595457553863525, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 29194 + }, + { + "epoch": 0.29195, + "grad_norm": 0.7988605499267578, + "learning_rate": 0.003, + "loss": 3.9478, + "step": 29195 + }, + { + "epoch": 0.29196, + "grad_norm": 0.8283690214157104, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 29196 + }, + { + "epoch": 0.29197, + "grad_norm": 0.8145237565040588, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 29197 + }, + { + "epoch": 0.29198, + "grad_norm": 0.8930994272232056, + "learning_rate": 0.003, + "loss": 3.9402, + "step": 29198 + }, + { + "epoch": 0.29199, + "grad_norm": 1.0436289310455322, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 29199 + }, + { + "epoch": 0.292, + "grad_norm": 0.8150096535682678, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 29200 + }, + { + "epoch": 0.29201, + "grad_norm": 0.8856778740882874, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 29201 + }, + { + "epoch": 0.29202, + "grad_norm": 0.9844528436660767, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 29202 + }, + { + "epoch": 0.29203, + "grad_norm": 1.1051750183105469, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 29203 + }, + { + "epoch": 0.29204, + "grad_norm": 1.0388286113739014, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 29204 + }, + { + "epoch": 0.29205, + "grad_norm": 1.042790412902832, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 29205 + }, + { + "epoch": 0.29206, + "grad_norm": 1.1539785861968994, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 29206 + }, + { + "epoch": 0.29207, + "grad_norm": 0.8248788714408875, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 29207 + }, + { + "epoch": 0.29208, + "grad_norm": 0.7372804284095764, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 29208 + }, + { + "epoch": 0.29209, + "grad_norm": 0.704511284828186, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 29209 + }, + { + "epoch": 0.2921, + "grad_norm": 0.7195176482200623, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 29210 + }, + { + "epoch": 0.29211, + "grad_norm": 0.7721216082572937, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 29211 + }, + { + "epoch": 0.29212, + "grad_norm": 0.7659585475921631, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 29212 + }, + { + "epoch": 0.29213, + "grad_norm": 0.9152370691299438, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 29213 + }, + { + "epoch": 0.29214, + "grad_norm": 1.307989478111267, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 29214 + }, + { + "epoch": 0.29215, + "grad_norm": 0.8232973217964172, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 29215 + }, + { + "epoch": 0.29216, + "grad_norm": 0.7361544966697693, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 29216 + }, + { + "epoch": 0.29217, + "grad_norm": 0.7464147806167603, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 29217 + }, + { + "epoch": 0.29218, + "grad_norm": 0.6581504940986633, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 29218 + }, + { + "epoch": 0.29219, + "grad_norm": 0.7712945342063904, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 29219 + }, + { + "epoch": 0.2922, + "grad_norm": 0.9574998617172241, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 29220 + }, + { + "epoch": 0.29221, + "grad_norm": 1.099255084991455, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 29221 + }, + { + "epoch": 0.29222, + "grad_norm": 1.12336003780365, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 29222 + }, + { + "epoch": 0.29223, + "grad_norm": 0.9597835540771484, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 29223 + }, + { + "epoch": 0.29224, + "grad_norm": 0.8007677793502808, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 29224 + }, + { + "epoch": 0.29225, + "grad_norm": 0.7391369938850403, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 29225 + }, + { + "epoch": 0.29226, + "grad_norm": 0.8533346056938171, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 29226 + }, + { + "epoch": 0.29227, + "grad_norm": 0.9875987768173218, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 29227 + }, + { + "epoch": 0.29228, + "grad_norm": 1.1680116653442383, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 29228 + }, + { + "epoch": 0.29229, + "grad_norm": 0.7859846353530884, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 29229 + }, + { + "epoch": 0.2923, + "grad_norm": 0.7610158920288086, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 29230 + }, + { + "epoch": 0.29231, + "grad_norm": 0.8984367251396179, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 29231 + }, + { + "epoch": 0.29232, + "grad_norm": 0.9831880331039429, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 29232 + }, + { + "epoch": 0.29233, + "grad_norm": 1.1804401874542236, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 29233 + }, + { + "epoch": 0.29234, + "grad_norm": 0.9006233215332031, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 29234 + }, + { + "epoch": 0.29235, + "grad_norm": 0.7933160066604614, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 29235 + }, + { + "epoch": 0.29236, + "grad_norm": 0.7082411050796509, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 29236 + }, + { + "epoch": 0.29237, + "grad_norm": 0.9357009530067444, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 29237 + }, + { + "epoch": 0.29238, + "grad_norm": 0.9452374577522278, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 29238 + }, + { + "epoch": 0.29239, + "grad_norm": 0.794632613658905, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 29239 + }, + { + "epoch": 0.2924, + "grad_norm": 0.6512444019317627, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 29240 + }, + { + "epoch": 0.29241, + "grad_norm": 0.6226893663406372, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 29241 + }, + { + "epoch": 0.29242, + "grad_norm": 0.6755322217941284, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 29242 + }, + { + "epoch": 0.29243, + "grad_norm": 0.9990161657333374, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 29243 + }, + { + "epoch": 0.29244, + "grad_norm": 1.444792628288269, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 29244 + }, + { + "epoch": 0.29245, + "grad_norm": 0.5044780373573303, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 29245 + }, + { + "epoch": 0.29246, + "grad_norm": 0.7684431076049805, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 29246 + }, + { + "epoch": 0.29247, + "grad_norm": 0.9647959470748901, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 29247 + }, + { + "epoch": 0.29248, + "grad_norm": 1.2201297283172607, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 29248 + }, + { + "epoch": 0.29249, + "grad_norm": 0.8724824786186218, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 29249 + }, + { + "epoch": 0.2925, + "grad_norm": 0.6886383295059204, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 29250 + }, + { + "epoch": 0.29251, + "grad_norm": 0.653754711151123, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 29251 + }, + { + "epoch": 0.29252, + "grad_norm": 0.7422829270362854, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 29252 + }, + { + "epoch": 0.29253, + "grad_norm": 0.8616805672645569, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 29253 + }, + { + "epoch": 0.29254, + "grad_norm": 0.8735698461532593, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 29254 + }, + { + "epoch": 0.29255, + "grad_norm": 0.9764900207519531, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 29255 + }, + { + "epoch": 0.29256, + "grad_norm": 1.2018746137619019, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 29256 + }, + { + "epoch": 0.29257, + "grad_norm": 0.7719864249229431, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 29257 + }, + { + "epoch": 0.29258, + "grad_norm": 0.7986122965812683, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 29258 + }, + { + "epoch": 0.29259, + "grad_norm": 0.885481059551239, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 29259 + }, + { + "epoch": 0.2926, + "grad_norm": 1.0476270914077759, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 29260 + }, + { + "epoch": 0.29261, + "grad_norm": 1.0565071105957031, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 29261 + }, + { + "epoch": 0.29262, + "grad_norm": 0.8654701709747314, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 29262 + }, + { + "epoch": 0.29263, + "grad_norm": 0.7555111646652222, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 29263 + }, + { + "epoch": 0.29264, + "grad_norm": 0.7232216596603394, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 29264 + }, + { + "epoch": 0.29265, + "grad_norm": 0.6900628805160522, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 29265 + }, + { + "epoch": 0.29266, + "grad_norm": 0.6933633089065552, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 29266 + }, + { + "epoch": 0.29267, + "grad_norm": 0.7504774332046509, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 29267 + }, + { + "epoch": 0.29268, + "grad_norm": 0.9046775698661804, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 29268 + }, + { + "epoch": 0.29269, + "grad_norm": 0.911592960357666, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 29269 + }, + { + "epoch": 0.2927, + "grad_norm": 0.8777236938476562, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 29270 + }, + { + "epoch": 0.29271, + "grad_norm": 1.0550326108932495, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 29271 + }, + { + "epoch": 0.29272, + "grad_norm": 1.112715482711792, + "learning_rate": 0.003, + "loss": 3.981, + "step": 29272 + }, + { + "epoch": 0.29273, + "grad_norm": 1.103261113166809, + "learning_rate": 0.003, + "loss": 3.992, + "step": 29273 + }, + { + "epoch": 0.29274, + "grad_norm": 0.7478103637695312, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 29274 + }, + { + "epoch": 0.29275, + "grad_norm": 0.8776088953018188, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 29275 + }, + { + "epoch": 0.29276, + "grad_norm": 0.9592344760894775, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 29276 + }, + { + "epoch": 0.29277, + "grad_norm": 1.1287157535552979, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 29277 + }, + { + "epoch": 0.29278, + "grad_norm": 0.9521082043647766, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 29278 + }, + { + "epoch": 0.29279, + "grad_norm": 0.9368396401405334, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 29279 + }, + { + "epoch": 0.2928, + "grad_norm": 0.9996592998504639, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 29280 + }, + { + "epoch": 0.29281, + "grad_norm": 0.9669919610023499, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 29281 + }, + { + "epoch": 0.29282, + "grad_norm": 0.8890926241874695, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 29282 + }, + { + "epoch": 0.29283, + "grad_norm": 0.7899990081787109, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 29283 + }, + { + "epoch": 0.29284, + "grad_norm": 0.7909675240516663, + "learning_rate": 0.003, + "loss": 3.973, + "step": 29284 + }, + { + "epoch": 0.29285, + "grad_norm": 0.7556570768356323, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 29285 + }, + { + "epoch": 0.29286, + "grad_norm": 0.741875410079956, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 29286 + }, + { + "epoch": 0.29287, + "grad_norm": 0.7488574981689453, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 29287 + }, + { + "epoch": 0.29288, + "grad_norm": 0.6786359548568726, + "learning_rate": 0.003, + "loss": 3.973, + "step": 29288 + }, + { + "epoch": 0.29289, + "grad_norm": 0.7307427525520325, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 29289 + }, + { + "epoch": 0.2929, + "grad_norm": 0.7799848914146423, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 29290 + }, + { + "epoch": 0.29291, + "grad_norm": 0.9664953947067261, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 29291 + }, + { + "epoch": 0.29292, + "grad_norm": 1.1708741188049316, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 29292 + }, + { + "epoch": 0.29293, + "grad_norm": 0.9785391092300415, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 29293 + }, + { + "epoch": 0.29294, + "grad_norm": 0.9946236610412598, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 29294 + }, + { + "epoch": 0.29295, + "grad_norm": 0.8694761991500854, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 29295 + }, + { + "epoch": 0.29296, + "grad_norm": 0.8015434741973877, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 29296 + }, + { + "epoch": 0.29297, + "grad_norm": 0.8718987703323364, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 29297 + }, + { + "epoch": 0.29298, + "grad_norm": 0.8006660342216492, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 29298 + }, + { + "epoch": 0.29299, + "grad_norm": 0.7811927795410156, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 29299 + }, + { + "epoch": 0.293, + "grad_norm": 0.8821768760681152, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 29300 + }, + { + "epoch": 0.29301, + "grad_norm": 0.9367978572845459, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 29301 + }, + { + "epoch": 0.29302, + "grad_norm": 0.8181772828102112, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 29302 + }, + { + "epoch": 0.29303, + "grad_norm": 0.8177184462547302, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 29303 + }, + { + "epoch": 0.29304, + "grad_norm": 0.8471592664718628, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 29304 + }, + { + "epoch": 0.29305, + "grad_norm": 0.8865258097648621, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 29305 + }, + { + "epoch": 0.29306, + "grad_norm": 0.9488347768783569, + "learning_rate": 0.003, + "loss": 3.988, + "step": 29306 + }, + { + "epoch": 0.29307, + "grad_norm": 1.0728765726089478, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 29307 + }, + { + "epoch": 0.29308, + "grad_norm": 0.9546968936920166, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 29308 + }, + { + "epoch": 0.29309, + "grad_norm": 0.8857902884483337, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 29309 + }, + { + "epoch": 0.2931, + "grad_norm": 0.9656665921211243, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 29310 + }, + { + "epoch": 0.29311, + "grad_norm": 1.028521180152893, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 29311 + }, + { + "epoch": 0.29312, + "grad_norm": 0.8224505186080933, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 29312 + }, + { + "epoch": 0.29313, + "grad_norm": 0.8240838050842285, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 29313 + }, + { + "epoch": 0.29314, + "grad_norm": 0.7057287096977234, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 29314 + }, + { + "epoch": 0.29315, + "grad_norm": 0.8484646677970886, + "learning_rate": 0.003, + "loss": 4.011, + "step": 29315 + }, + { + "epoch": 0.29316, + "grad_norm": 0.8153771162033081, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 29316 + }, + { + "epoch": 0.29317, + "grad_norm": 0.8124666810035706, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 29317 + }, + { + "epoch": 0.29318, + "grad_norm": 0.7132452726364136, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 29318 + }, + { + "epoch": 0.29319, + "grad_norm": 0.8547071218490601, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 29319 + }, + { + "epoch": 0.2932, + "grad_norm": 0.9821109771728516, + "learning_rate": 0.003, + "loss": 3.994, + "step": 29320 + }, + { + "epoch": 0.29321, + "grad_norm": 1.2050490379333496, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 29321 + }, + { + "epoch": 0.29322, + "grad_norm": 0.9138466715812683, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 29322 + }, + { + "epoch": 0.29323, + "grad_norm": 0.9306283593177795, + "learning_rate": 0.003, + "loss": 3.969, + "step": 29323 + }, + { + "epoch": 0.29324, + "grad_norm": 0.987065851688385, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 29324 + }, + { + "epoch": 0.29325, + "grad_norm": 0.915293276309967, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 29325 + }, + { + "epoch": 0.29326, + "grad_norm": 0.7338712215423584, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 29326 + }, + { + "epoch": 0.29327, + "grad_norm": 0.641572117805481, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 29327 + }, + { + "epoch": 0.29328, + "grad_norm": 0.6119932532310486, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 29328 + }, + { + "epoch": 0.29329, + "grad_norm": 0.5940769910812378, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 29329 + }, + { + "epoch": 0.2933, + "grad_norm": 0.65184485912323, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 29330 + }, + { + "epoch": 0.29331, + "grad_norm": 0.8962559700012207, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 29331 + }, + { + "epoch": 0.29332, + "grad_norm": 1.1467703580856323, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 29332 + }, + { + "epoch": 0.29333, + "grad_norm": 1.007598876953125, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 29333 + }, + { + "epoch": 0.29334, + "grad_norm": 0.8257408142089844, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 29334 + }, + { + "epoch": 0.29335, + "grad_norm": 0.7550776600837708, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 29335 + }, + { + "epoch": 0.29336, + "grad_norm": 0.8530635833740234, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 29336 + }, + { + "epoch": 0.29337, + "grad_norm": 0.7313423156738281, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 29337 + }, + { + "epoch": 0.29338, + "grad_norm": 0.6883638501167297, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 29338 + }, + { + "epoch": 0.29339, + "grad_norm": 0.7446149587631226, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 29339 + }, + { + "epoch": 0.2934, + "grad_norm": 0.9066970348358154, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 29340 + }, + { + "epoch": 0.29341, + "grad_norm": 1.100481629371643, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 29341 + }, + { + "epoch": 0.29342, + "grad_norm": 0.9821372032165527, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 29342 + }, + { + "epoch": 0.29343, + "grad_norm": 1.0033522844314575, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 29343 + }, + { + "epoch": 0.29344, + "grad_norm": 1.0465612411499023, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 29344 + }, + { + "epoch": 0.29345, + "grad_norm": 1.0303397178649902, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 29345 + }, + { + "epoch": 0.29346, + "grad_norm": 0.9658594727516174, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 29346 + }, + { + "epoch": 0.29347, + "grad_norm": 0.9909182786941528, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 29347 + }, + { + "epoch": 0.29348, + "grad_norm": 1.0295697450637817, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 29348 + }, + { + "epoch": 0.29349, + "grad_norm": 0.8586769104003906, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 29349 + }, + { + "epoch": 0.2935, + "grad_norm": 0.8966729044914246, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 29350 + }, + { + "epoch": 0.29351, + "grad_norm": 0.824487030506134, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 29351 + }, + { + "epoch": 0.29352, + "grad_norm": 0.7553017735481262, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 29352 + }, + { + "epoch": 0.29353, + "grad_norm": 0.6734135150909424, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 29353 + }, + { + "epoch": 0.29354, + "grad_norm": 0.6902981400489807, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 29354 + }, + { + "epoch": 0.29355, + "grad_norm": 0.7422550320625305, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 29355 + }, + { + "epoch": 0.29356, + "grad_norm": 0.7456428408622742, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 29356 + }, + { + "epoch": 0.29357, + "grad_norm": 0.702684760093689, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 29357 + }, + { + "epoch": 0.29358, + "grad_norm": 0.8083476424217224, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 29358 + }, + { + "epoch": 0.29359, + "grad_norm": 1.0360956192016602, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 29359 + }, + { + "epoch": 0.2936, + "grad_norm": 0.9674257636070251, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 29360 + }, + { + "epoch": 0.29361, + "grad_norm": 1.105712890625, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 29361 + }, + { + "epoch": 0.29362, + "grad_norm": 0.9481111764907837, + "learning_rate": 0.003, + "loss": 3.971, + "step": 29362 + }, + { + "epoch": 0.29363, + "grad_norm": 0.882268488407135, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 29363 + }, + { + "epoch": 0.29364, + "grad_norm": 0.8746227622032166, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 29364 + }, + { + "epoch": 0.29365, + "grad_norm": 0.886730968952179, + "learning_rate": 0.003, + "loss": 3.976, + "step": 29365 + }, + { + "epoch": 0.29366, + "grad_norm": 0.9215877652168274, + "learning_rate": 0.003, + "loss": 4.004, + "step": 29366 + }, + { + "epoch": 0.29367, + "grad_norm": 1.0151828527450562, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 29367 + }, + { + "epoch": 0.29368, + "grad_norm": 1.0615054368972778, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 29368 + }, + { + "epoch": 0.29369, + "grad_norm": 0.844161331653595, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 29369 + }, + { + "epoch": 0.2937, + "grad_norm": 0.9694164991378784, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 29370 + }, + { + "epoch": 0.29371, + "grad_norm": 1.2660253047943115, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 29371 + }, + { + "epoch": 0.29372, + "grad_norm": 0.5459420680999756, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 29372 + }, + { + "epoch": 0.29373, + "grad_norm": 0.7477525472640991, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 29373 + }, + { + "epoch": 0.29374, + "grad_norm": 0.786474883556366, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 29374 + }, + { + "epoch": 0.29375, + "grad_norm": 0.7205208539962769, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 29375 + }, + { + "epoch": 0.29376, + "grad_norm": 0.8531522154808044, + "learning_rate": 0.003, + "loss": 3.9164, + "step": 29376 + }, + { + "epoch": 0.29377, + "grad_norm": 0.8495098948478699, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 29377 + }, + { + "epoch": 0.29378, + "grad_norm": 0.8920244574546814, + "learning_rate": 0.003, + "loss": 3.963, + "step": 29378 + }, + { + "epoch": 0.29379, + "grad_norm": 0.8130154609680176, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 29379 + }, + { + "epoch": 0.2938, + "grad_norm": 0.7649970650672913, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 29380 + }, + { + "epoch": 0.29381, + "grad_norm": 0.7626245617866516, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 29381 + }, + { + "epoch": 0.29382, + "grad_norm": 0.816891074180603, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 29382 + }, + { + "epoch": 0.29383, + "grad_norm": 0.8472288846969604, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 29383 + }, + { + "epoch": 0.29384, + "grad_norm": 1.0428357124328613, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 29384 + }, + { + "epoch": 0.29385, + "grad_norm": 0.9986610412597656, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 29385 + }, + { + "epoch": 0.29386, + "grad_norm": 1.0163211822509766, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 29386 + }, + { + "epoch": 0.29387, + "grad_norm": 0.8433914184570312, + "learning_rate": 0.003, + "loss": 4.0, + "step": 29387 + }, + { + "epoch": 0.29388, + "grad_norm": 0.9769079089164734, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 29388 + }, + { + "epoch": 0.29389, + "grad_norm": 1.0912336111068726, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 29389 + }, + { + "epoch": 0.2939, + "grad_norm": 0.8398504853248596, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 29390 + }, + { + "epoch": 0.29391, + "grad_norm": 0.8236151337623596, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 29391 + }, + { + "epoch": 0.29392, + "grad_norm": 1.0148974657058716, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 29392 + }, + { + "epoch": 0.29393, + "grad_norm": 1.3317327499389648, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 29393 + }, + { + "epoch": 0.29394, + "grad_norm": 0.574306070804596, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 29394 + }, + { + "epoch": 0.29395, + "grad_norm": 0.7591708302497864, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 29395 + }, + { + "epoch": 0.29396, + "grad_norm": 0.919202446937561, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 29396 + }, + { + "epoch": 0.29397, + "grad_norm": 0.8230723142623901, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 29397 + }, + { + "epoch": 0.29398, + "grad_norm": 0.9167483448982239, + "learning_rate": 0.003, + "loss": 4.001, + "step": 29398 + }, + { + "epoch": 0.29399, + "grad_norm": 0.9203762412071228, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 29399 + }, + { + "epoch": 0.294, + "grad_norm": 0.8343217968940735, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 29400 + }, + { + "epoch": 0.29401, + "grad_norm": 0.8445692658424377, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 29401 + }, + { + "epoch": 0.29402, + "grad_norm": 0.8812264204025269, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 29402 + }, + { + "epoch": 0.29403, + "grad_norm": 0.9557315111160278, + "learning_rate": 0.003, + "loss": 4.005, + "step": 29403 + }, + { + "epoch": 0.29404, + "grad_norm": 1.066479206085205, + "learning_rate": 0.003, + "loss": 4.017, + "step": 29404 + }, + { + "epoch": 0.29405, + "grad_norm": 1.0261693000793457, + "learning_rate": 0.003, + "loss": 3.982, + "step": 29405 + }, + { + "epoch": 0.29406, + "grad_norm": 0.870642900466919, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 29406 + }, + { + "epoch": 0.29407, + "grad_norm": 0.7431668639183044, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 29407 + }, + { + "epoch": 0.29408, + "grad_norm": 0.644280731678009, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 29408 + }, + { + "epoch": 0.29409, + "grad_norm": 0.6458126902580261, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 29409 + }, + { + "epoch": 0.2941, + "grad_norm": 0.7772267460823059, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 29410 + }, + { + "epoch": 0.29411, + "grad_norm": 0.8633269667625427, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 29411 + }, + { + "epoch": 0.29412, + "grad_norm": 0.9064978957176208, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 29412 + }, + { + "epoch": 0.29413, + "grad_norm": 0.9872669577598572, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 29413 + }, + { + "epoch": 0.29414, + "grad_norm": 1.1904133558273315, + "learning_rate": 0.003, + "loss": 3.981, + "step": 29414 + }, + { + "epoch": 0.29415, + "grad_norm": 0.8279935121536255, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 29415 + }, + { + "epoch": 0.29416, + "grad_norm": 0.7774080634117126, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 29416 + }, + { + "epoch": 0.29417, + "grad_norm": 0.7333403825759888, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 29417 + }, + { + "epoch": 0.29418, + "grad_norm": 0.7363682985305786, + "learning_rate": 0.003, + "loss": 3.984, + "step": 29418 + }, + { + "epoch": 0.29419, + "grad_norm": 0.8407112956047058, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 29419 + }, + { + "epoch": 0.2942, + "grad_norm": 1.0754133462905884, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 29420 + }, + { + "epoch": 0.29421, + "grad_norm": 0.9951717853546143, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 29421 + }, + { + "epoch": 0.29422, + "grad_norm": 1.0224894285202026, + "learning_rate": 0.003, + "loss": 3.978, + "step": 29422 + }, + { + "epoch": 0.29423, + "grad_norm": 0.7910917401313782, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 29423 + }, + { + "epoch": 0.29424, + "grad_norm": 0.6298081278800964, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 29424 + }, + { + "epoch": 0.29425, + "grad_norm": 0.6252541542053223, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 29425 + }, + { + "epoch": 0.29426, + "grad_norm": 0.6886066794395447, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 29426 + }, + { + "epoch": 0.29427, + "grad_norm": 0.9520053267478943, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 29427 + }, + { + "epoch": 0.29428, + "grad_norm": 1.1914699077606201, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 29428 + }, + { + "epoch": 0.29429, + "grad_norm": 0.8406829833984375, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 29429 + }, + { + "epoch": 0.2943, + "grad_norm": 0.7798550724983215, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 29430 + }, + { + "epoch": 0.29431, + "grad_norm": 0.7320646643638611, + "learning_rate": 0.003, + "loss": 3.977, + "step": 29431 + }, + { + "epoch": 0.29432, + "grad_norm": 0.8317462205886841, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 29432 + }, + { + "epoch": 0.29433, + "grad_norm": 0.9207707047462463, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 29433 + }, + { + "epoch": 0.29434, + "grad_norm": 0.9213225245475769, + "learning_rate": 0.003, + "loss": 3.994, + "step": 29434 + }, + { + "epoch": 0.29435, + "grad_norm": 0.9248151183128357, + "learning_rate": 0.003, + "loss": 3.97, + "step": 29435 + }, + { + "epoch": 0.29436, + "grad_norm": 0.8685757517814636, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 29436 + }, + { + "epoch": 0.29437, + "grad_norm": 0.8004436492919922, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 29437 + }, + { + "epoch": 0.29438, + "grad_norm": 0.7417466044425964, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 29438 + }, + { + "epoch": 0.29439, + "grad_norm": 0.7571178078651428, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 29439 + }, + { + "epoch": 0.2944, + "grad_norm": 0.7205926179885864, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 29440 + }, + { + "epoch": 0.29441, + "grad_norm": 0.7405290007591248, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 29441 + }, + { + "epoch": 0.29442, + "grad_norm": 0.8475127816200256, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 29442 + }, + { + "epoch": 0.29443, + "grad_norm": 0.7538436651229858, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 29443 + }, + { + "epoch": 0.29444, + "grad_norm": 0.7439383268356323, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 29444 + }, + { + "epoch": 0.29445, + "grad_norm": 0.9432771801948547, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 29445 + }, + { + "epoch": 0.29446, + "grad_norm": 1.0450979471206665, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 29446 + }, + { + "epoch": 0.29447, + "grad_norm": 1.1705185174942017, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 29447 + }, + { + "epoch": 0.29448, + "grad_norm": 0.9743937253952026, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 29448 + }, + { + "epoch": 0.29449, + "grad_norm": 1.0387613773345947, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 29449 + }, + { + "epoch": 0.2945, + "grad_norm": 0.9424250721931458, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 29450 + }, + { + "epoch": 0.29451, + "grad_norm": 0.9737036824226379, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 29451 + }, + { + "epoch": 0.29452, + "grad_norm": 1.0393654108047485, + "learning_rate": 0.003, + "loss": 3.986, + "step": 29452 + }, + { + "epoch": 0.29453, + "grad_norm": 1.0304518938064575, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 29453 + }, + { + "epoch": 0.29454, + "grad_norm": 1.0731675624847412, + "learning_rate": 0.003, + "loss": 3.986, + "step": 29454 + }, + { + "epoch": 0.29455, + "grad_norm": 1.2530088424682617, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 29455 + }, + { + "epoch": 0.29456, + "grad_norm": 0.8360188603401184, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 29456 + }, + { + "epoch": 0.29457, + "grad_norm": 0.8797240853309631, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 29457 + }, + { + "epoch": 0.29458, + "grad_norm": 1.1756221055984497, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 29458 + }, + { + "epoch": 0.29459, + "grad_norm": 1.0922672748565674, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 29459 + }, + { + "epoch": 0.2946, + "grad_norm": 0.6397703289985657, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 29460 + }, + { + "epoch": 0.29461, + "grad_norm": 0.6645256876945496, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 29461 + }, + { + "epoch": 0.29462, + "grad_norm": 0.8470885157585144, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 29462 + }, + { + "epoch": 0.29463, + "grad_norm": 1.0032199621200562, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 29463 + }, + { + "epoch": 0.29464, + "grad_norm": 1.151637077331543, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 29464 + }, + { + "epoch": 0.29465, + "grad_norm": 0.6663187146186829, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 29465 + }, + { + "epoch": 0.29466, + "grad_norm": 0.6387907266616821, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 29466 + }, + { + "epoch": 0.29467, + "grad_norm": 0.6934433579444885, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 29467 + }, + { + "epoch": 0.29468, + "grad_norm": 0.6397847533226013, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 29468 + }, + { + "epoch": 0.29469, + "grad_norm": 0.7027409076690674, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 29469 + }, + { + "epoch": 0.2947, + "grad_norm": 0.8423077464103699, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 29470 + }, + { + "epoch": 0.29471, + "grad_norm": 1.0069019794464111, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 29471 + }, + { + "epoch": 0.29472, + "grad_norm": 0.9783880710601807, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 29472 + }, + { + "epoch": 0.29473, + "grad_norm": 0.734711766242981, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 29473 + }, + { + "epoch": 0.29474, + "grad_norm": 0.6755634546279907, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 29474 + }, + { + "epoch": 0.29475, + "grad_norm": 0.8146964311599731, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 29475 + }, + { + "epoch": 0.29476, + "grad_norm": 0.9028863906860352, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 29476 + }, + { + "epoch": 0.29477, + "grad_norm": 0.7522125244140625, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 29477 + }, + { + "epoch": 0.29478, + "grad_norm": 0.6880446672439575, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 29478 + }, + { + "epoch": 0.29479, + "grad_norm": 0.7696476578712463, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 29479 + }, + { + "epoch": 0.2948, + "grad_norm": 0.7596817016601562, + "learning_rate": 0.003, + "loss": 3.961, + "step": 29480 + }, + { + "epoch": 0.29481, + "grad_norm": 0.9789131879806519, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 29481 + }, + { + "epoch": 0.29482, + "grad_norm": 1.4145268201828003, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 29482 + }, + { + "epoch": 0.29483, + "grad_norm": 0.7816336750984192, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 29483 + }, + { + "epoch": 0.29484, + "grad_norm": 0.7807256579399109, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 29484 + }, + { + "epoch": 0.29485, + "grad_norm": 0.8141565322875977, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 29485 + }, + { + "epoch": 0.29486, + "grad_norm": 0.8169483542442322, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 29486 + }, + { + "epoch": 0.29487, + "grad_norm": 0.730822741985321, + "learning_rate": 0.003, + "loss": 3.98, + "step": 29487 + }, + { + "epoch": 0.29488, + "grad_norm": 0.7629803419113159, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 29488 + }, + { + "epoch": 0.29489, + "grad_norm": 0.7626176476478577, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 29489 + }, + { + "epoch": 0.2949, + "grad_norm": 0.7405943274497986, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 29490 + }, + { + "epoch": 0.29491, + "grad_norm": 0.8009028434753418, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 29491 + }, + { + "epoch": 0.29492, + "grad_norm": 1.0272531509399414, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 29492 + }, + { + "epoch": 0.29493, + "grad_norm": 1.0617762804031372, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 29493 + }, + { + "epoch": 0.29494, + "grad_norm": 0.9065073728561401, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 29494 + }, + { + "epoch": 0.29495, + "grad_norm": 0.9408895969390869, + "learning_rate": 0.003, + "loss": 3.966, + "step": 29495 + }, + { + "epoch": 0.29496, + "grad_norm": 0.9805231690406799, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 29496 + }, + { + "epoch": 0.29497, + "grad_norm": 0.9432904124259949, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 29497 + }, + { + "epoch": 0.29498, + "grad_norm": 0.8331028819084167, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 29498 + }, + { + "epoch": 0.29499, + "grad_norm": 0.7905882000923157, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 29499 + }, + { + "epoch": 0.295, + "grad_norm": 0.9288694262504578, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 29500 + }, + { + "epoch": 0.29501, + "grad_norm": 1.1118485927581787, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 29501 + }, + { + "epoch": 0.29502, + "grad_norm": 0.8594845533370972, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 29502 + }, + { + "epoch": 0.29503, + "grad_norm": 0.7038068771362305, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 29503 + }, + { + "epoch": 0.29504, + "grad_norm": 0.748149573802948, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 29504 + }, + { + "epoch": 0.29505, + "grad_norm": 0.6862558126449585, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 29505 + }, + { + "epoch": 0.29506, + "grad_norm": 0.7263050079345703, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 29506 + }, + { + "epoch": 0.29507, + "grad_norm": 0.7836843132972717, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 29507 + }, + { + "epoch": 0.29508, + "grad_norm": 1.0202431678771973, + "learning_rate": 0.003, + "loss": 3.973, + "step": 29508 + }, + { + "epoch": 0.29509, + "grad_norm": 1.2579073905944824, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 29509 + }, + { + "epoch": 0.2951, + "grad_norm": 0.72624272108078, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 29510 + }, + { + "epoch": 0.29511, + "grad_norm": 0.6857823133468628, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 29511 + }, + { + "epoch": 0.29512, + "grad_norm": 0.7770724892616272, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 29512 + }, + { + "epoch": 0.29513, + "grad_norm": 0.8341947197914124, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 29513 + }, + { + "epoch": 0.29514, + "grad_norm": 0.8868319988250732, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 29514 + }, + { + "epoch": 0.29515, + "grad_norm": 0.8945943713188171, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 29515 + }, + { + "epoch": 0.29516, + "grad_norm": 0.939319372177124, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 29516 + }, + { + "epoch": 0.29517, + "grad_norm": 1.1063039302825928, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 29517 + }, + { + "epoch": 0.29518, + "grad_norm": 0.9834089875221252, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 29518 + }, + { + "epoch": 0.29519, + "grad_norm": 1.207808494567871, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 29519 + }, + { + "epoch": 0.2952, + "grad_norm": 0.9563003778457642, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 29520 + }, + { + "epoch": 0.29521, + "grad_norm": 0.7663787603378296, + "learning_rate": 0.003, + "loss": 3.967, + "step": 29521 + }, + { + "epoch": 0.29522, + "grad_norm": 0.7900487780570984, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 29522 + }, + { + "epoch": 0.29523, + "grad_norm": 0.8739162087440491, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 29523 + }, + { + "epoch": 0.29524, + "grad_norm": 0.9294065833091736, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 29524 + }, + { + "epoch": 0.29525, + "grad_norm": 0.8715143799781799, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 29525 + }, + { + "epoch": 0.29526, + "grad_norm": 0.8158550262451172, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 29526 + }, + { + "epoch": 0.29527, + "grad_norm": 0.8496813774108887, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 29527 + }, + { + "epoch": 0.29528, + "grad_norm": 0.836165726184845, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 29528 + }, + { + "epoch": 0.29529, + "grad_norm": 0.7689876556396484, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 29529 + }, + { + "epoch": 0.2953, + "grad_norm": 0.8182174563407898, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 29530 + }, + { + "epoch": 0.29531, + "grad_norm": 0.8892466425895691, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 29531 + }, + { + "epoch": 0.29532, + "grad_norm": 0.8972439169883728, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 29532 + }, + { + "epoch": 0.29533, + "grad_norm": 0.8818033933639526, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 29533 + }, + { + "epoch": 0.29534, + "grad_norm": 0.7209142446517944, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 29534 + }, + { + "epoch": 0.29535, + "grad_norm": 0.644940197467804, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 29535 + }, + { + "epoch": 0.29536, + "grad_norm": 0.5728176236152649, + "learning_rate": 0.003, + "loss": 3.9352, + "step": 29536 + }, + { + "epoch": 0.29537, + "grad_norm": 0.6152737140655518, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 29537 + }, + { + "epoch": 0.29538, + "grad_norm": 0.5709540247917175, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 29538 + }, + { + "epoch": 0.29539, + "grad_norm": 0.6175227165222168, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 29539 + }, + { + "epoch": 0.2954, + "grad_norm": 0.7302018404006958, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 29540 + }, + { + "epoch": 0.29541, + "grad_norm": 0.9423792362213135, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 29541 + }, + { + "epoch": 0.29542, + "grad_norm": 1.224637508392334, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 29542 + }, + { + "epoch": 0.29543, + "grad_norm": 0.8775757551193237, + "learning_rate": 0.003, + "loss": 3.959, + "step": 29543 + }, + { + "epoch": 0.29544, + "grad_norm": 1.0761919021606445, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 29544 + }, + { + "epoch": 0.29545, + "grad_norm": 1.4207384586334229, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 29545 + }, + { + "epoch": 0.29546, + "grad_norm": 0.6588701605796814, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 29546 + }, + { + "epoch": 0.29547, + "grad_norm": 0.8516614437103271, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 29547 + }, + { + "epoch": 0.29548, + "grad_norm": 1.2689292430877686, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 29548 + }, + { + "epoch": 0.29549, + "grad_norm": 0.4965902864933014, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 29549 + }, + { + "epoch": 0.2955, + "grad_norm": 0.7374591827392578, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 29550 + }, + { + "epoch": 0.29551, + "grad_norm": 0.9045299291610718, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 29551 + }, + { + "epoch": 0.29552, + "grad_norm": 0.9769382476806641, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 29552 + }, + { + "epoch": 0.29553, + "grad_norm": 0.9225655794143677, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 29553 + }, + { + "epoch": 0.29554, + "grad_norm": 0.8218221068382263, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 29554 + }, + { + "epoch": 0.29555, + "grad_norm": 0.7080115675926208, + "learning_rate": 0.003, + "loss": 3.9423, + "step": 29555 + }, + { + "epoch": 0.29556, + "grad_norm": 0.7873331904411316, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 29556 + }, + { + "epoch": 0.29557, + "grad_norm": 0.8303126692771912, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 29557 + }, + { + "epoch": 0.29558, + "grad_norm": 0.8772971630096436, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 29558 + }, + { + "epoch": 0.29559, + "grad_norm": 0.8748769164085388, + "learning_rate": 0.003, + "loss": 3.9457, + "step": 29559 + }, + { + "epoch": 0.2956, + "grad_norm": 0.8554930090904236, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 29560 + }, + { + "epoch": 0.29561, + "grad_norm": 0.9860345125198364, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 29561 + }, + { + "epoch": 0.29562, + "grad_norm": 1.1569358110427856, + "learning_rate": 0.003, + "loss": 4.003, + "step": 29562 + }, + { + "epoch": 0.29563, + "grad_norm": 0.9217943549156189, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 29563 + }, + { + "epoch": 0.29564, + "grad_norm": 1.0569361448287964, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 29564 + }, + { + "epoch": 0.29565, + "grad_norm": 1.0972470045089722, + "learning_rate": 0.003, + "loss": 3.993, + "step": 29565 + }, + { + "epoch": 0.29566, + "grad_norm": 1.0330699682235718, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 29566 + }, + { + "epoch": 0.29567, + "grad_norm": 1.019675850868225, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 29567 + }, + { + "epoch": 0.29568, + "grad_norm": 0.8915186524391174, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 29568 + }, + { + "epoch": 0.29569, + "grad_norm": 0.8462814092636108, + "learning_rate": 0.003, + "loss": 4.015, + "step": 29569 + }, + { + "epoch": 0.2957, + "grad_norm": 0.9513604640960693, + "learning_rate": 0.003, + "loss": 3.9428, + "step": 29570 + }, + { + "epoch": 0.29571, + "grad_norm": 1.0340825319290161, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 29571 + }, + { + "epoch": 0.29572, + "grad_norm": 0.8625932931900024, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 29572 + }, + { + "epoch": 0.29573, + "grad_norm": 0.812528669834137, + "learning_rate": 0.003, + "loss": 3.991, + "step": 29573 + }, + { + "epoch": 0.29574, + "grad_norm": 0.9883827567100525, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 29574 + }, + { + "epoch": 0.29575, + "grad_norm": 1.088459849357605, + "learning_rate": 0.003, + "loss": 3.99, + "step": 29575 + }, + { + "epoch": 0.29576, + "grad_norm": 1.1589136123657227, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 29576 + }, + { + "epoch": 0.29577, + "grad_norm": 1.0364735126495361, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 29577 + }, + { + "epoch": 0.29578, + "grad_norm": 0.9383551478385925, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 29578 + }, + { + "epoch": 0.29579, + "grad_norm": 0.9229227900505066, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 29579 + }, + { + "epoch": 0.2958, + "grad_norm": 0.8289278745651245, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 29580 + }, + { + "epoch": 0.29581, + "grad_norm": 0.8282666802406311, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 29581 + }, + { + "epoch": 0.29582, + "grad_norm": 0.8662962317466736, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 29582 + }, + { + "epoch": 0.29583, + "grad_norm": 0.8545817136764526, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 29583 + }, + { + "epoch": 0.29584, + "grad_norm": 0.9037402868270874, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 29584 + }, + { + "epoch": 0.29585, + "grad_norm": 0.994565486907959, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 29585 + }, + { + "epoch": 0.29586, + "grad_norm": 1.0399020910263062, + "learning_rate": 0.003, + "loss": 3.986, + "step": 29586 + }, + { + "epoch": 0.29587, + "grad_norm": 0.8149809837341309, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 29587 + }, + { + "epoch": 0.29588, + "grad_norm": 0.7505362033843994, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 29588 + }, + { + "epoch": 0.29589, + "grad_norm": 0.6876791715621948, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 29589 + }, + { + "epoch": 0.2959, + "grad_norm": 0.6758566498756409, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 29590 + }, + { + "epoch": 0.29591, + "grad_norm": 0.7282861471176147, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 29591 + }, + { + "epoch": 0.29592, + "grad_norm": 0.7820356488227844, + "learning_rate": 0.003, + "loss": 3.991, + "step": 29592 + }, + { + "epoch": 0.29593, + "grad_norm": 0.9156736731529236, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 29593 + }, + { + "epoch": 0.29594, + "grad_norm": 0.9895374178886414, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 29594 + }, + { + "epoch": 0.29595, + "grad_norm": 1.1853673458099365, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 29595 + }, + { + "epoch": 0.29596, + "grad_norm": 1.0289767980575562, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 29596 + }, + { + "epoch": 0.29597, + "grad_norm": 0.7877872586250305, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 29597 + }, + { + "epoch": 0.29598, + "grad_norm": 0.598124623298645, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 29598 + }, + { + "epoch": 0.29599, + "grad_norm": 0.5849763751029968, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 29599 + }, + { + "epoch": 0.296, + "grad_norm": 0.5528271794319153, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 29600 + }, + { + "epoch": 0.29601, + "grad_norm": 0.6161496639251709, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 29601 + }, + { + "epoch": 0.29602, + "grad_norm": 0.5795060396194458, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 29602 + }, + { + "epoch": 0.29603, + "grad_norm": 0.5201045274734497, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 29603 + }, + { + "epoch": 0.29604, + "grad_norm": 0.5199407339096069, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 29604 + }, + { + "epoch": 0.29605, + "grad_norm": 0.5928400754928589, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 29605 + }, + { + "epoch": 0.29606, + "grad_norm": 0.6696777939796448, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 29606 + }, + { + "epoch": 0.29607, + "grad_norm": 0.9539058208465576, + "learning_rate": 0.003, + "loss": 3.966, + "step": 29607 + }, + { + "epoch": 0.29608, + "grad_norm": 1.2144259214401245, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 29608 + }, + { + "epoch": 0.29609, + "grad_norm": 0.6802286505699158, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 29609 + }, + { + "epoch": 0.2961, + "grad_norm": 0.6969552636146545, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 29610 + }, + { + "epoch": 0.29611, + "grad_norm": 0.9950405955314636, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 29611 + }, + { + "epoch": 0.29612, + "grad_norm": 1.124452829360962, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 29612 + }, + { + "epoch": 0.29613, + "grad_norm": 0.7266449332237244, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 29613 + }, + { + "epoch": 0.29614, + "grad_norm": 0.8623940944671631, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 29614 + }, + { + "epoch": 0.29615, + "grad_norm": 0.9214829206466675, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 29615 + }, + { + "epoch": 0.29616, + "grad_norm": 1.0470106601715088, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 29616 + }, + { + "epoch": 0.29617, + "grad_norm": 1.1625028848648071, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 29617 + }, + { + "epoch": 0.29618, + "grad_norm": 0.783739447593689, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 29618 + }, + { + "epoch": 0.29619, + "grad_norm": 0.8116397261619568, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 29619 + }, + { + "epoch": 0.2962, + "grad_norm": 0.9034227132797241, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 29620 + }, + { + "epoch": 0.29621, + "grad_norm": 0.9260092973709106, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 29621 + }, + { + "epoch": 0.29622, + "grad_norm": 1.1667368412017822, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 29622 + }, + { + "epoch": 0.29623, + "grad_norm": 1.1631075143814087, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 29623 + }, + { + "epoch": 0.29624, + "grad_norm": 0.8111530542373657, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 29624 + }, + { + "epoch": 0.29625, + "grad_norm": 0.8461827039718628, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 29625 + }, + { + "epoch": 0.29626, + "grad_norm": 0.9317545890808105, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 29626 + }, + { + "epoch": 0.29627, + "grad_norm": 0.8790347576141357, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 29627 + }, + { + "epoch": 0.29628, + "grad_norm": 0.8641663789749146, + "learning_rate": 0.003, + "loss": 3.985, + "step": 29628 + }, + { + "epoch": 0.29629, + "grad_norm": 0.8844912052154541, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 29629 + }, + { + "epoch": 0.2963, + "grad_norm": 1.0306646823883057, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 29630 + }, + { + "epoch": 0.29631, + "grad_norm": 1.2106250524520874, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 29631 + }, + { + "epoch": 0.29632, + "grad_norm": 0.9000135660171509, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 29632 + }, + { + "epoch": 0.29633, + "grad_norm": 0.887087345123291, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 29633 + }, + { + "epoch": 0.29634, + "grad_norm": 0.8413495421409607, + "learning_rate": 0.003, + "loss": 3.989, + "step": 29634 + }, + { + "epoch": 0.29635, + "grad_norm": 0.9314548969268799, + "learning_rate": 0.003, + "loss": 4.004, + "step": 29635 + }, + { + "epoch": 0.29636, + "grad_norm": 1.0316063165664673, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 29636 + }, + { + "epoch": 0.29637, + "grad_norm": 0.8167059421539307, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 29637 + }, + { + "epoch": 0.29638, + "grad_norm": 0.8492022156715393, + "learning_rate": 0.003, + "loss": 3.989, + "step": 29638 + }, + { + "epoch": 0.29639, + "grad_norm": 0.7673171758651733, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 29639 + }, + { + "epoch": 0.2964, + "grad_norm": 0.6688352227210999, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 29640 + }, + { + "epoch": 0.29641, + "grad_norm": 0.6392949223518372, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 29641 + }, + { + "epoch": 0.29642, + "grad_norm": 0.6368036866188049, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 29642 + }, + { + "epoch": 0.29643, + "grad_norm": 0.6977731585502625, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 29643 + }, + { + "epoch": 0.29644, + "grad_norm": 0.6870149970054626, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 29644 + }, + { + "epoch": 0.29645, + "grad_norm": 0.7170557975769043, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 29645 + }, + { + "epoch": 0.29646, + "grad_norm": 0.7729990482330322, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 29646 + }, + { + "epoch": 0.29647, + "grad_norm": 0.7862572073936462, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 29647 + }, + { + "epoch": 0.29648, + "grad_norm": 0.9703529477119446, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 29648 + }, + { + "epoch": 0.29649, + "grad_norm": 1.4968109130859375, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 29649 + }, + { + "epoch": 0.2965, + "grad_norm": 0.6462939977645874, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 29650 + }, + { + "epoch": 0.29651, + "grad_norm": 0.8198158740997314, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 29651 + }, + { + "epoch": 0.29652, + "grad_norm": 0.9528974890708923, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 29652 + }, + { + "epoch": 0.29653, + "grad_norm": 1.056094765663147, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 29653 + }, + { + "epoch": 0.29654, + "grad_norm": 1.0934232473373413, + "learning_rate": 0.003, + "loss": 3.9401, + "step": 29654 + }, + { + "epoch": 0.29655, + "grad_norm": 0.7166484594345093, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 29655 + }, + { + "epoch": 0.29656, + "grad_norm": 0.5817852020263672, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 29656 + }, + { + "epoch": 0.29657, + "grad_norm": 0.6450607180595398, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 29657 + }, + { + "epoch": 0.29658, + "grad_norm": 0.7723235487937927, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 29658 + }, + { + "epoch": 0.29659, + "grad_norm": 0.8622956275939941, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 29659 + }, + { + "epoch": 0.2966, + "grad_norm": 0.8361473083496094, + "learning_rate": 0.003, + "loss": 3.993, + "step": 29660 + }, + { + "epoch": 0.29661, + "grad_norm": 0.8041814565658569, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 29661 + }, + { + "epoch": 0.29662, + "grad_norm": 0.8842101097106934, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 29662 + }, + { + "epoch": 0.29663, + "grad_norm": 0.8350965976715088, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 29663 + }, + { + "epoch": 0.29664, + "grad_norm": 0.9171514511108398, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 29664 + }, + { + "epoch": 0.29665, + "grad_norm": 1.0585627555847168, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 29665 + }, + { + "epoch": 0.29666, + "grad_norm": 0.993595540523529, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 29666 + }, + { + "epoch": 0.29667, + "grad_norm": 0.9997632503509521, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 29667 + }, + { + "epoch": 0.29668, + "grad_norm": 0.8913691639900208, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 29668 + }, + { + "epoch": 0.29669, + "grad_norm": 0.9008095264434814, + "learning_rate": 0.003, + "loss": 4.0, + "step": 29669 + }, + { + "epoch": 0.2967, + "grad_norm": 1.0142158269882202, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 29670 + }, + { + "epoch": 0.29671, + "grad_norm": 0.9174388647079468, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 29671 + }, + { + "epoch": 0.29672, + "grad_norm": 0.8925811648368835, + "learning_rate": 0.003, + "loss": 4.021, + "step": 29672 + }, + { + "epoch": 0.29673, + "grad_norm": 0.9385372400283813, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 29673 + }, + { + "epoch": 0.29674, + "grad_norm": 0.9592816829681396, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 29674 + }, + { + "epoch": 0.29675, + "grad_norm": 0.9779744148254395, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 29675 + }, + { + "epoch": 0.29676, + "grad_norm": 1.1066632270812988, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 29676 + }, + { + "epoch": 0.29677, + "grad_norm": 1.0460600852966309, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 29677 + }, + { + "epoch": 0.29678, + "grad_norm": 0.9741427302360535, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 29678 + }, + { + "epoch": 0.29679, + "grad_norm": 1.0756850242614746, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 29679 + }, + { + "epoch": 0.2968, + "grad_norm": 0.9474878311157227, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 29680 + }, + { + "epoch": 0.29681, + "grad_norm": 0.771369218826294, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 29681 + }, + { + "epoch": 0.29682, + "grad_norm": 0.698830783367157, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 29682 + }, + { + "epoch": 0.29683, + "grad_norm": 0.7976289987564087, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 29683 + }, + { + "epoch": 0.29684, + "grad_norm": 0.7273269295692444, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 29684 + }, + { + "epoch": 0.29685, + "grad_norm": 0.7225176692008972, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 29685 + }, + { + "epoch": 0.29686, + "grad_norm": 0.8227348327636719, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 29686 + }, + { + "epoch": 0.29687, + "grad_norm": 0.820213794708252, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 29687 + }, + { + "epoch": 0.29688, + "grad_norm": 0.726661741733551, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 29688 + }, + { + "epoch": 0.29689, + "grad_norm": 0.7326152324676514, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 29689 + }, + { + "epoch": 0.2969, + "grad_norm": 0.8784520030021667, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 29690 + }, + { + "epoch": 0.29691, + "grad_norm": 0.9887343049049377, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 29691 + }, + { + "epoch": 0.29692, + "grad_norm": 1.0343443155288696, + "learning_rate": 0.003, + "loss": 3.993, + "step": 29692 + }, + { + "epoch": 0.29693, + "grad_norm": 1.1490634679794312, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 29693 + }, + { + "epoch": 0.29694, + "grad_norm": 0.8397719860076904, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 29694 + }, + { + "epoch": 0.29695, + "grad_norm": 0.6545901298522949, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 29695 + }, + { + "epoch": 0.29696, + "grad_norm": 0.7751273512840271, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 29696 + }, + { + "epoch": 0.29697, + "grad_norm": 1.0305339097976685, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 29697 + }, + { + "epoch": 0.29698, + "grad_norm": 1.1981133222579956, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 29698 + }, + { + "epoch": 0.29699, + "grad_norm": 0.534037709236145, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 29699 + }, + { + "epoch": 0.297, + "grad_norm": 0.7277973890304565, + "learning_rate": 0.003, + "loss": 3.99, + "step": 29700 + }, + { + "epoch": 0.29701, + "grad_norm": 0.9162527918815613, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 29701 + }, + { + "epoch": 0.29702, + "grad_norm": 0.9446585178375244, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 29702 + }, + { + "epoch": 0.29703, + "grad_norm": 0.9889883399009705, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 29703 + }, + { + "epoch": 0.29704, + "grad_norm": 0.9324970245361328, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 29704 + }, + { + "epoch": 0.29705, + "grad_norm": 0.9616522789001465, + "learning_rate": 0.003, + "loss": 3.968, + "step": 29705 + }, + { + "epoch": 0.29706, + "grad_norm": 1.0922437906265259, + "learning_rate": 0.003, + "loss": 3.994, + "step": 29706 + }, + { + "epoch": 0.29707, + "grad_norm": 0.8124606609344482, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 29707 + }, + { + "epoch": 0.29708, + "grad_norm": 0.7833040356636047, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 29708 + }, + { + "epoch": 0.29709, + "grad_norm": 0.8334892392158508, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 29709 + }, + { + "epoch": 0.2971, + "grad_norm": 0.7967464327812195, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 29710 + }, + { + "epoch": 0.29711, + "grad_norm": 0.8251787424087524, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 29711 + }, + { + "epoch": 0.29712, + "grad_norm": 0.8668203949928284, + "learning_rate": 0.003, + "loss": 3.972, + "step": 29712 + }, + { + "epoch": 0.29713, + "grad_norm": 0.85626620054245, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 29713 + }, + { + "epoch": 0.29714, + "grad_norm": 0.903086245059967, + "learning_rate": 0.003, + "loss": 3.995, + "step": 29714 + }, + { + "epoch": 0.29715, + "grad_norm": 1.0029584169387817, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 29715 + }, + { + "epoch": 0.29716, + "grad_norm": 0.9410133957862854, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 29716 + }, + { + "epoch": 0.29717, + "grad_norm": 0.8901590704917908, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 29717 + }, + { + "epoch": 0.29718, + "grad_norm": 0.7959598302841187, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 29718 + }, + { + "epoch": 0.29719, + "grad_norm": 0.7391949892044067, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 29719 + }, + { + "epoch": 0.2972, + "grad_norm": 0.7642425298690796, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 29720 + }, + { + "epoch": 0.29721, + "grad_norm": 0.8257391452789307, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 29721 + }, + { + "epoch": 0.29722, + "grad_norm": 0.8378844261169434, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 29722 + }, + { + "epoch": 0.29723, + "grad_norm": 0.8906729817390442, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 29723 + }, + { + "epoch": 0.29724, + "grad_norm": 1.0044156312942505, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 29724 + }, + { + "epoch": 0.29725, + "grad_norm": 1.1747190952301025, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 29725 + }, + { + "epoch": 0.29726, + "grad_norm": 0.7163479328155518, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 29726 + }, + { + "epoch": 0.29727, + "grad_norm": 0.5920770168304443, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 29727 + }, + { + "epoch": 0.29728, + "grad_norm": 0.7314079999923706, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 29728 + }, + { + "epoch": 0.29729, + "grad_norm": 0.7636217474937439, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 29729 + }, + { + "epoch": 0.2973, + "grad_norm": 0.7837602496147156, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 29730 + }, + { + "epoch": 0.29731, + "grad_norm": 0.8630800843238831, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 29731 + }, + { + "epoch": 0.29732, + "grad_norm": 0.9534946084022522, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 29732 + }, + { + "epoch": 0.29733, + "grad_norm": 0.9010741710662842, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 29733 + }, + { + "epoch": 0.29734, + "grad_norm": 0.6109136343002319, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 29734 + }, + { + "epoch": 0.29735, + "grad_norm": 0.718002200126648, + "learning_rate": 0.003, + "loss": 3.971, + "step": 29735 + }, + { + "epoch": 0.29736, + "grad_norm": 1.0014768838882446, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 29736 + }, + { + "epoch": 0.29737, + "grad_norm": 1.183933973312378, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 29737 + }, + { + "epoch": 0.29738, + "grad_norm": 0.6389482617378235, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 29738 + }, + { + "epoch": 0.29739, + "grad_norm": 0.6314660310745239, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 29739 + }, + { + "epoch": 0.2974, + "grad_norm": 0.749783992767334, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 29740 + }, + { + "epoch": 0.29741, + "grad_norm": 0.798271894454956, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 29741 + }, + { + "epoch": 0.29742, + "grad_norm": 1.0317788124084473, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 29742 + }, + { + "epoch": 0.29743, + "grad_norm": 1.1192528009414673, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 29743 + }, + { + "epoch": 0.29744, + "grad_norm": 0.7582260370254517, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 29744 + }, + { + "epoch": 0.29745, + "grad_norm": 0.6019008159637451, + "learning_rate": 0.003, + "loss": 3.97, + "step": 29745 + }, + { + "epoch": 0.29746, + "grad_norm": 0.7865267992019653, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 29746 + }, + { + "epoch": 0.29747, + "grad_norm": 0.9011349678039551, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 29747 + }, + { + "epoch": 0.29748, + "grad_norm": 0.851690411567688, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 29748 + }, + { + "epoch": 0.29749, + "grad_norm": 0.8518191576004028, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 29749 + }, + { + "epoch": 0.2975, + "grad_norm": 0.8943049311637878, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 29750 + }, + { + "epoch": 0.29751, + "grad_norm": 0.9826757907867432, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 29751 + }, + { + "epoch": 0.29752, + "grad_norm": 0.9871321320533752, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 29752 + }, + { + "epoch": 0.29753, + "grad_norm": 0.8190510272979736, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 29753 + }, + { + "epoch": 0.29754, + "grad_norm": 0.7009322047233582, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 29754 + }, + { + "epoch": 0.29755, + "grad_norm": 0.8769269585609436, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 29755 + }, + { + "epoch": 0.29756, + "grad_norm": 1.0613000392913818, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 29756 + }, + { + "epoch": 0.29757, + "grad_norm": 0.7464430332183838, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 29757 + }, + { + "epoch": 0.29758, + "grad_norm": 0.7481516599655151, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 29758 + }, + { + "epoch": 0.29759, + "grad_norm": 0.8317455053329468, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 29759 + }, + { + "epoch": 0.2976, + "grad_norm": 0.9457995295524597, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 29760 + }, + { + "epoch": 0.29761, + "grad_norm": 0.8088728189468384, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 29761 + }, + { + "epoch": 0.29762, + "grad_norm": 0.7852393984794617, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 29762 + }, + { + "epoch": 0.29763, + "grad_norm": 0.8608635663986206, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 29763 + }, + { + "epoch": 0.29764, + "grad_norm": 1.0472168922424316, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 29764 + }, + { + "epoch": 0.29765, + "grad_norm": 1.2125816345214844, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 29765 + }, + { + "epoch": 0.29766, + "grad_norm": 0.8687702417373657, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 29766 + }, + { + "epoch": 0.29767, + "grad_norm": 0.8163714408874512, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 29767 + }, + { + "epoch": 0.29768, + "grad_norm": 0.7073726654052734, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 29768 + }, + { + "epoch": 0.29769, + "grad_norm": 0.8097246289253235, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 29769 + }, + { + "epoch": 0.2977, + "grad_norm": 0.8063693046569824, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 29770 + }, + { + "epoch": 0.29771, + "grad_norm": 0.8232944011688232, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 29771 + }, + { + "epoch": 0.29772, + "grad_norm": 1.0224273204803467, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 29772 + }, + { + "epoch": 0.29773, + "grad_norm": 1.3430874347686768, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 29773 + }, + { + "epoch": 0.29774, + "grad_norm": 1.1287938356399536, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 29774 + }, + { + "epoch": 0.29775, + "grad_norm": 0.9546907544136047, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 29775 + }, + { + "epoch": 0.29776, + "grad_norm": 0.9649332165718079, + "learning_rate": 0.003, + "loss": 3.981, + "step": 29776 + }, + { + "epoch": 0.29777, + "grad_norm": 1.1058704853057861, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 29777 + }, + { + "epoch": 0.29778, + "grad_norm": 1.0941568613052368, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 29778 + }, + { + "epoch": 0.29779, + "grad_norm": 0.7883844375610352, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 29779 + }, + { + "epoch": 0.2978, + "grad_norm": 0.7056304216384888, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 29780 + }, + { + "epoch": 0.29781, + "grad_norm": 0.8604498505592346, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 29781 + }, + { + "epoch": 0.29782, + "grad_norm": 0.8427525758743286, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 29782 + }, + { + "epoch": 0.29783, + "grad_norm": 0.8143576383590698, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 29783 + }, + { + "epoch": 0.29784, + "grad_norm": 0.9198464751243591, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 29784 + }, + { + "epoch": 0.29785, + "grad_norm": 0.8614175319671631, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 29785 + }, + { + "epoch": 0.29786, + "grad_norm": 0.8174096345901489, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 29786 + }, + { + "epoch": 0.29787, + "grad_norm": 0.8219383955001831, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 29787 + }, + { + "epoch": 0.29788, + "grad_norm": 0.9726031422615051, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 29788 + }, + { + "epoch": 0.29789, + "grad_norm": 1.1935296058654785, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 29789 + }, + { + "epoch": 0.2979, + "grad_norm": 0.8465375900268555, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 29790 + }, + { + "epoch": 0.29791, + "grad_norm": 0.729188859462738, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 29791 + }, + { + "epoch": 0.29792, + "grad_norm": 0.6375157833099365, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 29792 + }, + { + "epoch": 0.29793, + "grad_norm": 0.600857675075531, + "learning_rate": 0.003, + "loss": 3.979, + "step": 29793 + }, + { + "epoch": 0.29794, + "grad_norm": 0.6652498245239258, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 29794 + }, + { + "epoch": 0.29795, + "grad_norm": 0.848651111125946, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 29795 + }, + { + "epoch": 0.29796, + "grad_norm": 1.102792739868164, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 29796 + }, + { + "epoch": 0.29797, + "grad_norm": 1.0328304767608643, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 29797 + }, + { + "epoch": 0.29798, + "grad_norm": 0.7900938391685486, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 29798 + }, + { + "epoch": 0.29799, + "grad_norm": 0.5503284931182861, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 29799 + }, + { + "epoch": 0.298, + "grad_norm": 0.5818873047828674, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 29800 + }, + { + "epoch": 0.29801, + "grad_norm": 0.6118683815002441, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 29801 + }, + { + "epoch": 0.29802, + "grad_norm": 0.6348412036895752, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 29802 + }, + { + "epoch": 0.29803, + "grad_norm": 0.6458591222763062, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 29803 + }, + { + "epoch": 0.29804, + "grad_norm": 0.6659438610076904, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 29804 + }, + { + "epoch": 0.29805, + "grad_norm": 0.6947615146636963, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 29805 + }, + { + "epoch": 0.29806, + "grad_norm": 0.7901530861854553, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 29806 + }, + { + "epoch": 0.29807, + "grad_norm": 0.8252387046813965, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 29807 + }, + { + "epoch": 0.29808, + "grad_norm": 0.8768091201782227, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 29808 + }, + { + "epoch": 0.29809, + "grad_norm": 0.9805389046669006, + "learning_rate": 0.003, + "loss": 3.973, + "step": 29809 + }, + { + "epoch": 0.2981, + "grad_norm": 1.307969093322754, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 29810 + }, + { + "epoch": 0.29811, + "grad_norm": 0.8339400887489319, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 29811 + }, + { + "epoch": 0.29812, + "grad_norm": 0.7800565361976624, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 29812 + }, + { + "epoch": 0.29813, + "grad_norm": 0.9247667193412781, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 29813 + }, + { + "epoch": 0.29814, + "grad_norm": 1.1148399114608765, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 29814 + }, + { + "epoch": 0.29815, + "grad_norm": 0.903077244758606, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 29815 + }, + { + "epoch": 0.29816, + "grad_norm": 0.8003380298614502, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 29816 + }, + { + "epoch": 0.29817, + "grad_norm": 0.826287567615509, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 29817 + }, + { + "epoch": 0.29818, + "grad_norm": 1.0630240440368652, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 29818 + }, + { + "epoch": 0.29819, + "grad_norm": 1.1658202409744263, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 29819 + }, + { + "epoch": 0.2982, + "grad_norm": 0.8464640378952026, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 29820 + }, + { + "epoch": 0.29821, + "grad_norm": 0.8198652267456055, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 29821 + }, + { + "epoch": 0.29822, + "grad_norm": 0.8647462129592896, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 29822 + }, + { + "epoch": 0.29823, + "grad_norm": 0.8520760536193848, + "learning_rate": 0.003, + "loss": 3.975, + "step": 29823 + }, + { + "epoch": 0.29824, + "grad_norm": 0.7547090649604797, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 29824 + }, + { + "epoch": 0.29825, + "grad_norm": 0.5965854525566101, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 29825 + }, + { + "epoch": 0.29826, + "grad_norm": 0.558817982673645, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 29826 + }, + { + "epoch": 0.29827, + "grad_norm": 0.6407290101051331, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 29827 + }, + { + "epoch": 0.29828, + "grad_norm": 0.6450330018997192, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 29828 + }, + { + "epoch": 0.29829, + "grad_norm": 0.6990267038345337, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 29829 + }, + { + "epoch": 0.2983, + "grad_norm": 0.7831560373306274, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 29830 + }, + { + "epoch": 0.29831, + "grad_norm": 0.9068610072135925, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 29831 + }, + { + "epoch": 0.29832, + "grad_norm": 0.997868537902832, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 29832 + }, + { + "epoch": 0.29833, + "grad_norm": 1.2821260690689087, + "learning_rate": 0.003, + "loss": 3.985, + "step": 29833 + }, + { + "epoch": 0.29834, + "grad_norm": 0.8058915734291077, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 29834 + }, + { + "epoch": 0.29835, + "grad_norm": 0.6452077627182007, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 29835 + }, + { + "epoch": 0.29836, + "grad_norm": 0.6677200198173523, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 29836 + }, + { + "epoch": 0.29837, + "grad_norm": 0.7437049746513367, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 29837 + }, + { + "epoch": 0.29838, + "grad_norm": 0.9222956299781799, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 29838 + }, + { + "epoch": 0.29839, + "grad_norm": 1.1394660472869873, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 29839 + }, + { + "epoch": 0.2984, + "grad_norm": 0.962175190448761, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 29840 + }, + { + "epoch": 0.29841, + "grad_norm": 1.141413927078247, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 29841 + }, + { + "epoch": 0.29842, + "grad_norm": 0.9822302460670471, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 29842 + }, + { + "epoch": 0.29843, + "grad_norm": 0.939382791519165, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 29843 + }, + { + "epoch": 0.29844, + "grad_norm": 0.902665376663208, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 29844 + }, + { + "epoch": 0.29845, + "grad_norm": 0.8079807758331299, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 29845 + }, + { + "epoch": 0.29846, + "grad_norm": 0.8347387313842773, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 29846 + }, + { + "epoch": 0.29847, + "grad_norm": 0.9297822117805481, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 29847 + }, + { + "epoch": 0.29848, + "grad_norm": 0.7676069736480713, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 29848 + }, + { + "epoch": 0.29849, + "grad_norm": 0.780820906162262, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 29849 + }, + { + "epoch": 0.2985, + "grad_norm": 0.9186661243438721, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 29850 + }, + { + "epoch": 0.29851, + "grad_norm": 0.8660373091697693, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 29851 + }, + { + "epoch": 0.29852, + "grad_norm": 0.7583914995193481, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 29852 + }, + { + "epoch": 0.29853, + "grad_norm": 0.7049664855003357, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 29853 + }, + { + "epoch": 0.29854, + "grad_norm": 0.7680448293685913, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 29854 + }, + { + "epoch": 0.29855, + "grad_norm": 0.8294632434844971, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 29855 + }, + { + "epoch": 0.29856, + "grad_norm": 1.1094530820846558, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 29856 + }, + { + "epoch": 0.29857, + "grad_norm": 1.2309154272079468, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 29857 + }, + { + "epoch": 0.29858, + "grad_norm": 0.8412700891494751, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 29858 + }, + { + "epoch": 0.29859, + "grad_norm": 0.9265427589416504, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 29859 + }, + { + "epoch": 0.2986, + "grad_norm": 1.1622416973114014, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 29860 + }, + { + "epoch": 0.29861, + "grad_norm": 0.8345713019371033, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 29861 + }, + { + "epoch": 0.29862, + "grad_norm": 0.7225048542022705, + "learning_rate": 0.003, + "loss": 4.002, + "step": 29862 + }, + { + "epoch": 0.29863, + "grad_norm": 0.6259239315986633, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 29863 + }, + { + "epoch": 0.29864, + "grad_norm": 0.6951460242271423, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 29864 + }, + { + "epoch": 0.29865, + "grad_norm": 0.812473714351654, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 29865 + }, + { + "epoch": 0.29866, + "grad_norm": 1.0680967569351196, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 29866 + }, + { + "epoch": 0.29867, + "grad_norm": 1.1497949361801147, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 29867 + }, + { + "epoch": 0.29868, + "grad_norm": 0.7571655511856079, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 29868 + }, + { + "epoch": 0.29869, + "grad_norm": 0.6814003586769104, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 29869 + }, + { + "epoch": 0.2987, + "grad_norm": 0.6778420805931091, + "learning_rate": 0.003, + "loss": 3.98, + "step": 29870 + }, + { + "epoch": 0.29871, + "grad_norm": 0.8610602617263794, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 29871 + }, + { + "epoch": 0.29872, + "grad_norm": 0.740324854850769, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 29872 + }, + { + "epoch": 0.29873, + "grad_norm": 0.7461093664169312, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 29873 + }, + { + "epoch": 0.29874, + "grad_norm": 0.9316704273223877, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 29874 + }, + { + "epoch": 0.29875, + "grad_norm": 1.425276756286621, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 29875 + }, + { + "epoch": 0.29876, + "grad_norm": 0.5666950345039368, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 29876 + }, + { + "epoch": 0.29877, + "grad_norm": 0.7394827604293823, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 29877 + }, + { + "epoch": 0.29878, + "grad_norm": 0.8542216420173645, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 29878 + }, + { + "epoch": 0.29879, + "grad_norm": 0.9952075481414795, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 29879 + }, + { + "epoch": 0.2988, + "grad_norm": 0.9750637412071228, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 29880 + }, + { + "epoch": 0.29881, + "grad_norm": 0.7808839678764343, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 29881 + }, + { + "epoch": 0.29882, + "grad_norm": 0.7214662432670593, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 29882 + }, + { + "epoch": 0.29883, + "grad_norm": 0.8061822056770325, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 29883 + }, + { + "epoch": 0.29884, + "grad_norm": 0.8918421864509583, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 29884 + }, + { + "epoch": 0.29885, + "grad_norm": 0.7537373900413513, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 29885 + }, + { + "epoch": 0.29886, + "grad_norm": 0.8265218734741211, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 29886 + }, + { + "epoch": 0.29887, + "grad_norm": 0.8843728303909302, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 29887 + }, + { + "epoch": 0.29888, + "grad_norm": 1.1339071989059448, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 29888 + }, + { + "epoch": 0.29889, + "grad_norm": 0.9074516296386719, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 29889 + }, + { + "epoch": 0.2989, + "grad_norm": 0.8799979090690613, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 29890 + }, + { + "epoch": 0.29891, + "grad_norm": 0.8443678021430969, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 29891 + }, + { + "epoch": 0.29892, + "grad_norm": 0.9082952737808228, + "learning_rate": 0.003, + "loss": 3.989, + "step": 29892 + }, + { + "epoch": 0.29893, + "grad_norm": 0.9738061428070068, + "learning_rate": 0.003, + "loss": 3.997, + "step": 29893 + }, + { + "epoch": 0.29894, + "grad_norm": 1.0368942022323608, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 29894 + }, + { + "epoch": 0.29895, + "grad_norm": 1.2658419609069824, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 29895 + }, + { + "epoch": 0.29896, + "grad_norm": 0.936763346195221, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 29896 + }, + { + "epoch": 0.29897, + "grad_norm": 0.8710474967956543, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 29897 + }, + { + "epoch": 0.29898, + "grad_norm": 0.8997664451599121, + "learning_rate": 0.003, + "loss": 4.001, + "step": 29898 + }, + { + "epoch": 0.29899, + "grad_norm": 0.9231278300285339, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 29899 + }, + { + "epoch": 0.299, + "grad_norm": 1.0372713804244995, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 29900 + }, + { + "epoch": 0.29901, + "grad_norm": 0.8776853680610657, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 29901 + }, + { + "epoch": 0.29902, + "grad_norm": 0.8602263927459717, + "learning_rate": 0.003, + "loss": 3.966, + "step": 29902 + }, + { + "epoch": 0.29903, + "grad_norm": 0.9229939579963684, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 29903 + }, + { + "epoch": 0.29904, + "grad_norm": 0.897021472454071, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 29904 + }, + { + "epoch": 0.29905, + "grad_norm": 0.9034028053283691, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 29905 + }, + { + "epoch": 0.29906, + "grad_norm": 0.7475103735923767, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 29906 + }, + { + "epoch": 0.29907, + "grad_norm": 0.6711054444313049, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 29907 + }, + { + "epoch": 0.29908, + "grad_norm": 0.5694817900657654, + "learning_rate": 0.003, + "loss": 3.976, + "step": 29908 + }, + { + "epoch": 0.29909, + "grad_norm": 0.6087790727615356, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 29909 + }, + { + "epoch": 0.2991, + "grad_norm": 0.5950635671615601, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 29910 + }, + { + "epoch": 0.29911, + "grad_norm": 0.7800112366676331, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 29911 + }, + { + "epoch": 0.29912, + "grad_norm": 1.2342994213104248, + "learning_rate": 0.003, + "loss": 4.009, + "step": 29912 + }, + { + "epoch": 0.29913, + "grad_norm": 1.15948486328125, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 29913 + }, + { + "epoch": 0.29914, + "grad_norm": 0.7838550209999084, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 29914 + }, + { + "epoch": 0.29915, + "grad_norm": 0.7574610710144043, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 29915 + }, + { + "epoch": 0.29916, + "grad_norm": 0.8705871105194092, + "learning_rate": 0.003, + "loss": 3.9303, + "step": 29916 + }, + { + "epoch": 0.29917, + "grad_norm": 0.8776805400848389, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 29917 + }, + { + "epoch": 0.29918, + "grad_norm": 0.8041352033615112, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 29918 + }, + { + "epoch": 0.29919, + "grad_norm": 0.9741889238357544, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 29919 + }, + { + "epoch": 0.2992, + "grad_norm": 1.031924843788147, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 29920 + }, + { + "epoch": 0.29921, + "grad_norm": 0.9230374097824097, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 29921 + }, + { + "epoch": 0.29922, + "grad_norm": 0.878333568572998, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 29922 + }, + { + "epoch": 0.29923, + "grad_norm": 1.0893750190734863, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 29923 + }, + { + "epoch": 0.29924, + "grad_norm": 1.3460417985916138, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 29924 + }, + { + "epoch": 0.29925, + "grad_norm": 0.7866143584251404, + "learning_rate": 0.003, + "loss": 4.002, + "step": 29925 + }, + { + "epoch": 0.29926, + "grad_norm": 0.7408838868141174, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 29926 + }, + { + "epoch": 0.29927, + "grad_norm": 0.6309239864349365, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 29927 + }, + { + "epoch": 0.29928, + "grad_norm": 0.7635916471481323, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 29928 + }, + { + "epoch": 0.29929, + "grad_norm": 0.8663201928138733, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 29929 + }, + { + "epoch": 0.2993, + "grad_norm": 0.990358829498291, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 29930 + }, + { + "epoch": 0.29931, + "grad_norm": 1.0431926250457764, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 29931 + }, + { + "epoch": 0.29932, + "grad_norm": 0.9537420272827148, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 29932 + }, + { + "epoch": 0.29933, + "grad_norm": 0.7446601390838623, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 29933 + }, + { + "epoch": 0.29934, + "grad_norm": 0.7512003779411316, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 29934 + }, + { + "epoch": 0.29935, + "grad_norm": 0.845818281173706, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 29935 + }, + { + "epoch": 0.29936, + "grad_norm": 0.8143840432167053, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 29936 + }, + { + "epoch": 0.29937, + "grad_norm": 0.942068874835968, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 29937 + }, + { + "epoch": 0.29938, + "grad_norm": 1.2116124629974365, + "learning_rate": 0.003, + "loss": 3.959, + "step": 29938 + }, + { + "epoch": 0.29939, + "grad_norm": 0.7923477292060852, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 29939 + }, + { + "epoch": 0.2994, + "grad_norm": 0.7845305800437927, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 29940 + }, + { + "epoch": 0.29941, + "grad_norm": 0.7702876329421997, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 29941 + }, + { + "epoch": 0.29942, + "grad_norm": 0.7150661945343018, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 29942 + }, + { + "epoch": 0.29943, + "grad_norm": 0.7746699452400208, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 29943 + }, + { + "epoch": 0.29944, + "grad_norm": 0.7820342183113098, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 29944 + }, + { + "epoch": 0.29945, + "grad_norm": 0.7896479964256287, + "learning_rate": 0.003, + "loss": 3.993, + "step": 29945 + }, + { + "epoch": 0.29946, + "grad_norm": 1.0003989934921265, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 29946 + }, + { + "epoch": 0.29947, + "grad_norm": 1.3566621541976929, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 29947 + }, + { + "epoch": 0.29948, + "grad_norm": 0.7074670195579529, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 29948 + }, + { + "epoch": 0.29949, + "grad_norm": 0.6568766832351685, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 29949 + }, + { + "epoch": 0.2995, + "grad_norm": 0.7043598890304565, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 29950 + }, + { + "epoch": 0.29951, + "grad_norm": 0.7433149814605713, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 29951 + }, + { + "epoch": 0.29952, + "grad_norm": 0.7945345044136047, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 29952 + }, + { + "epoch": 0.29953, + "grad_norm": 0.7517597675323486, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 29953 + }, + { + "epoch": 0.29954, + "grad_norm": 0.9090352058410645, + "learning_rate": 0.003, + "loss": 4.009, + "step": 29954 + }, + { + "epoch": 0.29955, + "grad_norm": 1.0515285730361938, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 29955 + }, + { + "epoch": 0.29956, + "grad_norm": 0.9571055769920349, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 29956 + }, + { + "epoch": 0.29957, + "grad_norm": 0.9939196705818176, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 29957 + }, + { + "epoch": 0.29958, + "grad_norm": 0.9134706854820251, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 29958 + }, + { + "epoch": 0.29959, + "grad_norm": 0.8269138336181641, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 29959 + }, + { + "epoch": 0.2996, + "grad_norm": 0.7732140421867371, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 29960 + }, + { + "epoch": 0.29961, + "grad_norm": 0.8421785831451416, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 29961 + }, + { + "epoch": 0.29962, + "grad_norm": 1.0248817205429077, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 29962 + }, + { + "epoch": 0.29963, + "grad_norm": 1.1286312341690063, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 29963 + }, + { + "epoch": 0.29964, + "grad_norm": 1.1287399530410767, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 29964 + }, + { + "epoch": 0.29965, + "grad_norm": 0.9874267578125, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 29965 + }, + { + "epoch": 0.29966, + "grad_norm": 1.064638614654541, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 29966 + }, + { + "epoch": 0.29967, + "grad_norm": 0.9722912907600403, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 29967 + }, + { + "epoch": 0.29968, + "grad_norm": 1.075437068939209, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 29968 + }, + { + "epoch": 0.29969, + "grad_norm": 0.9694464802742004, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 29969 + }, + { + "epoch": 0.2997, + "grad_norm": 0.9734860062599182, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 29970 + }, + { + "epoch": 0.29971, + "grad_norm": 0.8778698444366455, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 29971 + }, + { + "epoch": 0.29972, + "grad_norm": 0.8479580879211426, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 29972 + }, + { + "epoch": 0.29973, + "grad_norm": 0.8249054551124573, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 29973 + }, + { + "epoch": 0.29974, + "grad_norm": 0.7735233902931213, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 29974 + }, + { + "epoch": 0.29975, + "grad_norm": 0.7731571197509766, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 29975 + }, + { + "epoch": 0.29976, + "grad_norm": 0.8539149761199951, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 29976 + }, + { + "epoch": 0.29977, + "grad_norm": 0.773512065410614, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 29977 + }, + { + "epoch": 0.29978, + "grad_norm": 0.7225778698921204, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 29978 + }, + { + "epoch": 0.29979, + "grad_norm": 0.7287128567695618, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 29979 + }, + { + "epoch": 0.2998, + "grad_norm": 0.6057713031768799, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 29980 + }, + { + "epoch": 0.29981, + "grad_norm": 0.6498987078666687, + "learning_rate": 0.003, + "loss": 3.974, + "step": 29981 + }, + { + "epoch": 0.29982, + "grad_norm": 0.796296238899231, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 29982 + }, + { + "epoch": 0.29983, + "grad_norm": 0.9391615390777588, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 29983 + }, + { + "epoch": 0.29984, + "grad_norm": 1.1355677843093872, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 29984 + }, + { + "epoch": 0.29985, + "grad_norm": 0.916980504989624, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 29985 + }, + { + "epoch": 0.29986, + "grad_norm": 0.8715880513191223, + "learning_rate": 0.003, + "loss": 3.994, + "step": 29986 + }, + { + "epoch": 0.29987, + "grad_norm": 0.6248953342437744, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 29987 + }, + { + "epoch": 0.29988, + "grad_norm": 0.7417610883712769, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 29988 + }, + { + "epoch": 0.29989, + "grad_norm": 0.9351855516433716, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 29989 + }, + { + "epoch": 0.2999, + "grad_norm": 1.095752477645874, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 29990 + }, + { + "epoch": 0.29991, + "grad_norm": 0.938480794429779, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 29991 + }, + { + "epoch": 0.29992, + "grad_norm": 0.9365402460098267, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 29992 + }, + { + "epoch": 0.29993, + "grad_norm": 0.9121213555335999, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 29993 + }, + { + "epoch": 0.29994, + "grad_norm": 0.897790253162384, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 29994 + }, + { + "epoch": 0.29995, + "grad_norm": 0.9634882807731628, + "learning_rate": 0.003, + "loss": 3.962, + "step": 29995 + }, + { + "epoch": 0.29996, + "grad_norm": 1.0030559301376343, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 29996 + }, + { + "epoch": 0.29997, + "grad_norm": 0.949841320514679, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 29997 + }, + { + "epoch": 0.29998, + "grad_norm": 0.9278163909912109, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 29998 + }, + { + "epoch": 0.29999, + "grad_norm": 0.9078219532966614, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 29999 + }, + { + "epoch": 0.3, + "grad_norm": 0.8065701723098755, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 30000 + }, + { + "epoch": 0.30001, + "grad_norm": 0.8625451922416687, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 30001 + }, + { + "epoch": 0.30002, + "grad_norm": 0.7884432077407837, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 30002 + }, + { + "epoch": 0.30003, + "grad_norm": 0.7552029490470886, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 30003 + }, + { + "epoch": 0.30004, + "grad_norm": 0.690104067325592, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 30004 + }, + { + "epoch": 0.30005, + "grad_norm": 0.7086097002029419, + "learning_rate": 0.003, + "loss": 3.9456, + "step": 30005 + }, + { + "epoch": 0.30006, + "grad_norm": 0.8162664771080017, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 30006 + }, + { + "epoch": 0.30007, + "grad_norm": 1.1970733404159546, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 30007 + }, + { + "epoch": 0.30008, + "grad_norm": 1.2510138750076294, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 30008 + }, + { + "epoch": 0.30009, + "grad_norm": 0.9969817996025085, + "learning_rate": 0.003, + "loss": 3.988, + "step": 30009 + }, + { + "epoch": 0.3001, + "grad_norm": 0.9584653377532959, + "learning_rate": 0.003, + "loss": 3.998, + "step": 30010 + }, + { + "epoch": 0.30011, + "grad_norm": 0.9930746555328369, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 30011 + }, + { + "epoch": 0.30012, + "grad_norm": 1.2280558347702026, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 30012 + }, + { + "epoch": 0.30013, + "grad_norm": 0.7480072975158691, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 30013 + }, + { + "epoch": 0.30014, + "grad_norm": 0.7142095565795898, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 30014 + }, + { + "epoch": 0.30015, + "grad_norm": 0.8255698680877686, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 30015 + }, + { + "epoch": 0.30016, + "grad_norm": 0.8260887861251831, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 30016 + }, + { + "epoch": 0.30017, + "grad_norm": 0.8093573451042175, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 30017 + }, + { + "epoch": 0.30018, + "grad_norm": 0.6904893517494202, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 30018 + }, + { + "epoch": 0.30019, + "grad_norm": 0.6204531788825989, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 30019 + }, + { + "epoch": 0.3002, + "grad_norm": 0.70346599817276, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 30020 + }, + { + "epoch": 0.30021, + "grad_norm": 0.8427658081054688, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 30021 + }, + { + "epoch": 0.30022, + "grad_norm": 0.8703791499137878, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 30022 + }, + { + "epoch": 0.30023, + "grad_norm": 0.7133869528770447, + "learning_rate": 0.003, + "loss": 3.974, + "step": 30023 + }, + { + "epoch": 0.30024, + "grad_norm": 0.710033118724823, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 30024 + }, + { + "epoch": 0.30025, + "grad_norm": 0.7898790240287781, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 30025 + }, + { + "epoch": 0.30026, + "grad_norm": 0.7271448373794556, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 30026 + }, + { + "epoch": 0.30027, + "grad_norm": 0.6501927375793457, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 30027 + }, + { + "epoch": 0.30028, + "grad_norm": 0.6511788964271545, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 30028 + }, + { + "epoch": 0.30029, + "grad_norm": 0.657411515712738, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 30029 + }, + { + "epoch": 0.3003, + "grad_norm": 0.9336095452308655, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 30030 + }, + { + "epoch": 0.30031, + "grad_norm": 1.2842094898223877, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 30031 + }, + { + "epoch": 0.30032, + "grad_norm": 0.6977804899215698, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 30032 + }, + { + "epoch": 0.30033, + "grad_norm": 0.7139055132865906, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 30033 + }, + { + "epoch": 0.30034, + "grad_norm": 0.8749229907989502, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 30034 + }, + { + "epoch": 0.30035, + "grad_norm": 0.975334882736206, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 30035 + }, + { + "epoch": 0.30036, + "grad_norm": 1.0658411979675293, + "learning_rate": 0.003, + "loss": 3.986, + "step": 30036 + }, + { + "epoch": 0.30037, + "grad_norm": 1.0140398740768433, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 30037 + }, + { + "epoch": 0.30038, + "grad_norm": 0.9082064032554626, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 30038 + }, + { + "epoch": 0.30039, + "grad_norm": 0.9409763216972351, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 30039 + }, + { + "epoch": 0.3004, + "grad_norm": 0.8320639729499817, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 30040 + }, + { + "epoch": 0.30041, + "grad_norm": 0.782095193862915, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 30041 + }, + { + "epoch": 0.30042, + "grad_norm": 0.6615424752235413, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 30042 + }, + { + "epoch": 0.30043, + "grad_norm": 0.8003228902816772, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 30043 + }, + { + "epoch": 0.30044, + "grad_norm": 0.9351726770401001, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 30044 + }, + { + "epoch": 0.30045, + "grad_norm": 1.08657705783844, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 30045 + }, + { + "epoch": 0.30046, + "grad_norm": 1.1283643245697021, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 30046 + }, + { + "epoch": 0.30047, + "grad_norm": 0.7673572897911072, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 30047 + }, + { + "epoch": 0.30048, + "grad_norm": 0.6854022741317749, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 30048 + }, + { + "epoch": 0.30049, + "grad_norm": 0.7318994998931885, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 30049 + }, + { + "epoch": 0.3005, + "grad_norm": 0.6633291840553284, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 30050 + }, + { + "epoch": 0.30051, + "grad_norm": 0.7120344638824463, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 30051 + }, + { + "epoch": 0.30052, + "grad_norm": 1.0714799165725708, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 30052 + }, + { + "epoch": 0.30053, + "grad_norm": 1.1291557550430298, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 30053 + }, + { + "epoch": 0.30054, + "grad_norm": 0.9070069789886475, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 30054 + }, + { + "epoch": 0.30055, + "grad_norm": 0.7999599575996399, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 30055 + }, + { + "epoch": 0.30056, + "grad_norm": 0.8344054222106934, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 30056 + }, + { + "epoch": 0.30057, + "grad_norm": 0.7618876099586487, + "learning_rate": 0.003, + "loss": 3.974, + "step": 30057 + }, + { + "epoch": 0.30058, + "grad_norm": 0.7579072713851929, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 30058 + }, + { + "epoch": 0.30059, + "grad_norm": 0.9053711295127869, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 30059 + }, + { + "epoch": 0.3006, + "grad_norm": 1.091832160949707, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 30060 + }, + { + "epoch": 0.30061, + "grad_norm": 1.0880329608917236, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 30061 + }, + { + "epoch": 0.30062, + "grad_norm": 0.8385860323905945, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 30062 + }, + { + "epoch": 0.30063, + "grad_norm": 0.8769740462303162, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 30063 + }, + { + "epoch": 0.30064, + "grad_norm": 0.8327873349189758, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 30064 + }, + { + "epoch": 0.30065, + "grad_norm": 0.7745826244354248, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 30065 + }, + { + "epoch": 0.30066, + "grad_norm": 0.76980060338974, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 30066 + }, + { + "epoch": 0.30067, + "grad_norm": 0.7592105269432068, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 30067 + }, + { + "epoch": 0.30068, + "grad_norm": 0.7846325039863586, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 30068 + }, + { + "epoch": 0.30069, + "grad_norm": 0.9249014854431152, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 30069 + }, + { + "epoch": 0.3007, + "grad_norm": 1.132541298866272, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 30070 + }, + { + "epoch": 0.30071, + "grad_norm": 1.2604901790618896, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 30071 + }, + { + "epoch": 0.30072, + "grad_norm": 0.9784888029098511, + "learning_rate": 0.003, + "loss": 3.999, + "step": 30072 + }, + { + "epoch": 0.30073, + "grad_norm": 0.9072867631912231, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 30073 + }, + { + "epoch": 0.30074, + "grad_norm": 0.9356440901756287, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 30074 + }, + { + "epoch": 0.30075, + "grad_norm": 0.8647052645683289, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 30075 + }, + { + "epoch": 0.30076, + "grad_norm": 0.8362512588500977, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 30076 + }, + { + "epoch": 0.30077, + "grad_norm": 1.0388696193695068, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 30077 + }, + { + "epoch": 0.30078, + "grad_norm": 1.0706446170806885, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 30078 + }, + { + "epoch": 0.30079, + "grad_norm": 0.9441853165626526, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 30079 + }, + { + "epoch": 0.3008, + "grad_norm": 0.9213715195655823, + "learning_rate": 0.003, + "loss": 3.996, + "step": 30080 + }, + { + "epoch": 0.30081, + "grad_norm": 0.8341843485832214, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 30081 + }, + { + "epoch": 0.30082, + "grad_norm": 0.7605437636375427, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 30082 + }, + { + "epoch": 0.30083, + "grad_norm": 0.7582386136054993, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 30083 + }, + { + "epoch": 0.30084, + "grad_norm": 0.8573883175849915, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 30084 + }, + { + "epoch": 0.30085, + "grad_norm": 0.7673612236976624, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 30085 + }, + { + "epoch": 0.30086, + "grad_norm": 0.7227311730384827, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 30086 + }, + { + "epoch": 0.30087, + "grad_norm": 0.631710946559906, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 30087 + }, + { + "epoch": 0.30088, + "grad_norm": 0.7334915399551392, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 30088 + }, + { + "epoch": 0.30089, + "grad_norm": 0.8096153140068054, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 30089 + }, + { + "epoch": 0.3009, + "grad_norm": 0.7332887053489685, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 30090 + }, + { + "epoch": 0.30091, + "grad_norm": 0.7058311700820923, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 30091 + }, + { + "epoch": 0.30092, + "grad_norm": 0.7290036082267761, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 30092 + }, + { + "epoch": 0.30093, + "grad_norm": 0.7351526618003845, + "learning_rate": 0.003, + "loss": 3.981, + "step": 30093 + }, + { + "epoch": 0.30094, + "grad_norm": 0.8050387501716614, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 30094 + }, + { + "epoch": 0.30095, + "grad_norm": 0.7678536176681519, + "learning_rate": 0.003, + "loss": 3.964, + "step": 30095 + }, + { + "epoch": 0.30096, + "grad_norm": 0.7614619731903076, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 30096 + }, + { + "epoch": 0.30097, + "grad_norm": 0.8450554609298706, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 30097 + }, + { + "epoch": 0.30098, + "grad_norm": 1.0655944347381592, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 30098 + }, + { + "epoch": 0.30099, + "grad_norm": 1.0738990306854248, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 30099 + }, + { + "epoch": 0.301, + "grad_norm": 1.042629361152649, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 30100 + }, + { + "epoch": 0.30101, + "grad_norm": 1.0754369497299194, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 30101 + }, + { + "epoch": 0.30102, + "grad_norm": 0.9848163723945618, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 30102 + }, + { + "epoch": 0.30103, + "grad_norm": 0.942868173122406, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 30103 + }, + { + "epoch": 0.30104, + "grad_norm": 0.9812747240066528, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 30104 + }, + { + "epoch": 0.30105, + "grad_norm": 0.9426110982894897, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 30105 + }, + { + "epoch": 0.30106, + "grad_norm": 0.9842859506607056, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 30106 + }, + { + "epoch": 0.30107, + "grad_norm": 0.9787732362747192, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 30107 + }, + { + "epoch": 0.30108, + "grad_norm": 0.8049921989440918, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 30108 + }, + { + "epoch": 0.30109, + "grad_norm": 0.7150912284851074, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 30109 + }, + { + "epoch": 0.3011, + "grad_norm": 0.7039679884910583, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 30110 + }, + { + "epoch": 0.30111, + "grad_norm": 0.7068908214569092, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 30111 + }, + { + "epoch": 0.30112, + "grad_norm": 0.7115660905838013, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 30112 + }, + { + "epoch": 0.30113, + "grad_norm": 0.8008670210838318, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 30113 + }, + { + "epoch": 0.30114, + "grad_norm": 0.980202853679657, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 30114 + }, + { + "epoch": 0.30115, + "grad_norm": 1.220043659210205, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 30115 + }, + { + "epoch": 0.30116, + "grad_norm": 0.9642595648765564, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 30116 + }, + { + "epoch": 0.30117, + "grad_norm": 1.0809905529022217, + "learning_rate": 0.003, + "loss": 3.992, + "step": 30117 + }, + { + "epoch": 0.30118, + "grad_norm": 1.1123946905136108, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 30118 + }, + { + "epoch": 0.30119, + "grad_norm": 0.8607103824615479, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 30119 + }, + { + "epoch": 0.3012, + "grad_norm": 0.6477658152580261, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 30120 + }, + { + "epoch": 0.30121, + "grad_norm": 0.7092880010604858, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 30121 + }, + { + "epoch": 0.30122, + "grad_norm": 0.7088544368743896, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 30122 + }, + { + "epoch": 0.30123, + "grad_norm": 0.6974127292633057, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 30123 + }, + { + "epoch": 0.30124, + "grad_norm": 0.6846452355384827, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 30124 + }, + { + "epoch": 0.30125, + "grad_norm": 0.6657893061637878, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 30125 + }, + { + "epoch": 0.30126, + "grad_norm": 0.6490287184715271, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 30126 + }, + { + "epoch": 0.30127, + "grad_norm": 0.6458324790000916, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 30127 + }, + { + "epoch": 0.30128, + "grad_norm": 0.8407251238822937, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 30128 + }, + { + "epoch": 0.30129, + "grad_norm": 1.1672179698944092, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 30129 + }, + { + "epoch": 0.3013, + "grad_norm": 1.2326774597167969, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 30130 + }, + { + "epoch": 0.30131, + "grad_norm": 0.8593541383743286, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 30131 + }, + { + "epoch": 0.30132, + "grad_norm": 0.8164668679237366, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 30132 + }, + { + "epoch": 0.30133, + "grad_norm": 1.1315991878509521, + "learning_rate": 0.003, + "loss": 3.992, + "step": 30133 + }, + { + "epoch": 0.30134, + "grad_norm": 1.16073477268219, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 30134 + }, + { + "epoch": 0.30135, + "grad_norm": 0.5769139528274536, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 30135 + }, + { + "epoch": 0.30136, + "grad_norm": 0.6323615312576294, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 30136 + }, + { + "epoch": 0.30137, + "grad_norm": 0.7352566123008728, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 30137 + }, + { + "epoch": 0.30138, + "grad_norm": 0.7768644690513611, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 30138 + }, + { + "epoch": 0.30139, + "grad_norm": 0.7829878926277161, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 30139 + }, + { + "epoch": 0.3014, + "grad_norm": 0.8192238807678223, + "learning_rate": 0.003, + "loss": 3.9444, + "step": 30140 + }, + { + "epoch": 0.30141, + "grad_norm": 0.8512728214263916, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 30141 + }, + { + "epoch": 0.30142, + "grad_norm": 0.742274820804596, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 30142 + }, + { + "epoch": 0.30143, + "grad_norm": 0.7394730448722839, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 30143 + }, + { + "epoch": 0.30144, + "grad_norm": 0.7136843800544739, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 30144 + }, + { + "epoch": 0.30145, + "grad_norm": 0.6769461631774902, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 30145 + }, + { + "epoch": 0.30146, + "grad_norm": 0.6456104516983032, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 30146 + }, + { + "epoch": 0.30147, + "grad_norm": 0.8683112263679504, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 30147 + }, + { + "epoch": 0.30148, + "grad_norm": 1.2638649940490723, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 30148 + }, + { + "epoch": 0.30149, + "grad_norm": 0.8888698220252991, + "learning_rate": 0.003, + "loss": 3.997, + "step": 30149 + }, + { + "epoch": 0.3015, + "grad_norm": 0.6345845460891724, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 30150 + }, + { + "epoch": 0.30151, + "grad_norm": 0.6463819742202759, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 30151 + }, + { + "epoch": 0.30152, + "grad_norm": 0.6624374389648438, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 30152 + }, + { + "epoch": 0.30153, + "grad_norm": 0.7461624145507812, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 30153 + }, + { + "epoch": 0.30154, + "grad_norm": 1.0862773656845093, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 30154 + }, + { + "epoch": 0.30155, + "grad_norm": 1.0890814065933228, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 30155 + }, + { + "epoch": 0.30156, + "grad_norm": 0.9427317976951599, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 30156 + }, + { + "epoch": 0.30157, + "grad_norm": 0.9929181337356567, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 30157 + }, + { + "epoch": 0.30158, + "grad_norm": 0.9653158187866211, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 30158 + }, + { + "epoch": 0.30159, + "grad_norm": 0.8313683271408081, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 30159 + }, + { + "epoch": 0.3016, + "grad_norm": 0.7334973216056824, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 30160 + }, + { + "epoch": 0.30161, + "grad_norm": 0.8397002220153809, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 30161 + }, + { + "epoch": 0.30162, + "grad_norm": 0.9599658846855164, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 30162 + }, + { + "epoch": 0.30163, + "grad_norm": 1.209795355796814, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 30163 + }, + { + "epoch": 0.30164, + "grad_norm": 0.8126457333564758, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 30164 + }, + { + "epoch": 0.30165, + "grad_norm": 0.8885881304740906, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 30165 + }, + { + "epoch": 0.30166, + "grad_norm": 0.9843958616256714, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 30166 + }, + { + "epoch": 0.30167, + "grad_norm": 0.899753212928772, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 30167 + }, + { + "epoch": 0.30168, + "grad_norm": 0.9796943664550781, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 30168 + }, + { + "epoch": 0.30169, + "grad_norm": 1.15346360206604, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 30169 + }, + { + "epoch": 0.3017, + "grad_norm": 0.7540034055709839, + "learning_rate": 0.003, + "loss": 3.982, + "step": 30170 + }, + { + "epoch": 0.30171, + "grad_norm": 0.8037791848182678, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 30171 + }, + { + "epoch": 0.30172, + "grad_norm": 0.9041877388954163, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 30172 + }, + { + "epoch": 0.30173, + "grad_norm": 0.828494131565094, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 30173 + }, + { + "epoch": 0.30174, + "grad_norm": 0.8965911865234375, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 30174 + }, + { + "epoch": 0.30175, + "grad_norm": 0.8915791511535645, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 30175 + }, + { + "epoch": 0.30176, + "grad_norm": 0.7786785364151001, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 30176 + }, + { + "epoch": 0.30177, + "grad_norm": 0.734568178653717, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 30177 + }, + { + "epoch": 0.30178, + "grad_norm": 0.8215957880020142, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 30178 + }, + { + "epoch": 0.30179, + "grad_norm": 0.8617557287216187, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 30179 + }, + { + "epoch": 0.3018, + "grad_norm": 1.269036054611206, + "learning_rate": 0.003, + "loss": 4.024, + "step": 30180 + }, + { + "epoch": 0.30181, + "grad_norm": 1.0422440767288208, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 30181 + }, + { + "epoch": 0.30182, + "grad_norm": 1.0597362518310547, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 30182 + }, + { + "epoch": 0.30183, + "grad_norm": 0.9966402053833008, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 30183 + }, + { + "epoch": 0.30184, + "grad_norm": 0.9943029880523682, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 30184 + }, + { + "epoch": 0.30185, + "grad_norm": 0.9219706654548645, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 30185 + }, + { + "epoch": 0.30186, + "grad_norm": 0.885248601436615, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 30186 + }, + { + "epoch": 0.30187, + "grad_norm": 0.943252444267273, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 30187 + }, + { + "epoch": 0.30188, + "grad_norm": 0.9793200492858887, + "learning_rate": 0.003, + "loss": 4.03, + "step": 30188 + }, + { + "epoch": 0.30189, + "grad_norm": 0.853391170501709, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 30189 + }, + { + "epoch": 0.3019, + "grad_norm": 0.7386717796325684, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 30190 + }, + { + "epoch": 0.30191, + "grad_norm": 0.7621029615402222, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 30191 + }, + { + "epoch": 0.30192, + "grad_norm": 0.8036871552467346, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 30192 + }, + { + "epoch": 0.30193, + "grad_norm": 1.0040621757507324, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 30193 + }, + { + "epoch": 0.30194, + "grad_norm": 1.0711164474487305, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 30194 + }, + { + "epoch": 0.30195, + "grad_norm": 1.0152817964553833, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 30195 + }, + { + "epoch": 0.30196, + "grad_norm": 1.0529993772506714, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 30196 + }, + { + "epoch": 0.30197, + "grad_norm": 0.9189193844795227, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 30197 + }, + { + "epoch": 0.30198, + "grad_norm": 0.7565774321556091, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 30198 + }, + { + "epoch": 0.30199, + "grad_norm": 0.772654116153717, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 30199 + }, + { + "epoch": 0.302, + "grad_norm": 0.8694997429847717, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 30200 + }, + { + "epoch": 0.30201, + "grad_norm": 1.0984700918197632, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 30201 + }, + { + "epoch": 0.30202, + "grad_norm": 0.9813957214355469, + "learning_rate": 0.003, + "loss": 3.977, + "step": 30202 + }, + { + "epoch": 0.30203, + "grad_norm": 0.9099338054656982, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 30203 + }, + { + "epoch": 0.30204, + "grad_norm": 0.8468520641326904, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 30204 + }, + { + "epoch": 0.30205, + "grad_norm": 0.6965942978858948, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 30205 + }, + { + "epoch": 0.30206, + "grad_norm": 0.5863549709320068, + "learning_rate": 0.003, + "loss": 3.987, + "step": 30206 + }, + { + "epoch": 0.30207, + "grad_norm": 0.530653178691864, + "learning_rate": 0.003, + "loss": 3.998, + "step": 30207 + }, + { + "epoch": 0.30208, + "grad_norm": 0.5877682566642761, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 30208 + }, + { + "epoch": 0.30209, + "grad_norm": 0.7072252631187439, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 30209 + }, + { + "epoch": 0.3021, + "grad_norm": 0.7412365674972534, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 30210 + }, + { + "epoch": 0.30211, + "grad_norm": 0.705809473991394, + "learning_rate": 0.003, + "loss": 3.969, + "step": 30211 + }, + { + "epoch": 0.30212, + "grad_norm": 0.838628888130188, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 30212 + }, + { + "epoch": 0.30213, + "grad_norm": 1.1106610298156738, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 30213 + }, + { + "epoch": 0.30214, + "grad_norm": 1.1140817403793335, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 30214 + }, + { + "epoch": 0.30215, + "grad_norm": 0.8705548048019409, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 30215 + }, + { + "epoch": 0.30216, + "grad_norm": 0.8175590634346008, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 30216 + }, + { + "epoch": 0.30217, + "grad_norm": 0.8588775396347046, + "learning_rate": 0.003, + "loss": 3.966, + "step": 30217 + }, + { + "epoch": 0.30218, + "grad_norm": 0.9957062602043152, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 30218 + }, + { + "epoch": 0.30219, + "grad_norm": 0.9825361967086792, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 30219 + }, + { + "epoch": 0.3022, + "grad_norm": 1.12387216091156, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 30220 + }, + { + "epoch": 0.30221, + "grad_norm": 1.0322357416152954, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 30221 + }, + { + "epoch": 0.30222, + "grad_norm": 0.9849070310592651, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 30222 + }, + { + "epoch": 0.30223, + "grad_norm": 0.8280496597290039, + "learning_rate": 0.003, + "loss": 3.998, + "step": 30223 + }, + { + "epoch": 0.30224, + "grad_norm": 0.7078755497932434, + "learning_rate": 0.003, + "loss": 3.971, + "step": 30224 + }, + { + "epoch": 0.30225, + "grad_norm": 0.834050178527832, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 30225 + }, + { + "epoch": 0.30226, + "grad_norm": 0.8735666275024414, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 30226 + }, + { + "epoch": 0.30227, + "grad_norm": 0.9653239846229553, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 30227 + }, + { + "epoch": 0.30228, + "grad_norm": 1.049216866493225, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 30228 + }, + { + "epoch": 0.30229, + "grad_norm": 0.9415720701217651, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 30229 + }, + { + "epoch": 0.3023, + "grad_norm": 0.8177942037582397, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 30230 + }, + { + "epoch": 0.30231, + "grad_norm": 0.7304908037185669, + "learning_rate": 0.003, + "loss": 3.9334, + "step": 30231 + }, + { + "epoch": 0.30232, + "grad_norm": 0.683897078037262, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 30232 + }, + { + "epoch": 0.30233, + "grad_norm": 0.7323724627494812, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 30233 + }, + { + "epoch": 0.30234, + "grad_norm": 0.8167960047721863, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 30234 + }, + { + "epoch": 0.30235, + "grad_norm": 0.9478397369384766, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 30235 + }, + { + "epoch": 0.30236, + "grad_norm": 1.1945768594741821, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 30236 + }, + { + "epoch": 0.30237, + "grad_norm": 0.6515331864356995, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 30237 + }, + { + "epoch": 0.30238, + "grad_norm": 0.5916810035705566, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 30238 + }, + { + "epoch": 0.30239, + "grad_norm": 0.6537622809410095, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 30239 + }, + { + "epoch": 0.3024, + "grad_norm": 0.7715683579444885, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 30240 + }, + { + "epoch": 0.30241, + "grad_norm": 0.8983054161071777, + "learning_rate": 0.003, + "loss": 3.988, + "step": 30241 + }, + { + "epoch": 0.30242, + "grad_norm": 1.007004976272583, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 30242 + }, + { + "epoch": 0.30243, + "grad_norm": 1.0031712055206299, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 30243 + }, + { + "epoch": 0.30244, + "grad_norm": 0.968390941619873, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 30244 + }, + { + "epoch": 0.30245, + "grad_norm": 1.1322963237762451, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 30245 + }, + { + "epoch": 0.30246, + "grad_norm": 1.0056508779525757, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 30246 + }, + { + "epoch": 0.30247, + "grad_norm": 1.0045267343521118, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 30247 + }, + { + "epoch": 0.30248, + "grad_norm": 1.0507842302322388, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 30248 + }, + { + "epoch": 0.30249, + "grad_norm": 0.9432129263877869, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 30249 + }, + { + "epoch": 0.3025, + "grad_norm": 0.987248957157135, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 30250 + }, + { + "epoch": 0.30251, + "grad_norm": 1.13827645778656, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 30251 + }, + { + "epoch": 0.30252, + "grad_norm": 0.8720024228096008, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 30252 + }, + { + "epoch": 0.30253, + "grad_norm": 0.8775963187217712, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 30253 + }, + { + "epoch": 0.30254, + "grad_norm": 0.8547742366790771, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 30254 + }, + { + "epoch": 0.30255, + "grad_norm": 0.8464097380638123, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 30255 + }, + { + "epoch": 0.30256, + "grad_norm": 0.8163641691207886, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 30256 + }, + { + "epoch": 0.30257, + "grad_norm": 0.9668068885803223, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 30257 + }, + { + "epoch": 0.30258, + "grad_norm": 1.2231512069702148, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 30258 + }, + { + "epoch": 0.30259, + "grad_norm": 0.8461035490036011, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 30259 + }, + { + "epoch": 0.3026, + "grad_norm": 0.8074164390563965, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 30260 + }, + { + "epoch": 0.30261, + "grad_norm": 0.8834480047225952, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 30261 + }, + { + "epoch": 0.30262, + "grad_norm": 0.9935200810432434, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 30262 + }, + { + "epoch": 0.30263, + "grad_norm": 0.9422881007194519, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 30263 + }, + { + "epoch": 0.30264, + "grad_norm": 0.6559075713157654, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 30264 + }, + { + "epoch": 0.30265, + "grad_norm": 0.6439586877822876, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 30265 + }, + { + "epoch": 0.30266, + "grad_norm": 0.5963873267173767, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 30266 + }, + { + "epoch": 0.30267, + "grad_norm": 0.59284508228302, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 30267 + }, + { + "epoch": 0.30268, + "grad_norm": 0.5823668837547302, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 30268 + }, + { + "epoch": 0.30269, + "grad_norm": 0.6388369202613831, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 30269 + }, + { + "epoch": 0.3027, + "grad_norm": 0.6970305442810059, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 30270 + }, + { + "epoch": 0.30271, + "grad_norm": 0.718646228313446, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 30271 + }, + { + "epoch": 0.30272, + "grad_norm": 0.8192839622497559, + "learning_rate": 0.003, + "loss": 4.029, + "step": 30272 + }, + { + "epoch": 0.30273, + "grad_norm": 0.9334774613380432, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 30273 + }, + { + "epoch": 0.30274, + "grad_norm": 0.9300666451454163, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 30274 + }, + { + "epoch": 0.30275, + "grad_norm": 1.1687853336334229, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 30275 + }, + { + "epoch": 0.30276, + "grad_norm": 0.9622959494590759, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 30276 + }, + { + "epoch": 0.30277, + "grad_norm": 0.9957545399665833, + "learning_rate": 0.003, + "loss": 3.987, + "step": 30277 + }, + { + "epoch": 0.30278, + "grad_norm": 1.1958448886871338, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 30278 + }, + { + "epoch": 0.30279, + "grad_norm": 0.8666081428527832, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 30279 + }, + { + "epoch": 0.3028, + "grad_norm": 0.7442982196807861, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 30280 + }, + { + "epoch": 0.30281, + "grad_norm": 0.6730168461799622, + "learning_rate": 0.003, + "loss": 3.9391, + "step": 30281 + }, + { + "epoch": 0.30282, + "grad_norm": 0.6875019669532776, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 30282 + }, + { + "epoch": 0.30283, + "grad_norm": 0.676116943359375, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 30283 + }, + { + "epoch": 0.30284, + "grad_norm": 0.6623132228851318, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 30284 + }, + { + "epoch": 0.30285, + "grad_norm": 0.740074872970581, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 30285 + }, + { + "epoch": 0.30286, + "grad_norm": 0.7388473749160767, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 30286 + }, + { + "epoch": 0.30287, + "grad_norm": 0.7327272295951843, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 30287 + }, + { + "epoch": 0.30288, + "grad_norm": 0.9095813035964966, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 30288 + }, + { + "epoch": 0.30289, + "grad_norm": 1.1258975267410278, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 30289 + }, + { + "epoch": 0.3029, + "grad_norm": 0.9522033333778381, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 30290 + }, + { + "epoch": 0.30291, + "grad_norm": 0.896200954914093, + "learning_rate": 0.003, + "loss": 4.003, + "step": 30291 + }, + { + "epoch": 0.30292, + "grad_norm": 1.0581750869750977, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 30292 + }, + { + "epoch": 0.30293, + "grad_norm": 1.053808569908142, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 30293 + }, + { + "epoch": 0.30294, + "grad_norm": 1.092095136642456, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 30294 + }, + { + "epoch": 0.30295, + "grad_norm": 1.0238161087036133, + "learning_rate": 0.003, + "loss": 3.991, + "step": 30295 + }, + { + "epoch": 0.30296, + "grad_norm": 1.1544692516326904, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 30296 + }, + { + "epoch": 0.30297, + "grad_norm": 0.7969047427177429, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 30297 + }, + { + "epoch": 0.30298, + "grad_norm": 0.6529414653778076, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 30298 + }, + { + "epoch": 0.30299, + "grad_norm": 0.6814016699790955, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 30299 + }, + { + "epoch": 0.303, + "grad_norm": 0.7160606384277344, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 30300 + }, + { + "epoch": 0.30301, + "grad_norm": 0.8287773728370667, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 30301 + }, + { + "epoch": 0.30302, + "grad_norm": 0.8782233595848083, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 30302 + }, + { + "epoch": 0.30303, + "grad_norm": 1.018998622894287, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 30303 + }, + { + "epoch": 0.30304, + "grad_norm": 1.0426489114761353, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 30304 + }, + { + "epoch": 0.30305, + "grad_norm": 0.9040126204490662, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 30305 + }, + { + "epoch": 0.30306, + "grad_norm": 0.9290229082107544, + "learning_rate": 0.003, + "loss": 4.011, + "step": 30306 + }, + { + "epoch": 0.30307, + "grad_norm": 1.12712824344635, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 30307 + }, + { + "epoch": 0.30308, + "grad_norm": 0.8862696886062622, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 30308 + }, + { + "epoch": 0.30309, + "grad_norm": 0.7538245916366577, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 30309 + }, + { + "epoch": 0.3031, + "grad_norm": 0.7690710425376892, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 30310 + }, + { + "epoch": 0.30311, + "grad_norm": 0.7860538959503174, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 30311 + }, + { + "epoch": 0.30312, + "grad_norm": 0.8599854707717896, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 30312 + }, + { + "epoch": 0.30313, + "grad_norm": 0.9369581341743469, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 30313 + }, + { + "epoch": 0.30314, + "grad_norm": 0.8865220546722412, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 30314 + }, + { + "epoch": 0.30315, + "grad_norm": 0.8452551960945129, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 30315 + }, + { + "epoch": 0.30316, + "grad_norm": 0.8175135254859924, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 30316 + }, + { + "epoch": 0.30317, + "grad_norm": 0.9032526612281799, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 30317 + }, + { + "epoch": 0.30318, + "grad_norm": 0.8174006938934326, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 30318 + }, + { + "epoch": 0.30319, + "grad_norm": 0.7728899717330933, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 30319 + }, + { + "epoch": 0.3032, + "grad_norm": 0.9277541041374207, + "learning_rate": 0.003, + "loss": 4.002, + "step": 30320 + }, + { + "epoch": 0.30321, + "grad_norm": 1.0379595756530762, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 30321 + }, + { + "epoch": 0.30322, + "grad_norm": 0.9739081263542175, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 30322 + }, + { + "epoch": 0.30323, + "grad_norm": 1.0283160209655762, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 30323 + }, + { + "epoch": 0.30324, + "grad_norm": 0.9107156991958618, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 30324 + }, + { + "epoch": 0.30325, + "grad_norm": 0.7640366554260254, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 30325 + }, + { + "epoch": 0.30326, + "grad_norm": 0.9104044437408447, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 30326 + }, + { + "epoch": 0.30327, + "grad_norm": 0.8267000913619995, + "learning_rate": 0.003, + "loss": 3.988, + "step": 30327 + }, + { + "epoch": 0.30328, + "grad_norm": 0.6858292818069458, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 30328 + }, + { + "epoch": 0.30329, + "grad_norm": 0.6727209687232971, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 30329 + }, + { + "epoch": 0.3033, + "grad_norm": 0.7084246873855591, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 30330 + }, + { + "epoch": 0.30331, + "grad_norm": 0.6978114247322083, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 30331 + }, + { + "epoch": 0.30332, + "grad_norm": 0.6696245074272156, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 30332 + }, + { + "epoch": 0.30333, + "grad_norm": 0.7507653832435608, + "learning_rate": 0.003, + "loss": 3.993, + "step": 30333 + }, + { + "epoch": 0.30334, + "grad_norm": 0.9417454600334167, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 30334 + }, + { + "epoch": 0.30335, + "grad_norm": 1.1512012481689453, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 30335 + }, + { + "epoch": 0.30336, + "grad_norm": 0.883955717086792, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 30336 + }, + { + "epoch": 0.30337, + "grad_norm": 0.8430861830711365, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 30337 + }, + { + "epoch": 0.30338, + "grad_norm": 0.9104881882667542, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 30338 + }, + { + "epoch": 0.30339, + "grad_norm": 0.9280827045440674, + "learning_rate": 0.003, + "loss": 3.988, + "step": 30339 + }, + { + "epoch": 0.3034, + "grad_norm": 0.846112072467804, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 30340 + }, + { + "epoch": 0.30341, + "grad_norm": 0.7893561720848083, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 30341 + }, + { + "epoch": 0.30342, + "grad_norm": 0.7822876572608948, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 30342 + }, + { + "epoch": 0.30343, + "grad_norm": 0.6683368682861328, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 30343 + }, + { + "epoch": 0.30344, + "grad_norm": 0.6545978784561157, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 30344 + }, + { + "epoch": 0.30345, + "grad_norm": 0.751083493232727, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 30345 + }, + { + "epoch": 0.30346, + "grad_norm": 1.0199860334396362, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 30346 + }, + { + "epoch": 0.30347, + "grad_norm": 1.0735636949539185, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 30347 + }, + { + "epoch": 0.30348, + "grad_norm": 1.110585331916809, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 30348 + }, + { + "epoch": 0.30349, + "grad_norm": 1.0312978029251099, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 30349 + }, + { + "epoch": 0.3035, + "grad_norm": 1.330212116241455, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 30350 + }, + { + "epoch": 0.30351, + "grad_norm": 0.6948171854019165, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 30351 + }, + { + "epoch": 0.30352, + "grad_norm": 0.7467683553695679, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 30352 + }, + { + "epoch": 0.30353, + "grad_norm": 0.8510563969612122, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 30353 + }, + { + "epoch": 0.30354, + "grad_norm": 0.8115134835243225, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 30354 + }, + { + "epoch": 0.30355, + "grad_norm": 0.9251977801322937, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 30355 + }, + { + "epoch": 0.30356, + "grad_norm": 0.9468076229095459, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 30356 + }, + { + "epoch": 0.30357, + "grad_norm": 0.935030996799469, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 30357 + }, + { + "epoch": 0.30358, + "grad_norm": 0.8760204911231995, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 30358 + }, + { + "epoch": 0.30359, + "grad_norm": 0.7657546997070312, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 30359 + }, + { + "epoch": 0.3036, + "grad_norm": 0.7461541295051575, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 30360 + }, + { + "epoch": 0.30361, + "grad_norm": 0.872285008430481, + "learning_rate": 0.003, + "loss": 3.955, + "step": 30361 + }, + { + "epoch": 0.30362, + "grad_norm": 1.0580230951309204, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 30362 + }, + { + "epoch": 0.30363, + "grad_norm": 1.1550146341323853, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 30363 + }, + { + "epoch": 0.30364, + "grad_norm": 0.8858004212379456, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 30364 + }, + { + "epoch": 0.30365, + "grad_norm": 0.849958598613739, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 30365 + }, + { + "epoch": 0.30366, + "grad_norm": 0.7315015196800232, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 30366 + }, + { + "epoch": 0.30367, + "grad_norm": 0.7496376633644104, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 30367 + }, + { + "epoch": 0.30368, + "grad_norm": 0.7840912342071533, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 30368 + }, + { + "epoch": 0.30369, + "grad_norm": 0.8360124826431274, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 30369 + }, + { + "epoch": 0.3037, + "grad_norm": 0.9152970910072327, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 30370 + }, + { + "epoch": 0.30371, + "grad_norm": 1.0833603143692017, + "learning_rate": 0.003, + "loss": 3.9438, + "step": 30371 + }, + { + "epoch": 0.30372, + "grad_norm": 0.9836386442184448, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 30372 + }, + { + "epoch": 0.30373, + "grad_norm": 1.2573151588439941, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 30373 + }, + { + "epoch": 0.30374, + "grad_norm": 0.7872157096862793, + "learning_rate": 0.003, + "loss": 3.9411, + "step": 30374 + }, + { + "epoch": 0.30375, + "grad_norm": 0.782706081867218, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 30375 + }, + { + "epoch": 0.30376, + "grad_norm": 0.8658480644226074, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 30376 + }, + { + "epoch": 0.30377, + "grad_norm": 1.050161361694336, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 30377 + }, + { + "epoch": 0.30378, + "grad_norm": 1.0117853879928589, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 30378 + }, + { + "epoch": 0.30379, + "grad_norm": 1.0371570587158203, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 30379 + }, + { + "epoch": 0.3038, + "grad_norm": 0.7909414172172546, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 30380 + }, + { + "epoch": 0.30381, + "grad_norm": 0.7520966529846191, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 30381 + }, + { + "epoch": 0.30382, + "grad_norm": 0.7825096845626831, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 30382 + }, + { + "epoch": 0.30383, + "grad_norm": 0.7395613789558411, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 30383 + }, + { + "epoch": 0.30384, + "grad_norm": 0.8062317371368408, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 30384 + }, + { + "epoch": 0.30385, + "grad_norm": 0.80616694688797, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 30385 + }, + { + "epoch": 0.30386, + "grad_norm": 0.7647818922996521, + "learning_rate": 0.003, + "loss": 4.004, + "step": 30386 + }, + { + "epoch": 0.30387, + "grad_norm": 0.8726249933242798, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 30387 + }, + { + "epoch": 0.30388, + "grad_norm": 0.8078866600990295, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 30388 + }, + { + "epoch": 0.30389, + "grad_norm": 0.7049688100814819, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 30389 + }, + { + "epoch": 0.3039, + "grad_norm": 0.7068129181861877, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 30390 + }, + { + "epoch": 0.30391, + "grad_norm": 0.7048782706260681, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 30391 + }, + { + "epoch": 0.30392, + "grad_norm": 0.8119542598724365, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 30392 + }, + { + "epoch": 0.30393, + "grad_norm": 0.8419193029403687, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 30393 + }, + { + "epoch": 0.30394, + "grad_norm": 0.9211612343788147, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 30394 + }, + { + "epoch": 0.30395, + "grad_norm": 0.9578372240066528, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 30395 + }, + { + "epoch": 0.30396, + "grad_norm": 1.3032948970794678, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 30396 + }, + { + "epoch": 0.30397, + "grad_norm": 0.8814945220947266, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 30397 + }, + { + "epoch": 0.30398, + "grad_norm": 0.8843804001808167, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 30398 + }, + { + "epoch": 0.30399, + "grad_norm": 0.9005182981491089, + "learning_rate": 0.003, + "loss": 3.962, + "step": 30399 + }, + { + "epoch": 0.304, + "grad_norm": 0.9073457717895508, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 30400 + }, + { + "epoch": 0.30401, + "grad_norm": 0.9085062742233276, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 30401 + }, + { + "epoch": 0.30402, + "grad_norm": 1.0890225172042847, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 30402 + }, + { + "epoch": 0.30403, + "grad_norm": 1.0235743522644043, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 30403 + }, + { + "epoch": 0.30404, + "grad_norm": 0.8662883639335632, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 30404 + }, + { + "epoch": 0.30405, + "grad_norm": 0.8001978993415833, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 30405 + }, + { + "epoch": 0.30406, + "grad_norm": 0.9236955046653748, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 30406 + }, + { + "epoch": 0.30407, + "grad_norm": 1.0693122148513794, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 30407 + }, + { + "epoch": 0.30408, + "grad_norm": 1.0588600635528564, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 30408 + }, + { + "epoch": 0.30409, + "grad_norm": 0.8371297121047974, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 30409 + }, + { + "epoch": 0.3041, + "grad_norm": 0.8291128277778625, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 30410 + }, + { + "epoch": 0.30411, + "grad_norm": 1.0019060373306274, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 30411 + }, + { + "epoch": 0.30412, + "grad_norm": 1.1443995237350464, + "learning_rate": 0.003, + "loss": 3.991, + "step": 30412 + }, + { + "epoch": 0.30413, + "grad_norm": 0.6598800420761108, + "learning_rate": 0.003, + "loss": 3.975, + "step": 30413 + }, + { + "epoch": 0.30414, + "grad_norm": 0.6251203417778015, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 30414 + }, + { + "epoch": 0.30415, + "grad_norm": 0.5753851532936096, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 30415 + }, + { + "epoch": 0.30416, + "grad_norm": 0.5802266597747803, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 30416 + }, + { + "epoch": 0.30417, + "grad_norm": 0.6083036661148071, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 30417 + }, + { + "epoch": 0.30418, + "grad_norm": 0.7024540305137634, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 30418 + }, + { + "epoch": 0.30419, + "grad_norm": 0.8453680276870728, + "learning_rate": 0.003, + "loss": 3.972, + "step": 30419 + }, + { + "epoch": 0.3042, + "grad_norm": 1.0181288719177246, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 30420 + }, + { + "epoch": 0.30421, + "grad_norm": 0.9517219066619873, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 30421 + }, + { + "epoch": 0.30422, + "grad_norm": 0.8908277750015259, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 30422 + }, + { + "epoch": 0.30423, + "grad_norm": 0.9402300715446472, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 30423 + }, + { + "epoch": 0.30424, + "grad_norm": 0.8686082363128662, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 30424 + }, + { + "epoch": 0.30425, + "grad_norm": 0.8262990117073059, + "learning_rate": 0.003, + "loss": 3.975, + "step": 30425 + }, + { + "epoch": 0.30426, + "grad_norm": 0.945346474647522, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 30426 + }, + { + "epoch": 0.30427, + "grad_norm": 0.8240861892700195, + "learning_rate": 0.003, + "loss": 4.002, + "step": 30427 + }, + { + "epoch": 0.30428, + "grad_norm": 0.6996665000915527, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 30428 + }, + { + "epoch": 0.30429, + "grad_norm": 0.6547027826309204, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 30429 + }, + { + "epoch": 0.3043, + "grad_norm": 0.748733401298523, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 30430 + }, + { + "epoch": 0.30431, + "grad_norm": 0.7600756287574768, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 30431 + }, + { + "epoch": 0.30432, + "grad_norm": 0.6672586798667908, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 30432 + }, + { + "epoch": 0.30433, + "grad_norm": 0.6173412799835205, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 30433 + }, + { + "epoch": 0.30434, + "grad_norm": 0.618618905544281, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 30434 + }, + { + "epoch": 0.30435, + "grad_norm": 0.9690936207771301, + "learning_rate": 0.003, + "loss": 3.998, + "step": 30435 + }, + { + "epoch": 0.30436, + "grad_norm": 1.417122483253479, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 30436 + }, + { + "epoch": 0.30437, + "grad_norm": 0.7688219547271729, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 30437 + }, + { + "epoch": 0.30438, + "grad_norm": 0.7708353400230408, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 30438 + }, + { + "epoch": 0.30439, + "grad_norm": 0.91774982213974, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 30439 + }, + { + "epoch": 0.3044, + "grad_norm": 0.8581909537315369, + "learning_rate": 0.003, + "loss": 3.9343, + "step": 30440 + }, + { + "epoch": 0.30441, + "grad_norm": 0.6650703549385071, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 30441 + }, + { + "epoch": 0.30442, + "grad_norm": 0.7440251708030701, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 30442 + }, + { + "epoch": 0.30443, + "grad_norm": 0.8805598616600037, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 30443 + }, + { + "epoch": 0.30444, + "grad_norm": 1.1519081592559814, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 30444 + }, + { + "epoch": 0.30445, + "grad_norm": 1.0719715356826782, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 30445 + }, + { + "epoch": 0.30446, + "grad_norm": 0.7880178093910217, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 30446 + }, + { + "epoch": 0.30447, + "grad_norm": 0.6771432757377625, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 30447 + }, + { + "epoch": 0.30448, + "grad_norm": 0.6853340268135071, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 30448 + }, + { + "epoch": 0.30449, + "grad_norm": 0.8004649877548218, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 30449 + }, + { + "epoch": 0.3045, + "grad_norm": 1.0659260749816895, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 30450 + }, + { + "epoch": 0.30451, + "grad_norm": 1.1473972797393799, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 30451 + }, + { + "epoch": 0.30452, + "grad_norm": 0.7445893883705139, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 30452 + }, + { + "epoch": 0.30453, + "grad_norm": 0.714892566204071, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 30453 + }, + { + "epoch": 0.30454, + "grad_norm": 0.7449681758880615, + "learning_rate": 0.003, + "loss": 3.98, + "step": 30454 + }, + { + "epoch": 0.30455, + "grad_norm": 0.8328741788864136, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 30455 + }, + { + "epoch": 0.30456, + "grad_norm": 0.9247656464576721, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 30456 + }, + { + "epoch": 0.30457, + "grad_norm": 1.0316188335418701, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 30457 + }, + { + "epoch": 0.30458, + "grad_norm": 1.065641164779663, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 30458 + }, + { + "epoch": 0.30459, + "grad_norm": 0.9002147316932678, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 30459 + }, + { + "epoch": 0.3046, + "grad_norm": 0.8882553577423096, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 30460 + }, + { + "epoch": 0.30461, + "grad_norm": 0.9306603074073792, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 30461 + }, + { + "epoch": 0.30462, + "grad_norm": 1.0451393127441406, + "learning_rate": 0.003, + "loss": 3.999, + "step": 30462 + }, + { + "epoch": 0.30463, + "grad_norm": 1.1494266986846924, + "learning_rate": 0.003, + "loss": 3.966, + "step": 30463 + }, + { + "epoch": 0.30464, + "grad_norm": 0.8066320419311523, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 30464 + }, + { + "epoch": 0.30465, + "grad_norm": 0.9459854364395142, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 30465 + }, + { + "epoch": 0.30466, + "grad_norm": 0.9340614080429077, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 30466 + }, + { + "epoch": 0.30467, + "grad_norm": 1.0651161670684814, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 30467 + }, + { + "epoch": 0.30468, + "grad_norm": 0.9782781600952148, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 30468 + }, + { + "epoch": 0.30469, + "grad_norm": 0.8731213808059692, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 30469 + }, + { + "epoch": 0.3047, + "grad_norm": 0.812019407749176, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 30470 + }, + { + "epoch": 0.30471, + "grad_norm": 0.7774103879928589, + "learning_rate": 0.003, + "loss": 4.01, + "step": 30471 + }, + { + "epoch": 0.30472, + "grad_norm": 0.6420204043388367, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 30472 + }, + { + "epoch": 0.30473, + "grad_norm": 0.5688521862030029, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 30473 + }, + { + "epoch": 0.30474, + "grad_norm": 0.6244319677352905, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 30474 + }, + { + "epoch": 0.30475, + "grad_norm": 0.7256326675415039, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 30475 + }, + { + "epoch": 0.30476, + "grad_norm": 0.7150586843490601, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 30476 + }, + { + "epoch": 0.30477, + "grad_norm": 0.7211515307426453, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 30477 + }, + { + "epoch": 0.30478, + "grad_norm": 0.846062183380127, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 30478 + }, + { + "epoch": 0.30479, + "grad_norm": 1.105612874031067, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 30479 + }, + { + "epoch": 0.3048, + "grad_norm": 1.0909448862075806, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 30480 + }, + { + "epoch": 0.30481, + "grad_norm": 0.8004578351974487, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 30481 + }, + { + "epoch": 0.30482, + "grad_norm": 0.6309317946434021, + "learning_rate": 0.003, + "loss": 3.998, + "step": 30482 + }, + { + "epoch": 0.30483, + "grad_norm": 0.5971251726150513, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 30483 + }, + { + "epoch": 0.30484, + "grad_norm": 0.6250884532928467, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 30484 + }, + { + "epoch": 0.30485, + "grad_norm": 0.6943829655647278, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 30485 + }, + { + "epoch": 0.30486, + "grad_norm": 0.7837355732917786, + "learning_rate": 0.003, + "loss": 3.986, + "step": 30486 + }, + { + "epoch": 0.30487, + "grad_norm": 0.7870044708251953, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 30487 + }, + { + "epoch": 0.30488, + "grad_norm": 0.8624109029769897, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 30488 + }, + { + "epoch": 0.30489, + "grad_norm": 1.0303211212158203, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 30489 + }, + { + "epoch": 0.3049, + "grad_norm": 0.9132218956947327, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 30490 + }, + { + "epoch": 0.30491, + "grad_norm": 0.8284927010536194, + "learning_rate": 0.003, + "loss": 3.9458, + "step": 30491 + }, + { + "epoch": 0.30492, + "grad_norm": 0.8763497471809387, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 30492 + }, + { + "epoch": 0.30493, + "grad_norm": 0.8493266105651855, + "learning_rate": 0.003, + "loss": 3.958, + "step": 30493 + }, + { + "epoch": 0.30494, + "grad_norm": 0.8238226175308228, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 30494 + }, + { + "epoch": 0.30495, + "grad_norm": 0.8363548517227173, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 30495 + }, + { + "epoch": 0.30496, + "grad_norm": 1.0465199947357178, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 30496 + }, + { + "epoch": 0.30497, + "grad_norm": 1.166298508644104, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 30497 + }, + { + "epoch": 0.30498, + "grad_norm": 0.8258225917816162, + "learning_rate": 0.003, + "loss": 3.982, + "step": 30498 + }, + { + "epoch": 0.30499, + "grad_norm": 0.8082234859466553, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 30499 + }, + { + "epoch": 0.305, + "grad_norm": 0.8300761580467224, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 30500 + }, + { + "epoch": 0.30501, + "grad_norm": 0.9048359990119934, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 30501 + }, + { + "epoch": 0.30502, + "grad_norm": 0.839637815952301, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 30502 + }, + { + "epoch": 0.30503, + "grad_norm": 1.0831258296966553, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 30503 + }, + { + "epoch": 0.30504, + "grad_norm": 1.2963016033172607, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 30504 + }, + { + "epoch": 0.30505, + "grad_norm": 0.8445757031440735, + "learning_rate": 0.003, + "loss": 3.987, + "step": 30505 + }, + { + "epoch": 0.30506, + "grad_norm": 0.7638489603996277, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 30506 + }, + { + "epoch": 0.30507, + "grad_norm": 0.7848820090293884, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 30507 + }, + { + "epoch": 0.30508, + "grad_norm": 0.9011987447738647, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 30508 + }, + { + "epoch": 0.30509, + "grad_norm": 0.9988778829574585, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 30509 + }, + { + "epoch": 0.3051, + "grad_norm": 0.9100135564804077, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 30510 + }, + { + "epoch": 0.30511, + "grad_norm": 0.7960159778594971, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 30511 + }, + { + "epoch": 0.30512, + "grad_norm": 0.9097263216972351, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 30512 + }, + { + "epoch": 0.30513, + "grad_norm": 1.0588728189468384, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 30513 + }, + { + "epoch": 0.30514, + "grad_norm": 0.8763190507888794, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 30514 + }, + { + "epoch": 0.30515, + "grad_norm": 1.0033458471298218, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 30515 + }, + { + "epoch": 0.30516, + "grad_norm": 1.2956793308258057, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 30516 + }, + { + "epoch": 0.30517, + "grad_norm": 0.995089590549469, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 30517 + }, + { + "epoch": 0.30518, + "grad_norm": 0.8067682981491089, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 30518 + }, + { + "epoch": 0.30519, + "grad_norm": 0.778660237789154, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 30519 + }, + { + "epoch": 0.3052, + "grad_norm": 0.8473300933837891, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 30520 + }, + { + "epoch": 0.30521, + "grad_norm": 0.9629321694374084, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 30521 + }, + { + "epoch": 0.30522, + "grad_norm": 1.2478245496749878, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 30522 + }, + { + "epoch": 0.30523, + "grad_norm": 0.7133013010025024, + "learning_rate": 0.003, + "loss": 3.992, + "step": 30523 + }, + { + "epoch": 0.30524, + "grad_norm": 0.7227146625518799, + "learning_rate": 0.003, + "loss": 3.988, + "step": 30524 + }, + { + "epoch": 0.30525, + "grad_norm": 0.8078490495681763, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 30525 + }, + { + "epoch": 0.30526, + "grad_norm": 0.9179291129112244, + "learning_rate": 0.003, + "loss": 3.99, + "step": 30526 + }, + { + "epoch": 0.30527, + "grad_norm": 0.9237551093101501, + "learning_rate": 0.003, + "loss": 3.9394, + "step": 30527 + }, + { + "epoch": 0.30528, + "grad_norm": 0.9715039134025574, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 30528 + }, + { + "epoch": 0.30529, + "grad_norm": 0.9236636757850647, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 30529 + }, + { + "epoch": 0.3053, + "grad_norm": 0.8432654142379761, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 30530 + }, + { + "epoch": 0.30531, + "grad_norm": 0.7837088108062744, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 30531 + }, + { + "epoch": 0.30532, + "grad_norm": 0.8400914072990417, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 30532 + }, + { + "epoch": 0.30533, + "grad_norm": 0.9610514044761658, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 30533 + }, + { + "epoch": 0.30534, + "grad_norm": 1.0681147575378418, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 30534 + }, + { + "epoch": 0.30535, + "grad_norm": 0.8678343296051025, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 30535 + }, + { + "epoch": 0.30536, + "grad_norm": 0.8463944792747498, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 30536 + }, + { + "epoch": 0.30537, + "grad_norm": 0.7381572723388672, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 30537 + }, + { + "epoch": 0.30538, + "grad_norm": 0.7218008041381836, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 30538 + }, + { + "epoch": 0.30539, + "grad_norm": 0.7692341208457947, + "learning_rate": 0.003, + "loss": 3.982, + "step": 30539 + }, + { + "epoch": 0.3054, + "grad_norm": 0.8740194439888, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 30540 + }, + { + "epoch": 0.30541, + "grad_norm": 0.9832338690757751, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 30541 + }, + { + "epoch": 0.30542, + "grad_norm": 1.3624318838119507, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 30542 + }, + { + "epoch": 0.30543, + "grad_norm": 0.6052711606025696, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 30543 + }, + { + "epoch": 0.30544, + "grad_norm": 0.8193807005882263, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 30544 + }, + { + "epoch": 0.30545, + "grad_norm": 1.035422921180725, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 30545 + }, + { + "epoch": 0.30546, + "grad_norm": 0.9633319973945618, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 30546 + }, + { + "epoch": 0.30547, + "grad_norm": 0.9895157814025879, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 30547 + }, + { + "epoch": 0.30548, + "grad_norm": 1.057220697402954, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 30548 + }, + { + "epoch": 0.30549, + "grad_norm": 0.9040523767471313, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 30549 + }, + { + "epoch": 0.3055, + "grad_norm": 0.7558218240737915, + "learning_rate": 0.003, + "loss": 3.994, + "step": 30550 + }, + { + "epoch": 0.30551, + "grad_norm": 0.8455974459648132, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 30551 + }, + { + "epoch": 0.30552, + "grad_norm": 0.781519889831543, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 30552 + }, + { + "epoch": 0.30553, + "grad_norm": 0.695599377155304, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 30553 + }, + { + "epoch": 0.30554, + "grad_norm": 0.748912513256073, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 30554 + }, + { + "epoch": 0.30555, + "grad_norm": 0.8814196586608887, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 30555 + }, + { + "epoch": 0.30556, + "grad_norm": 0.7522661089897156, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 30556 + }, + { + "epoch": 0.30557, + "grad_norm": 0.8589982986450195, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 30557 + }, + { + "epoch": 0.30558, + "grad_norm": 1.0504270792007446, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 30558 + }, + { + "epoch": 0.30559, + "grad_norm": 0.9183588624000549, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 30559 + }, + { + "epoch": 0.3056, + "grad_norm": 0.9478943347930908, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 30560 + }, + { + "epoch": 0.30561, + "grad_norm": 0.9203839302062988, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 30561 + }, + { + "epoch": 0.30562, + "grad_norm": 1.2402474880218506, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 30562 + }, + { + "epoch": 0.30563, + "grad_norm": 1.0284563302993774, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 30563 + }, + { + "epoch": 0.30564, + "grad_norm": 0.8479577302932739, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 30564 + }, + { + "epoch": 0.30565, + "grad_norm": 0.9096984267234802, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 30565 + }, + { + "epoch": 0.30566, + "grad_norm": 1.1136798858642578, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 30566 + }, + { + "epoch": 0.30567, + "grad_norm": 0.9273459315299988, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 30567 + }, + { + "epoch": 0.30568, + "grad_norm": 0.7964998483657837, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 30568 + }, + { + "epoch": 0.30569, + "grad_norm": 0.8477041125297546, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 30569 + }, + { + "epoch": 0.3057, + "grad_norm": 0.8123272061347961, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 30570 + }, + { + "epoch": 0.30571, + "grad_norm": 0.7795884013175964, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 30571 + }, + { + "epoch": 0.30572, + "grad_norm": 0.6733548045158386, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 30572 + }, + { + "epoch": 0.30573, + "grad_norm": 0.6201255917549133, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 30573 + }, + { + "epoch": 0.30574, + "grad_norm": 0.5391696691513062, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 30574 + }, + { + "epoch": 0.30575, + "grad_norm": 0.6170613169670105, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 30575 + }, + { + "epoch": 0.30576, + "grad_norm": 0.6998900771141052, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 30576 + }, + { + "epoch": 0.30577, + "grad_norm": 0.768905758857727, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 30577 + }, + { + "epoch": 0.30578, + "grad_norm": 1.000012993812561, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 30578 + }, + { + "epoch": 0.30579, + "grad_norm": 1.3634874820709229, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 30579 + }, + { + "epoch": 0.3058, + "grad_norm": 0.6634529829025269, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 30580 + }, + { + "epoch": 0.30581, + "grad_norm": 0.6959792971611023, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 30581 + }, + { + "epoch": 0.30582, + "grad_norm": 0.7159451842308044, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 30582 + }, + { + "epoch": 0.30583, + "grad_norm": 0.7664344310760498, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 30583 + }, + { + "epoch": 0.30584, + "grad_norm": 0.9371193051338196, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 30584 + }, + { + "epoch": 0.30585, + "grad_norm": 1.1118688583374023, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 30585 + }, + { + "epoch": 0.30586, + "grad_norm": 0.9556582570075989, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 30586 + }, + { + "epoch": 0.30587, + "grad_norm": 1.124391794204712, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 30587 + }, + { + "epoch": 0.30588, + "grad_norm": 0.9187823534011841, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 30588 + }, + { + "epoch": 0.30589, + "grad_norm": 0.6912382245063782, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 30589 + }, + { + "epoch": 0.3059, + "grad_norm": 0.7585684061050415, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 30590 + }, + { + "epoch": 0.30591, + "grad_norm": 0.9306550025939941, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 30591 + }, + { + "epoch": 0.30592, + "grad_norm": 1.091058611869812, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 30592 + }, + { + "epoch": 0.30593, + "grad_norm": 1.0130480527877808, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 30593 + }, + { + "epoch": 0.30594, + "grad_norm": 0.9245081543922424, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 30594 + }, + { + "epoch": 0.30595, + "grad_norm": 0.9567170143127441, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 30595 + }, + { + "epoch": 0.30596, + "grad_norm": 1.092675805091858, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 30596 + }, + { + "epoch": 0.30597, + "grad_norm": 0.8847861289978027, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 30597 + }, + { + "epoch": 0.30598, + "grad_norm": 0.7441532611846924, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 30598 + }, + { + "epoch": 0.30599, + "grad_norm": 0.7178089618682861, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 30599 + }, + { + "epoch": 0.306, + "grad_norm": 0.6196877956390381, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 30600 + }, + { + "epoch": 0.30601, + "grad_norm": 0.5587158203125, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 30601 + }, + { + "epoch": 0.30602, + "grad_norm": 0.6882948279380798, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 30602 + }, + { + "epoch": 0.30603, + "grad_norm": 0.7731837034225464, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 30603 + }, + { + "epoch": 0.30604, + "grad_norm": 0.9342118501663208, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 30604 + }, + { + "epoch": 0.30605, + "grad_norm": 0.9070847630500793, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 30605 + }, + { + "epoch": 0.30606, + "grad_norm": 0.8916009068489075, + "learning_rate": 0.003, + "loss": 4.026, + "step": 30606 + }, + { + "epoch": 0.30607, + "grad_norm": 0.958459734916687, + "learning_rate": 0.003, + "loss": 3.983, + "step": 30607 + }, + { + "epoch": 0.30608, + "grad_norm": 0.9858197569847107, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 30608 + }, + { + "epoch": 0.30609, + "grad_norm": 1.1876393556594849, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 30609 + }, + { + "epoch": 0.3061, + "grad_norm": 1.036871314048767, + "learning_rate": 0.003, + "loss": 4.014, + "step": 30610 + }, + { + "epoch": 0.30611, + "grad_norm": 0.9125224351882935, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 30611 + }, + { + "epoch": 0.30612, + "grad_norm": 1.0596343278884888, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 30612 + }, + { + "epoch": 0.30613, + "grad_norm": 1.1583744287490845, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 30613 + }, + { + "epoch": 0.30614, + "grad_norm": 0.816627025604248, + "learning_rate": 0.003, + "loss": 3.959, + "step": 30614 + }, + { + "epoch": 0.30615, + "grad_norm": 0.8075582981109619, + "learning_rate": 0.003, + "loss": 4.016, + "step": 30615 + }, + { + "epoch": 0.30616, + "grad_norm": 1.0060006380081177, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 30616 + }, + { + "epoch": 0.30617, + "grad_norm": 1.104568600654602, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 30617 + }, + { + "epoch": 0.30618, + "grad_norm": 0.9270997643470764, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 30618 + }, + { + "epoch": 0.30619, + "grad_norm": 0.8035250902175903, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 30619 + }, + { + "epoch": 0.3062, + "grad_norm": 0.8798080086708069, + "learning_rate": 0.003, + "loss": 3.974, + "step": 30620 + }, + { + "epoch": 0.30621, + "grad_norm": 0.7744510769844055, + "learning_rate": 0.003, + "loss": 3.981, + "step": 30621 + }, + { + "epoch": 0.30622, + "grad_norm": 0.8767858147621155, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 30622 + }, + { + "epoch": 0.30623, + "grad_norm": 1.2187179327011108, + "learning_rate": 0.003, + "loss": 4.012, + "step": 30623 + }, + { + "epoch": 0.30624, + "grad_norm": 0.7951979041099548, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 30624 + }, + { + "epoch": 0.30625, + "grad_norm": 0.7299324870109558, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 30625 + }, + { + "epoch": 0.30626, + "grad_norm": 0.8382022976875305, + "learning_rate": 0.003, + "loss": 3.997, + "step": 30626 + }, + { + "epoch": 0.30627, + "grad_norm": 0.8708194494247437, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 30627 + }, + { + "epoch": 0.30628, + "grad_norm": 0.8286699652671814, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 30628 + }, + { + "epoch": 0.30629, + "grad_norm": 0.8103384971618652, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 30629 + }, + { + "epoch": 0.3063, + "grad_norm": 0.7321674227714539, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 30630 + }, + { + "epoch": 0.30631, + "grad_norm": 0.7942976951599121, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 30631 + }, + { + "epoch": 0.30632, + "grad_norm": 0.8226797580718994, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 30632 + }, + { + "epoch": 0.30633, + "grad_norm": 0.942704975605011, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 30633 + }, + { + "epoch": 0.30634, + "grad_norm": 0.9711281061172485, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 30634 + }, + { + "epoch": 0.30635, + "grad_norm": 0.9732150435447693, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 30635 + }, + { + "epoch": 0.30636, + "grad_norm": 1.2646981477737427, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 30636 + }, + { + "epoch": 0.30637, + "grad_norm": 0.9270835518836975, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 30637 + }, + { + "epoch": 0.30638, + "grad_norm": 0.8100348114967346, + "learning_rate": 0.003, + "loss": 3.9369, + "step": 30638 + }, + { + "epoch": 0.30639, + "grad_norm": 0.8455439209938049, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 30639 + }, + { + "epoch": 0.3064, + "grad_norm": 0.936989426612854, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 30640 + }, + { + "epoch": 0.30641, + "grad_norm": 0.9502411484718323, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 30641 + }, + { + "epoch": 0.30642, + "grad_norm": 0.7969107031822205, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 30642 + }, + { + "epoch": 0.30643, + "grad_norm": 0.6728684902191162, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 30643 + }, + { + "epoch": 0.30644, + "grad_norm": 0.6918038129806519, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 30644 + }, + { + "epoch": 0.30645, + "grad_norm": 0.7473955750465393, + "learning_rate": 0.003, + "loss": 3.973, + "step": 30645 + }, + { + "epoch": 0.30646, + "grad_norm": 0.9189848899841309, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 30646 + }, + { + "epoch": 0.30647, + "grad_norm": 1.1519391536712646, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 30647 + }, + { + "epoch": 0.30648, + "grad_norm": 0.9308950901031494, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 30648 + }, + { + "epoch": 0.30649, + "grad_norm": 0.8803507685661316, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 30649 + }, + { + "epoch": 0.3065, + "grad_norm": 0.8469889760017395, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 30650 + }, + { + "epoch": 0.30651, + "grad_norm": 0.8393092155456543, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 30651 + }, + { + "epoch": 0.30652, + "grad_norm": 0.7520397901535034, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 30652 + }, + { + "epoch": 0.30653, + "grad_norm": 0.8572109341621399, + "learning_rate": 0.003, + "loss": 3.955, + "step": 30653 + }, + { + "epoch": 0.30654, + "grad_norm": 1.0819790363311768, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 30654 + }, + { + "epoch": 0.30655, + "grad_norm": 0.9937855005264282, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 30655 + }, + { + "epoch": 0.30656, + "grad_norm": 0.891447126865387, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 30656 + }, + { + "epoch": 0.30657, + "grad_norm": 0.746498703956604, + "learning_rate": 0.003, + "loss": 3.957, + "step": 30657 + }, + { + "epoch": 0.30658, + "grad_norm": 0.8198862075805664, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 30658 + }, + { + "epoch": 0.30659, + "grad_norm": 0.8650001883506775, + "learning_rate": 0.003, + "loss": 4.005, + "step": 30659 + }, + { + "epoch": 0.3066, + "grad_norm": 0.7953513264656067, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 30660 + }, + { + "epoch": 0.30661, + "grad_norm": 0.9175745248794556, + "learning_rate": 0.003, + "loss": 3.991, + "step": 30661 + }, + { + "epoch": 0.30662, + "grad_norm": 1.091300368309021, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 30662 + }, + { + "epoch": 0.30663, + "grad_norm": 0.9431161284446716, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 30663 + }, + { + "epoch": 0.30664, + "grad_norm": 0.7637342214584351, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 30664 + }, + { + "epoch": 0.30665, + "grad_norm": 0.6082112193107605, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 30665 + }, + { + "epoch": 0.30666, + "grad_norm": 0.6159123182296753, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 30666 + }, + { + "epoch": 0.30667, + "grad_norm": 0.6559363007545471, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 30667 + }, + { + "epoch": 0.30668, + "grad_norm": 0.6518560647964478, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 30668 + }, + { + "epoch": 0.30669, + "grad_norm": 0.8356534242630005, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 30669 + }, + { + "epoch": 0.3067, + "grad_norm": 1.0654710531234741, + "learning_rate": 0.003, + "loss": 3.992, + "step": 30670 + }, + { + "epoch": 0.30671, + "grad_norm": 1.0638222694396973, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 30671 + }, + { + "epoch": 0.30672, + "grad_norm": 0.9449580907821655, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 30672 + }, + { + "epoch": 0.30673, + "grad_norm": 0.9074576497077942, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 30673 + }, + { + "epoch": 0.30674, + "grad_norm": 0.8361562490463257, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 30674 + }, + { + "epoch": 0.30675, + "grad_norm": 0.7773516178131104, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 30675 + }, + { + "epoch": 0.30676, + "grad_norm": 0.697791337966919, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 30676 + }, + { + "epoch": 0.30677, + "grad_norm": 0.77277010679245, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 30677 + }, + { + "epoch": 0.30678, + "grad_norm": 0.7813935279846191, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 30678 + }, + { + "epoch": 0.30679, + "grad_norm": 0.8213193416595459, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 30679 + }, + { + "epoch": 0.3068, + "grad_norm": 0.9167961478233337, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 30680 + }, + { + "epoch": 0.30681, + "grad_norm": 1.09382963180542, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 30681 + }, + { + "epoch": 0.30682, + "grad_norm": 0.9981308579444885, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 30682 + }, + { + "epoch": 0.30683, + "grad_norm": 1.0407847166061401, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 30683 + }, + { + "epoch": 0.30684, + "grad_norm": 0.9478374719619751, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 30684 + }, + { + "epoch": 0.30685, + "grad_norm": 0.9192802906036377, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 30685 + }, + { + "epoch": 0.30686, + "grad_norm": 1.141384482383728, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 30686 + }, + { + "epoch": 0.30687, + "grad_norm": 0.7456248998641968, + "learning_rate": 0.003, + "loss": 3.972, + "step": 30687 + }, + { + "epoch": 0.30688, + "grad_norm": 0.6289696097373962, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 30688 + }, + { + "epoch": 0.30689, + "grad_norm": 0.7373767495155334, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 30689 + }, + { + "epoch": 0.3069, + "grad_norm": 0.7935293912887573, + "learning_rate": 0.003, + "loss": 3.975, + "step": 30690 + }, + { + "epoch": 0.30691, + "grad_norm": 0.870015561580658, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 30691 + }, + { + "epoch": 0.30692, + "grad_norm": 0.9297496676445007, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 30692 + }, + { + "epoch": 0.30693, + "grad_norm": 1.1669538021087646, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 30693 + }, + { + "epoch": 0.30694, + "grad_norm": 0.9870584011077881, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 30694 + }, + { + "epoch": 0.30695, + "grad_norm": 0.972052812576294, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 30695 + }, + { + "epoch": 0.30696, + "grad_norm": 0.9249961972236633, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 30696 + }, + { + "epoch": 0.30697, + "grad_norm": 0.9585703015327454, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 30697 + }, + { + "epoch": 0.30698, + "grad_norm": 0.7980860471725464, + "learning_rate": 0.003, + "loss": 3.984, + "step": 30698 + }, + { + "epoch": 0.30699, + "grad_norm": 0.6694889664649963, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 30699 + }, + { + "epoch": 0.307, + "grad_norm": 0.721317708492279, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 30700 + }, + { + "epoch": 0.30701, + "grad_norm": 0.6855815649032593, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 30701 + }, + { + "epoch": 0.30702, + "grad_norm": 0.7577670812606812, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 30702 + }, + { + "epoch": 0.30703, + "grad_norm": 0.7654701471328735, + "learning_rate": 0.003, + "loss": 3.995, + "step": 30703 + }, + { + "epoch": 0.30704, + "grad_norm": 0.850576639175415, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 30704 + }, + { + "epoch": 0.30705, + "grad_norm": 0.9081757664680481, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 30705 + }, + { + "epoch": 0.30706, + "grad_norm": 0.8980525135993958, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 30706 + }, + { + "epoch": 0.30707, + "grad_norm": 1.0770753622055054, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 30707 + }, + { + "epoch": 0.30708, + "grad_norm": 1.1091724634170532, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 30708 + }, + { + "epoch": 0.30709, + "grad_norm": 0.9213112592697144, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 30709 + }, + { + "epoch": 0.3071, + "grad_norm": 0.8756240010261536, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 30710 + }, + { + "epoch": 0.30711, + "grad_norm": 0.9981350302696228, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 30711 + }, + { + "epoch": 0.30712, + "grad_norm": 1.078291654586792, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 30712 + }, + { + "epoch": 0.30713, + "grad_norm": 1.163339376449585, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 30713 + }, + { + "epoch": 0.30714, + "grad_norm": 0.8176140785217285, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 30714 + }, + { + "epoch": 0.30715, + "grad_norm": 0.8117172718048096, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 30715 + }, + { + "epoch": 0.30716, + "grad_norm": 0.941407322883606, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 30716 + }, + { + "epoch": 0.30717, + "grad_norm": 1.0049095153808594, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 30717 + }, + { + "epoch": 0.30718, + "grad_norm": 0.8918133974075317, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 30718 + }, + { + "epoch": 0.30719, + "grad_norm": 0.8062427639961243, + "learning_rate": 0.003, + "loss": 3.987, + "step": 30719 + }, + { + "epoch": 0.3072, + "grad_norm": 0.6939215064048767, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 30720 + }, + { + "epoch": 0.30721, + "grad_norm": 0.8913199305534363, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 30721 + }, + { + "epoch": 0.30722, + "grad_norm": 0.9262449741363525, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 30722 + }, + { + "epoch": 0.30723, + "grad_norm": 1.1616439819335938, + "learning_rate": 0.003, + "loss": 3.967, + "step": 30723 + }, + { + "epoch": 0.30724, + "grad_norm": 0.9698355793952942, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 30724 + }, + { + "epoch": 0.30725, + "grad_norm": 0.8564360737800598, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 30725 + }, + { + "epoch": 0.30726, + "grad_norm": 0.7706262469291687, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 30726 + }, + { + "epoch": 0.30727, + "grad_norm": 0.8154146671295166, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 30727 + }, + { + "epoch": 0.30728, + "grad_norm": 0.8263653516769409, + "learning_rate": 0.003, + "loss": 3.97, + "step": 30728 + }, + { + "epoch": 0.30729, + "grad_norm": 0.8068422079086304, + "learning_rate": 0.003, + "loss": 4.022, + "step": 30729 + }, + { + "epoch": 0.3073, + "grad_norm": 0.8868446946144104, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 30730 + }, + { + "epoch": 0.30731, + "grad_norm": 0.9689352512359619, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 30731 + }, + { + "epoch": 0.30732, + "grad_norm": 0.9709058403968811, + "learning_rate": 0.003, + "loss": 3.9292, + "step": 30732 + }, + { + "epoch": 0.30733, + "grad_norm": 1.0263891220092773, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 30733 + }, + { + "epoch": 0.30734, + "grad_norm": 0.9162255525588989, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 30734 + }, + { + "epoch": 0.30735, + "grad_norm": 0.8312210440635681, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 30735 + }, + { + "epoch": 0.30736, + "grad_norm": 0.709874153137207, + "learning_rate": 0.003, + "loss": 4.004, + "step": 30736 + }, + { + "epoch": 0.30737, + "grad_norm": 0.7518811821937561, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 30737 + }, + { + "epoch": 0.30738, + "grad_norm": 0.6948967576026917, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 30738 + }, + { + "epoch": 0.30739, + "grad_norm": 0.6590483784675598, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 30739 + }, + { + "epoch": 0.3074, + "grad_norm": 0.7028836011886597, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 30740 + }, + { + "epoch": 0.30741, + "grad_norm": 0.7641788721084595, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 30741 + }, + { + "epoch": 0.30742, + "grad_norm": 0.9370028376579285, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 30742 + }, + { + "epoch": 0.30743, + "grad_norm": 0.8916642665863037, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 30743 + }, + { + "epoch": 0.30744, + "grad_norm": 0.9345336556434631, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 30744 + }, + { + "epoch": 0.30745, + "grad_norm": 1.3236756324768066, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 30745 + }, + { + "epoch": 0.30746, + "grad_norm": 1.007238745689392, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 30746 + }, + { + "epoch": 0.30747, + "grad_norm": 0.8696395754814148, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 30747 + }, + { + "epoch": 0.30748, + "grad_norm": 0.8959425687789917, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 30748 + }, + { + "epoch": 0.30749, + "grad_norm": 0.8492132425308228, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 30749 + }, + { + "epoch": 0.3075, + "grad_norm": 0.7420181632041931, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 30750 + }, + { + "epoch": 0.30751, + "grad_norm": 0.665791928768158, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 30751 + }, + { + "epoch": 0.30752, + "grad_norm": 0.6952983736991882, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 30752 + }, + { + "epoch": 0.30753, + "grad_norm": 0.7583016753196716, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 30753 + }, + { + "epoch": 0.30754, + "grad_norm": 0.7442736625671387, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 30754 + }, + { + "epoch": 0.30755, + "grad_norm": 0.6917513608932495, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 30755 + }, + { + "epoch": 0.30756, + "grad_norm": 0.7232774496078491, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 30756 + }, + { + "epoch": 0.30757, + "grad_norm": 0.8982821702957153, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 30757 + }, + { + "epoch": 0.30758, + "grad_norm": 1.1486623287200928, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 30758 + }, + { + "epoch": 0.30759, + "grad_norm": 0.9366014003753662, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 30759 + }, + { + "epoch": 0.3076, + "grad_norm": 0.8373289108276367, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 30760 + }, + { + "epoch": 0.30761, + "grad_norm": 0.8777595162391663, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 30761 + }, + { + "epoch": 0.30762, + "grad_norm": 0.8704143166542053, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 30762 + }, + { + "epoch": 0.30763, + "grad_norm": 0.6569860577583313, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 30763 + }, + { + "epoch": 0.30764, + "grad_norm": 0.6012580394744873, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 30764 + }, + { + "epoch": 0.30765, + "grad_norm": 0.5986373424530029, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 30765 + }, + { + "epoch": 0.30766, + "grad_norm": 0.5636370182037354, + "learning_rate": 0.003, + "loss": 3.967, + "step": 30766 + }, + { + "epoch": 0.30767, + "grad_norm": 0.5599316358566284, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 30767 + }, + { + "epoch": 0.30768, + "grad_norm": 0.6067478656768799, + "learning_rate": 0.003, + "loss": 3.999, + "step": 30768 + }, + { + "epoch": 0.30769, + "grad_norm": 0.7281287908554077, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 30769 + }, + { + "epoch": 0.3077, + "grad_norm": 0.835479736328125, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 30770 + }, + { + "epoch": 0.30771, + "grad_norm": 1.0763078927993774, + "learning_rate": 0.003, + "loss": 3.953, + "step": 30771 + }, + { + "epoch": 0.30772, + "grad_norm": 1.0973634719848633, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 30772 + }, + { + "epoch": 0.30773, + "grad_norm": 0.8174333572387695, + "learning_rate": 0.003, + "loss": 3.973, + "step": 30773 + }, + { + "epoch": 0.30774, + "grad_norm": 0.8216995000839233, + "learning_rate": 0.003, + "loss": 3.988, + "step": 30774 + }, + { + "epoch": 0.30775, + "grad_norm": 0.7696967720985413, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 30775 + }, + { + "epoch": 0.30776, + "grad_norm": 0.8547626733779907, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 30776 + }, + { + "epoch": 0.30777, + "grad_norm": 1.138253927230835, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 30777 + }, + { + "epoch": 0.30778, + "grad_norm": 0.9814075231552124, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 30778 + }, + { + "epoch": 0.30779, + "grad_norm": 0.9518291354179382, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 30779 + }, + { + "epoch": 0.3078, + "grad_norm": 0.9754531383514404, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 30780 + }, + { + "epoch": 0.30781, + "grad_norm": 0.9863748550415039, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 30781 + }, + { + "epoch": 0.30782, + "grad_norm": 1.0354273319244385, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 30782 + }, + { + "epoch": 0.30783, + "grad_norm": 0.9466879963874817, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 30783 + }, + { + "epoch": 0.30784, + "grad_norm": 0.9126093983650208, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 30784 + }, + { + "epoch": 0.30785, + "grad_norm": 1.1524488925933838, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 30785 + }, + { + "epoch": 0.30786, + "grad_norm": 1.1938556432724, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 30786 + }, + { + "epoch": 0.30787, + "grad_norm": 0.9225050806999207, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 30787 + }, + { + "epoch": 0.30788, + "grad_norm": 0.8980876803398132, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 30788 + }, + { + "epoch": 0.30789, + "grad_norm": 0.922467052936554, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 30789 + }, + { + "epoch": 0.3079, + "grad_norm": 0.9036548733711243, + "learning_rate": 0.003, + "loss": 3.984, + "step": 30790 + }, + { + "epoch": 0.30791, + "grad_norm": 0.8449410796165466, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 30791 + }, + { + "epoch": 0.30792, + "grad_norm": 1.024229645729065, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 30792 + }, + { + "epoch": 0.30793, + "grad_norm": 1.0609626770019531, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 30793 + }, + { + "epoch": 0.30794, + "grad_norm": 0.9126906394958496, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 30794 + }, + { + "epoch": 0.30795, + "grad_norm": 0.9202067255973816, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 30795 + }, + { + "epoch": 0.30796, + "grad_norm": 1.041786789894104, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 30796 + }, + { + "epoch": 0.30797, + "grad_norm": 1.29253351688385, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 30797 + }, + { + "epoch": 0.30798, + "grad_norm": 0.7027266025543213, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 30798 + }, + { + "epoch": 0.30799, + "grad_norm": 0.661356508731842, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 30799 + }, + { + "epoch": 0.308, + "grad_norm": 0.8871188759803772, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 30800 + }, + { + "epoch": 0.30801, + "grad_norm": 0.9151998162269592, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 30801 + }, + { + "epoch": 0.30802, + "grad_norm": 1.0257346630096436, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 30802 + }, + { + "epoch": 0.30803, + "grad_norm": 1.0671848058700562, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 30803 + }, + { + "epoch": 0.30804, + "grad_norm": 0.7683122158050537, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 30804 + }, + { + "epoch": 0.30805, + "grad_norm": 0.7095206379890442, + "learning_rate": 0.003, + "loss": 3.978, + "step": 30805 + }, + { + "epoch": 0.30806, + "grad_norm": 0.8385022878646851, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 30806 + }, + { + "epoch": 0.30807, + "grad_norm": 1.0319606065750122, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 30807 + }, + { + "epoch": 0.30808, + "grad_norm": 1.0124037265777588, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 30808 + }, + { + "epoch": 0.30809, + "grad_norm": 0.9301835894584656, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 30809 + }, + { + "epoch": 0.3081, + "grad_norm": 0.9929497838020325, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 30810 + }, + { + "epoch": 0.30811, + "grad_norm": 1.0492149591445923, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 30811 + }, + { + "epoch": 0.30812, + "grad_norm": 1.142856478691101, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 30812 + }, + { + "epoch": 0.30813, + "grad_norm": 0.7310068011283875, + "learning_rate": 0.003, + "loss": 3.979, + "step": 30813 + }, + { + "epoch": 0.30814, + "grad_norm": 0.6632123589515686, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 30814 + }, + { + "epoch": 0.30815, + "grad_norm": 0.8477546572685242, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 30815 + }, + { + "epoch": 0.30816, + "grad_norm": 0.9337522387504578, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 30816 + }, + { + "epoch": 0.30817, + "grad_norm": 0.9468395113945007, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 30817 + }, + { + "epoch": 0.30818, + "grad_norm": 0.7905578017234802, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 30818 + }, + { + "epoch": 0.30819, + "grad_norm": 0.737535297870636, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 30819 + }, + { + "epoch": 0.3082, + "grad_norm": 0.8040305376052856, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 30820 + }, + { + "epoch": 0.30821, + "grad_norm": 0.8418785333633423, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 30821 + }, + { + "epoch": 0.30822, + "grad_norm": 0.9753386378288269, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 30822 + }, + { + "epoch": 0.30823, + "grad_norm": 1.1201002597808838, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 30823 + }, + { + "epoch": 0.30824, + "grad_norm": 0.9152430295944214, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 30824 + }, + { + "epoch": 0.30825, + "grad_norm": 0.8478184938430786, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 30825 + }, + { + "epoch": 0.30826, + "grad_norm": 0.8158889412879944, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 30826 + }, + { + "epoch": 0.30827, + "grad_norm": 0.8474761843681335, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 30827 + }, + { + "epoch": 0.30828, + "grad_norm": 0.8060882091522217, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 30828 + }, + { + "epoch": 0.30829, + "grad_norm": 0.7927574515342712, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 30829 + }, + { + "epoch": 0.3083, + "grad_norm": 0.8730419874191284, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 30830 + }, + { + "epoch": 0.30831, + "grad_norm": 0.9823596477508545, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 30831 + }, + { + "epoch": 0.30832, + "grad_norm": 1.02117919921875, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 30832 + }, + { + "epoch": 0.30833, + "grad_norm": 0.8114142417907715, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 30833 + }, + { + "epoch": 0.30834, + "grad_norm": 0.8020756244659424, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 30834 + }, + { + "epoch": 0.30835, + "grad_norm": 0.9440411329269409, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 30835 + }, + { + "epoch": 0.30836, + "grad_norm": 0.9526855945587158, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 30836 + }, + { + "epoch": 0.30837, + "grad_norm": 0.8930078148841858, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 30837 + }, + { + "epoch": 0.30838, + "grad_norm": 1.0869745016098022, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 30838 + }, + { + "epoch": 0.30839, + "grad_norm": 1.1024293899536133, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 30839 + }, + { + "epoch": 0.3084, + "grad_norm": 0.857626736164093, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 30840 + }, + { + "epoch": 0.30841, + "grad_norm": 0.7289055585861206, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 30841 + }, + { + "epoch": 0.30842, + "grad_norm": 0.6768589615821838, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 30842 + }, + { + "epoch": 0.30843, + "grad_norm": 0.6834684610366821, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 30843 + }, + { + "epoch": 0.30844, + "grad_norm": 0.8336805105209351, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 30844 + }, + { + "epoch": 0.30845, + "grad_norm": 0.8948435187339783, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 30845 + }, + { + "epoch": 0.30846, + "grad_norm": 0.9395321607589722, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 30846 + }, + { + "epoch": 0.30847, + "grad_norm": 1.0715928077697754, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 30847 + }, + { + "epoch": 0.30848, + "grad_norm": 0.9883683323860168, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 30848 + }, + { + "epoch": 0.30849, + "grad_norm": 0.8295038938522339, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 30849 + }, + { + "epoch": 0.3085, + "grad_norm": 0.7715690732002258, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 30850 + }, + { + "epoch": 0.30851, + "grad_norm": 0.7741749286651611, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 30851 + }, + { + "epoch": 0.30852, + "grad_norm": 0.7417683005332947, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 30852 + }, + { + "epoch": 0.30853, + "grad_norm": 0.8107874393463135, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 30853 + }, + { + "epoch": 0.30854, + "grad_norm": 0.8936319351196289, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 30854 + }, + { + "epoch": 0.30855, + "grad_norm": 0.9610341787338257, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 30855 + }, + { + "epoch": 0.30856, + "grad_norm": 1.4761669635772705, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 30856 + }, + { + "epoch": 0.30857, + "grad_norm": 0.731768012046814, + "learning_rate": 0.003, + "loss": 4.002, + "step": 30857 + }, + { + "epoch": 0.30858, + "grad_norm": 0.6133517026901245, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 30858 + }, + { + "epoch": 0.30859, + "grad_norm": 0.778045117855072, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 30859 + }, + { + "epoch": 0.3086, + "grad_norm": 0.6594052314758301, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 30860 + }, + { + "epoch": 0.30861, + "grad_norm": 0.6072326898574829, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 30861 + }, + { + "epoch": 0.30862, + "grad_norm": 0.7369550466537476, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 30862 + }, + { + "epoch": 0.30863, + "grad_norm": 0.6927176117897034, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 30863 + }, + { + "epoch": 0.30864, + "grad_norm": 0.6587769389152527, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 30864 + }, + { + "epoch": 0.30865, + "grad_norm": 0.7761625051498413, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 30865 + }, + { + "epoch": 0.30866, + "grad_norm": 0.9280504584312439, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 30866 + }, + { + "epoch": 0.30867, + "grad_norm": 1.3131680488586426, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 30867 + }, + { + "epoch": 0.30868, + "grad_norm": 0.832633376121521, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 30868 + }, + { + "epoch": 0.30869, + "grad_norm": 0.7639395594596863, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 30869 + }, + { + "epoch": 0.3087, + "grad_norm": 0.7404260039329529, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 30870 + }, + { + "epoch": 0.30871, + "grad_norm": 0.7073952555656433, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 30871 + }, + { + "epoch": 0.30872, + "grad_norm": 0.7493676543235779, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 30872 + }, + { + "epoch": 0.30873, + "grad_norm": 0.8757566213607788, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 30873 + }, + { + "epoch": 0.30874, + "grad_norm": 0.9659149646759033, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 30874 + }, + { + "epoch": 0.30875, + "grad_norm": 1.1489213705062866, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 30875 + }, + { + "epoch": 0.30876, + "grad_norm": 0.9251287579536438, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 30876 + }, + { + "epoch": 0.30877, + "grad_norm": 0.8200145363807678, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 30877 + }, + { + "epoch": 0.30878, + "grad_norm": 0.7325860261917114, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 30878 + }, + { + "epoch": 0.30879, + "grad_norm": 0.7247792482376099, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 30879 + }, + { + "epoch": 0.3088, + "grad_norm": 0.7962667346000671, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 30880 + }, + { + "epoch": 0.30881, + "grad_norm": 0.7682673931121826, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 30881 + }, + { + "epoch": 0.30882, + "grad_norm": 0.7465470433235168, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 30882 + }, + { + "epoch": 0.30883, + "grad_norm": 0.9776432514190674, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 30883 + }, + { + "epoch": 0.30884, + "grad_norm": 1.087443232536316, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 30884 + }, + { + "epoch": 0.30885, + "grad_norm": 0.9176280498504639, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 30885 + }, + { + "epoch": 0.30886, + "grad_norm": 0.8743244409561157, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 30886 + }, + { + "epoch": 0.30887, + "grad_norm": 0.730586588382721, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 30887 + }, + { + "epoch": 0.30888, + "grad_norm": 0.6315854787826538, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 30888 + }, + { + "epoch": 0.30889, + "grad_norm": 0.621346116065979, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 30889 + }, + { + "epoch": 0.3089, + "grad_norm": 0.6593024730682373, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 30890 + }, + { + "epoch": 0.30891, + "grad_norm": 0.6536547541618347, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 30891 + }, + { + "epoch": 0.30892, + "grad_norm": 0.7169095277786255, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 30892 + }, + { + "epoch": 0.30893, + "grad_norm": 1.0016939640045166, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 30893 + }, + { + "epoch": 0.30894, + "grad_norm": 1.2001861333847046, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 30894 + }, + { + "epoch": 0.30895, + "grad_norm": 0.8227499127388, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 30895 + }, + { + "epoch": 0.30896, + "grad_norm": 0.9125561714172363, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 30896 + }, + { + "epoch": 0.30897, + "grad_norm": 0.8758599758148193, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 30897 + }, + { + "epoch": 0.30898, + "grad_norm": 0.8722800016403198, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 30898 + }, + { + "epoch": 0.30899, + "grad_norm": 0.9590011239051819, + "learning_rate": 0.003, + "loss": 3.9389, + "step": 30899 + }, + { + "epoch": 0.309, + "grad_norm": 0.9985750913619995, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 30900 + }, + { + "epoch": 0.30901, + "grad_norm": 1.3745064735412598, + "learning_rate": 0.003, + "loss": 4.008, + "step": 30901 + }, + { + "epoch": 0.30902, + "grad_norm": 0.7419667840003967, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 30902 + }, + { + "epoch": 0.30903, + "grad_norm": 0.6378922462463379, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 30903 + }, + { + "epoch": 0.30904, + "grad_norm": 0.7854878306388855, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 30904 + }, + { + "epoch": 0.30905, + "grad_norm": 0.876837968826294, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 30905 + }, + { + "epoch": 0.30906, + "grad_norm": 1.0031284093856812, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 30906 + }, + { + "epoch": 0.30907, + "grad_norm": 1.3510195016860962, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 30907 + }, + { + "epoch": 0.30908, + "grad_norm": 0.851047158241272, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 30908 + }, + { + "epoch": 0.30909, + "grad_norm": 0.7446119785308838, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 30909 + }, + { + "epoch": 0.3091, + "grad_norm": 0.9427217245101929, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 30910 + }, + { + "epoch": 0.30911, + "grad_norm": 1.3650840520858765, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 30911 + }, + { + "epoch": 0.30912, + "grad_norm": 0.7256133556365967, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 30912 + }, + { + "epoch": 0.30913, + "grad_norm": 0.7032670378684998, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 30913 + }, + { + "epoch": 0.30914, + "grad_norm": 0.7236923575401306, + "learning_rate": 0.003, + "loss": 3.9435, + "step": 30914 + }, + { + "epoch": 0.30915, + "grad_norm": 0.7407827377319336, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 30915 + }, + { + "epoch": 0.30916, + "grad_norm": 0.9022109508514404, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 30916 + }, + { + "epoch": 0.30917, + "grad_norm": 1.0237579345703125, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 30917 + }, + { + "epoch": 0.30918, + "grad_norm": 1.1757525205612183, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 30918 + }, + { + "epoch": 0.30919, + "grad_norm": 0.8383086323738098, + "learning_rate": 0.003, + "loss": 3.987, + "step": 30919 + }, + { + "epoch": 0.3092, + "grad_norm": 0.7692552804946899, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 30920 + }, + { + "epoch": 0.30921, + "grad_norm": 0.7720386385917664, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 30921 + }, + { + "epoch": 0.30922, + "grad_norm": 0.8688241839408875, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 30922 + }, + { + "epoch": 0.30923, + "grad_norm": 0.9057418704032898, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 30923 + }, + { + "epoch": 0.30924, + "grad_norm": 0.9113157391548157, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 30924 + }, + { + "epoch": 0.30925, + "grad_norm": 1.0878233909606934, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 30925 + }, + { + "epoch": 0.30926, + "grad_norm": 1.149067759513855, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 30926 + }, + { + "epoch": 0.30927, + "grad_norm": 0.7788525819778442, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 30927 + }, + { + "epoch": 0.30928, + "grad_norm": 0.746897280216217, + "learning_rate": 0.003, + "loss": 3.981, + "step": 30928 + }, + { + "epoch": 0.30929, + "grad_norm": 0.8232402205467224, + "learning_rate": 0.003, + "loss": 3.944, + "step": 30929 + }, + { + "epoch": 0.3093, + "grad_norm": 1.0002985000610352, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 30930 + }, + { + "epoch": 0.30931, + "grad_norm": 1.049239993095398, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 30931 + }, + { + "epoch": 0.30932, + "grad_norm": 0.9537498354911804, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 30932 + }, + { + "epoch": 0.30933, + "grad_norm": 0.9364720582962036, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 30933 + }, + { + "epoch": 0.30934, + "grad_norm": 0.8069536685943604, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 30934 + }, + { + "epoch": 0.30935, + "grad_norm": 0.7318139672279358, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 30935 + }, + { + "epoch": 0.30936, + "grad_norm": 0.7163798809051514, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 30936 + }, + { + "epoch": 0.30937, + "grad_norm": 0.8208103775978088, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 30937 + }, + { + "epoch": 0.30938, + "grad_norm": 0.8081580996513367, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 30938 + }, + { + "epoch": 0.30939, + "grad_norm": 0.8887249827384949, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 30939 + }, + { + "epoch": 0.3094, + "grad_norm": 1.1092829704284668, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 30940 + }, + { + "epoch": 0.30941, + "grad_norm": 1.017830491065979, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 30941 + }, + { + "epoch": 0.30942, + "grad_norm": 1.0504926443099976, + "learning_rate": 0.003, + "loss": 3.975, + "step": 30942 + }, + { + "epoch": 0.30943, + "grad_norm": 0.9657722115516663, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 30943 + }, + { + "epoch": 0.30944, + "grad_norm": 0.8380178809165955, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 30944 + }, + { + "epoch": 0.30945, + "grad_norm": 0.8412208557128906, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 30945 + }, + { + "epoch": 0.30946, + "grad_norm": 0.8611930012702942, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 30946 + }, + { + "epoch": 0.30947, + "grad_norm": 0.7942565083503723, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 30947 + }, + { + "epoch": 0.30948, + "grad_norm": 0.7671312093734741, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 30948 + }, + { + "epoch": 0.30949, + "grad_norm": 0.8001713156700134, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 30949 + }, + { + "epoch": 0.3095, + "grad_norm": 0.7979334592819214, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 30950 + }, + { + "epoch": 0.30951, + "grad_norm": 0.7309614419937134, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 30951 + }, + { + "epoch": 0.30952, + "grad_norm": 0.6950778961181641, + "learning_rate": 0.003, + "loss": 3.962, + "step": 30952 + }, + { + "epoch": 0.30953, + "grad_norm": 0.735960841178894, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 30953 + }, + { + "epoch": 0.30954, + "grad_norm": 0.8212732076644897, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 30954 + }, + { + "epoch": 0.30955, + "grad_norm": 1.2007713317871094, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 30955 + }, + { + "epoch": 0.30956, + "grad_norm": 1.1494503021240234, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 30956 + }, + { + "epoch": 0.30957, + "grad_norm": 0.7825203537940979, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 30957 + }, + { + "epoch": 0.30958, + "grad_norm": 0.7343149781227112, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 30958 + }, + { + "epoch": 0.30959, + "grad_norm": 0.738079845905304, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 30959 + }, + { + "epoch": 0.3096, + "grad_norm": 0.8391587138175964, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 30960 + }, + { + "epoch": 0.30961, + "grad_norm": 0.8329426050186157, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 30961 + }, + { + "epoch": 0.30962, + "grad_norm": 1.0566493272781372, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 30962 + }, + { + "epoch": 0.30963, + "grad_norm": 1.2389732599258423, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 30963 + }, + { + "epoch": 0.30964, + "grad_norm": 0.9868148565292358, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 30964 + }, + { + "epoch": 0.30965, + "grad_norm": 1.1375012397766113, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 30965 + }, + { + "epoch": 0.30966, + "grad_norm": 0.8549085855484009, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 30966 + }, + { + "epoch": 0.30967, + "grad_norm": 0.7379135489463806, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 30967 + }, + { + "epoch": 0.30968, + "grad_norm": 0.8331286311149597, + "learning_rate": 0.003, + "loss": 3.963, + "step": 30968 + }, + { + "epoch": 0.30969, + "grad_norm": 0.899120569229126, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 30969 + }, + { + "epoch": 0.3097, + "grad_norm": 0.9566233158111572, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 30970 + }, + { + "epoch": 0.30971, + "grad_norm": 0.9181606769561768, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 30971 + }, + { + "epoch": 0.30972, + "grad_norm": 0.8362905979156494, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 30972 + }, + { + "epoch": 0.30973, + "grad_norm": 0.8711179494857788, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 30973 + }, + { + "epoch": 0.30974, + "grad_norm": 0.9471859335899353, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 30974 + }, + { + "epoch": 0.30975, + "grad_norm": 1.1806086301803589, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 30975 + }, + { + "epoch": 0.30976, + "grad_norm": 0.8900488018989563, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 30976 + }, + { + "epoch": 0.30977, + "grad_norm": 0.8443350791931152, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 30977 + }, + { + "epoch": 0.30978, + "grad_norm": 0.8960421085357666, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 30978 + }, + { + "epoch": 0.30979, + "grad_norm": 1.0800766944885254, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 30979 + }, + { + "epoch": 0.3098, + "grad_norm": 1.0280297994613647, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 30980 + }, + { + "epoch": 0.30981, + "grad_norm": 0.864987850189209, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 30981 + }, + { + "epoch": 0.30982, + "grad_norm": 0.8029942512512207, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 30982 + }, + { + "epoch": 0.30983, + "grad_norm": 0.766416609287262, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 30983 + }, + { + "epoch": 0.30984, + "grad_norm": 0.7081720232963562, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 30984 + }, + { + "epoch": 0.30985, + "grad_norm": 0.7827467322349548, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 30985 + }, + { + "epoch": 0.30986, + "grad_norm": 0.7845280170440674, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 30986 + }, + { + "epoch": 0.30987, + "grad_norm": 0.8895677328109741, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 30987 + }, + { + "epoch": 0.30988, + "grad_norm": 0.9675378203392029, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 30988 + }, + { + "epoch": 0.30989, + "grad_norm": 0.9929342269897461, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 30989 + }, + { + "epoch": 0.3099, + "grad_norm": 1.1774400472640991, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 30990 + }, + { + "epoch": 0.30991, + "grad_norm": 0.7923581004142761, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 30991 + }, + { + "epoch": 0.30992, + "grad_norm": 0.7073946595191956, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 30992 + }, + { + "epoch": 0.30993, + "grad_norm": 0.710835874080658, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 30993 + }, + { + "epoch": 0.30994, + "grad_norm": 0.649730920791626, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 30994 + }, + { + "epoch": 0.30995, + "grad_norm": 0.6385684013366699, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 30995 + }, + { + "epoch": 0.30996, + "grad_norm": 0.7375938296318054, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 30996 + }, + { + "epoch": 0.30997, + "grad_norm": 0.8679918050765991, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 30997 + }, + { + "epoch": 0.30998, + "grad_norm": 1.0643709897994995, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 30998 + }, + { + "epoch": 0.30999, + "grad_norm": 0.9517081379890442, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 30999 + }, + { + "epoch": 0.31, + "grad_norm": 1.0096635818481445, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 31000 + }, + { + "epoch": 0.31001, + "grad_norm": 1.1224404573440552, + "learning_rate": 0.003, + "loss": 3.983, + "step": 31001 + }, + { + "epoch": 0.31002, + "grad_norm": 0.8119789361953735, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 31002 + }, + { + "epoch": 0.31003, + "grad_norm": 0.700625479221344, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 31003 + }, + { + "epoch": 0.31004, + "grad_norm": 0.7133662104606628, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 31004 + }, + { + "epoch": 0.31005, + "grad_norm": 0.9758654832839966, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 31005 + }, + { + "epoch": 0.31006, + "grad_norm": 0.9434835910797119, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 31006 + }, + { + "epoch": 0.31007, + "grad_norm": 0.7757193446159363, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 31007 + }, + { + "epoch": 0.31008, + "grad_norm": 0.7861281633377075, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 31008 + }, + { + "epoch": 0.31009, + "grad_norm": 0.8851728439331055, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 31009 + }, + { + "epoch": 0.3101, + "grad_norm": 0.8107077479362488, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 31010 + }, + { + "epoch": 0.31011, + "grad_norm": 0.9422912001609802, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 31011 + }, + { + "epoch": 0.31012, + "grad_norm": 1.193611979484558, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 31012 + }, + { + "epoch": 0.31013, + "grad_norm": 0.8879119753837585, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 31013 + }, + { + "epoch": 0.31014, + "grad_norm": 0.7490536570549011, + "learning_rate": 0.003, + "loss": 3.989, + "step": 31014 + }, + { + "epoch": 0.31015, + "grad_norm": 0.7697620987892151, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 31015 + }, + { + "epoch": 0.31016, + "grad_norm": 0.747331440448761, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 31016 + }, + { + "epoch": 0.31017, + "grad_norm": 0.8451557755470276, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 31017 + }, + { + "epoch": 0.31018, + "grad_norm": 1.0530297756195068, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 31018 + }, + { + "epoch": 0.31019, + "grad_norm": 0.937260627746582, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 31019 + }, + { + "epoch": 0.3102, + "grad_norm": 0.7762146592140198, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 31020 + }, + { + "epoch": 0.31021, + "grad_norm": 0.692551851272583, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 31021 + }, + { + "epoch": 0.31022, + "grad_norm": 0.6892310380935669, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 31022 + }, + { + "epoch": 0.31023, + "grad_norm": 0.7278590202331543, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 31023 + }, + { + "epoch": 0.31024, + "grad_norm": 0.8904547095298767, + "learning_rate": 0.003, + "loss": 3.977, + "step": 31024 + }, + { + "epoch": 0.31025, + "grad_norm": 1.147729516029358, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 31025 + }, + { + "epoch": 0.31026, + "grad_norm": 0.9791727066040039, + "learning_rate": 0.003, + "loss": 3.988, + "step": 31026 + }, + { + "epoch": 0.31027, + "grad_norm": 0.9623188972473145, + "learning_rate": 0.003, + "loss": 3.996, + "step": 31027 + }, + { + "epoch": 0.31028, + "grad_norm": 0.9856346845626831, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 31028 + }, + { + "epoch": 0.31029, + "grad_norm": 1.1689181327819824, + "learning_rate": 0.003, + "loss": 3.98, + "step": 31029 + }, + { + "epoch": 0.3103, + "grad_norm": 0.8035436272621155, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 31030 + }, + { + "epoch": 0.31031, + "grad_norm": 0.7386220097541809, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 31031 + }, + { + "epoch": 0.31032, + "grad_norm": 0.742174506187439, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 31032 + }, + { + "epoch": 0.31033, + "grad_norm": 0.7251983284950256, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 31033 + }, + { + "epoch": 0.31034, + "grad_norm": 0.6614605188369751, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 31034 + }, + { + "epoch": 0.31035, + "grad_norm": 0.5905389785766602, + "learning_rate": 0.003, + "loss": 3.984, + "step": 31035 + }, + { + "epoch": 0.31036, + "grad_norm": 0.6688835024833679, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 31036 + }, + { + "epoch": 0.31037, + "grad_norm": 0.7823680639266968, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 31037 + }, + { + "epoch": 0.31038, + "grad_norm": 0.8825580477714539, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 31038 + }, + { + "epoch": 0.31039, + "grad_norm": 0.8459956645965576, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 31039 + }, + { + "epoch": 0.3104, + "grad_norm": 0.8116077184677124, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 31040 + }, + { + "epoch": 0.31041, + "grad_norm": 0.8449925184249878, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 31041 + }, + { + "epoch": 0.31042, + "grad_norm": 0.9948757886886597, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 31042 + }, + { + "epoch": 0.31043, + "grad_norm": 1.3889992237091064, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 31043 + }, + { + "epoch": 0.31044, + "grad_norm": 0.8011574149131775, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 31044 + }, + { + "epoch": 0.31045, + "grad_norm": 0.7368257641792297, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 31045 + }, + { + "epoch": 0.31046, + "grad_norm": 0.7752619981765747, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 31046 + }, + { + "epoch": 0.31047, + "grad_norm": 0.9642040729522705, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 31047 + }, + { + "epoch": 0.31048, + "grad_norm": 1.1010830402374268, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 31048 + }, + { + "epoch": 0.31049, + "grad_norm": 0.9506378173828125, + "learning_rate": 0.003, + "loss": 3.983, + "step": 31049 + }, + { + "epoch": 0.3105, + "grad_norm": 0.9478399753570557, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 31050 + }, + { + "epoch": 0.31051, + "grad_norm": 1.1100181341171265, + "learning_rate": 0.003, + "loss": 4.006, + "step": 31051 + }, + { + "epoch": 0.31052, + "grad_norm": 0.8177278637886047, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 31052 + }, + { + "epoch": 0.31053, + "grad_norm": 0.6501681804656982, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 31053 + }, + { + "epoch": 0.31054, + "grad_norm": 0.6588812470436096, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 31054 + }, + { + "epoch": 0.31055, + "grad_norm": 0.6845262050628662, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 31055 + }, + { + "epoch": 0.31056, + "grad_norm": 0.7345082759857178, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 31056 + }, + { + "epoch": 0.31057, + "grad_norm": 0.7383362054824829, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 31057 + }, + { + "epoch": 0.31058, + "grad_norm": 0.7008395195007324, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 31058 + }, + { + "epoch": 0.31059, + "grad_norm": 0.8314328789710999, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 31059 + }, + { + "epoch": 0.3106, + "grad_norm": 1.1818510293960571, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 31060 + }, + { + "epoch": 0.31061, + "grad_norm": 1.131486415863037, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 31061 + }, + { + "epoch": 0.31062, + "grad_norm": 0.817224383354187, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 31062 + }, + { + "epoch": 0.31063, + "grad_norm": 0.6096135377883911, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 31063 + }, + { + "epoch": 0.31064, + "grad_norm": 0.6916061043739319, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 31064 + }, + { + "epoch": 0.31065, + "grad_norm": 0.7657514810562134, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 31065 + }, + { + "epoch": 0.31066, + "grad_norm": 1.0308797359466553, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 31066 + }, + { + "epoch": 0.31067, + "grad_norm": 1.0070970058441162, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 31067 + }, + { + "epoch": 0.31068, + "grad_norm": 1.0685611963272095, + "learning_rate": 0.003, + "loss": 3.975, + "step": 31068 + }, + { + "epoch": 0.31069, + "grad_norm": 1.0918611288070679, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 31069 + }, + { + "epoch": 0.3107, + "grad_norm": 0.9739444255828857, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 31070 + }, + { + "epoch": 0.31071, + "grad_norm": 1.017590880393982, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 31071 + }, + { + "epoch": 0.31072, + "grad_norm": 1.1298353672027588, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 31072 + }, + { + "epoch": 0.31073, + "grad_norm": 1.153358817100525, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 31073 + }, + { + "epoch": 0.31074, + "grad_norm": 0.7779840230941772, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 31074 + }, + { + "epoch": 0.31075, + "grad_norm": 0.7930902242660522, + "learning_rate": 0.003, + "loss": 3.987, + "step": 31075 + }, + { + "epoch": 0.31076, + "grad_norm": 0.8280520439147949, + "learning_rate": 0.003, + "loss": 4.0, + "step": 31076 + }, + { + "epoch": 0.31077, + "grad_norm": 0.9038435220718384, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 31077 + }, + { + "epoch": 0.31078, + "grad_norm": 0.8968542218208313, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 31078 + }, + { + "epoch": 0.31079, + "grad_norm": 0.9571460485458374, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 31079 + }, + { + "epoch": 0.3108, + "grad_norm": 1.0983102321624756, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 31080 + }, + { + "epoch": 0.31081, + "grad_norm": 0.8243315815925598, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 31081 + }, + { + "epoch": 0.31082, + "grad_norm": 0.826403796672821, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 31082 + }, + { + "epoch": 0.31083, + "grad_norm": 1.1011091470718384, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 31083 + }, + { + "epoch": 0.31084, + "grad_norm": 1.1626007556915283, + "learning_rate": 0.003, + "loss": 3.968, + "step": 31084 + }, + { + "epoch": 0.31085, + "grad_norm": 0.6295737624168396, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 31085 + }, + { + "epoch": 0.31086, + "grad_norm": 0.5726386904716492, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 31086 + }, + { + "epoch": 0.31087, + "grad_norm": 0.6058558821678162, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 31087 + }, + { + "epoch": 0.31088, + "grad_norm": 0.5638027787208557, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 31088 + }, + { + "epoch": 0.31089, + "grad_norm": 0.8156859278678894, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 31089 + }, + { + "epoch": 0.3109, + "grad_norm": 1.047931432723999, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 31090 + }, + { + "epoch": 0.31091, + "grad_norm": 1.068554401397705, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 31091 + }, + { + "epoch": 0.31092, + "grad_norm": 0.8138704895973206, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 31092 + }, + { + "epoch": 0.31093, + "grad_norm": 0.7202593088150024, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 31093 + }, + { + "epoch": 0.31094, + "grad_norm": 0.8097513914108276, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 31094 + }, + { + "epoch": 0.31095, + "grad_norm": 0.9131002426147461, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 31095 + }, + { + "epoch": 0.31096, + "grad_norm": 1.1938605308532715, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 31096 + }, + { + "epoch": 0.31097, + "grad_norm": 0.7580569386482239, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 31097 + }, + { + "epoch": 0.31098, + "grad_norm": 0.6723451018333435, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 31098 + }, + { + "epoch": 0.31099, + "grad_norm": 0.7479480504989624, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 31099 + }, + { + "epoch": 0.311, + "grad_norm": 0.9915566444396973, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 31100 + }, + { + "epoch": 0.31101, + "grad_norm": 0.9668180346488953, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 31101 + }, + { + "epoch": 0.31102, + "grad_norm": 0.961155891418457, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 31102 + }, + { + "epoch": 0.31103, + "grad_norm": 1.0231808423995972, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 31103 + }, + { + "epoch": 0.31104, + "grad_norm": 1.115075945854187, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 31104 + }, + { + "epoch": 0.31105, + "grad_norm": 0.8911434412002563, + "learning_rate": 0.003, + "loss": 3.956, + "step": 31105 + }, + { + "epoch": 0.31106, + "grad_norm": 0.8154585361480713, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 31106 + }, + { + "epoch": 0.31107, + "grad_norm": 0.7798078060150146, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 31107 + }, + { + "epoch": 0.31108, + "grad_norm": 0.8141049742698669, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 31108 + }, + { + "epoch": 0.31109, + "grad_norm": 0.7264660000801086, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 31109 + }, + { + "epoch": 0.3111, + "grad_norm": 0.7973153591156006, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 31110 + }, + { + "epoch": 0.31111, + "grad_norm": 0.8001403212547302, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 31111 + }, + { + "epoch": 0.31112, + "grad_norm": 0.772428035736084, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 31112 + }, + { + "epoch": 0.31113, + "grad_norm": 0.7438635230064392, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 31113 + }, + { + "epoch": 0.31114, + "grad_norm": 0.7226239442825317, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 31114 + }, + { + "epoch": 0.31115, + "grad_norm": 0.7275788187980652, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 31115 + }, + { + "epoch": 0.31116, + "grad_norm": 0.6487720012664795, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 31116 + }, + { + "epoch": 0.31117, + "grad_norm": 0.7410804033279419, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 31117 + }, + { + "epoch": 0.31118, + "grad_norm": 1.1604832410812378, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 31118 + }, + { + "epoch": 0.31119, + "grad_norm": 1.3595629930496216, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 31119 + }, + { + "epoch": 0.3112, + "grad_norm": 0.5176845192909241, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 31120 + }, + { + "epoch": 0.31121, + "grad_norm": 0.6841402649879456, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 31121 + }, + { + "epoch": 0.31122, + "grad_norm": 0.8317727446556091, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 31122 + }, + { + "epoch": 0.31123, + "grad_norm": 1.0295401811599731, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 31123 + }, + { + "epoch": 0.31124, + "grad_norm": 1.264285683631897, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 31124 + }, + { + "epoch": 0.31125, + "grad_norm": 0.5679516196250916, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 31125 + }, + { + "epoch": 0.31126, + "grad_norm": 0.7959771752357483, + "learning_rate": 0.003, + "loss": 3.984, + "step": 31126 + }, + { + "epoch": 0.31127, + "grad_norm": 1.1192301511764526, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 31127 + }, + { + "epoch": 0.31128, + "grad_norm": 0.7685637474060059, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 31128 + }, + { + "epoch": 0.31129, + "grad_norm": 0.6711486577987671, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 31129 + }, + { + "epoch": 0.3113, + "grad_norm": 0.7388922572135925, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 31130 + }, + { + "epoch": 0.31131, + "grad_norm": 0.8555168509483337, + "learning_rate": 0.003, + "loss": 3.974, + "step": 31131 + }, + { + "epoch": 0.31132, + "grad_norm": 0.9118188619613647, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 31132 + }, + { + "epoch": 0.31133, + "grad_norm": 0.9879167675971985, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 31133 + }, + { + "epoch": 0.31134, + "grad_norm": 0.9806547164916992, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 31134 + }, + { + "epoch": 0.31135, + "grad_norm": 0.9994316101074219, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 31135 + }, + { + "epoch": 0.31136, + "grad_norm": 1.0122648477554321, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 31136 + }, + { + "epoch": 0.31137, + "grad_norm": 0.9708930850028992, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 31137 + }, + { + "epoch": 0.31138, + "grad_norm": 1.1452738046646118, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 31138 + }, + { + "epoch": 0.31139, + "grad_norm": 0.9842559099197388, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 31139 + }, + { + "epoch": 0.3114, + "grad_norm": 0.9701203107833862, + "learning_rate": 0.003, + "loss": 4.002, + "step": 31140 + }, + { + "epoch": 0.31141, + "grad_norm": 1.1244276762008667, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 31141 + }, + { + "epoch": 0.31142, + "grad_norm": 0.8801722526550293, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 31142 + }, + { + "epoch": 0.31143, + "grad_norm": 1.015092372894287, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 31143 + }, + { + "epoch": 0.31144, + "grad_norm": 1.009128451347351, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 31144 + }, + { + "epoch": 0.31145, + "grad_norm": 0.9586864113807678, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 31145 + }, + { + "epoch": 0.31146, + "grad_norm": 0.908545970916748, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 31146 + }, + { + "epoch": 0.31147, + "grad_norm": 0.9417285919189453, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 31147 + }, + { + "epoch": 0.31148, + "grad_norm": 1.1449049711227417, + "learning_rate": 0.003, + "loss": 4.01, + "step": 31148 + }, + { + "epoch": 0.31149, + "grad_norm": 1.1621876955032349, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 31149 + }, + { + "epoch": 0.3115, + "grad_norm": 0.9992176294326782, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 31150 + }, + { + "epoch": 0.31151, + "grad_norm": 1.1684057712554932, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 31151 + }, + { + "epoch": 0.31152, + "grad_norm": 0.85197514295578, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 31152 + }, + { + "epoch": 0.31153, + "grad_norm": 0.6936026811599731, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 31153 + }, + { + "epoch": 0.31154, + "grad_norm": 0.6304795145988464, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 31154 + }, + { + "epoch": 0.31155, + "grad_norm": 0.7411498427391052, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 31155 + }, + { + "epoch": 0.31156, + "grad_norm": 1.0309104919433594, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 31156 + }, + { + "epoch": 0.31157, + "grad_norm": 1.02849280834198, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 31157 + }, + { + "epoch": 0.31158, + "grad_norm": 0.6933895349502563, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 31158 + }, + { + "epoch": 0.31159, + "grad_norm": 0.7217861413955688, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 31159 + }, + { + "epoch": 0.3116, + "grad_norm": 0.7043586373329163, + "learning_rate": 0.003, + "loss": 3.992, + "step": 31160 + }, + { + "epoch": 0.31161, + "grad_norm": 0.7454587817192078, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 31161 + }, + { + "epoch": 0.31162, + "grad_norm": 0.8173890709877014, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 31162 + }, + { + "epoch": 0.31163, + "grad_norm": 0.9654850363731384, + "learning_rate": 0.003, + "loss": 3.986, + "step": 31163 + }, + { + "epoch": 0.31164, + "grad_norm": 1.0433603525161743, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 31164 + }, + { + "epoch": 0.31165, + "grad_norm": 0.7966647744178772, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 31165 + }, + { + "epoch": 0.31166, + "grad_norm": 0.6278672814369202, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 31166 + }, + { + "epoch": 0.31167, + "grad_norm": 0.6640574932098389, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 31167 + }, + { + "epoch": 0.31168, + "grad_norm": 0.6988121271133423, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 31168 + }, + { + "epoch": 0.31169, + "grad_norm": 0.7950130701065063, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 31169 + }, + { + "epoch": 0.3117, + "grad_norm": 0.8357150554656982, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 31170 + }, + { + "epoch": 0.31171, + "grad_norm": 0.8700249195098877, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 31171 + }, + { + "epoch": 0.31172, + "grad_norm": 0.7998148202896118, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 31172 + }, + { + "epoch": 0.31173, + "grad_norm": 0.7523528337478638, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 31173 + }, + { + "epoch": 0.31174, + "grad_norm": 0.862156867980957, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 31174 + }, + { + "epoch": 0.31175, + "grad_norm": 0.8454000949859619, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 31175 + }, + { + "epoch": 0.31176, + "grad_norm": 0.8931593298912048, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 31176 + }, + { + "epoch": 0.31177, + "grad_norm": 0.9670479893684387, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 31177 + }, + { + "epoch": 0.31178, + "grad_norm": 1.040085792541504, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 31178 + }, + { + "epoch": 0.31179, + "grad_norm": 1.1292089223861694, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 31179 + }, + { + "epoch": 0.3118, + "grad_norm": 0.8383427262306213, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 31180 + }, + { + "epoch": 0.31181, + "grad_norm": 0.7850264310836792, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 31181 + }, + { + "epoch": 0.31182, + "grad_norm": 0.9414145350456238, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 31182 + }, + { + "epoch": 0.31183, + "grad_norm": 1.2161206007003784, + "learning_rate": 0.003, + "loss": 3.964, + "step": 31183 + }, + { + "epoch": 0.31184, + "grad_norm": 0.7119455337524414, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 31184 + }, + { + "epoch": 0.31185, + "grad_norm": 0.6677044630050659, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 31185 + }, + { + "epoch": 0.31186, + "grad_norm": 0.6794235110282898, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 31186 + }, + { + "epoch": 0.31187, + "grad_norm": 0.6848109364509583, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 31187 + }, + { + "epoch": 0.31188, + "grad_norm": 0.7162713408470154, + "learning_rate": 0.003, + "loss": 3.929, + "step": 31188 + }, + { + "epoch": 0.31189, + "grad_norm": 0.7696809768676758, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 31189 + }, + { + "epoch": 0.3119, + "grad_norm": 0.9756592512130737, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 31190 + }, + { + "epoch": 0.31191, + "grad_norm": 1.188006043434143, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 31191 + }, + { + "epoch": 0.31192, + "grad_norm": 0.8090013265609741, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 31192 + }, + { + "epoch": 0.31193, + "grad_norm": 0.860957682132721, + "learning_rate": 0.003, + "loss": 4.015, + "step": 31193 + }, + { + "epoch": 0.31194, + "grad_norm": 0.8759660720825195, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 31194 + }, + { + "epoch": 0.31195, + "grad_norm": 0.8114926815032959, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 31195 + }, + { + "epoch": 0.31196, + "grad_norm": 0.8058712482452393, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 31196 + }, + { + "epoch": 0.31197, + "grad_norm": 0.9239717721939087, + "learning_rate": 0.003, + "loss": 3.97, + "step": 31197 + }, + { + "epoch": 0.31198, + "grad_norm": 1.0976028442382812, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 31198 + }, + { + "epoch": 0.31199, + "grad_norm": 1.1001397371292114, + "learning_rate": 0.003, + "loss": 3.993, + "step": 31199 + }, + { + "epoch": 0.312, + "grad_norm": 0.9672970175743103, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 31200 + }, + { + "epoch": 0.31201, + "grad_norm": 1.216080904006958, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 31201 + }, + { + "epoch": 0.31202, + "grad_norm": 1.2046945095062256, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 31202 + }, + { + "epoch": 0.31203, + "grad_norm": 0.7490370273590088, + "learning_rate": 0.003, + "loss": 4.002, + "step": 31203 + }, + { + "epoch": 0.31204, + "grad_norm": 0.7546696066856384, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 31204 + }, + { + "epoch": 0.31205, + "grad_norm": 0.7940678000450134, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 31205 + }, + { + "epoch": 0.31206, + "grad_norm": 0.8018761277198792, + "learning_rate": 0.003, + "loss": 3.992, + "step": 31206 + }, + { + "epoch": 0.31207, + "grad_norm": 0.7382722496986389, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 31207 + }, + { + "epoch": 0.31208, + "grad_norm": 0.7172749638557434, + "learning_rate": 0.003, + "loss": 3.999, + "step": 31208 + }, + { + "epoch": 0.31209, + "grad_norm": 0.773297131061554, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 31209 + }, + { + "epoch": 0.3121, + "grad_norm": 0.9073452949523926, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 31210 + }, + { + "epoch": 0.31211, + "grad_norm": 1.065244436264038, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 31211 + }, + { + "epoch": 0.31212, + "grad_norm": 1.0224369764328003, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 31212 + }, + { + "epoch": 0.31213, + "grad_norm": 0.9646946787834167, + "learning_rate": 0.003, + "loss": 4.008, + "step": 31213 + }, + { + "epoch": 0.31214, + "grad_norm": 0.896306574344635, + "learning_rate": 0.003, + "loss": 3.979, + "step": 31214 + }, + { + "epoch": 0.31215, + "grad_norm": 0.7169904112815857, + "learning_rate": 0.003, + "loss": 3.972, + "step": 31215 + }, + { + "epoch": 0.31216, + "grad_norm": 0.7255986332893372, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 31216 + }, + { + "epoch": 0.31217, + "grad_norm": 0.8283522129058838, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 31217 + }, + { + "epoch": 0.31218, + "grad_norm": 0.8307432532310486, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 31218 + }, + { + "epoch": 0.31219, + "grad_norm": 0.7282951474189758, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 31219 + }, + { + "epoch": 0.3122, + "grad_norm": 0.6700785756111145, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 31220 + }, + { + "epoch": 0.31221, + "grad_norm": 0.7545685172080994, + "learning_rate": 0.003, + "loss": 3.985, + "step": 31221 + }, + { + "epoch": 0.31222, + "grad_norm": 0.936159610748291, + "learning_rate": 0.003, + "loss": 3.9394, + "step": 31222 + }, + { + "epoch": 0.31223, + "grad_norm": 1.188838243484497, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 31223 + }, + { + "epoch": 0.31224, + "grad_norm": 0.7994606494903564, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 31224 + }, + { + "epoch": 0.31225, + "grad_norm": 0.6985687613487244, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 31225 + }, + { + "epoch": 0.31226, + "grad_norm": 0.9044394493103027, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 31226 + }, + { + "epoch": 0.31227, + "grad_norm": 0.9603644609451294, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 31227 + }, + { + "epoch": 0.31228, + "grad_norm": 1.0557714700698853, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 31228 + }, + { + "epoch": 0.31229, + "grad_norm": 1.085221290588379, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 31229 + }, + { + "epoch": 0.3123, + "grad_norm": 0.8646539449691772, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 31230 + }, + { + "epoch": 0.31231, + "grad_norm": 0.8425296545028687, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 31231 + }, + { + "epoch": 0.31232, + "grad_norm": 0.8113564252853394, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 31232 + }, + { + "epoch": 0.31233, + "grad_norm": 0.8993610739707947, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 31233 + }, + { + "epoch": 0.31234, + "grad_norm": 0.8656307458877563, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 31234 + }, + { + "epoch": 0.31235, + "grad_norm": 0.9412980675697327, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 31235 + }, + { + "epoch": 0.31236, + "grad_norm": 0.9027689695358276, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 31236 + }, + { + "epoch": 0.31237, + "grad_norm": 0.89676833152771, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 31237 + }, + { + "epoch": 0.31238, + "grad_norm": 0.9188695549964905, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 31238 + }, + { + "epoch": 0.31239, + "grad_norm": 1.0294517278671265, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 31239 + }, + { + "epoch": 0.3124, + "grad_norm": 1.126757264137268, + "learning_rate": 0.003, + "loss": 3.991, + "step": 31240 + }, + { + "epoch": 0.31241, + "grad_norm": 1.1050031185150146, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 31241 + }, + { + "epoch": 0.31242, + "grad_norm": 1.0373437404632568, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 31242 + }, + { + "epoch": 0.31243, + "grad_norm": 0.9558225274085999, + "learning_rate": 0.003, + "loss": 3.979, + "step": 31243 + }, + { + "epoch": 0.31244, + "grad_norm": 0.6910544037818909, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 31244 + }, + { + "epoch": 0.31245, + "grad_norm": 0.7050145268440247, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 31245 + }, + { + "epoch": 0.31246, + "grad_norm": 0.6880144476890564, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 31246 + }, + { + "epoch": 0.31247, + "grad_norm": 0.7430996298789978, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 31247 + }, + { + "epoch": 0.31248, + "grad_norm": 0.7314741015434265, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 31248 + }, + { + "epoch": 0.31249, + "grad_norm": 0.8649519681930542, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 31249 + }, + { + "epoch": 0.3125, + "grad_norm": 1.094720482826233, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 31250 + }, + { + "epoch": 0.31251, + "grad_norm": 0.9483115673065186, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 31251 + }, + { + "epoch": 0.31252, + "grad_norm": 0.9388488531112671, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 31252 + }, + { + "epoch": 0.31253, + "grad_norm": 0.8754322528839111, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 31253 + }, + { + "epoch": 0.31254, + "grad_norm": 1.0646311044692993, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 31254 + }, + { + "epoch": 0.31255, + "grad_norm": 1.001735806465149, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 31255 + }, + { + "epoch": 0.31256, + "grad_norm": 0.9590053558349609, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 31256 + }, + { + "epoch": 0.31257, + "grad_norm": 0.7667624950408936, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 31257 + }, + { + "epoch": 0.31258, + "grad_norm": 0.7840853333473206, + "learning_rate": 0.003, + "loss": 3.976, + "step": 31258 + }, + { + "epoch": 0.31259, + "grad_norm": 0.7983201146125793, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 31259 + }, + { + "epoch": 0.3126, + "grad_norm": 0.7947219014167786, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 31260 + }, + { + "epoch": 0.31261, + "grad_norm": 0.7658385038375854, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 31261 + }, + { + "epoch": 0.31262, + "grad_norm": 0.8305807709693909, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 31262 + }, + { + "epoch": 0.31263, + "grad_norm": 1.0941888093948364, + "learning_rate": 0.003, + "loss": 4.002, + "step": 31263 + }, + { + "epoch": 0.31264, + "grad_norm": 1.2195295095443726, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 31264 + }, + { + "epoch": 0.31265, + "grad_norm": 0.6940476298332214, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 31265 + }, + { + "epoch": 0.31266, + "grad_norm": 0.5660890340805054, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 31266 + }, + { + "epoch": 0.31267, + "grad_norm": 0.7037683129310608, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 31267 + }, + { + "epoch": 0.31268, + "grad_norm": 0.6536495089530945, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 31268 + }, + { + "epoch": 0.31269, + "grad_norm": 0.7302444577217102, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 31269 + }, + { + "epoch": 0.3127, + "grad_norm": 0.7013571858406067, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 31270 + }, + { + "epoch": 0.31271, + "grad_norm": 0.6009429693222046, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 31271 + }, + { + "epoch": 0.31272, + "grad_norm": 0.6996914744377136, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 31272 + }, + { + "epoch": 0.31273, + "grad_norm": 0.7733346819877625, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 31273 + }, + { + "epoch": 0.31274, + "grad_norm": 0.718083381652832, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 31274 + }, + { + "epoch": 0.31275, + "grad_norm": 0.6683833599090576, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 31275 + }, + { + "epoch": 0.31276, + "grad_norm": 0.7649011015892029, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 31276 + }, + { + "epoch": 0.31277, + "grad_norm": 1.132703185081482, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 31277 + }, + { + "epoch": 0.31278, + "grad_norm": 1.0154306888580322, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 31278 + }, + { + "epoch": 0.31279, + "grad_norm": 1.0591881275177002, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 31279 + }, + { + "epoch": 0.3128, + "grad_norm": 0.8985984325408936, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 31280 + }, + { + "epoch": 0.31281, + "grad_norm": 0.843172550201416, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 31281 + }, + { + "epoch": 0.31282, + "grad_norm": 0.8644895553588867, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 31282 + }, + { + "epoch": 0.31283, + "grad_norm": 0.9091817736625671, + "learning_rate": 0.003, + "loss": 3.966, + "step": 31283 + }, + { + "epoch": 0.31284, + "grad_norm": 0.9283657073974609, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 31284 + }, + { + "epoch": 0.31285, + "grad_norm": 0.8668957352638245, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 31285 + }, + { + "epoch": 0.31286, + "grad_norm": 0.7516258955001831, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 31286 + }, + { + "epoch": 0.31287, + "grad_norm": 1.015925645828247, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 31287 + }, + { + "epoch": 0.31288, + "grad_norm": 1.2769092321395874, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 31288 + }, + { + "epoch": 0.31289, + "grad_norm": 0.8613848686218262, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 31289 + }, + { + "epoch": 0.3129, + "grad_norm": 1.057500958442688, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 31290 + }, + { + "epoch": 0.31291, + "grad_norm": 1.2217587232589722, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 31291 + }, + { + "epoch": 0.31292, + "grad_norm": 0.776614785194397, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 31292 + }, + { + "epoch": 0.31293, + "grad_norm": 0.7306621670722961, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 31293 + }, + { + "epoch": 0.31294, + "grad_norm": 0.8119484186172485, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 31294 + }, + { + "epoch": 0.31295, + "grad_norm": 0.8775831460952759, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 31295 + }, + { + "epoch": 0.31296, + "grad_norm": 0.9394083619117737, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 31296 + }, + { + "epoch": 0.31297, + "grad_norm": 0.9988691210746765, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 31297 + }, + { + "epoch": 0.31298, + "grad_norm": 1.0786423683166504, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 31298 + }, + { + "epoch": 0.31299, + "grad_norm": 0.9597800374031067, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 31299 + }, + { + "epoch": 0.313, + "grad_norm": 1.0306519269943237, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 31300 + }, + { + "epoch": 0.31301, + "grad_norm": 1.060413122177124, + "learning_rate": 0.003, + "loss": 3.999, + "step": 31301 + }, + { + "epoch": 0.31302, + "grad_norm": 1.0879158973693848, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 31302 + }, + { + "epoch": 0.31303, + "grad_norm": 1.0639002323150635, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 31303 + }, + { + "epoch": 0.31304, + "grad_norm": 0.9321784973144531, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 31304 + }, + { + "epoch": 0.31305, + "grad_norm": 0.8075738549232483, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 31305 + }, + { + "epoch": 0.31306, + "grad_norm": 0.7427290081977844, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 31306 + }, + { + "epoch": 0.31307, + "grad_norm": 0.7290706038475037, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 31307 + }, + { + "epoch": 0.31308, + "grad_norm": 0.7243801951408386, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 31308 + }, + { + "epoch": 0.31309, + "grad_norm": 0.7682090401649475, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 31309 + }, + { + "epoch": 0.3131, + "grad_norm": 0.9443854689598083, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 31310 + }, + { + "epoch": 0.31311, + "grad_norm": 1.0157785415649414, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 31311 + }, + { + "epoch": 0.31312, + "grad_norm": 0.9225245118141174, + "learning_rate": 0.003, + "loss": 4.011, + "step": 31312 + }, + { + "epoch": 0.31313, + "grad_norm": 0.8209887742996216, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 31313 + }, + { + "epoch": 0.31314, + "grad_norm": 0.7582195997238159, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 31314 + }, + { + "epoch": 0.31315, + "grad_norm": 0.7014158368110657, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 31315 + }, + { + "epoch": 0.31316, + "grad_norm": 0.6973161101341248, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 31316 + }, + { + "epoch": 0.31317, + "grad_norm": 0.697063684463501, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 31317 + }, + { + "epoch": 0.31318, + "grad_norm": 0.8635243773460388, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 31318 + }, + { + "epoch": 0.31319, + "grad_norm": 0.9805964827537537, + "learning_rate": 0.003, + "loss": 3.988, + "step": 31319 + }, + { + "epoch": 0.3132, + "grad_norm": 1.0616095066070557, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 31320 + }, + { + "epoch": 0.31321, + "grad_norm": 1.0897717475891113, + "learning_rate": 0.003, + "loss": 3.996, + "step": 31321 + }, + { + "epoch": 0.31322, + "grad_norm": 0.9927883744239807, + "learning_rate": 0.003, + "loss": 3.995, + "step": 31322 + }, + { + "epoch": 0.31323, + "grad_norm": 0.8970299959182739, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 31323 + }, + { + "epoch": 0.31324, + "grad_norm": 0.9203004837036133, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 31324 + }, + { + "epoch": 0.31325, + "grad_norm": 1.0411338806152344, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 31325 + }, + { + "epoch": 0.31326, + "grad_norm": 0.9179123044013977, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 31326 + }, + { + "epoch": 0.31327, + "grad_norm": 0.8109869360923767, + "learning_rate": 0.003, + "loss": 3.991, + "step": 31327 + }, + { + "epoch": 0.31328, + "grad_norm": 1.2667601108551025, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 31328 + }, + { + "epoch": 0.31329, + "grad_norm": 0.7679722309112549, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 31329 + }, + { + "epoch": 0.3133, + "grad_norm": 0.7397465109825134, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 31330 + }, + { + "epoch": 0.31331, + "grad_norm": 0.8443636894226074, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 31331 + }, + { + "epoch": 0.31332, + "grad_norm": 1.0091114044189453, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 31332 + }, + { + "epoch": 0.31333, + "grad_norm": 0.9494238495826721, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 31333 + }, + { + "epoch": 0.31334, + "grad_norm": 0.8166471123695374, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 31334 + }, + { + "epoch": 0.31335, + "grad_norm": 0.8208678364753723, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 31335 + }, + { + "epoch": 0.31336, + "grad_norm": 0.8390712141990662, + "learning_rate": 0.003, + "loss": 3.9378, + "step": 31336 + }, + { + "epoch": 0.31337, + "grad_norm": 0.7920916080474854, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 31337 + }, + { + "epoch": 0.31338, + "grad_norm": 0.7823023796081543, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 31338 + }, + { + "epoch": 0.31339, + "grad_norm": 0.71385258436203, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 31339 + }, + { + "epoch": 0.3134, + "grad_norm": 0.6696440577507019, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 31340 + }, + { + "epoch": 0.31341, + "grad_norm": 0.700568675994873, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 31341 + }, + { + "epoch": 0.31342, + "grad_norm": 0.8417391180992126, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 31342 + }, + { + "epoch": 0.31343, + "grad_norm": 1.0863373279571533, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 31343 + }, + { + "epoch": 0.31344, + "grad_norm": 1.1779485940933228, + "learning_rate": 0.003, + "loss": 3.999, + "step": 31344 + }, + { + "epoch": 0.31345, + "grad_norm": 0.6310502886772156, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 31345 + }, + { + "epoch": 0.31346, + "grad_norm": 0.6302571892738342, + "learning_rate": 0.003, + "loss": 4.004, + "step": 31346 + }, + { + "epoch": 0.31347, + "grad_norm": 0.7016510963439941, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 31347 + }, + { + "epoch": 0.31348, + "grad_norm": 0.7027056217193604, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 31348 + }, + { + "epoch": 0.31349, + "grad_norm": 0.764622688293457, + "learning_rate": 0.003, + "loss": 3.974, + "step": 31349 + }, + { + "epoch": 0.3135, + "grad_norm": 0.8219848871231079, + "learning_rate": 0.003, + "loss": 3.99, + "step": 31350 + }, + { + "epoch": 0.31351, + "grad_norm": 0.90004563331604, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 31351 + }, + { + "epoch": 0.31352, + "grad_norm": 1.0431373119354248, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 31352 + }, + { + "epoch": 0.31353, + "grad_norm": 0.8609665632247925, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 31353 + }, + { + "epoch": 0.31354, + "grad_norm": 0.793486475944519, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 31354 + }, + { + "epoch": 0.31355, + "grad_norm": 0.9968968033790588, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 31355 + }, + { + "epoch": 0.31356, + "grad_norm": 1.2120070457458496, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 31356 + }, + { + "epoch": 0.31357, + "grad_norm": 0.7855210304260254, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 31357 + }, + { + "epoch": 0.31358, + "grad_norm": 0.6776762008666992, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 31358 + }, + { + "epoch": 0.31359, + "grad_norm": 0.7637615203857422, + "learning_rate": 0.003, + "loss": 3.988, + "step": 31359 + }, + { + "epoch": 0.3136, + "grad_norm": 0.9005922675132751, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 31360 + }, + { + "epoch": 0.31361, + "grad_norm": 0.9484728574752808, + "learning_rate": 0.003, + "loss": 3.982, + "step": 31361 + }, + { + "epoch": 0.31362, + "grad_norm": 0.9870548248291016, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 31362 + }, + { + "epoch": 0.31363, + "grad_norm": 1.091963768005371, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 31363 + }, + { + "epoch": 0.31364, + "grad_norm": 0.9955143332481384, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 31364 + }, + { + "epoch": 0.31365, + "grad_norm": 0.8592554926872253, + "learning_rate": 0.003, + "loss": 3.989, + "step": 31365 + }, + { + "epoch": 0.31366, + "grad_norm": 0.8167763948440552, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 31366 + }, + { + "epoch": 0.31367, + "grad_norm": 0.7681087851524353, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 31367 + }, + { + "epoch": 0.31368, + "grad_norm": 0.8022711873054504, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 31368 + }, + { + "epoch": 0.31369, + "grad_norm": 0.920181393623352, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 31369 + }, + { + "epoch": 0.3137, + "grad_norm": 0.8823000192642212, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 31370 + }, + { + "epoch": 0.31371, + "grad_norm": 0.8768589496612549, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 31371 + }, + { + "epoch": 0.31372, + "grad_norm": 1.0269001722335815, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 31372 + }, + { + "epoch": 0.31373, + "grad_norm": 0.9906975626945496, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 31373 + }, + { + "epoch": 0.31374, + "grad_norm": 0.745241105556488, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 31374 + }, + { + "epoch": 0.31375, + "grad_norm": 0.7713347673416138, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 31375 + }, + { + "epoch": 0.31376, + "grad_norm": 0.861531674861908, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 31376 + }, + { + "epoch": 0.31377, + "grad_norm": 0.9879153370857239, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 31377 + }, + { + "epoch": 0.31378, + "grad_norm": 1.3498996496200562, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 31378 + }, + { + "epoch": 0.31379, + "grad_norm": 0.7451598048210144, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 31379 + }, + { + "epoch": 0.3138, + "grad_norm": 0.6367138028144836, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 31380 + }, + { + "epoch": 0.31381, + "grad_norm": 0.6524907946586609, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 31381 + }, + { + "epoch": 0.31382, + "grad_norm": 0.6226068735122681, + "learning_rate": 0.003, + "loss": 3.966, + "step": 31382 + }, + { + "epoch": 0.31383, + "grad_norm": 0.725540280342102, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 31383 + }, + { + "epoch": 0.31384, + "grad_norm": 0.8213433027267456, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 31384 + }, + { + "epoch": 0.31385, + "grad_norm": 0.8668132424354553, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 31385 + }, + { + "epoch": 0.31386, + "grad_norm": 0.8802310824394226, + "learning_rate": 0.003, + "loss": 3.986, + "step": 31386 + }, + { + "epoch": 0.31387, + "grad_norm": 1.345528244972229, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 31387 + }, + { + "epoch": 0.31388, + "grad_norm": 0.7986776232719421, + "learning_rate": 0.003, + "loss": 4.001, + "step": 31388 + }, + { + "epoch": 0.31389, + "grad_norm": 0.6840220093727112, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 31389 + }, + { + "epoch": 0.3139, + "grad_norm": 0.773623526096344, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 31390 + }, + { + "epoch": 0.31391, + "grad_norm": 0.7418786287307739, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 31391 + }, + { + "epoch": 0.31392, + "grad_norm": 0.7576735615730286, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 31392 + }, + { + "epoch": 0.31393, + "grad_norm": 0.7864428758621216, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 31393 + }, + { + "epoch": 0.31394, + "grad_norm": 0.8202427625656128, + "learning_rate": 0.003, + "loss": 3.999, + "step": 31394 + }, + { + "epoch": 0.31395, + "grad_norm": 0.9752988219261169, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 31395 + }, + { + "epoch": 0.31396, + "grad_norm": 1.207958459854126, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 31396 + }, + { + "epoch": 0.31397, + "grad_norm": 0.7851834893226624, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 31397 + }, + { + "epoch": 0.31398, + "grad_norm": 0.5944402813911438, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 31398 + }, + { + "epoch": 0.31399, + "grad_norm": 0.6202830076217651, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 31399 + }, + { + "epoch": 0.314, + "grad_norm": 0.7020778656005859, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 31400 + }, + { + "epoch": 0.31401, + "grad_norm": 0.795920729637146, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 31401 + }, + { + "epoch": 0.31402, + "grad_norm": 0.9158527851104736, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 31402 + }, + { + "epoch": 0.31403, + "grad_norm": 0.8321881294250488, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 31403 + }, + { + "epoch": 0.31404, + "grad_norm": 0.8159991502761841, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 31404 + }, + { + "epoch": 0.31405, + "grad_norm": 0.8613864779472351, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 31405 + }, + { + "epoch": 0.31406, + "grad_norm": 0.8170823454856873, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 31406 + }, + { + "epoch": 0.31407, + "grad_norm": 0.8830284476280212, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 31407 + }, + { + "epoch": 0.31408, + "grad_norm": 0.978163480758667, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 31408 + }, + { + "epoch": 0.31409, + "grad_norm": 1.343705415725708, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 31409 + }, + { + "epoch": 0.3141, + "grad_norm": 1.037132978439331, + "learning_rate": 0.003, + "loss": 3.986, + "step": 31410 + }, + { + "epoch": 0.31411, + "grad_norm": 1.1840696334838867, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 31411 + }, + { + "epoch": 0.31412, + "grad_norm": 0.8445314764976501, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 31412 + }, + { + "epoch": 0.31413, + "grad_norm": 0.6534401178359985, + "learning_rate": 0.003, + "loss": 3.9434, + "step": 31413 + }, + { + "epoch": 0.31414, + "grad_norm": 0.6753685474395752, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 31414 + }, + { + "epoch": 0.31415, + "grad_norm": 0.6737540364265442, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 31415 + }, + { + "epoch": 0.31416, + "grad_norm": 0.6660255789756775, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 31416 + }, + { + "epoch": 0.31417, + "grad_norm": 0.7752023339271545, + "learning_rate": 0.003, + "loss": 3.974, + "step": 31417 + }, + { + "epoch": 0.31418, + "grad_norm": 0.8739541172981262, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 31418 + }, + { + "epoch": 0.31419, + "grad_norm": 0.8695088028907776, + "learning_rate": 0.003, + "loss": 3.998, + "step": 31419 + }, + { + "epoch": 0.3142, + "grad_norm": 0.9262238144874573, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 31420 + }, + { + "epoch": 0.31421, + "grad_norm": 0.9244855642318726, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 31421 + }, + { + "epoch": 0.31422, + "grad_norm": 1.0133798122406006, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 31422 + }, + { + "epoch": 0.31423, + "grad_norm": 1.0495332479476929, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 31423 + }, + { + "epoch": 0.31424, + "grad_norm": 0.9132116436958313, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 31424 + }, + { + "epoch": 0.31425, + "grad_norm": 0.8704999089241028, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 31425 + }, + { + "epoch": 0.31426, + "grad_norm": 0.9265024065971375, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 31426 + }, + { + "epoch": 0.31427, + "grad_norm": 1.187462329864502, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 31427 + }, + { + "epoch": 0.31428, + "grad_norm": 0.9301339983940125, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 31428 + }, + { + "epoch": 0.31429, + "grad_norm": 0.8249815106391907, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 31429 + }, + { + "epoch": 0.3143, + "grad_norm": 0.9112917184829712, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 31430 + }, + { + "epoch": 0.31431, + "grad_norm": 1.0672436952590942, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 31431 + }, + { + "epoch": 0.31432, + "grad_norm": 1.0266845226287842, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 31432 + }, + { + "epoch": 0.31433, + "grad_norm": 0.9664485454559326, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 31433 + }, + { + "epoch": 0.31434, + "grad_norm": 0.9882024526596069, + "learning_rate": 0.003, + "loss": 3.996, + "step": 31434 + }, + { + "epoch": 0.31435, + "grad_norm": 0.9654036164283752, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 31435 + }, + { + "epoch": 0.31436, + "grad_norm": 0.9558396935462952, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 31436 + }, + { + "epoch": 0.31437, + "grad_norm": 1.0502467155456543, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 31437 + }, + { + "epoch": 0.31438, + "grad_norm": 1.0105630159378052, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 31438 + }, + { + "epoch": 0.31439, + "grad_norm": 0.9389442205429077, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 31439 + }, + { + "epoch": 0.3144, + "grad_norm": 0.9138251543045044, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 31440 + }, + { + "epoch": 0.31441, + "grad_norm": 0.9111635088920593, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 31441 + }, + { + "epoch": 0.31442, + "grad_norm": 0.821111261844635, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 31442 + }, + { + "epoch": 0.31443, + "grad_norm": 0.7541088461875916, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 31443 + }, + { + "epoch": 0.31444, + "grad_norm": 0.7861844897270203, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 31444 + }, + { + "epoch": 0.31445, + "grad_norm": 0.8646368980407715, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 31445 + }, + { + "epoch": 0.31446, + "grad_norm": 0.8043690919876099, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 31446 + }, + { + "epoch": 0.31447, + "grad_norm": 0.8005145788192749, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 31447 + }, + { + "epoch": 0.31448, + "grad_norm": 0.8852628469467163, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 31448 + }, + { + "epoch": 0.31449, + "grad_norm": 1.0913528203964233, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 31449 + }, + { + "epoch": 0.3145, + "grad_norm": 1.1195290088653564, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 31450 + }, + { + "epoch": 0.31451, + "grad_norm": 1.0720741748809814, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 31451 + }, + { + "epoch": 0.31452, + "grad_norm": 0.9820308089256287, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 31452 + }, + { + "epoch": 0.31453, + "grad_norm": 0.892271101474762, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 31453 + }, + { + "epoch": 0.31454, + "grad_norm": 0.7932468056678772, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 31454 + }, + { + "epoch": 0.31455, + "grad_norm": 0.7637684345245361, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 31455 + }, + { + "epoch": 0.31456, + "grad_norm": 0.7911094427108765, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 31456 + }, + { + "epoch": 0.31457, + "grad_norm": 0.8689959645271301, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 31457 + }, + { + "epoch": 0.31458, + "grad_norm": 0.8620853424072266, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 31458 + }, + { + "epoch": 0.31459, + "grad_norm": 0.8492268919944763, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 31459 + }, + { + "epoch": 0.3146, + "grad_norm": 0.8578975200653076, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 31460 + }, + { + "epoch": 0.31461, + "grad_norm": 1.3736985921859741, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 31461 + }, + { + "epoch": 0.31462, + "grad_norm": 0.8169393539428711, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 31462 + }, + { + "epoch": 0.31463, + "grad_norm": 0.6680393218994141, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 31463 + }, + { + "epoch": 0.31464, + "grad_norm": 0.5907289385795593, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 31464 + }, + { + "epoch": 0.31465, + "grad_norm": 0.6108154058456421, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 31465 + }, + { + "epoch": 0.31466, + "grad_norm": 0.6354832053184509, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 31466 + }, + { + "epoch": 0.31467, + "grad_norm": 0.6309311985969543, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 31467 + }, + { + "epoch": 0.31468, + "grad_norm": 0.7065508365631104, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 31468 + }, + { + "epoch": 0.31469, + "grad_norm": 0.721655547618866, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 31469 + }, + { + "epoch": 0.3147, + "grad_norm": 0.869932234287262, + "learning_rate": 0.003, + "loss": 3.986, + "step": 31470 + }, + { + "epoch": 0.31471, + "grad_norm": 1.0696085691452026, + "learning_rate": 0.003, + "loss": 3.976, + "step": 31471 + }, + { + "epoch": 0.31472, + "grad_norm": 0.9036632180213928, + "learning_rate": 0.003, + "loss": 3.988, + "step": 31472 + }, + { + "epoch": 0.31473, + "grad_norm": 0.7589657306671143, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 31473 + }, + { + "epoch": 0.31474, + "grad_norm": 0.6267876625061035, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 31474 + }, + { + "epoch": 0.31475, + "grad_norm": 0.643936038017273, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 31475 + }, + { + "epoch": 0.31476, + "grad_norm": 0.7966037392616272, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 31476 + }, + { + "epoch": 0.31477, + "grad_norm": 1.0547406673431396, + "learning_rate": 0.003, + "loss": 4.003, + "step": 31477 + }, + { + "epoch": 0.31478, + "grad_norm": 0.974925696849823, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 31478 + }, + { + "epoch": 0.31479, + "grad_norm": 1.083832859992981, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 31479 + }, + { + "epoch": 0.3148, + "grad_norm": 1.0686428546905518, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 31480 + }, + { + "epoch": 0.31481, + "grad_norm": 1.0262690782546997, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 31481 + }, + { + "epoch": 0.31482, + "grad_norm": 1.0029557943344116, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 31482 + }, + { + "epoch": 0.31483, + "grad_norm": 1.038375973701477, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 31483 + }, + { + "epoch": 0.31484, + "grad_norm": 1.086342692375183, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 31484 + }, + { + "epoch": 0.31485, + "grad_norm": 0.9384824633598328, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 31485 + }, + { + "epoch": 0.31486, + "grad_norm": 1.0884160995483398, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 31486 + }, + { + "epoch": 0.31487, + "grad_norm": 1.1613703966140747, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 31487 + }, + { + "epoch": 0.31488, + "grad_norm": 1.0197044610977173, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 31488 + }, + { + "epoch": 0.31489, + "grad_norm": 0.9866082668304443, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 31489 + }, + { + "epoch": 0.3149, + "grad_norm": 0.8209992051124573, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 31490 + }, + { + "epoch": 0.31491, + "grad_norm": 0.766494631767273, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 31491 + }, + { + "epoch": 0.31492, + "grad_norm": 0.8257040977478027, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 31492 + }, + { + "epoch": 0.31493, + "grad_norm": 1.0813149213790894, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 31493 + }, + { + "epoch": 0.31494, + "grad_norm": 1.0175695419311523, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 31494 + }, + { + "epoch": 0.31495, + "grad_norm": 0.9250473380088806, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 31495 + }, + { + "epoch": 0.31496, + "grad_norm": 0.9119294881820679, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 31496 + }, + { + "epoch": 0.31497, + "grad_norm": 0.8759008049964905, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 31497 + }, + { + "epoch": 0.31498, + "grad_norm": 0.7983118295669556, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 31498 + }, + { + "epoch": 0.31499, + "grad_norm": 0.7476514577865601, + "learning_rate": 0.003, + "loss": 3.995, + "step": 31499 + }, + { + "epoch": 0.315, + "grad_norm": 0.710420548915863, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 31500 + }, + { + "epoch": 0.31501, + "grad_norm": 0.6545474529266357, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 31501 + }, + { + "epoch": 0.31502, + "grad_norm": 0.6959617137908936, + "learning_rate": 0.003, + "loss": 3.983, + "step": 31502 + }, + { + "epoch": 0.31503, + "grad_norm": 0.8850348591804504, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 31503 + }, + { + "epoch": 0.31504, + "grad_norm": 1.0837652683258057, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 31504 + }, + { + "epoch": 0.31505, + "grad_norm": 0.9772597551345825, + "learning_rate": 0.003, + "loss": 4.005, + "step": 31505 + }, + { + "epoch": 0.31506, + "grad_norm": 0.8813307285308838, + "learning_rate": 0.003, + "loss": 4.011, + "step": 31506 + }, + { + "epoch": 0.31507, + "grad_norm": 0.815606951713562, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 31507 + }, + { + "epoch": 0.31508, + "grad_norm": 0.9152340292930603, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 31508 + }, + { + "epoch": 0.31509, + "grad_norm": 1.2149466276168823, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 31509 + }, + { + "epoch": 0.3151, + "grad_norm": 0.8144436478614807, + "learning_rate": 0.003, + "loss": 3.989, + "step": 31510 + }, + { + "epoch": 0.31511, + "grad_norm": 0.755057156085968, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 31511 + }, + { + "epoch": 0.31512, + "grad_norm": 0.6785875558853149, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 31512 + }, + { + "epoch": 0.31513, + "grad_norm": 0.717448353767395, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 31513 + }, + { + "epoch": 0.31514, + "grad_norm": 0.769439697265625, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 31514 + }, + { + "epoch": 0.31515, + "grad_norm": 0.7719340324401855, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 31515 + }, + { + "epoch": 0.31516, + "grad_norm": 0.8034616708755493, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 31516 + }, + { + "epoch": 0.31517, + "grad_norm": 0.8920993208885193, + "learning_rate": 0.003, + "loss": 3.98, + "step": 31517 + }, + { + "epoch": 0.31518, + "grad_norm": 1.0045137405395508, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 31518 + }, + { + "epoch": 0.31519, + "grad_norm": 1.1665575504302979, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 31519 + }, + { + "epoch": 0.3152, + "grad_norm": 0.8525962233543396, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 31520 + }, + { + "epoch": 0.31521, + "grad_norm": 0.7489387392997742, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 31521 + }, + { + "epoch": 0.31522, + "grad_norm": 0.7487748265266418, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 31522 + }, + { + "epoch": 0.31523, + "grad_norm": 0.7282821536064148, + "learning_rate": 0.003, + "loss": 3.96, + "step": 31523 + }, + { + "epoch": 0.31524, + "grad_norm": 0.7085949182510376, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 31524 + }, + { + "epoch": 0.31525, + "grad_norm": 0.7785043716430664, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 31525 + }, + { + "epoch": 0.31526, + "grad_norm": 0.8140043020248413, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 31526 + }, + { + "epoch": 0.31527, + "grad_norm": 0.9199599623680115, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 31527 + }, + { + "epoch": 0.31528, + "grad_norm": 1.1687812805175781, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 31528 + }, + { + "epoch": 0.31529, + "grad_norm": 1.1358506679534912, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 31529 + }, + { + "epoch": 0.3153, + "grad_norm": 0.9846063256263733, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 31530 + }, + { + "epoch": 0.31531, + "grad_norm": 0.8458243012428284, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 31531 + }, + { + "epoch": 0.31532, + "grad_norm": 0.814402163028717, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 31532 + }, + { + "epoch": 0.31533, + "grad_norm": 0.8446137309074402, + "learning_rate": 0.003, + "loss": 3.995, + "step": 31533 + }, + { + "epoch": 0.31534, + "grad_norm": 0.9621380567550659, + "learning_rate": 0.003, + "loss": 4.026, + "step": 31534 + }, + { + "epoch": 0.31535, + "grad_norm": 0.9192095994949341, + "learning_rate": 0.003, + "loss": 3.969, + "step": 31535 + }, + { + "epoch": 0.31536, + "grad_norm": 1.0785046815872192, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 31536 + }, + { + "epoch": 0.31537, + "grad_norm": 0.9556738138198853, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 31537 + }, + { + "epoch": 0.31538, + "grad_norm": 0.9870012402534485, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 31538 + }, + { + "epoch": 0.31539, + "grad_norm": 1.0146006345748901, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 31539 + }, + { + "epoch": 0.3154, + "grad_norm": 0.9239429831504822, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 31540 + }, + { + "epoch": 0.31541, + "grad_norm": 1.1348484754562378, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 31541 + }, + { + "epoch": 0.31542, + "grad_norm": 1.0905593633651733, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 31542 + }, + { + "epoch": 0.31543, + "grad_norm": 1.11601984500885, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 31543 + }, + { + "epoch": 0.31544, + "grad_norm": 1.0919580459594727, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 31544 + }, + { + "epoch": 0.31545, + "grad_norm": 1.005882978439331, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 31545 + }, + { + "epoch": 0.31546, + "grad_norm": 1.009232521057129, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 31546 + }, + { + "epoch": 0.31547, + "grad_norm": 1.0525789260864258, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 31547 + }, + { + "epoch": 0.31548, + "grad_norm": 0.8100241422653198, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 31548 + }, + { + "epoch": 0.31549, + "grad_norm": 0.7967370748519897, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 31549 + }, + { + "epoch": 0.3155, + "grad_norm": 0.9764791131019592, + "learning_rate": 0.003, + "loss": 4.026, + "step": 31550 + }, + { + "epoch": 0.31551, + "grad_norm": 1.045299768447876, + "learning_rate": 0.003, + "loss": 4.003, + "step": 31551 + }, + { + "epoch": 0.31552, + "grad_norm": 0.9120970964431763, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 31552 + }, + { + "epoch": 0.31553, + "grad_norm": 0.78750079870224, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 31553 + }, + { + "epoch": 0.31554, + "grad_norm": 0.7891731262207031, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 31554 + }, + { + "epoch": 0.31555, + "grad_norm": 0.6240924596786499, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 31555 + }, + { + "epoch": 0.31556, + "grad_norm": 0.7752370834350586, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 31556 + }, + { + "epoch": 0.31557, + "grad_norm": 0.9535866379737854, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 31557 + }, + { + "epoch": 0.31558, + "grad_norm": 0.9239205718040466, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 31558 + }, + { + "epoch": 0.31559, + "grad_norm": 0.8807371258735657, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 31559 + }, + { + "epoch": 0.3156, + "grad_norm": 1.006279468536377, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 31560 + }, + { + "epoch": 0.31561, + "grad_norm": 0.9400851130485535, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 31561 + }, + { + "epoch": 0.31562, + "grad_norm": 0.9635488986968994, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 31562 + }, + { + "epoch": 0.31563, + "grad_norm": 0.9768532514572144, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 31563 + }, + { + "epoch": 0.31564, + "grad_norm": 0.9553535580635071, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 31564 + }, + { + "epoch": 0.31565, + "grad_norm": 0.9574935436248779, + "learning_rate": 0.003, + "loss": 3.995, + "step": 31565 + }, + { + "epoch": 0.31566, + "grad_norm": 0.9291415214538574, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 31566 + }, + { + "epoch": 0.31567, + "grad_norm": 0.8657254576683044, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 31567 + }, + { + "epoch": 0.31568, + "grad_norm": 0.7590442299842834, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 31568 + }, + { + "epoch": 0.31569, + "grad_norm": 0.6938315033912659, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 31569 + }, + { + "epoch": 0.3157, + "grad_norm": 0.8297334313392639, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 31570 + }, + { + "epoch": 0.31571, + "grad_norm": 0.9093274474143982, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 31571 + }, + { + "epoch": 0.31572, + "grad_norm": 0.821811854839325, + "learning_rate": 0.003, + "loss": 3.969, + "step": 31572 + }, + { + "epoch": 0.31573, + "grad_norm": 0.7767503261566162, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 31573 + }, + { + "epoch": 0.31574, + "grad_norm": 0.8154393434524536, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 31574 + }, + { + "epoch": 0.31575, + "grad_norm": 0.735800564289093, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 31575 + }, + { + "epoch": 0.31576, + "grad_norm": 0.8117727637290955, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 31576 + }, + { + "epoch": 0.31577, + "grad_norm": 1.0743675231933594, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 31577 + }, + { + "epoch": 0.31578, + "grad_norm": 1.3926831483840942, + "learning_rate": 0.003, + "loss": 3.967, + "step": 31578 + }, + { + "epoch": 0.31579, + "grad_norm": 0.5578306317329407, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 31579 + }, + { + "epoch": 0.3158, + "grad_norm": 0.8390445709228516, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 31580 + }, + { + "epoch": 0.31581, + "grad_norm": 0.8390048146247864, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 31581 + }, + { + "epoch": 0.31582, + "grad_norm": 0.8825708627700806, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 31582 + }, + { + "epoch": 0.31583, + "grad_norm": 1.1064887046813965, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 31583 + }, + { + "epoch": 0.31584, + "grad_norm": 1.0937680006027222, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 31584 + }, + { + "epoch": 0.31585, + "grad_norm": 0.7654660940170288, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 31585 + }, + { + "epoch": 0.31586, + "grad_norm": 0.5469478964805603, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 31586 + }, + { + "epoch": 0.31587, + "grad_norm": 0.5387517213821411, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 31587 + }, + { + "epoch": 0.31588, + "grad_norm": 0.6665916442871094, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 31588 + }, + { + "epoch": 0.31589, + "grad_norm": 1.0261543989181519, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 31589 + }, + { + "epoch": 0.3159, + "grad_norm": 1.2180557250976562, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 31590 + }, + { + "epoch": 0.31591, + "grad_norm": 0.6077673435211182, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 31591 + }, + { + "epoch": 0.31592, + "grad_norm": 0.8233144283294678, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 31592 + }, + { + "epoch": 0.31593, + "grad_norm": 1.0500850677490234, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 31593 + }, + { + "epoch": 0.31594, + "grad_norm": 0.9342678785324097, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 31594 + }, + { + "epoch": 0.31595, + "grad_norm": 0.789665162563324, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 31595 + }, + { + "epoch": 0.31596, + "grad_norm": 0.8492036461830139, + "learning_rate": 0.003, + "loss": 3.991, + "step": 31596 + }, + { + "epoch": 0.31597, + "grad_norm": 0.8951340317726135, + "learning_rate": 0.003, + "loss": 4.0, + "step": 31597 + }, + { + "epoch": 0.31598, + "grad_norm": 0.8556100726127625, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 31598 + }, + { + "epoch": 0.31599, + "grad_norm": 0.863976240158081, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 31599 + }, + { + "epoch": 0.316, + "grad_norm": 0.837589681148529, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 31600 + }, + { + "epoch": 0.31601, + "grad_norm": 0.8025705814361572, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 31601 + }, + { + "epoch": 0.31602, + "grad_norm": 0.9886983633041382, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 31602 + }, + { + "epoch": 0.31603, + "grad_norm": 1.2323392629623413, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 31603 + }, + { + "epoch": 0.31604, + "grad_norm": 0.8303616642951965, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 31604 + }, + { + "epoch": 0.31605, + "grad_norm": 1.029985785484314, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 31605 + }, + { + "epoch": 0.31606, + "grad_norm": 1.2088992595672607, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 31606 + }, + { + "epoch": 0.31607, + "grad_norm": 0.9132320880889893, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 31607 + }, + { + "epoch": 0.31608, + "grad_norm": 0.7305404543876648, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 31608 + }, + { + "epoch": 0.31609, + "grad_norm": 0.8295374512672424, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 31609 + }, + { + "epoch": 0.3161, + "grad_norm": 0.8211520314216614, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 31610 + }, + { + "epoch": 0.31611, + "grad_norm": 0.7141633629798889, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 31611 + }, + { + "epoch": 0.31612, + "grad_norm": 0.7661651968955994, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 31612 + }, + { + "epoch": 0.31613, + "grad_norm": 0.9492795467376709, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 31613 + }, + { + "epoch": 0.31614, + "grad_norm": 1.1179238557815552, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 31614 + }, + { + "epoch": 0.31615, + "grad_norm": 0.9906667470932007, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 31615 + }, + { + "epoch": 0.31616, + "grad_norm": 0.8321180939674377, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 31616 + }, + { + "epoch": 0.31617, + "grad_norm": 0.5525113344192505, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 31617 + }, + { + "epoch": 0.31618, + "grad_norm": 0.5144295692443848, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 31618 + }, + { + "epoch": 0.31619, + "grad_norm": 0.5826591849327087, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 31619 + }, + { + "epoch": 0.3162, + "grad_norm": 0.6966007351875305, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 31620 + }, + { + "epoch": 0.31621, + "grad_norm": 0.7940075993537903, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 31621 + }, + { + "epoch": 0.31622, + "grad_norm": 0.8238639831542969, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 31622 + }, + { + "epoch": 0.31623, + "grad_norm": 0.8302192091941833, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 31623 + }, + { + "epoch": 0.31624, + "grad_norm": 0.7493805885314941, + "learning_rate": 0.003, + "loss": 3.971, + "step": 31624 + }, + { + "epoch": 0.31625, + "grad_norm": 0.7922492027282715, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 31625 + }, + { + "epoch": 0.31626, + "grad_norm": 0.8943837285041809, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 31626 + }, + { + "epoch": 0.31627, + "grad_norm": 1.027125358581543, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 31627 + }, + { + "epoch": 0.31628, + "grad_norm": 1.0804615020751953, + "learning_rate": 0.003, + "loss": 3.96, + "step": 31628 + }, + { + "epoch": 0.31629, + "grad_norm": 0.9506551623344421, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 31629 + }, + { + "epoch": 0.3163, + "grad_norm": 0.9711284637451172, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 31630 + }, + { + "epoch": 0.31631, + "grad_norm": 1.08293616771698, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 31631 + }, + { + "epoch": 0.31632, + "grad_norm": 0.9576627016067505, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 31632 + }, + { + "epoch": 0.31633, + "grad_norm": 1.0039870738983154, + "learning_rate": 0.003, + "loss": 3.975, + "step": 31633 + }, + { + "epoch": 0.31634, + "grad_norm": 1.17881178855896, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 31634 + }, + { + "epoch": 0.31635, + "grad_norm": 0.8189864754676819, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 31635 + }, + { + "epoch": 0.31636, + "grad_norm": 0.7461751699447632, + "learning_rate": 0.003, + "loss": 3.99, + "step": 31636 + }, + { + "epoch": 0.31637, + "grad_norm": 0.9638386964797974, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 31637 + }, + { + "epoch": 0.31638, + "grad_norm": 0.9180126786231995, + "learning_rate": 0.003, + "loss": 3.997, + "step": 31638 + }, + { + "epoch": 0.31639, + "grad_norm": 0.8843728303909302, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 31639 + }, + { + "epoch": 0.3164, + "grad_norm": 0.8008260726928711, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 31640 + }, + { + "epoch": 0.31641, + "grad_norm": 0.836844801902771, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 31641 + }, + { + "epoch": 0.31642, + "grad_norm": 0.9437279105186462, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 31642 + }, + { + "epoch": 0.31643, + "grad_norm": 1.0438199043273926, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 31643 + }, + { + "epoch": 0.31644, + "grad_norm": 1.032981038093567, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 31644 + }, + { + "epoch": 0.31645, + "grad_norm": 0.9972506761550903, + "learning_rate": 0.003, + "loss": 3.997, + "step": 31645 + }, + { + "epoch": 0.31646, + "grad_norm": 1.0052753686904907, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 31646 + }, + { + "epoch": 0.31647, + "grad_norm": 0.9653274416923523, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 31647 + }, + { + "epoch": 0.31648, + "grad_norm": 0.9014284610748291, + "learning_rate": 0.003, + "loss": 3.981, + "step": 31648 + }, + { + "epoch": 0.31649, + "grad_norm": 1.0175485610961914, + "learning_rate": 0.003, + "loss": 3.976, + "step": 31649 + }, + { + "epoch": 0.3165, + "grad_norm": 1.0647904872894287, + "learning_rate": 0.003, + "loss": 3.981, + "step": 31650 + }, + { + "epoch": 0.31651, + "grad_norm": 0.8752798438072205, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 31651 + }, + { + "epoch": 0.31652, + "grad_norm": 0.798172652721405, + "learning_rate": 0.003, + "loss": 3.979, + "step": 31652 + }, + { + "epoch": 0.31653, + "grad_norm": 0.7230918407440186, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 31653 + }, + { + "epoch": 0.31654, + "grad_norm": 0.7099258899688721, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 31654 + }, + { + "epoch": 0.31655, + "grad_norm": 0.7175295352935791, + "learning_rate": 0.003, + "loss": 3.9478, + "step": 31655 + }, + { + "epoch": 0.31656, + "grad_norm": 0.6898815035820007, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 31656 + }, + { + "epoch": 0.31657, + "grad_norm": 0.7266850471496582, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 31657 + }, + { + "epoch": 0.31658, + "grad_norm": 0.8750666975975037, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 31658 + }, + { + "epoch": 0.31659, + "grad_norm": 1.1948661804199219, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 31659 + }, + { + "epoch": 0.3166, + "grad_norm": 0.8305080533027649, + "learning_rate": 0.003, + "loss": 3.9426, + "step": 31660 + }, + { + "epoch": 0.31661, + "grad_norm": 0.6259477138519287, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 31661 + }, + { + "epoch": 0.31662, + "grad_norm": 0.6959621906280518, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 31662 + }, + { + "epoch": 0.31663, + "grad_norm": 0.7640928626060486, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 31663 + }, + { + "epoch": 0.31664, + "grad_norm": 0.8830692768096924, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 31664 + }, + { + "epoch": 0.31665, + "grad_norm": 1.1780201196670532, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 31665 + }, + { + "epoch": 0.31666, + "grad_norm": 0.9359006285667419, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 31666 + }, + { + "epoch": 0.31667, + "grad_norm": 0.922167956829071, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 31667 + }, + { + "epoch": 0.31668, + "grad_norm": 1.0273782014846802, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 31668 + }, + { + "epoch": 0.31669, + "grad_norm": 1.3461475372314453, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 31669 + }, + { + "epoch": 0.3167, + "grad_norm": 0.724689245223999, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 31670 + }, + { + "epoch": 0.31671, + "grad_norm": 0.6703101396560669, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 31671 + }, + { + "epoch": 0.31672, + "grad_norm": 0.6333152651786804, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 31672 + }, + { + "epoch": 0.31673, + "grad_norm": 0.7066918611526489, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 31673 + }, + { + "epoch": 0.31674, + "grad_norm": 0.7862377166748047, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 31674 + }, + { + "epoch": 0.31675, + "grad_norm": 0.774613618850708, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 31675 + }, + { + "epoch": 0.31676, + "grad_norm": 0.833723247051239, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 31676 + }, + { + "epoch": 0.31677, + "grad_norm": 1.111862301826477, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 31677 + }, + { + "epoch": 0.31678, + "grad_norm": 0.869006872177124, + "learning_rate": 0.003, + "loss": 3.984, + "step": 31678 + }, + { + "epoch": 0.31679, + "grad_norm": 0.9793900847434998, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 31679 + }, + { + "epoch": 0.3168, + "grad_norm": 1.2541221380233765, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 31680 + }, + { + "epoch": 0.31681, + "grad_norm": 0.8061410784721375, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 31681 + }, + { + "epoch": 0.31682, + "grad_norm": 0.6587876081466675, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 31682 + }, + { + "epoch": 0.31683, + "grad_norm": 0.6758038401603699, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 31683 + }, + { + "epoch": 0.31684, + "grad_norm": 0.7809858918190002, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 31684 + }, + { + "epoch": 0.31685, + "grad_norm": 0.808178186416626, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 31685 + }, + { + "epoch": 0.31686, + "grad_norm": 0.752869188785553, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 31686 + }, + { + "epoch": 0.31687, + "grad_norm": 0.6905260682106018, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 31687 + }, + { + "epoch": 0.31688, + "grad_norm": 0.6585781574249268, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 31688 + }, + { + "epoch": 0.31689, + "grad_norm": 0.6920794248580933, + "learning_rate": 0.003, + "loss": 3.991, + "step": 31689 + }, + { + "epoch": 0.3169, + "grad_norm": 0.6998615264892578, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 31690 + }, + { + "epoch": 0.31691, + "grad_norm": 0.7228922843933105, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 31691 + }, + { + "epoch": 0.31692, + "grad_norm": 0.9704743027687073, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 31692 + }, + { + "epoch": 0.31693, + "grad_norm": 1.2422326803207397, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 31693 + }, + { + "epoch": 0.31694, + "grad_norm": 0.6637818217277527, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 31694 + }, + { + "epoch": 0.31695, + "grad_norm": 0.6624817252159119, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 31695 + }, + { + "epoch": 0.31696, + "grad_norm": 0.816455066204071, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 31696 + }, + { + "epoch": 0.31697, + "grad_norm": 0.9992376565933228, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 31697 + }, + { + "epoch": 0.31698, + "grad_norm": 1.1654653549194336, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 31698 + }, + { + "epoch": 0.31699, + "grad_norm": 0.819172739982605, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 31699 + }, + { + "epoch": 0.317, + "grad_norm": 0.771256685256958, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 31700 + }, + { + "epoch": 0.31701, + "grad_norm": 0.7711849808692932, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 31701 + }, + { + "epoch": 0.31702, + "grad_norm": 0.8628012537956238, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 31702 + }, + { + "epoch": 0.31703, + "grad_norm": 0.9250704050064087, + "learning_rate": 0.003, + "loss": 3.991, + "step": 31703 + }, + { + "epoch": 0.31704, + "grad_norm": 1.0327938795089722, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 31704 + }, + { + "epoch": 0.31705, + "grad_norm": 1.1841731071472168, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 31705 + }, + { + "epoch": 0.31706, + "grad_norm": 0.9261520504951477, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 31706 + }, + { + "epoch": 0.31707, + "grad_norm": 0.9975132346153259, + "learning_rate": 0.003, + "loss": 3.977, + "step": 31707 + }, + { + "epoch": 0.31708, + "grad_norm": 1.248265266418457, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 31708 + }, + { + "epoch": 0.31709, + "grad_norm": 0.791175127029419, + "learning_rate": 0.003, + "loss": 3.997, + "step": 31709 + }, + { + "epoch": 0.3171, + "grad_norm": 0.6334706544876099, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 31710 + }, + { + "epoch": 0.31711, + "grad_norm": 0.7502360939979553, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 31711 + }, + { + "epoch": 0.31712, + "grad_norm": 0.850875973701477, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 31712 + }, + { + "epoch": 0.31713, + "grad_norm": 0.8762100338935852, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 31713 + }, + { + "epoch": 0.31714, + "grad_norm": 0.916811466217041, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 31714 + }, + { + "epoch": 0.31715, + "grad_norm": 1.1223503351211548, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 31715 + }, + { + "epoch": 0.31716, + "grad_norm": 1.0630011558532715, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 31716 + }, + { + "epoch": 0.31717, + "grad_norm": 0.8261072039604187, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 31717 + }, + { + "epoch": 0.31718, + "grad_norm": 0.8008760809898376, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 31718 + }, + { + "epoch": 0.31719, + "grad_norm": 0.7562857866287231, + "learning_rate": 0.003, + "loss": 3.946, + "step": 31719 + }, + { + "epoch": 0.3172, + "grad_norm": 0.8164099454879761, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 31720 + }, + { + "epoch": 0.31721, + "grad_norm": 1.0283623933792114, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 31721 + }, + { + "epoch": 0.31722, + "grad_norm": 1.4021257162094116, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 31722 + }, + { + "epoch": 0.31723, + "grad_norm": 0.634458601474762, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 31723 + }, + { + "epoch": 0.31724, + "grad_norm": 0.6761046648025513, + "learning_rate": 0.003, + "loss": 3.984, + "step": 31724 + }, + { + "epoch": 0.31725, + "grad_norm": 0.8097880482673645, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 31725 + }, + { + "epoch": 0.31726, + "grad_norm": 0.9031808972358704, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 31726 + }, + { + "epoch": 0.31727, + "grad_norm": 1.03922438621521, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 31727 + }, + { + "epoch": 0.31728, + "grad_norm": 0.9882245063781738, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 31728 + }, + { + "epoch": 0.31729, + "grad_norm": 1.0473922491073608, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 31729 + }, + { + "epoch": 0.3173, + "grad_norm": 1.006161093711853, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 31730 + }, + { + "epoch": 0.31731, + "grad_norm": 1.2391293048858643, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 31731 + }, + { + "epoch": 0.31732, + "grad_norm": 0.8511433601379395, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 31732 + }, + { + "epoch": 0.31733, + "grad_norm": 0.856207013130188, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 31733 + }, + { + "epoch": 0.31734, + "grad_norm": 0.7343817949295044, + "learning_rate": 0.003, + "loss": 3.989, + "step": 31734 + }, + { + "epoch": 0.31735, + "grad_norm": 0.8542333841323853, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 31735 + }, + { + "epoch": 0.31736, + "grad_norm": 0.9797667860984802, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 31736 + }, + { + "epoch": 0.31737, + "grad_norm": 0.8505750298500061, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 31737 + }, + { + "epoch": 0.31738, + "grad_norm": 0.9176780581474304, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 31738 + }, + { + "epoch": 0.31739, + "grad_norm": 1.444811224937439, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 31739 + }, + { + "epoch": 0.3174, + "grad_norm": 1.0772298574447632, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 31740 + }, + { + "epoch": 0.31741, + "grad_norm": 0.8409402370452881, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 31741 + }, + { + "epoch": 0.31742, + "grad_norm": 0.7758532762527466, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 31742 + }, + { + "epoch": 0.31743, + "grad_norm": 0.8127393126487732, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 31743 + }, + { + "epoch": 0.31744, + "grad_norm": 0.8525290489196777, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 31744 + }, + { + "epoch": 0.31745, + "grad_norm": 0.8255677223205566, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 31745 + }, + { + "epoch": 0.31746, + "grad_norm": 0.9548231959342957, + "learning_rate": 0.003, + "loss": 3.979, + "step": 31746 + }, + { + "epoch": 0.31747, + "grad_norm": 1.244844913482666, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 31747 + }, + { + "epoch": 0.31748, + "grad_norm": 0.7245697379112244, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 31748 + }, + { + "epoch": 0.31749, + "grad_norm": 0.7187261581420898, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 31749 + }, + { + "epoch": 0.3175, + "grad_norm": 0.6988082528114319, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 31750 + }, + { + "epoch": 0.31751, + "grad_norm": 0.7626096606254578, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 31751 + }, + { + "epoch": 0.31752, + "grad_norm": 0.9735206365585327, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 31752 + }, + { + "epoch": 0.31753, + "grad_norm": 1.1300380229949951, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 31753 + }, + { + "epoch": 0.31754, + "grad_norm": 0.9549205899238586, + "learning_rate": 0.003, + "loss": 3.935, + "step": 31754 + }, + { + "epoch": 0.31755, + "grad_norm": 1.0835528373718262, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 31755 + }, + { + "epoch": 0.31756, + "grad_norm": 1.1719111204147339, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 31756 + }, + { + "epoch": 0.31757, + "grad_norm": 0.8344889879226685, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 31757 + }, + { + "epoch": 0.31758, + "grad_norm": 0.7057539820671082, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 31758 + }, + { + "epoch": 0.31759, + "grad_norm": 0.6889806389808655, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 31759 + }, + { + "epoch": 0.3176, + "grad_norm": 0.7083370089530945, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 31760 + }, + { + "epoch": 0.31761, + "grad_norm": 0.8082462549209595, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 31761 + }, + { + "epoch": 0.31762, + "grad_norm": 0.9503257274627686, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 31762 + }, + { + "epoch": 0.31763, + "grad_norm": 0.9357032179832458, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 31763 + }, + { + "epoch": 0.31764, + "grad_norm": 0.9003579616546631, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 31764 + }, + { + "epoch": 0.31765, + "grad_norm": 0.8130712509155273, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 31765 + }, + { + "epoch": 0.31766, + "grad_norm": 0.6421576142311096, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 31766 + }, + { + "epoch": 0.31767, + "grad_norm": 0.620219349861145, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 31767 + }, + { + "epoch": 0.31768, + "grad_norm": 0.612430989742279, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 31768 + }, + { + "epoch": 0.31769, + "grad_norm": 0.6855059266090393, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 31769 + }, + { + "epoch": 0.3177, + "grad_norm": 0.7761200666427612, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 31770 + }, + { + "epoch": 0.31771, + "grad_norm": 1.0346169471740723, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 31771 + }, + { + "epoch": 0.31772, + "grad_norm": 1.294021487236023, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 31772 + }, + { + "epoch": 0.31773, + "grad_norm": 0.6247462630271912, + "learning_rate": 0.003, + "loss": 3.991, + "step": 31773 + }, + { + "epoch": 0.31774, + "grad_norm": 0.6427240371704102, + "learning_rate": 0.003, + "loss": 4.004, + "step": 31774 + }, + { + "epoch": 0.31775, + "grad_norm": 0.7129053473472595, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 31775 + }, + { + "epoch": 0.31776, + "grad_norm": 0.6727876663208008, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 31776 + }, + { + "epoch": 0.31777, + "grad_norm": 0.6286892294883728, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 31777 + }, + { + "epoch": 0.31778, + "grad_norm": 0.740919291973114, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 31778 + }, + { + "epoch": 0.31779, + "grad_norm": 0.9947946071624756, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 31779 + }, + { + "epoch": 0.3178, + "grad_norm": 1.4694466590881348, + "learning_rate": 0.003, + "loss": 3.975, + "step": 31780 + }, + { + "epoch": 0.31781, + "grad_norm": 0.5902783274650574, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 31781 + }, + { + "epoch": 0.31782, + "grad_norm": 0.9994298219680786, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 31782 + }, + { + "epoch": 0.31783, + "grad_norm": 1.2252731323242188, + "learning_rate": 0.003, + "loss": 3.988, + "step": 31783 + }, + { + "epoch": 0.31784, + "grad_norm": 0.722396969795227, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 31784 + }, + { + "epoch": 0.31785, + "grad_norm": 0.6646032929420471, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 31785 + }, + { + "epoch": 0.31786, + "grad_norm": 0.7301430702209473, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 31786 + }, + { + "epoch": 0.31787, + "grad_norm": 0.7271693348884583, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 31787 + }, + { + "epoch": 0.31788, + "grad_norm": 0.8197700381278992, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 31788 + }, + { + "epoch": 0.31789, + "grad_norm": 1.097848653793335, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 31789 + }, + { + "epoch": 0.3179, + "grad_norm": 1.0115256309509277, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 31790 + }, + { + "epoch": 0.31791, + "grad_norm": 1.0423035621643066, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 31791 + }, + { + "epoch": 0.31792, + "grad_norm": 1.0030487775802612, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 31792 + }, + { + "epoch": 0.31793, + "grad_norm": 0.9996585249900818, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 31793 + }, + { + "epoch": 0.31794, + "grad_norm": 1.058650255203247, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 31794 + }, + { + "epoch": 0.31795, + "grad_norm": 0.9676689505577087, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 31795 + }, + { + "epoch": 0.31796, + "grad_norm": 0.8451946377754211, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 31796 + }, + { + "epoch": 0.31797, + "grad_norm": 0.908967137336731, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 31797 + }, + { + "epoch": 0.31798, + "grad_norm": 0.9195347428321838, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 31798 + }, + { + "epoch": 0.31799, + "grad_norm": 0.8544921278953552, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 31799 + }, + { + "epoch": 0.318, + "grad_norm": 0.9118391275405884, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 31800 + }, + { + "epoch": 0.31801, + "grad_norm": 1.137799859046936, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 31801 + }, + { + "epoch": 0.31802, + "grad_norm": 0.9834898114204407, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 31802 + }, + { + "epoch": 0.31803, + "grad_norm": 1.1282732486724854, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 31803 + }, + { + "epoch": 0.31804, + "grad_norm": 0.9380391836166382, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 31804 + }, + { + "epoch": 0.31805, + "grad_norm": 0.854851245880127, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 31805 + }, + { + "epoch": 0.31806, + "grad_norm": 0.919080913066864, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 31806 + }, + { + "epoch": 0.31807, + "grad_norm": 0.928793728351593, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 31807 + }, + { + "epoch": 0.31808, + "grad_norm": 1.0575717687606812, + "learning_rate": 0.003, + "loss": 3.999, + "step": 31808 + }, + { + "epoch": 0.31809, + "grad_norm": 1.250876784324646, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 31809 + }, + { + "epoch": 0.3181, + "grad_norm": 0.6330140829086304, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 31810 + }, + { + "epoch": 0.31811, + "grad_norm": 0.6328076720237732, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 31811 + }, + { + "epoch": 0.31812, + "grad_norm": 0.691418468952179, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 31812 + }, + { + "epoch": 0.31813, + "grad_norm": 0.8090693950653076, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 31813 + }, + { + "epoch": 0.31814, + "grad_norm": 1.01861572265625, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 31814 + }, + { + "epoch": 0.31815, + "grad_norm": 1.4345952272415161, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 31815 + }, + { + "epoch": 0.31816, + "grad_norm": 0.6068592667579651, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 31816 + }, + { + "epoch": 0.31817, + "grad_norm": 1.117012858390808, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 31817 + }, + { + "epoch": 0.31818, + "grad_norm": 1.0835974216461182, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 31818 + }, + { + "epoch": 0.31819, + "grad_norm": 0.6589281558990479, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 31819 + }, + { + "epoch": 0.3182, + "grad_norm": 0.8035961389541626, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 31820 + }, + { + "epoch": 0.31821, + "grad_norm": 1.1727505922317505, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 31821 + }, + { + "epoch": 0.31822, + "grad_norm": 0.9814245700836182, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 31822 + }, + { + "epoch": 0.31823, + "grad_norm": 0.8495355248451233, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 31823 + }, + { + "epoch": 0.31824, + "grad_norm": 0.7615958452224731, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 31824 + }, + { + "epoch": 0.31825, + "grad_norm": 0.7221822738647461, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 31825 + }, + { + "epoch": 0.31826, + "grad_norm": 0.6991267800331116, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 31826 + }, + { + "epoch": 0.31827, + "grad_norm": 0.761833906173706, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 31827 + }, + { + "epoch": 0.31828, + "grad_norm": 0.7387428283691406, + "learning_rate": 0.003, + "loss": 3.9498, + "step": 31828 + }, + { + "epoch": 0.31829, + "grad_norm": 0.6951528191566467, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 31829 + }, + { + "epoch": 0.3183, + "grad_norm": 0.6106337308883667, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 31830 + }, + { + "epoch": 0.31831, + "grad_norm": 0.6223343014717102, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 31831 + }, + { + "epoch": 0.31832, + "grad_norm": 0.66086345911026, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 31832 + }, + { + "epoch": 0.31833, + "grad_norm": 0.7564086318016052, + "learning_rate": 0.003, + "loss": 3.994, + "step": 31833 + }, + { + "epoch": 0.31834, + "grad_norm": 0.9723995923995972, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 31834 + }, + { + "epoch": 0.31835, + "grad_norm": 1.0781182050704956, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 31835 + }, + { + "epoch": 0.31836, + "grad_norm": 1.0347298383712769, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 31836 + }, + { + "epoch": 0.31837, + "grad_norm": 0.9686373472213745, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 31837 + }, + { + "epoch": 0.31838, + "grad_norm": 0.9500611424446106, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 31838 + }, + { + "epoch": 0.31839, + "grad_norm": 0.9354565739631653, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 31839 + }, + { + "epoch": 0.3184, + "grad_norm": 0.9826027154922485, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 31840 + }, + { + "epoch": 0.31841, + "grad_norm": 0.8233250379562378, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 31841 + }, + { + "epoch": 0.31842, + "grad_norm": 0.9076042771339417, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 31842 + }, + { + "epoch": 0.31843, + "grad_norm": 0.8545289635658264, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 31843 + }, + { + "epoch": 0.31844, + "grad_norm": 0.9894211292266846, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 31844 + }, + { + "epoch": 0.31845, + "grad_norm": 1.1263675689697266, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 31845 + }, + { + "epoch": 0.31846, + "grad_norm": 1.1251802444458008, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 31846 + }, + { + "epoch": 0.31847, + "grad_norm": 0.8648034930229187, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 31847 + }, + { + "epoch": 0.31848, + "grad_norm": 0.6889275908470154, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 31848 + }, + { + "epoch": 0.31849, + "grad_norm": 0.8119910955429077, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 31849 + }, + { + "epoch": 0.3185, + "grad_norm": 0.8435540795326233, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 31850 + }, + { + "epoch": 0.31851, + "grad_norm": 0.7579423785209656, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 31851 + }, + { + "epoch": 0.31852, + "grad_norm": 0.8567580580711365, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 31852 + }, + { + "epoch": 0.31853, + "grad_norm": 0.9540555477142334, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 31853 + }, + { + "epoch": 0.31854, + "grad_norm": 0.8542634844779968, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 31854 + }, + { + "epoch": 0.31855, + "grad_norm": 0.7486078143119812, + "learning_rate": 0.003, + "loss": 3.98, + "step": 31855 + }, + { + "epoch": 0.31856, + "grad_norm": 0.8155931830406189, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 31856 + }, + { + "epoch": 0.31857, + "grad_norm": 1.1752797365188599, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 31857 + }, + { + "epoch": 0.31858, + "grad_norm": 0.9710087776184082, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 31858 + }, + { + "epoch": 0.31859, + "grad_norm": 0.8305866718292236, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 31859 + }, + { + "epoch": 0.3186, + "grad_norm": 0.8492990136146545, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 31860 + }, + { + "epoch": 0.31861, + "grad_norm": 0.8320987820625305, + "learning_rate": 0.003, + "loss": 3.945, + "step": 31861 + }, + { + "epoch": 0.31862, + "grad_norm": 0.869533360004425, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 31862 + }, + { + "epoch": 0.31863, + "grad_norm": 0.9678048491477966, + "learning_rate": 0.003, + "loss": 3.986, + "step": 31863 + }, + { + "epoch": 0.31864, + "grad_norm": 1.227085828781128, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 31864 + }, + { + "epoch": 0.31865, + "grad_norm": 1.1210212707519531, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 31865 + }, + { + "epoch": 0.31866, + "grad_norm": 0.8381785750389099, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 31866 + }, + { + "epoch": 0.31867, + "grad_norm": 0.6773449778556824, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 31867 + }, + { + "epoch": 0.31868, + "grad_norm": 0.746036946773529, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 31868 + }, + { + "epoch": 0.31869, + "grad_norm": 0.7157782316207886, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 31869 + }, + { + "epoch": 0.3187, + "grad_norm": 0.6878886818885803, + "learning_rate": 0.003, + "loss": 3.965, + "step": 31870 + }, + { + "epoch": 0.31871, + "grad_norm": 0.7600991725921631, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 31871 + }, + { + "epoch": 0.31872, + "grad_norm": 1.0150800943374634, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 31872 + }, + { + "epoch": 0.31873, + "grad_norm": 1.1399106979370117, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 31873 + }, + { + "epoch": 0.31874, + "grad_norm": 0.7033064365386963, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 31874 + }, + { + "epoch": 0.31875, + "grad_norm": 0.608832061290741, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 31875 + }, + { + "epoch": 0.31876, + "grad_norm": 0.6436609625816345, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 31876 + }, + { + "epoch": 0.31877, + "grad_norm": 0.6003946661949158, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 31877 + }, + { + "epoch": 0.31878, + "grad_norm": 0.6700440049171448, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 31878 + }, + { + "epoch": 0.31879, + "grad_norm": 0.6995030641555786, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 31879 + }, + { + "epoch": 0.3188, + "grad_norm": 0.8816395998001099, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 31880 + }, + { + "epoch": 0.31881, + "grad_norm": 1.128891110420227, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 31881 + }, + { + "epoch": 0.31882, + "grad_norm": 1.4259902238845825, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 31882 + }, + { + "epoch": 0.31883, + "grad_norm": 0.5541914701461792, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 31883 + }, + { + "epoch": 0.31884, + "grad_norm": 0.8740122318267822, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 31884 + }, + { + "epoch": 0.31885, + "grad_norm": 1.2284598350524902, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 31885 + }, + { + "epoch": 0.31886, + "grad_norm": 0.7731668949127197, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 31886 + }, + { + "epoch": 0.31887, + "grad_norm": 0.9066620469093323, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 31887 + }, + { + "epoch": 0.31888, + "grad_norm": 1.1326112747192383, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 31888 + }, + { + "epoch": 0.31889, + "grad_norm": 1.0159990787506104, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 31889 + }, + { + "epoch": 0.3189, + "grad_norm": 0.8653330206871033, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 31890 + }, + { + "epoch": 0.31891, + "grad_norm": 0.8110001087188721, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 31891 + }, + { + "epoch": 0.31892, + "grad_norm": 0.8207891583442688, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 31892 + }, + { + "epoch": 0.31893, + "grad_norm": 0.876427948474884, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 31893 + }, + { + "epoch": 0.31894, + "grad_norm": 0.9153008460998535, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 31894 + }, + { + "epoch": 0.31895, + "grad_norm": 1.1140238046646118, + "learning_rate": 0.003, + "loss": 4.012, + "step": 31895 + }, + { + "epoch": 0.31896, + "grad_norm": 1.0174435377120972, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 31896 + }, + { + "epoch": 0.31897, + "grad_norm": 0.9680898785591125, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 31897 + }, + { + "epoch": 0.31898, + "grad_norm": 1.083661437034607, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 31898 + }, + { + "epoch": 0.31899, + "grad_norm": 1.105623483657837, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 31899 + }, + { + "epoch": 0.319, + "grad_norm": 1.1407814025878906, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 31900 + }, + { + "epoch": 0.31901, + "grad_norm": 0.9970688223838806, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 31901 + }, + { + "epoch": 0.31902, + "grad_norm": 0.89243084192276, + "learning_rate": 0.003, + "loss": 3.983, + "step": 31902 + }, + { + "epoch": 0.31903, + "grad_norm": 0.7973852753639221, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 31903 + }, + { + "epoch": 0.31904, + "grad_norm": 0.8219707012176514, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 31904 + }, + { + "epoch": 0.31905, + "grad_norm": 0.7502878904342651, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 31905 + }, + { + "epoch": 0.31906, + "grad_norm": 0.8150677680969238, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 31906 + }, + { + "epoch": 0.31907, + "grad_norm": 0.9132183194160461, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 31907 + }, + { + "epoch": 0.31908, + "grad_norm": 0.9461619257926941, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 31908 + }, + { + "epoch": 0.31909, + "grad_norm": 0.9853398203849792, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 31909 + }, + { + "epoch": 0.3191, + "grad_norm": 1.1081980466842651, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 31910 + }, + { + "epoch": 0.31911, + "grad_norm": 0.9336178302764893, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 31911 + }, + { + "epoch": 0.31912, + "grad_norm": 0.8408560156822205, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 31912 + }, + { + "epoch": 0.31913, + "grad_norm": 0.7662078142166138, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 31913 + }, + { + "epoch": 0.31914, + "grad_norm": 0.7586273550987244, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 31914 + }, + { + "epoch": 0.31915, + "grad_norm": 0.784388542175293, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 31915 + }, + { + "epoch": 0.31916, + "grad_norm": 0.7081760764122009, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 31916 + }, + { + "epoch": 0.31917, + "grad_norm": 0.8358636498451233, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 31917 + }, + { + "epoch": 0.31918, + "grad_norm": 0.8944282531738281, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 31918 + }, + { + "epoch": 0.31919, + "grad_norm": 1.190651297569275, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 31919 + }, + { + "epoch": 0.3192, + "grad_norm": 0.9930619597434998, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 31920 + }, + { + "epoch": 0.31921, + "grad_norm": 1.0406173467636108, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 31921 + }, + { + "epoch": 0.31922, + "grad_norm": 1.0267120599746704, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 31922 + }, + { + "epoch": 0.31923, + "grad_norm": 0.9404387474060059, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 31923 + }, + { + "epoch": 0.31924, + "grad_norm": 0.9527109265327454, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 31924 + }, + { + "epoch": 0.31925, + "grad_norm": 0.9819561839103699, + "learning_rate": 0.003, + "loss": 3.998, + "step": 31925 + }, + { + "epoch": 0.31926, + "grad_norm": 0.9993308186531067, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 31926 + }, + { + "epoch": 0.31927, + "grad_norm": 1.0661183595657349, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 31927 + }, + { + "epoch": 0.31928, + "grad_norm": 0.8764485120773315, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 31928 + }, + { + "epoch": 0.31929, + "grad_norm": 0.7623227834701538, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 31929 + }, + { + "epoch": 0.3193, + "grad_norm": 0.7056547403335571, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 31930 + }, + { + "epoch": 0.31931, + "grad_norm": 0.6098735928535461, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 31931 + }, + { + "epoch": 0.31932, + "grad_norm": 0.6026955842971802, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 31932 + }, + { + "epoch": 0.31933, + "grad_norm": 0.6267229318618774, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 31933 + }, + { + "epoch": 0.31934, + "grad_norm": 0.6613836884498596, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 31934 + }, + { + "epoch": 0.31935, + "grad_norm": 0.7172445058822632, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 31935 + }, + { + "epoch": 0.31936, + "grad_norm": 0.7846229076385498, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 31936 + }, + { + "epoch": 0.31937, + "grad_norm": 0.9621963500976562, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 31937 + }, + { + "epoch": 0.31938, + "grad_norm": 1.2367007732391357, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 31938 + }, + { + "epoch": 0.31939, + "grad_norm": 0.9796053767204285, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 31939 + }, + { + "epoch": 0.3194, + "grad_norm": 0.8778676390647888, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 31940 + }, + { + "epoch": 0.31941, + "grad_norm": 0.8051837682723999, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 31941 + }, + { + "epoch": 0.31942, + "grad_norm": 0.9197815656661987, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 31942 + }, + { + "epoch": 0.31943, + "grad_norm": 0.9260678291320801, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 31943 + }, + { + "epoch": 0.31944, + "grad_norm": 0.893118679523468, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 31944 + }, + { + "epoch": 0.31945, + "grad_norm": 0.8913757801055908, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 31945 + }, + { + "epoch": 0.31946, + "grad_norm": 0.9368793368339539, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 31946 + }, + { + "epoch": 0.31947, + "grad_norm": 1.0138922929763794, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 31947 + }, + { + "epoch": 0.31948, + "grad_norm": 1.126649260520935, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 31948 + }, + { + "epoch": 0.31949, + "grad_norm": 0.8848251700401306, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 31949 + }, + { + "epoch": 0.3195, + "grad_norm": 0.7945773005485535, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 31950 + }, + { + "epoch": 0.31951, + "grad_norm": 0.7909538149833679, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 31951 + }, + { + "epoch": 0.31952, + "grad_norm": 0.9891659617424011, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 31952 + }, + { + "epoch": 0.31953, + "grad_norm": 1.1875312328338623, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 31953 + }, + { + "epoch": 0.31954, + "grad_norm": 1.0523157119750977, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 31954 + }, + { + "epoch": 0.31955, + "grad_norm": 0.9876320362091064, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 31955 + }, + { + "epoch": 0.31956, + "grad_norm": 0.9307363033294678, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 31956 + }, + { + "epoch": 0.31957, + "grad_norm": 0.9263972043991089, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 31957 + }, + { + "epoch": 0.31958, + "grad_norm": 1.0037956237792969, + "learning_rate": 0.003, + "loss": 3.989, + "step": 31958 + }, + { + "epoch": 0.31959, + "grad_norm": 0.8445433974266052, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 31959 + }, + { + "epoch": 0.3196, + "grad_norm": 0.8871384263038635, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 31960 + }, + { + "epoch": 0.31961, + "grad_norm": 1.0519720315933228, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 31961 + }, + { + "epoch": 0.31962, + "grad_norm": 0.9678213596343994, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 31962 + }, + { + "epoch": 0.31963, + "grad_norm": 0.9001585841178894, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 31963 + }, + { + "epoch": 0.31964, + "grad_norm": 1.0312457084655762, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 31964 + }, + { + "epoch": 0.31965, + "grad_norm": 1.382035493850708, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 31965 + }, + { + "epoch": 0.31966, + "grad_norm": 0.9601774215698242, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 31966 + }, + { + "epoch": 0.31967, + "grad_norm": 0.8596053123474121, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 31967 + }, + { + "epoch": 0.31968, + "grad_norm": 0.7014753818511963, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 31968 + }, + { + "epoch": 0.31969, + "grad_norm": 0.7134129405021667, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 31969 + }, + { + "epoch": 0.3197, + "grad_norm": 0.6813555359840393, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 31970 + }, + { + "epoch": 0.31971, + "grad_norm": 0.6111615896224976, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 31971 + }, + { + "epoch": 0.31972, + "grad_norm": 0.6742223501205444, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 31972 + }, + { + "epoch": 0.31973, + "grad_norm": 0.8781028389930725, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 31973 + }, + { + "epoch": 0.31974, + "grad_norm": 1.1583718061447144, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 31974 + }, + { + "epoch": 0.31975, + "grad_norm": 1.1759109497070312, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 31975 + }, + { + "epoch": 0.31976, + "grad_norm": 0.7372112274169922, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 31976 + }, + { + "epoch": 0.31977, + "grad_norm": 0.6202447414398193, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 31977 + }, + { + "epoch": 0.31978, + "grad_norm": 0.8121834397315979, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 31978 + }, + { + "epoch": 0.31979, + "grad_norm": 1.0579547882080078, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 31979 + }, + { + "epoch": 0.3198, + "grad_norm": 1.106669306755066, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 31980 + }, + { + "epoch": 0.31981, + "grad_norm": 0.8236245512962341, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 31981 + }, + { + "epoch": 0.31982, + "grad_norm": 0.7657919526100159, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 31982 + }, + { + "epoch": 0.31983, + "grad_norm": 0.7493160963058472, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 31983 + }, + { + "epoch": 0.31984, + "grad_norm": 0.805999755859375, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 31984 + }, + { + "epoch": 0.31985, + "grad_norm": 0.740569531917572, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 31985 + }, + { + "epoch": 0.31986, + "grad_norm": 0.766746461391449, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 31986 + }, + { + "epoch": 0.31987, + "grad_norm": 0.8620671033859253, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 31987 + }, + { + "epoch": 0.31988, + "grad_norm": 1.0276273488998413, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 31988 + }, + { + "epoch": 0.31989, + "grad_norm": 0.9191228151321411, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 31989 + }, + { + "epoch": 0.3199, + "grad_norm": 0.929714560508728, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 31990 + }, + { + "epoch": 0.31991, + "grad_norm": 1.0309498310089111, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 31991 + }, + { + "epoch": 0.31992, + "grad_norm": 1.191603183746338, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 31992 + }, + { + "epoch": 0.31993, + "grad_norm": 0.8476164937019348, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 31993 + }, + { + "epoch": 0.31994, + "grad_norm": 0.798712968826294, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 31994 + }, + { + "epoch": 0.31995, + "grad_norm": 0.8979770541191101, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 31995 + }, + { + "epoch": 0.31996, + "grad_norm": 1.2367608547210693, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 31996 + }, + { + "epoch": 0.31997, + "grad_norm": 0.9919266104698181, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 31997 + }, + { + "epoch": 0.31998, + "grad_norm": 1.0319197177886963, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 31998 + }, + { + "epoch": 0.31999, + "grad_norm": 1.0581855773925781, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 31999 + }, + { + "epoch": 0.32, + "grad_norm": 0.9638508558273315, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 32000 + }, + { + "epoch": 0.32001, + "grad_norm": 0.8881365656852722, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 32001 + }, + { + "epoch": 0.32002, + "grad_norm": 1.0006725788116455, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 32002 + }, + { + "epoch": 0.32003, + "grad_norm": 1.1027144193649292, + "learning_rate": 0.003, + "loss": 3.992, + "step": 32003 + }, + { + "epoch": 0.32004, + "grad_norm": 0.8521268367767334, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 32004 + }, + { + "epoch": 0.32005, + "grad_norm": 0.7168695330619812, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 32005 + }, + { + "epoch": 0.32006, + "grad_norm": 0.7950488924980164, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 32006 + }, + { + "epoch": 0.32007, + "grad_norm": 0.8807989358901978, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 32007 + }, + { + "epoch": 0.32008, + "grad_norm": 1.0010309219360352, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 32008 + }, + { + "epoch": 0.32009, + "grad_norm": 1.000480055809021, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 32009 + }, + { + "epoch": 0.3201, + "grad_norm": 0.9359336495399475, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 32010 + }, + { + "epoch": 0.32011, + "grad_norm": 0.918449342250824, + "learning_rate": 0.003, + "loss": 3.989, + "step": 32011 + }, + { + "epoch": 0.32012, + "grad_norm": 1.0295767784118652, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 32012 + }, + { + "epoch": 0.32013, + "grad_norm": 1.1043448448181152, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 32013 + }, + { + "epoch": 0.32014, + "grad_norm": 0.8271875381469727, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 32014 + }, + { + "epoch": 0.32015, + "grad_norm": 0.8598615527153015, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 32015 + }, + { + "epoch": 0.32016, + "grad_norm": 1.0244101285934448, + "learning_rate": 0.003, + "loss": 4.006, + "step": 32016 + }, + { + "epoch": 0.32017, + "grad_norm": 1.2048462629318237, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 32017 + }, + { + "epoch": 0.32018, + "grad_norm": 0.947196900844574, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 32018 + }, + { + "epoch": 0.32019, + "grad_norm": 0.8753867149353027, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 32019 + }, + { + "epoch": 0.3202, + "grad_norm": 0.7672975659370422, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 32020 + }, + { + "epoch": 0.32021, + "grad_norm": 0.6597415208816528, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 32021 + }, + { + "epoch": 0.32022, + "grad_norm": 0.675281822681427, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 32022 + }, + { + "epoch": 0.32023, + "grad_norm": 0.7094748020172119, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 32023 + }, + { + "epoch": 0.32024, + "grad_norm": 0.6927016973495483, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 32024 + }, + { + "epoch": 0.32025, + "grad_norm": 0.8949888348579407, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 32025 + }, + { + "epoch": 0.32026, + "grad_norm": 1.225685477256775, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 32026 + }, + { + "epoch": 0.32027, + "grad_norm": 0.8576503992080688, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 32027 + }, + { + "epoch": 0.32028, + "grad_norm": 0.5893262028694153, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 32028 + }, + { + "epoch": 0.32029, + "grad_norm": 0.6179031729698181, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 32029 + }, + { + "epoch": 0.3203, + "grad_norm": 0.6709585189819336, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 32030 + }, + { + "epoch": 0.32031, + "grad_norm": 0.6312055587768555, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 32031 + }, + { + "epoch": 0.32032, + "grad_norm": 0.6012716889381409, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 32032 + }, + { + "epoch": 0.32033, + "grad_norm": 0.5388962626457214, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 32033 + }, + { + "epoch": 0.32034, + "grad_norm": 0.4804224669933319, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 32034 + }, + { + "epoch": 0.32035, + "grad_norm": 0.49112576246261597, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 32035 + }, + { + "epoch": 0.32036, + "grad_norm": 0.6058269143104553, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 32036 + }, + { + "epoch": 0.32037, + "grad_norm": 0.9101688265800476, + "learning_rate": 0.003, + "loss": 3.9489, + "step": 32037 + }, + { + "epoch": 0.32038, + "grad_norm": 1.3960705995559692, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 32038 + }, + { + "epoch": 0.32039, + "grad_norm": 0.7695090770721436, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 32039 + }, + { + "epoch": 0.3204, + "grad_norm": 0.6576930284500122, + "learning_rate": 0.003, + "loss": 3.963, + "step": 32040 + }, + { + "epoch": 0.32041, + "grad_norm": 0.8217552900314331, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 32041 + }, + { + "epoch": 0.32042, + "grad_norm": 0.8888188600540161, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 32042 + }, + { + "epoch": 0.32043, + "grad_norm": 0.9816799163818359, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 32043 + }, + { + "epoch": 0.32044, + "grad_norm": 1.3380223512649536, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 32044 + }, + { + "epoch": 0.32045, + "grad_norm": 0.6751304864883423, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 32045 + }, + { + "epoch": 0.32046, + "grad_norm": 0.7318328022956848, + "learning_rate": 0.003, + "loss": 3.977, + "step": 32046 + }, + { + "epoch": 0.32047, + "grad_norm": 0.8708478212356567, + "learning_rate": 0.003, + "loss": 3.968, + "step": 32047 + }, + { + "epoch": 0.32048, + "grad_norm": 0.8206448554992676, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 32048 + }, + { + "epoch": 0.32049, + "grad_norm": 0.8908599019050598, + "learning_rate": 0.003, + "loss": 3.985, + "step": 32049 + }, + { + "epoch": 0.3205, + "grad_norm": 1.0488476753234863, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 32050 + }, + { + "epoch": 0.32051, + "grad_norm": 1.1786977052688599, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 32051 + }, + { + "epoch": 0.32052, + "grad_norm": 0.7773699164390564, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 32052 + }, + { + "epoch": 0.32053, + "grad_norm": 0.6974233984947205, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 32053 + }, + { + "epoch": 0.32054, + "grad_norm": 0.637326717376709, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 32054 + }, + { + "epoch": 0.32055, + "grad_norm": 0.7553818225860596, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 32055 + }, + { + "epoch": 0.32056, + "grad_norm": 0.8503643870353699, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 32056 + }, + { + "epoch": 0.32057, + "grad_norm": 0.8968907594680786, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 32057 + }, + { + "epoch": 0.32058, + "grad_norm": 0.9432955980300903, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 32058 + }, + { + "epoch": 0.32059, + "grad_norm": 0.9564017653465271, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 32059 + }, + { + "epoch": 0.3206, + "grad_norm": 0.9587700963020325, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 32060 + }, + { + "epoch": 0.32061, + "grad_norm": 1.1753751039505005, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 32061 + }, + { + "epoch": 0.32062, + "grad_norm": 0.9436876177787781, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 32062 + }, + { + "epoch": 0.32063, + "grad_norm": 1.015836477279663, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 32063 + }, + { + "epoch": 0.32064, + "grad_norm": 1.1180943250656128, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 32064 + }, + { + "epoch": 0.32065, + "grad_norm": 0.9623943567276001, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 32065 + }, + { + "epoch": 0.32066, + "grad_norm": 0.8645842671394348, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 32066 + }, + { + "epoch": 0.32067, + "grad_norm": 0.8607925772666931, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 32067 + }, + { + "epoch": 0.32068, + "grad_norm": 0.8778126835823059, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 32068 + }, + { + "epoch": 0.32069, + "grad_norm": 0.9677315950393677, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 32069 + }, + { + "epoch": 0.3207, + "grad_norm": 0.9333333373069763, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 32070 + }, + { + "epoch": 0.32071, + "grad_norm": 0.9463524222373962, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 32071 + }, + { + "epoch": 0.32072, + "grad_norm": 0.960856556892395, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 32072 + }, + { + "epoch": 0.32073, + "grad_norm": 0.9155762791633606, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 32073 + }, + { + "epoch": 0.32074, + "grad_norm": 0.8109015822410583, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 32074 + }, + { + "epoch": 0.32075, + "grad_norm": 0.7097962498664856, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 32075 + }, + { + "epoch": 0.32076, + "grad_norm": 0.6270773410797119, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 32076 + }, + { + "epoch": 0.32077, + "grad_norm": 0.6736939549446106, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 32077 + }, + { + "epoch": 0.32078, + "grad_norm": 0.8834468126296997, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 32078 + }, + { + "epoch": 0.32079, + "grad_norm": 0.8992695212364197, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 32079 + }, + { + "epoch": 0.3208, + "grad_norm": 0.8740280866622925, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 32080 + }, + { + "epoch": 0.32081, + "grad_norm": 0.9242718815803528, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 32081 + }, + { + "epoch": 0.32082, + "grad_norm": 1.0411876440048218, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 32082 + }, + { + "epoch": 0.32083, + "grad_norm": 0.9944360852241516, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 32083 + }, + { + "epoch": 0.32084, + "grad_norm": 1.0596736669540405, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 32084 + }, + { + "epoch": 0.32085, + "grad_norm": 0.9781474471092224, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 32085 + }, + { + "epoch": 0.32086, + "grad_norm": 1.0314379930496216, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 32086 + }, + { + "epoch": 0.32087, + "grad_norm": 1.136717438697815, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 32087 + }, + { + "epoch": 0.32088, + "grad_norm": 1.3176428079605103, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 32088 + }, + { + "epoch": 0.32089, + "grad_norm": 0.7489807605743408, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 32089 + }, + { + "epoch": 0.3209, + "grad_norm": 0.6652177572250366, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 32090 + }, + { + "epoch": 0.32091, + "grad_norm": 0.779028594493866, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 32091 + }, + { + "epoch": 0.32092, + "grad_norm": 0.8479487895965576, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 32092 + }, + { + "epoch": 0.32093, + "grad_norm": 0.9498677253723145, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 32093 + }, + { + "epoch": 0.32094, + "grad_norm": 1.1844836473464966, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 32094 + }, + { + "epoch": 0.32095, + "grad_norm": 0.9266604781150818, + "learning_rate": 0.003, + "loss": 4.012, + "step": 32095 + }, + { + "epoch": 0.32096, + "grad_norm": 0.9776513576507568, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 32096 + }, + { + "epoch": 0.32097, + "grad_norm": 0.9743935465812683, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 32097 + }, + { + "epoch": 0.32098, + "grad_norm": 0.9046016335487366, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 32098 + }, + { + "epoch": 0.32099, + "grad_norm": 0.889572262763977, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 32099 + }, + { + "epoch": 0.321, + "grad_norm": 0.8822141289710999, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 32100 + }, + { + "epoch": 0.32101, + "grad_norm": 0.8343291282653809, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 32101 + }, + { + "epoch": 0.32102, + "grad_norm": 0.7870604395866394, + "learning_rate": 0.003, + "loss": 3.947, + "step": 32102 + }, + { + "epoch": 0.32103, + "grad_norm": 0.6847182512283325, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 32103 + }, + { + "epoch": 0.32104, + "grad_norm": 0.931690514087677, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 32104 + }, + { + "epoch": 0.32105, + "grad_norm": 1.3415247201919556, + "learning_rate": 0.003, + "loss": 4.004, + "step": 32105 + }, + { + "epoch": 0.32106, + "grad_norm": 0.5526766777038574, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 32106 + }, + { + "epoch": 0.32107, + "grad_norm": 0.6983759999275208, + "learning_rate": 0.003, + "loss": 4.015, + "step": 32107 + }, + { + "epoch": 0.32108, + "grad_norm": 0.9777393341064453, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 32108 + }, + { + "epoch": 0.32109, + "grad_norm": 1.1710542440414429, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 32109 + }, + { + "epoch": 0.3211, + "grad_norm": 0.7220216393470764, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 32110 + }, + { + "epoch": 0.32111, + "grad_norm": 0.5530011653900146, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 32111 + }, + { + "epoch": 0.32112, + "grad_norm": 0.5857248902320862, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 32112 + }, + { + "epoch": 0.32113, + "grad_norm": 0.6346360445022583, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 32113 + }, + { + "epoch": 0.32114, + "grad_norm": 0.6644352674484253, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 32114 + }, + { + "epoch": 0.32115, + "grad_norm": 0.7527903318405151, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 32115 + }, + { + "epoch": 0.32116, + "grad_norm": 0.88616544008255, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 32116 + }, + { + "epoch": 0.32117, + "grad_norm": 0.9311866760253906, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 32117 + }, + { + "epoch": 0.32118, + "grad_norm": 0.9333279132843018, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 32118 + }, + { + "epoch": 0.32119, + "grad_norm": 0.9234660267829895, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 32119 + }, + { + "epoch": 0.3212, + "grad_norm": 0.9395178556442261, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 32120 + }, + { + "epoch": 0.32121, + "grad_norm": 1.007749080657959, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 32121 + }, + { + "epoch": 0.32122, + "grad_norm": 0.8310779929161072, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 32122 + }, + { + "epoch": 0.32123, + "grad_norm": 0.9075390100479126, + "learning_rate": 0.003, + "loss": 3.967, + "step": 32123 + }, + { + "epoch": 0.32124, + "grad_norm": 1.0297671556472778, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 32124 + }, + { + "epoch": 0.32125, + "grad_norm": 1.1459410190582275, + "learning_rate": 0.003, + "loss": 4.015, + "step": 32125 + }, + { + "epoch": 0.32126, + "grad_norm": 0.9300817847251892, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 32126 + }, + { + "epoch": 0.32127, + "grad_norm": 1.0059821605682373, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 32127 + }, + { + "epoch": 0.32128, + "grad_norm": 0.9796208739280701, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 32128 + }, + { + "epoch": 0.32129, + "grad_norm": 0.9530287981033325, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 32129 + }, + { + "epoch": 0.3213, + "grad_norm": 0.9208945631980896, + "learning_rate": 0.003, + "loss": 3.951, + "step": 32130 + }, + { + "epoch": 0.32131, + "grad_norm": 1.0209599733352661, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 32131 + }, + { + "epoch": 0.32132, + "grad_norm": 1.0412724018096924, + "learning_rate": 0.003, + "loss": 4.009, + "step": 32132 + }, + { + "epoch": 0.32133, + "grad_norm": 0.9200287461280823, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 32133 + }, + { + "epoch": 0.32134, + "grad_norm": 1.048357367515564, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 32134 + }, + { + "epoch": 0.32135, + "grad_norm": 0.9670416116714478, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 32135 + }, + { + "epoch": 0.32136, + "grad_norm": 1.1542972326278687, + "learning_rate": 0.003, + "loss": 3.998, + "step": 32136 + }, + { + "epoch": 0.32137, + "grad_norm": 1.0319467782974243, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 32137 + }, + { + "epoch": 0.32138, + "grad_norm": 1.1577683687210083, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 32138 + }, + { + "epoch": 0.32139, + "grad_norm": 0.8710386753082275, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 32139 + }, + { + "epoch": 0.3214, + "grad_norm": 0.799198567867279, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 32140 + }, + { + "epoch": 0.32141, + "grad_norm": 0.7255551218986511, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 32141 + }, + { + "epoch": 0.32142, + "grad_norm": 0.6706990599632263, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 32142 + }, + { + "epoch": 0.32143, + "grad_norm": 0.6818363666534424, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 32143 + }, + { + "epoch": 0.32144, + "grad_norm": 0.6886780261993408, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 32144 + }, + { + "epoch": 0.32145, + "grad_norm": 0.7653381824493408, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 32145 + }, + { + "epoch": 0.32146, + "grad_norm": 0.7307772040367126, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 32146 + }, + { + "epoch": 0.32147, + "grad_norm": 0.7395780086517334, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 32147 + }, + { + "epoch": 0.32148, + "grad_norm": 0.7620351910591125, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 32148 + }, + { + "epoch": 0.32149, + "grad_norm": 0.8308371901512146, + "learning_rate": 0.003, + "loss": 3.9276, + "step": 32149 + }, + { + "epoch": 0.3215, + "grad_norm": 0.8445034623146057, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 32150 + }, + { + "epoch": 0.32151, + "grad_norm": 0.8301157355308533, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 32151 + }, + { + "epoch": 0.32152, + "grad_norm": 0.901214063167572, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 32152 + }, + { + "epoch": 0.32153, + "grad_norm": 0.9436374306678772, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 32153 + }, + { + "epoch": 0.32154, + "grad_norm": 0.9442698359489441, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 32154 + }, + { + "epoch": 0.32155, + "grad_norm": 1.0354136228561401, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 32155 + }, + { + "epoch": 0.32156, + "grad_norm": 1.0016286373138428, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 32156 + }, + { + "epoch": 0.32157, + "grad_norm": 1.0169380903244019, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 32157 + }, + { + "epoch": 0.32158, + "grad_norm": 0.9238132238388062, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 32158 + }, + { + "epoch": 0.32159, + "grad_norm": 0.8453273773193359, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 32159 + }, + { + "epoch": 0.3216, + "grad_norm": 0.8452710509300232, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 32160 + }, + { + "epoch": 0.32161, + "grad_norm": 0.8068813681602478, + "learning_rate": 0.003, + "loss": 3.997, + "step": 32161 + }, + { + "epoch": 0.32162, + "grad_norm": 0.797330379486084, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 32162 + }, + { + "epoch": 0.32163, + "grad_norm": 0.8399118185043335, + "learning_rate": 0.003, + "loss": 4.021, + "step": 32163 + }, + { + "epoch": 0.32164, + "grad_norm": 0.8134660720825195, + "learning_rate": 0.003, + "loss": 3.982, + "step": 32164 + }, + { + "epoch": 0.32165, + "grad_norm": 0.7816616296768188, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 32165 + }, + { + "epoch": 0.32166, + "grad_norm": 0.8920309543609619, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 32166 + }, + { + "epoch": 0.32167, + "grad_norm": 0.9671622514724731, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 32167 + }, + { + "epoch": 0.32168, + "grad_norm": 1.062778115272522, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 32168 + }, + { + "epoch": 0.32169, + "grad_norm": 0.9995540380477905, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 32169 + }, + { + "epoch": 0.3217, + "grad_norm": 0.9716194868087769, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 32170 + }, + { + "epoch": 0.32171, + "grad_norm": 0.7753605842590332, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 32171 + }, + { + "epoch": 0.32172, + "grad_norm": 0.7075203061103821, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 32172 + }, + { + "epoch": 0.32173, + "grad_norm": 0.6950739622116089, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 32173 + }, + { + "epoch": 0.32174, + "grad_norm": 0.7546954154968262, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 32174 + }, + { + "epoch": 0.32175, + "grad_norm": 0.7145999073982239, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 32175 + }, + { + "epoch": 0.32176, + "grad_norm": 0.6933294534683228, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 32176 + }, + { + "epoch": 0.32177, + "grad_norm": 0.874487042427063, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 32177 + }, + { + "epoch": 0.32178, + "grad_norm": 0.8701624870300293, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 32178 + }, + { + "epoch": 0.32179, + "grad_norm": 0.7529736161231995, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 32179 + }, + { + "epoch": 0.3218, + "grad_norm": 0.7300243377685547, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 32180 + }, + { + "epoch": 0.32181, + "grad_norm": 0.80812668800354, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 32181 + }, + { + "epoch": 0.32182, + "grad_norm": 1.0747077465057373, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 32182 + }, + { + "epoch": 0.32183, + "grad_norm": 1.3997248411178589, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 32183 + }, + { + "epoch": 0.32184, + "grad_norm": 0.7549858689308167, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 32184 + }, + { + "epoch": 0.32185, + "grad_norm": 0.7286356687545776, + "learning_rate": 0.003, + "loss": 3.969, + "step": 32185 + }, + { + "epoch": 0.32186, + "grad_norm": 0.8196125626564026, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 32186 + }, + { + "epoch": 0.32187, + "grad_norm": 0.8922078013420105, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 32187 + }, + { + "epoch": 0.32188, + "grad_norm": 0.9486750960350037, + "learning_rate": 0.003, + "loss": 4.013, + "step": 32188 + }, + { + "epoch": 0.32189, + "grad_norm": 1.1180956363677979, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 32189 + }, + { + "epoch": 0.3219, + "grad_norm": 1.0147185325622559, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 32190 + }, + { + "epoch": 0.32191, + "grad_norm": 0.916448175907135, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 32191 + }, + { + "epoch": 0.32192, + "grad_norm": 0.8543076515197754, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 32192 + }, + { + "epoch": 0.32193, + "grad_norm": 0.6600968837738037, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 32193 + }, + { + "epoch": 0.32194, + "grad_norm": 0.6526434421539307, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 32194 + }, + { + "epoch": 0.32195, + "grad_norm": 0.6639058589935303, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 32195 + }, + { + "epoch": 0.32196, + "grad_norm": 0.6575182676315308, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 32196 + }, + { + "epoch": 0.32197, + "grad_norm": 0.6928329467773438, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 32197 + }, + { + "epoch": 0.32198, + "grad_norm": 0.6661460995674133, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 32198 + }, + { + "epoch": 0.32199, + "grad_norm": 0.6022605299949646, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 32199 + }, + { + "epoch": 0.322, + "grad_norm": 0.6702955961227417, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 32200 + }, + { + "epoch": 0.32201, + "grad_norm": 0.8308436274528503, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 32201 + }, + { + "epoch": 0.32202, + "grad_norm": 1.0697872638702393, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 32202 + }, + { + "epoch": 0.32203, + "grad_norm": 1.0811126232147217, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 32203 + }, + { + "epoch": 0.32204, + "grad_norm": 1.1176178455352783, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 32204 + }, + { + "epoch": 0.32205, + "grad_norm": 0.9348847270011902, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 32205 + }, + { + "epoch": 0.32206, + "grad_norm": 0.8284643292427063, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 32206 + }, + { + "epoch": 0.32207, + "grad_norm": 0.8628534078598022, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 32207 + }, + { + "epoch": 0.32208, + "grad_norm": 1.0271297693252563, + "learning_rate": 0.003, + "loss": 3.998, + "step": 32208 + }, + { + "epoch": 0.32209, + "grad_norm": 1.0576344728469849, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 32209 + }, + { + "epoch": 0.3221, + "grad_norm": 0.8105870485305786, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 32210 + }, + { + "epoch": 0.32211, + "grad_norm": 0.7445099353790283, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 32211 + }, + { + "epoch": 0.32212, + "grad_norm": 0.7429304122924805, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 32212 + }, + { + "epoch": 0.32213, + "grad_norm": 0.7574753761291504, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 32213 + }, + { + "epoch": 0.32214, + "grad_norm": 0.8123064041137695, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 32214 + }, + { + "epoch": 0.32215, + "grad_norm": 1.0059483051300049, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 32215 + }, + { + "epoch": 0.32216, + "grad_norm": 1.2816588878631592, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 32216 + }, + { + "epoch": 0.32217, + "grad_norm": 0.7305879592895508, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 32217 + }, + { + "epoch": 0.32218, + "grad_norm": 0.6042340993881226, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 32218 + }, + { + "epoch": 0.32219, + "grad_norm": 0.6619648337364197, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 32219 + }, + { + "epoch": 0.3222, + "grad_norm": 0.6851823329925537, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 32220 + }, + { + "epoch": 0.32221, + "grad_norm": 0.6898722052574158, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 32221 + }, + { + "epoch": 0.32222, + "grad_norm": 0.7320903539657593, + "learning_rate": 0.003, + "loss": 3.988, + "step": 32222 + }, + { + "epoch": 0.32223, + "grad_norm": 0.9093299508094788, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 32223 + }, + { + "epoch": 0.32224, + "grad_norm": 1.07831871509552, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 32224 + }, + { + "epoch": 0.32225, + "grad_norm": 0.8075864315032959, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 32225 + }, + { + "epoch": 0.32226, + "grad_norm": 0.8251882195472717, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 32226 + }, + { + "epoch": 0.32227, + "grad_norm": 0.987983763217926, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 32227 + }, + { + "epoch": 0.32228, + "grad_norm": 1.1508266925811768, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 32228 + }, + { + "epoch": 0.32229, + "grad_norm": 0.9619185328483582, + "learning_rate": 0.003, + "loss": 3.986, + "step": 32229 + }, + { + "epoch": 0.3223, + "grad_norm": 1.1918784379959106, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 32230 + }, + { + "epoch": 0.32231, + "grad_norm": 0.7630748748779297, + "learning_rate": 0.003, + "loss": 3.961, + "step": 32231 + }, + { + "epoch": 0.32232, + "grad_norm": 0.7302007079124451, + "learning_rate": 0.003, + "loss": 3.996, + "step": 32232 + }, + { + "epoch": 0.32233, + "grad_norm": 0.9292137026786804, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 32233 + }, + { + "epoch": 0.32234, + "grad_norm": 1.2438193559646606, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 32234 + }, + { + "epoch": 0.32235, + "grad_norm": 1.2452045679092407, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 32235 + }, + { + "epoch": 0.32236, + "grad_norm": 0.8949970006942749, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 32236 + }, + { + "epoch": 0.32237, + "grad_norm": 0.8542140126228333, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 32237 + }, + { + "epoch": 0.32238, + "grad_norm": 0.8919786810874939, + "learning_rate": 0.003, + "loss": 3.96, + "step": 32238 + }, + { + "epoch": 0.32239, + "grad_norm": 0.9527594447135925, + "learning_rate": 0.003, + "loss": 4.001, + "step": 32239 + }, + { + "epoch": 0.3224, + "grad_norm": 0.7898276448249817, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 32240 + }, + { + "epoch": 0.32241, + "grad_norm": 0.6338796615600586, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 32241 + }, + { + "epoch": 0.32242, + "grad_norm": 0.6969757080078125, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 32242 + }, + { + "epoch": 0.32243, + "grad_norm": 0.8559563159942627, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 32243 + }, + { + "epoch": 0.32244, + "grad_norm": 1.2416833639144897, + "learning_rate": 0.003, + "loss": 4.019, + "step": 32244 + }, + { + "epoch": 0.32245, + "grad_norm": 0.9121217727661133, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 32245 + }, + { + "epoch": 0.32246, + "grad_norm": 0.8717151284217834, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 32246 + }, + { + "epoch": 0.32247, + "grad_norm": 0.9096789360046387, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 32247 + }, + { + "epoch": 0.32248, + "grad_norm": 1.064881682395935, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 32248 + }, + { + "epoch": 0.32249, + "grad_norm": 0.9089925289154053, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 32249 + }, + { + "epoch": 0.3225, + "grad_norm": 0.7414023280143738, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 32250 + }, + { + "epoch": 0.32251, + "grad_norm": 0.6914502382278442, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 32251 + }, + { + "epoch": 0.32252, + "grad_norm": 0.6581348180770874, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 32252 + }, + { + "epoch": 0.32253, + "grad_norm": 0.6479506492614746, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 32253 + }, + { + "epoch": 0.32254, + "grad_norm": 0.5545223951339722, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 32254 + }, + { + "epoch": 0.32255, + "grad_norm": 0.6083476543426514, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 32255 + }, + { + "epoch": 0.32256, + "grad_norm": 0.7738859057426453, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 32256 + }, + { + "epoch": 0.32257, + "grad_norm": 0.9229429364204407, + "learning_rate": 0.003, + "loss": 3.975, + "step": 32257 + }, + { + "epoch": 0.32258, + "grad_norm": 1.134451150894165, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 32258 + }, + { + "epoch": 0.32259, + "grad_norm": 0.8703568577766418, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 32259 + }, + { + "epoch": 0.3226, + "grad_norm": 0.7055097222328186, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 32260 + }, + { + "epoch": 0.32261, + "grad_norm": 0.8618442416191101, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 32261 + }, + { + "epoch": 0.32262, + "grad_norm": 0.9386074542999268, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 32262 + }, + { + "epoch": 0.32263, + "grad_norm": 0.9582805633544922, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 32263 + }, + { + "epoch": 0.32264, + "grad_norm": 1.0739742517471313, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 32264 + }, + { + "epoch": 0.32265, + "grad_norm": 0.9246262311935425, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 32265 + }, + { + "epoch": 0.32266, + "grad_norm": 1.0783287286758423, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 32266 + }, + { + "epoch": 0.32267, + "grad_norm": 1.0831187963485718, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 32267 + }, + { + "epoch": 0.32268, + "grad_norm": 1.0022269487380981, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 32268 + }, + { + "epoch": 0.32269, + "grad_norm": 0.9423023462295532, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 32269 + }, + { + "epoch": 0.3227, + "grad_norm": 0.9875876307487488, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 32270 + }, + { + "epoch": 0.32271, + "grad_norm": 1.4082411527633667, + "learning_rate": 0.003, + "loss": 3.992, + "step": 32271 + }, + { + "epoch": 0.32272, + "grad_norm": 0.5768482685089111, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 32272 + }, + { + "epoch": 0.32273, + "grad_norm": 0.638771653175354, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 32273 + }, + { + "epoch": 0.32274, + "grad_norm": 0.7771185040473938, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 32274 + }, + { + "epoch": 0.32275, + "grad_norm": 0.9093090295791626, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 32275 + }, + { + "epoch": 0.32276, + "grad_norm": 1.0398666858673096, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 32276 + }, + { + "epoch": 0.32277, + "grad_norm": 0.9857379198074341, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 32277 + }, + { + "epoch": 0.32278, + "grad_norm": 0.7963970303535461, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 32278 + }, + { + "epoch": 0.32279, + "grad_norm": 0.7230323553085327, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 32279 + }, + { + "epoch": 0.3228, + "grad_norm": 0.7916014790534973, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 32280 + }, + { + "epoch": 0.32281, + "grad_norm": 0.9555642604827881, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 32281 + }, + { + "epoch": 0.32282, + "grad_norm": 0.9469680786132812, + "learning_rate": 0.003, + "loss": 4.016, + "step": 32282 + }, + { + "epoch": 0.32283, + "grad_norm": 1.0832033157348633, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 32283 + }, + { + "epoch": 0.32284, + "grad_norm": 1.0983448028564453, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 32284 + }, + { + "epoch": 0.32285, + "grad_norm": 0.9183358550071716, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 32285 + }, + { + "epoch": 0.32286, + "grad_norm": 0.9704213738441467, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 32286 + }, + { + "epoch": 0.32287, + "grad_norm": 1.0426667928695679, + "learning_rate": 0.003, + "loss": 4.012, + "step": 32287 + }, + { + "epoch": 0.32288, + "grad_norm": 0.9370235800743103, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 32288 + }, + { + "epoch": 0.32289, + "grad_norm": 0.8204391598701477, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 32289 + }, + { + "epoch": 0.3229, + "grad_norm": 0.9482154846191406, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 32290 + }, + { + "epoch": 0.32291, + "grad_norm": 1.0033764839172363, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 32291 + }, + { + "epoch": 0.32292, + "grad_norm": 1.0096782445907593, + "learning_rate": 0.003, + "loss": 3.997, + "step": 32292 + }, + { + "epoch": 0.32293, + "grad_norm": 1.0364516973495483, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 32293 + }, + { + "epoch": 0.32294, + "grad_norm": 0.9117727279663086, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 32294 + }, + { + "epoch": 0.32295, + "grad_norm": 1.2075436115264893, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 32295 + }, + { + "epoch": 0.32296, + "grad_norm": 1.0027645826339722, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 32296 + }, + { + "epoch": 0.32297, + "grad_norm": 0.8825694918632507, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 32297 + }, + { + "epoch": 0.32298, + "grad_norm": 0.8066021800041199, + "learning_rate": 0.003, + "loss": 4.0, + "step": 32298 + }, + { + "epoch": 0.32299, + "grad_norm": 0.9161664843559265, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 32299 + }, + { + "epoch": 0.323, + "grad_norm": 0.8340259790420532, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 32300 + }, + { + "epoch": 0.32301, + "grad_norm": 0.6959447860717773, + "learning_rate": 0.003, + "loss": 3.967, + "step": 32301 + }, + { + "epoch": 0.32302, + "grad_norm": 0.680118203163147, + "learning_rate": 0.003, + "loss": 3.99, + "step": 32302 + }, + { + "epoch": 0.32303, + "grad_norm": 0.6992242932319641, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 32303 + }, + { + "epoch": 0.32304, + "grad_norm": 0.746486485004425, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 32304 + }, + { + "epoch": 0.32305, + "grad_norm": 0.8291348218917847, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 32305 + }, + { + "epoch": 0.32306, + "grad_norm": 1.007008671760559, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 32306 + }, + { + "epoch": 0.32307, + "grad_norm": 1.2149734497070312, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 32307 + }, + { + "epoch": 0.32308, + "grad_norm": 0.9130027294158936, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 32308 + }, + { + "epoch": 0.32309, + "grad_norm": 0.954329788684845, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 32309 + }, + { + "epoch": 0.3231, + "grad_norm": 0.8553028106689453, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 32310 + }, + { + "epoch": 0.32311, + "grad_norm": 0.829465925693512, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 32311 + }, + { + "epoch": 0.32312, + "grad_norm": 0.9485005736351013, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 32312 + }, + { + "epoch": 0.32313, + "grad_norm": 1.0531471967697144, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 32313 + }, + { + "epoch": 0.32314, + "grad_norm": 1.204620599746704, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 32314 + }, + { + "epoch": 0.32315, + "grad_norm": 0.8223780989646912, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 32315 + }, + { + "epoch": 0.32316, + "grad_norm": 0.8406969308853149, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 32316 + }, + { + "epoch": 0.32317, + "grad_norm": 0.822245717048645, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 32317 + }, + { + "epoch": 0.32318, + "grad_norm": 0.7420104146003723, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 32318 + }, + { + "epoch": 0.32319, + "grad_norm": 0.6499215364456177, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 32319 + }, + { + "epoch": 0.3232, + "grad_norm": 0.6286032795906067, + "learning_rate": 0.003, + "loss": 4.0, + "step": 32320 + }, + { + "epoch": 0.32321, + "grad_norm": 0.8257924318313599, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 32321 + }, + { + "epoch": 0.32322, + "grad_norm": 1.2029931545257568, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 32322 + }, + { + "epoch": 0.32323, + "grad_norm": 0.9160865545272827, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 32323 + }, + { + "epoch": 0.32324, + "grad_norm": 0.6568738222122192, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 32324 + }, + { + "epoch": 0.32325, + "grad_norm": 0.568204939365387, + "learning_rate": 0.003, + "loss": 3.9493, + "step": 32325 + }, + { + "epoch": 0.32326, + "grad_norm": 0.6395931243896484, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 32326 + }, + { + "epoch": 0.32327, + "grad_norm": 0.871443510055542, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 32327 + }, + { + "epoch": 0.32328, + "grad_norm": 1.079755187034607, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 32328 + }, + { + "epoch": 0.32329, + "grad_norm": 1.0661113262176514, + "learning_rate": 0.003, + "loss": 3.974, + "step": 32329 + }, + { + "epoch": 0.3233, + "grad_norm": 1.02070951461792, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 32330 + }, + { + "epoch": 0.32331, + "grad_norm": 1.0152618885040283, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 32331 + }, + { + "epoch": 0.32332, + "grad_norm": 0.8418744206428528, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 32332 + }, + { + "epoch": 0.32333, + "grad_norm": 0.7096357345581055, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 32333 + }, + { + "epoch": 0.32334, + "grad_norm": 0.8459305167198181, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 32334 + }, + { + "epoch": 0.32335, + "grad_norm": 0.8676722645759583, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 32335 + }, + { + "epoch": 0.32336, + "grad_norm": 0.813725471496582, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 32336 + }, + { + "epoch": 0.32337, + "grad_norm": 0.8303931355476379, + "learning_rate": 0.003, + "loss": 3.991, + "step": 32337 + }, + { + "epoch": 0.32338, + "grad_norm": 0.8910247087478638, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 32338 + }, + { + "epoch": 0.32339, + "grad_norm": 0.9713416695594788, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 32339 + }, + { + "epoch": 0.3234, + "grad_norm": 1.0637178421020508, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 32340 + }, + { + "epoch": 0.32341, + "grad_norm": 0.9398966431617737, + "learning_rate": 0.003, + "loss": 3.977, + "step": 32341 + }, + { + "epoch": 0.32342, + "grad_norm": 0.7631863951683044, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 32342 + }, + { + "epoch": 0.32343, + "grad_norm": 0.8152332305908203, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 32343 + }, + { + "epoch": 0.32344, + "grad_norm": 0.7518848180770874, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 32344 + }, + { + "epoch": 0.32345, + "grad_norm": 0.7838078141212463, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 32345 + }, + { + "epoch": 0.32346, + "grad_norm": 0.7782304286956787, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 32346 + }, + { + "epoch": 0.32347, + "grad_norm": 0.6964134573936462, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 32347 + }, + { + "epoch": 0.32348, + "grad_norm": 0.7466184496879578, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 32348 + }, + { + "epoch": 0.32349, + "grad_norm": 0.8547024130821228, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 32349 + }, + { + "epoch": 0.3235, + "grad_norm": 0.861370861530304, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 32350 + }, + { + "epoch": 0.32351, + "grad_norm": 0.8620891571044922, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 32351 + }, + { + "epoch": 0.32352, + "grad_norm": 1.2304534912109375, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 32352 + }, + { + "epoch": 0.32353, + "grad_norm": 0.9967436790466309, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 32353 + }, + { + "epoch": 0.32354, + "grad_norm": 0.8301941156387329, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 32354 + }, + { + "epoch": 0.32355, + "grad_norm": 0.787036120891571, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 32355 + }, + { + "epoch": 0.32356, + "grad_norm": 0.8031530976295471, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 32356 + }, + { + "epoch": 0.32357, + "grad_norm": 0.8607959151268005, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 32357 + }, + { + "epoch": 0.32358, + "grad_norm": 0.848793089389801, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 32358 + }, + { + "epoch": 0.32359, + "grad_norm": 0.919458270072937, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 32359 + }, + { + "epoch": 0.3236, + "grad_norm": 0.9305683970451355, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 32360 + }, + { + "epoch": 0.32361, + "grad_norm": 0.914644181728363, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 32361 + }, + { + "epoch": 0.32362, + "grad_norm": 0.9668870568275452, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 32362 + }, + { + "epoch": 0.32363, + "grad_norm": 1.1420881748199463, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 32363 + }, + { + "epoch": 0.32364, + "grad_norm": 0.9816522598266602, + "learning_rate": 0.003, + "loss": 3.971, + "step": 32364 + }, + { + "epoch": 0.32365, + "grad_norm": 1.103488564491272, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 32365 + }, + { + "epoch": 0.32366, + "grad_norm": 1.2164250612258911, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 32366 + }, + { + "epoch": 0.32367, + "grad_norm": 0.9420822262763977, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 32367 + }, + { + "epoch": 0.32368, + "grad_norm": 0.7487914562225342, + "learning_rate": 0.003, + "loss": 3.973, + "step": 32368 + }, + { + "epoch": 0.32369, + "grad_norm": 0.7210949659347534, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 32369 + }, + { + "epoch": 0.3237, + "grad_norm": 0.7946134805679321, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 32370 + }, + { + "epoch": 0.32371, + "grad_norm": 0.7713467478752136, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 32371 + }, + { + "epoch": 0.32372, + "grad_norm": 0.7363332509994507, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 32372 + }, + { + "epoch": 0.32373, + "grad_norm": 0.9714472889900208, + "learning_rate": 0.003, + "loss": 3.958, + "step": 32373 + }, + { + "epoch": 0.32374, + "grad_norm": 1.1275674104690552, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 32374 + }, + { + "epoch": 0.32375, + "grad_norm": 1.0079514980316162, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 32375 + }, + { + "epoch": 0.32376, + "grad_norm": 0.8604972958564758, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 32376 + }, + { + "epoch": 0.32377, + "grad_norm": 0.736783504486084, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 32377 + }, + { + "epoch": 0.32378, + "grad_norm": 0.7314930558204651, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 32378 + }, + { + "epoch": 0.32379, + "grad_norm": 0.8177298307418823, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 32379 + }, + { + "epoch": 0.3238, + "grad_norm": 0.8067805767059326, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 32380 + }, + { + "epoch": 0.32381, + "grad_norm": 0.8836551904678345, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 32381 + }, + { + "epoch": 0.32382, + "grad_norm": 1.0357013940811157, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 32382 + }, + { + "epoch": 0.32383, + "grad_norm": 0.9826351404190063, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 32383 + }, + { + "epoch": 0.32384, + "grad_norm": 1.071526050567627, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 32384 + }, + { + "epoch": 0.32385, + "grad_norm": 1.075355052947998, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 32385 + }, + { + "epoch": 0.32386, + "grad_norm": 0.8169649243354797, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 32386 + }, + { + "epoch": 0.32387, + "grad_norm": 0.7400320768356323, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 32387 + }, + { + "epoch": 0.32388, + "grad_norm": 0.759391725063324, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 32388 + }, + { + "epoch": 0.32389, + "grad_norm": 0.8145010471343994, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 32389 + }, + { + "epoch": 0.3239, + "grad_norm": 0.8103436827659607, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 32390 + }, + { + "epoch": 0.32391, + "grad_norm": 0.9339328408241272, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 32391 + }, + { + "epoch": 0.32392, + "grad_norm": 0.9615173935890198, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 32392 + }, + { + "epoch": 0.32393, + "grad_norm": 0.9145262837409973, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 32393 + }, + { + "epoch": 0.32394, + "grad_norm": 0.9737706184387207, + "learning_rate": 0.003, + "loss": 4.0, + "step": 32394 + }, + { + "epoch": 0.32395, + "grad_norm": 1.0059479475021362, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 32395 + }, + { + "epoch": 0.32396, + "grad_norm": 1.0570478439331055, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 32396 + }, + { + "epoch": 0.32397, + "grad_norm": 1.1724146604537964, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 32397 + }, + { + "epoch": 0.32398, + "grad_norm": 1.0043315887451172, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 32398 + }, + { + "epoch": 0.32399, + "grad_norm": 0.8153746724128723, + "learning_rate": 0.003, + "loss": 3.987, + "step": 32399 + }, + { + "epoch": 0.324, + "grad_norm": 0.7053890228271484, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 32400 + }, + { + "epoch": 0.32401, + "grad_norm": 0.6878343820571899, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 32401 + }, + { + "epoch": 0.32402, + "grad_norm": 0.7318712472915649, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 32402 + }, + { + "epoch": 0.32403, + "grad_norm": 0.79329514503479, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 32403 + }, + { + "epoch": 0.32404, + "grad_norm": 0.9848685264587402, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 32404 + }, + { + "epoch": 0.32405, + "grad_norm": 1.253307580947876, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 32405 + }, + { + "epoch": 0.32406, + "grad_norm": 1.0267142057418823, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 32406 + }, + { + "epoch": 0.32407, + "grad_norm": 0.9018604755401611, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 32407 + }, + { + "epoch": 0.32408, + "grad_norm": 0.6789781451225281, + "learning_rate": 0.003, + "loss": 3.991, + "step": 32408 + }, + { + "epoch": 0.32409, + "grad_norm": 0.648796021938324, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 32409 + }, + { + "epoch": 0.3241, + "grad_norm": 0.6519662141799927, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 32410 + }, + { + "epoch": 0.32411, + "grad_norm": 0.6108657717704773, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 32411 + }, + { + "epoch": 0.32412, + "grad_norm": 0.6896421313285828, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 32412 + }, + { + "epoch": 0.32413, + "grad_norm": 0.7221764326095581, + "learning_rate": 0.003, + "loss": 4.007, + "step": 32413 + }, + { + "epoch": 0.32414, + "grad_norm": 0.693912148475647, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 32414 + }, + { + "epoch": 0.32415, + "grad_norm": 0.7272496223449707, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 32415 + }, + { + "epoch": 0.32416, + "grad_norm": 0.8572666049003601, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 32416 + }, + { + "epoch": 0.32417, + "grad_norm": 1.103991985321045, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 32417 + }, + { + "epoch": 0.32418, + "grad_norm": 1.2339218854904175, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 32418 + }, + { + "epoch": 0.32419, + "grad_norm": 0.8128048181533813, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 32419 + }, + { + "epoch": 0.3242, + "grad_norm": 0.8113899230957031, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 32420 + }, + { + "epoch": 0.32421, + "grad_norm": 0.8569998145103455, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 32421 + }, + { + "epoch": 0.32422, + "grad_norm": 0.9089901447296143, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 32422 + }, + { + "epoch": 0.32423, + "grad_norm": 0.992407500743866, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 32423 + }, + { + "epoch": 0.32424, + "grad_norm": 1.0153896808624268, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 32424 + }, + { + "epoch": 0.32425, + "grad_norm": 0.9665413498878479, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 32425 + }, + { + "epoch": 0.32426, + "grad_norm": 0.9742746949195862, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 32426 + }, + { + "epoch": 0.32427, + "grad_norm": 1.0721241235733032, + "learning_rate": 0.003, + "loss": 3.99, + "step": 32427 + }, + { + "epoch": 0.32428, + "grad_norm": 1.1136698722839355, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 32428 + }, + { + "epoch": 0.32429, + "grad_norm": 0.852491021156311, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 32429 + }, + { + "epoch": 0.3243, + "grad_norm": 0.7609722018241882, + "learning_rate": 0.003, + "loss": 4.008, + "step": 32430 + }, + { + "epoch": 0.32431, + "grad_norm": 0.8527836799621582, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 32431 + }, + { + "epoch": 0.32432, + "grad_norm": 0.9638751149177551, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 32432 + }, + { + "epoch": 0.32433, + "grad_norm": 1.0909746885299683, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 32433 + }, + { + "epoch": 0.32434, + "grad_norm": 0.8552682399749756, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 32434 + }, + { + "epoch": 0.32435, + "grad_norm": 0.828174889087677, + "learning_rate": 0.003, + "loss": 3.964, + "step": 32435 + }, + { + "epoch": 0.32436, + "grad_norm": 0.6927368640899658, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 32436 + }, + { + "epoch": 0.32437, + "grad_norm": 0.7462885975837708, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 32437 + }, + { + "epoch": 0.32438, + "grad_norm": 0.7718014717102051, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 32438 + }, + { + "epoch": 0.32439, + "grad_norm": 0.6247260570526123, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 32439 + }, + { + "epoch": 0.3244, + "grad_norm": 0.7951813340187073, + "learning_rate": 0.003, + "loss": 3.985, + "step": 32440 + }, + { + "epoch": 0.32441, + "grad_norm": 0.8128353953361511, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 32441 + }, + { + "epoch": 0.32442, + "grad_norm": 0.7947268486022949, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 32442 + }, + { + "epoch": 0.32443, + "grad_norm": 0.9307166934013367, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 32443 + }, + { + "epoch": 0.32444, + "grad_norm": 1.160814642906189, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 32444 + }, + { + "epoch": 0.32445, + "grad_norm": 0.9503359198570251, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 32445 + }, + { + "epoch": 0.32446, + "grad_norm": 1.1751108169555664, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 32446 + }, + { + "epoch": 0.32447, + "grad_norm": 0.9505729079246521, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 32447 + }, + { + "epoch": 0.32448, + "grad_norm": 0.9534039497375488, + "learning_rate": 0.003, + "loss": 3.998, + "step": 32448 + }, + { + "epoch": 0.32449, + "grad_norm": 0.992527961730957, + "learning_rate": 0.003, + "loss": 3.995, + "step": 32449 + }, + { + "epoch": 0.3245, + "grad_norm": 1.0150166749954224, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 32450 + }, + { + "epoch": 0.32451, + "grad_norm": 0.9983183741569519, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 32451 + }, + { + "epoch": 0.32452, + "grad_norm": 1.0241029262542725, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 32452 + }, + { + "epoch": 0.32453, + "grad_norm": 0.9441582560539246, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 32453 + }, + { + "epoch": 0.32454, + "grad_norm": 0.8663548231124878, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 32454 + }, + { + "epoch": 0.32455, + "grad_norm": 0.887255072593689, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 32455 + }, + { + "epoch": 0.32456, + "grad_norm": 0.9426296949386597, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 32456 + }, + { + "epoch": 0.32457, + "grad_norm": 0.9346213936805725, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 32457 + }, + { + "epoch": 0.32458, + "grad_norm": 0.8860076069831848, + "learning_rate": 0.003, + "loss": 4.017, + "step": 32458 + }, + { + "epoch": 0.32459, + "grad_norm": 1.0007952451705933, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 32459 + }, + { + "epoch": 0.3246, + "grad_norm": 1.1156244277954102, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 32460 + }, + { + "epoch": 0.32461, + "grad_norm": 0.9125857949256897, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 32461 + }, + { + "epoch": 0.32462, + "grad_norm": 0.8135899901390076, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 32462 + }, + { + "epoch": 0.32463, + "grad_norm": 0.7852208018302917, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 32463 + }, + { + "epoch": 0.32464, + "grad_norm": 0.8360493779182434, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 32464 + }, + { + "epoch": 0.32465, + "grad_norm": 1.0424246788024902, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 32465 + }, + { + "epoch": 0.32466, + "grad_norm": 1.227095603942871, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 32466 + }, + { + "epoch": 0.32467, + "grad_norm": 0.8443180322647095, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 32467 + }, + { + "epoch": 0.32468, + "grad_norm": 0.8387784957885742, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 32468 + }, + { + "epoch": 0.32469, + "grad_norm": 0.8515106439590454, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 32469 + }, + { + "epoch": 0.3247, + "grad_norm": 0.8224942684173584, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 32470 + }, + { + "epoch": 0.32471, + "grad_norm": 0.8899878859519958, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 32471 + }, + { + "epoch": 0.32472, + "grad_norm": 0.7192010879516602, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 32472 + }, + { + "epoch": 0.32473, + "grad_norm": 0.6850727200508118, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 32473 + }, + { + "epoch": 0.32474, + "grad_norm": 0.7687437534332275, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 32474 + }, + { + "epoch": 0.32475, + "grad_norm": 0.9996352791786194, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 32475 + }, + { + "epoch": 0.32476, + "grad_norm": 1.1355944871902466, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 32476 + }, + { + "epoch": 0.32477, + "grad_norm": 0.8828991651535034, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 32477 + }, + { + "epoch": 0.32478, + "grad_norm": 0.7891404032707214, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 32478 + }, + { + "epoch": 0.32479, + "grad_norm": 0.7886939644813538, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 32479 + }, + { + "epoch": 0.3248, + "grad_norm": 0.7835010886192322, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 32480 + }, + { + "epoch": 0.32481, + "grad_norm": 0.8572627902030945, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 32481 + }, + { + "epoch": 0.32482, + "grad_norm": 0.9748180508613586, + "learning_rate": 0.003, + "loss": 3.978, + "step": 32482 + }, + { + "epoch": 0.32483, + "grad_norm": 0.9856172800064087, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 32483 + }, + { + "epoch": 0.32484, + "grad_norm": 1.0550568103790283, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 32484 + }, + { + "epoch": 0.32485, + "grad_norm": 0.9020842909812927, + "learning_rate": 0.003, + "loss": 4.0, + "step": 32485 + }, + { + "epoch": 0.32486, + "grad_norm": 0.7977316975593567, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 32486 + }, + { + "epoch": 0.32487, + "grad_norm": 0.7065430283546448, + "learning_rate": 0.003, + "loss": 3.974, + "step": 32487 + }, + { + "epoch": 0.32488, + "grad_norm": 0.7061853408813477, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 32488 + }, + { + "epoch": 0.32489, + "grad_norm": 0.9304752349853516, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 32489 + }, + { + "epoch": 0.3249, + "grad_norm": 1.0594961643218994, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 32490 + }, + { + "epoch": 0.32491, + "grad_norm": 0.9841848611831665, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 32491 + }, + { + "epoch": 0.32492, + "grad_norm": 0.9085679054260254, + "learning_rate": 0.003, + "loss": 3.97, + "step": 32492 + }, + { + "epoch": 0.32493, + "grad_norm": 0.7325956225395203, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 32493 + }, + { + "epoch": 0.32494, + "grad_norm": 0.6345090270042419, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 32494 + }, + { + "epoch": 0.32495, + "grad_norm": 0.7287845611572266, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 32495 + }, + { + "epoch": 0.32496, + "grad_norm": 0.7522757649421692, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 32496 + }, + { + "epoch": 0.32497, + "grad_norm": 0.7696222066879272, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 32497 + }, + { + "epoch": 0.32498, + "grad_norm": 0.7300878763198853, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 32498 + }, + { + "epoch": 0.32499, + "grad_norm": 0.7889106869697571, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 32499 + }, + { + "epoch": 0.325, + "grad_norm": 0.9589406251907349, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 32500 + }, + { + "epoch": 0.32501, + "grad_norm": 1.166284203529358, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 32501 + }, + { + "epoch": 0.32502, + "grad_norm": 1.0739814043045044, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 32502 + }, + { + "epoch": 0.32503, + "grad_norm": 0.978077232837677, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 32503 + }, + { + "epoch": 0.32504, + "grad_norm": 0.9205668568611145, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 32504 + }, + { + "epoch": 0.32505, + "grad_norm": 1.0002919435501099, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 32505 + }, + { + "epoch": 0.32506, + "grad_norm": 1.3610836267471313, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 32506 + }, + { + "epoch": 0.32507, + "grad_norm": 0.6793444752693176, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 32507 + }, + { + "epoch": 0.32508, + "grad_norm": 1.0329214334487915, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 32508 + }, + { + "epoch": 0.32509, + "grad_norm": 1.2525397539138794, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 32509 + }, + { + "epoch": 0.3251, + "grad_norm": 0.6166375875473022, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 32510 + }, + { + "epoch": 0.32511, + "grad_norm": 0.7102653384208679, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 32511 + }, + { + "epoch": 0.32512, + "grad_norm": 0.7744130492210388, + "learning_rate": 0.003, + "loss": 3.973, + "step": 32512 + }, + { + "epoch": 0.32513, + "grad_norm": 0.8295484781265259, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 32513 + }, + { + "epoch": 0.32514, + "grad_norm": 0.8636655807495117, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 32514 + }, + { + "epoch": 0.32515, + "grad_norm": 0.9787363409996033, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 32515 + }, + { + "epoch": 0.32516, + "grad_norm": 1.1663118600845337, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 32516 + }, + { + "epoch": 0.32517, + "grad_norm": 0.8378428220748901, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 32517 + }, + { + "epoch": 0.32518, + "grad_norm": 0.7270094156265259, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 32518 + }, + { + "epoch": 0.32519, + "grad_norm": 0.8020510673522949, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 32519 + }, + { + "epoch": 0.3252, + "grad_norm": 0.9433132410049438, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 32520 + }, + { + "epoch": 0.32521, + "grad_norm": 0.868827223777771, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 32521 + }, + { + "epoch": 0.32522, + "grad_norm": 0.9809443950653076, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 32522 + }, + { + "epoch": 0.32523, + "grad_norm": 1.1340868473052979, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 32523 + }, + { + "epoch": 0.32524, + "grad_norm": 0.869167685508728, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 32524 + }, + { + "epoch": 0.32525, + "grad_norm": 0.9135714173316956, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 32525 + }, + { + "epoch": 0.32526, + "grad_norm": 1.006718635559082, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 32526 + }, + { + "epoch": 0.32527, + "grad_norm": 0.9036099910736084, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 32527 + }, + { + "epoch": 0.32528, + "grad_norm": 0.8064888715744019, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 32528 + }, + { + "epoch": 0.32529, + "grad_norm": 0.8431469798088074, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 32529 + }, + { + "epoch": 0.3253, + "grad_norm": 0.8131701350212097, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 32530 + }, + { + "epoch": 0.32531, + "grad_norm": 0.9509857296943665, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 32531 + }, + { + "epoch": 0.32532, + "grad_norm": 1.0729550123214722, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 32532 + }, + { + "epoch": 0.32533, + "grad_norm": 1.0847480297088623, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 32533 + }, + { + "epoch": 0.32534, + "grad_norm": 1.1295086145401, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 32534 + }, + { + "epoch": 0.32535, + "grad_norm": 0.8757442235946655, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 32535 + }, + { + "epoch": 0.32536, + "grad_norm": 0.8627163171768188, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 32536 + }, + { + "epoch": 0.32537, + "grad_norm": 1.0107048749923706, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 32537 + }, + { + "epoch": 0.32538, + "grad_norm": 1.2814136743545532, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 32538 + }, + { + "epoch": 0.32539, + "grad_norm": 0.7934348583221436, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 32539 + }, + { + "epoch": 0.3254, + "grad_norm": 0.6985980272293091, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 32540 + }, + { + "epoch": 0.32541, + "grad_norm": 0.7069082260131836, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 32541 + }, + { + "epoch": 0.32542, + "grad_norm": 0.778586745262146, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 32542 + }, + { + "epoch": 0.32543, + "grad_norm": 0.9401214718818665, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 32543 + }, + { + "epoch": 0.32544, + "grad_norm": 1.2335535287857056, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 32544 + }, + { + "epoch": 0.32545, + "grad_norm": 0.9113735556602478, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 32545 + }, + { + "epoch": 0.32546, + "grad_norm": 0.8499879837036133, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 32546 + }, + { + "epoch": 0.32547, + "grad_norm": 0.8890844583511353, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 32547 + }, + { + "epoch": 0.32548, + "grad_norm": 0.9682372808456421, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 32548 + }, + { + "epoch": 0.32549, + "grad_norm": 0.9662706851959229, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 32549 + }, + { + "epoch": 0.3255, + "grad_norm": 0.9195154905319214, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 32550 + }, + { + "epoch": 0.32551, + "grad_norm": 0.8108278512954712, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 32551 + }, + { + "epoch": 0.32552, + "grad_norm": 0.6793009042739868, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 32552 + }, + { + "epoch": 0.32553, + "grad_norm": 0.6265491247177124, + "learning_rate": 0.003, + "loss": 3.972, + "step": 32553 + }, + { + "epoch": 0.32554, + "grad_norm": 0.5835283398628235, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 32554 + }, + { + "epoch": 0.32555, + "grad_norm": 0.6134976744651794, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 32555 + }, + { + "epoch": 0.32556, + "grad_norm": 0.6592286229133606, + "learning_rate": 0.003, + "loss": 3.991, + "step": 32556 + }, + { + "epoch": 0.32557, + "grad_norm": 0.7515560388565063, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 32557 + }, + { + "epoch": 0.32558, + "grad_norm": 0.8137169480323792, + "learning_rate": 0.003, + "loss": 3.985, + "step": 32558 + }, + { + "epoch": 0.32559, + "grad_norm": 0.8301559090614319, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 32559 + }, + { + "epoch": 0.3256, + "grad_norm": 1.1979173421859741, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 32560 + }, + { + "epoch": 0.32561, + "grad_norm": 1.0684479475021362, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 32561 + }, + { + "epoch": 0.32562, + "grad_norm": 0.8768090009689331, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 32562 + }, + { + "epoch": 0.32563, + "grad_norm": 0.6805611848831177, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 32563 + }, + { + "epoch": 0.32564, + "grad_norm": 0.5999616980552673, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 32564 + }, + { + "epoch": 0.32565, + "grad_norm": 0.6438753604888916, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 32565 + }, + { + "epoch": 0.32566, + "grad_norm": 0.7468984127044678, + "learning_rate": 0.003, + "loss": 3.986, + "step": 32566 + }, + { + "epoch": 0.32567, + "grad_norm": 0.8132427334785461, + "learning_rate": 0.003, + "loss": 3.979, + "step": 32567 + }, + { + "epoch": 0.32568, + "grad_norm": 0.6821421384811401, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 32568 + }, + { + "epoch": 0.32569, + "grad_norm": 0.6234792470932007, + "learning_rate": 0.003, + "loss": 3.974, + "step": 32569 + }, + { + "epoch": 0.3257, + "grad_norm": 0.8585687279701233, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 32570 + }, + { + "epoch": 0.32571, + "grad_norm": 1.1091886758804321, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 32571 + }, + { + "epoch": 0.32572, + "grad_norm": 1.0040985345840454, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 32572 + }, + { + "epoch": 0.32573, + "grad_norm": 1.0679906606674194, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 32573 + }, + { + "epoch": 0.32574, + "grad_norm": 0.9618886113166809, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 32574 + }, + { + "epoch": 0.32575, + "grad_norm": 0.8693458437919617, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 32575 + }, + { + "epoch": 0.32576, + "grad_norm": 0.7084919810295105, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 32576 + }, + { + "epoch": 0.32577, + "grad_norm": 0.6183563470840454, + "learning_rate": 0.003, + "loss": 3.9463, + "step": 32577 + }, + { + "epoch": 0.32578, + "grad_norm": 0.6480621695518494, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 32578 + }, + { + "epoch": 0.32579, + "grad_norm": 0.7791635990142822, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 32579 + }, + { + "epoch": 0.3258, + "grad_norm": 0.9340878129005432, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 32580 + }, + { + "epoch": 0.32581, + "grad_norm": 0.9535079598426819, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 32581 + }, + { + "epoch": 0.32582, + "grad_norm": 0.9579949975013733, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 32582 + }, + { + "epoch": 0.32583, + "grad_norm": 0.8326864838600159, + "learning_rate": 0.003, + "loss": 3.97, + "step": 32583 + }, + { + "epoch": 0.32584, + "grad_norm": 0.8904878497123718, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 32584 + }, + { + "epoch": 0.32585, + "grad_norm": 1.086823582649231, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 32585 + }, + { + "epoch": 0.32586, + "grad_norm": 1.0091261863708496, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 32586 + }, + { + "epoch": 0.32587, + "grad_norm": 1.265363335609436, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 32587 + }, + { + "epoch": 0.32588, + "grad_norm": 1.0284395217895508, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 32588 + }, + { + "epoch": 0.32589, + "grad_norm": 0.8306700587272644, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 32589 + }, + { + "epoch": 0.3259, + "grad_norm": 0.8400598764419556, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 32590 + }, + { + "epoch": 0.32591, + "grad_norm": 0.9942071437835693, + "learning_rate": 0.003, + "loss": 4.015, + "step": 32591 + }, + { + "epoch": 0.32592, + "grad_norm": 0.978355348110199, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 32592 + }, + { + "epoch": 0.32593, + "grad_norm": 1.2215123176574707, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 32593 + }, + { + "epoch": 0.32594, + "grad_norm": 1.25307297706604, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 32594 + }, + { + "epoch": 0.32595, + "grad_norm": 0.8293337821960449, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 32595 + }, + { + "epoch": 0.32596, + "grad_norm": 0.7735527753829956, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 32596 + }, + { + "epoch": 0.32597, + "grad_norm": 0.8621964454650879, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 32597 + }, + { + "epoch": 0.32598, + "grad_norm": 0.9827914237976074, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 32598 + }, + { + "epoch": 0.32599, + "grad_norm": 1.1624940633773804, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 32599 + }, + { + "epoch": 0.326, + "grad_norm": 0.8107041120529175, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 32600 + }, + { + "epoch": 0.32601, + "grad_norm": 0.5970117449760437, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 32601 + }, + { + "epoch": 0.32602, + "grad_norm": 0.5769016742706299, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 32602 + }, + { + "epoch": 0.32603, + "grad_norm": 0.7640804052352905, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 32603 + }, + { + "epoch": 0.32604, + "grad_norm": 1.0130397081375122, + "learning_rate": 0.003, + "loss": 4.002, + "step": 32604 + }, + { + "epoch": 0.32605, + "grad_norm": 1.306789517402649, + "learning_rate": 0.003, + "loss": 3.971, + "step": 32605 + }, + { + "epoch": 0.32606, + "grad_norm": 0.7935150861740112, + "learning_rate": 0.003, + "loss": 3.991, + "step": 32606 + }, + { + "epoch": 0.32607, + "grad_norm": 0.7662623524665833, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 32607 + }, + { + "epoch": 0.32608, + "grad_norm": 0.7886900901794434, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 32608 + }, + { + "epoch": 0.32609, + "grad_norm": 0.9660586714744568, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 32609 + }, + { + "epoch": 0.3261, + "grad_norm": 1.0723339319229126, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 32610 + }, + { + "epoch": 0.32611, + "grad_norm": 0.7247104048728943, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 32611 + }, + { + "epoch": 0.32612, + "grad_norm": 0.7389459609985352, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 32612 + }, + { + "epoch": 0.32613, + "grad_norm": 0.7714213132858276, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 32613 + }, + { + "epoch": 0.32614, + "grad_norm": 0.6870663166046143, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 32614 + }, + { + "epoch": 0.32615, + "grad_norm": 0.677933394908905, + "learning_rate": 0.003, + "loss": 3.998, + "step": 32615 + }, + { + "epoch": 0.32616, + "grad_norm": 0.6926342844963074, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 32616 + }, + { + "epoch": 0.32617, + "grad_norm": 0.8492617607116699, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 32617 + }, + { + "epoch": 0.32618, + "grad_norm": 1.0685712099075317, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 32618 + }, + { + "epoch": 0.32619, + "grad_norm": 1.166233777999878, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 32619 + }, + { + "epoch": 0.3262, + "grad_norm": 0.9578309059143066, + "learning_rate": 0.003, + "loss": 3.968, + "step": 32620 + }, + { + "epoch": 0.32621, + "grad_norm": 0.9564568400382996, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 32621 + }, + { + "epoch": 0.32622, + "grad_norm": 0.8658410906791687, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 32622 + }, + { + "epoch": 0.32623, + "grad_norm": 0.9681189656257629, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 32623 + }, + { + "epoch": 0.32624, + "grad_norm": 1.127404808998108, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 32624 + }, + { + "epoch": 0.32625, + "grad_norm": 0.907767117023468, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 32625 + }, + { + "epoch": 0.32626, + "grad_norm": 0.9187948107719421, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 32626 + }, + { + "epoch": 0.32627, + "grad_norm": 0.8486701250076294, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 32627 + }, + { + "epoch": 0.32628, + "grad_norm": 0.9882554411888123, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 32628 + }, + { + "epoch": 0.32629, + "grad_norm": 0.9940965175628662, + "learning_rate": 0.003, + "loss": 4.003, + "step": 32629 + }, + { + "epoch": 0.3263, + "grad_norm": 1.0183311700820923, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 32630 + }, + { + "epoch": 0.32631, + "grad_norm": 0.9084317088127136, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 32631 + }, + { + "epoch": 0.32632, + "grad_norm": 0.9853277206420898, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 32632 + }, + { + "epoch": 0.32633, + "grad_norm": 1.1276832818984985, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 32633 + }, + { + "epoch": 0.32634, + "grad_norm": 0.9792917966842651, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 32634 + }, + { + "epoch": 0.32635, + "grad_norm": 1.216816782951355, + "learning_rate": 0.003, + "loss": 4.019, + "step": 32635 + }, + { + "epoch": 0.32636, + "grad_norm": 1.1300883293151855, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 32636 + }, + { + "epoch": 0.32637, + "grad_norm": 0.9464437365531921, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 32637 + }, + { + "epoch": 0.32638, + "grad_norm": 1.0227962732315063, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 32638 + }, + { + "epoch": 0.32639, + "grad_norm": 1.0229580402374268, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 32639 + }, + { + "epoch": 0.3264, + "grad_norm": 0.7943432331085205, + "learning_rate": 0.003, + "loss": 3.996, + "step": 32640 + }, + { + "epoch": 0.32641, + "grad_norm": 0.6794039607048035, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 32641 + }, + { + "epoch": 0.32642, + "grad_norm": 0.6880098581314087, + "learning_rate": 0.003, + "loss": 3.974, + "step": 32642 + }, + { + "epoch": 0.32643, + "grad_norm": 0.7688810229301453, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 32643 + }, + { + "epoch": 0.32644, + "grad_norm": 0.8504377603530884, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 32644 + }, + { + "epoch": 0.32645, + "grad_norm": 0.8483926653862, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 32645 + }, + { + "epoch": 0.32646, + "grad_norm": 0.9341667890548706, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 32646 + }, + { + "epoch": 0.32647, + "grad_norm": 1.003975510597229, + "learning_rate": 0.003, + "loss": 3.964, + "step": 32647 + }, + { + "epoch": 0.32648, + "grad_norm": 1.109023928642273, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 32648 + }, + { + "epoch": 0.32649, + "grad_norm": 0.9672076106071472, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 32649 + }, + { + "epoch": 0.3265, + "grad_norm": 0.8053532242774963, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 32650 + }, + { + "epoch": 0.32651, + "grad_norm": 0.8136510848999023, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 32651 + }, + { + "epoch": 0.32652, + "grad_norm": 0.9078527688980103, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 32652 + }, + { + "epoch": 0.32653, + "grad_norm": 0.8249297142028809, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 32653 + }, + { + "epoch": 0.32654, + "grad_norm": 0.7923644185066223, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 32654 + }, + { + "epoch": 0.32655, + "grad_norm": 0.9373129606246948, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 32655 + }, + { + "epoch": 0.32656, + "grad_norm": 0.9603829383850098, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 32656 + }, + { + "epoch": 0.32657, + "grad_norm": 0.8529765605926514, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 32657 + }, + { + "epoch": 0.32658, + "grad_norm": 1.02993905544281, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 32658 + }, + { + "epoch": 0.32659, + "grad_norm": 1.4052790403366089, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 32659 + }, + { + "epoch": 0.3266, + "grad_norm": 0.6839971542358398, + "learning_rate": 0.003, + "loss": 3.967, + "step": 32660 + }, + { + "epoch": 0.32661, + "grad_norm": 0.5536689758300781, + "learning_rate": 0.003, + "loss": 3.974, + "step": 32661 + }, + { + "epoch": 0.32662, + "grad_norm": 0.5456752181053162, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 32662 + }, + { + "epoch": 0.32663, + "grad_norm": 0.6196727156639099, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 32663 + }, + { + "epoch": 0.32664, + "grad_norm": 0.7300091981887817, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 32664 + }, + { + "epoch": 0.32665, + "grad_norm": 0.9741966128349304, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 32665 + }, + { + "epoch": 0.32666, + "grad_norm": 1.0156868696212769, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 32666 + }, + { + "epoch": 0.32667, + "grad_norm": 0.8857244849205017, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 32667 + }, + { + "epoch": 0.32668, + "grad_norm": 0.7587184906005859, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 32668 + }, + { + "epoch": 0.32669, + "grad_norm": 0.728314995765686, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 32669 + }, + { + "epoch": 0.3267, + "grad_norm": 0.6885585784912109, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 32670 + }, + { + "epoch": 0.32671, + "grad_norm": 0.8091707825660706, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 32671 + }, + { + "epoch": 0.32672, + "grad_norm": 1.0514538288116455, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 32672 + }, + { + "epoch": 0.32673, + "grad_norm": 1.04509437084198, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 32673 + }, + { + "epoch": 0.32674, + "grad_norm": 0.9990705251693726, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 32674 + }, + { + "epoch": 0.32675, + "grad_norm": 0.8779171109199524, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 32675 + }, + { + "epoch": 0.32676, + "grad_norm": 0.9433568120002747, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 32676 + }, + { + "epoch": 0.32677, + "grad_norm": 0.8630367517471313, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 32677 + }, + { + "epoch": 0.32678, + "grad_norm": 0.6300763487815857, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 32678 + }, + { + "epoch": 0.32679, + "grad_norm": 0.7701874375343323, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 32679 + }, + { + "epoch": 0.3268, + "grad_norm": 0.8937686085700989, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 32680 + }, + { + "epoch": 0.32681, + "grad_norm": 1.0496277809143066, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 32681 + }, + { + "epoch": 0.32682, + "grad_norm": 0.9315981864929199, + "learning_rate": 0.003, + "loss": 3.992, + "step": 32682 + }, + { + "epoch": 0.32683, + "grad_norm": 1.0832232236862183, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 32683 + }, + { + "epoch": 0.32684, + "grad_norm": 0.9758277535438538, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 32684 + }, + { + "epoch": 0.32685, + "grad_norm": 0.8527059555053711, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 32685 + }, + { + "epoch": 0.32686, + "grad_norm": 0.9752069115638733, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 32686 + }, + { + "epoch": 0.32687, + "grad_norm": 1.028231143951416, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 32687 + }, + { + "epoch": 0.32688, + "grad_norm": 1.079309105873108, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 32688 + }, + { + "epoch": 0.32689, + "grad_norm": 1.0103952884674072, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 32689 + }, + { + "epoch": 0.3269, + "grad_norm": 1.139124870300293, + "learning_rate": 0.003, + "loss": 3.983, + "step": 32690 + }, + { + "epoch": 0.32691, + "grad_norm": 0.9731606245040894, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 32691 + }, + { + "epoch": 0.32692, + "grad_norm": 0.886932373046875, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 32692 + }, + { + "epoch": 0.32693, + "grad_norm": 0.8464771509170532, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 32693 + }, + { + "epoch": 0.32694, + "grad_norm": 0.8368782997131348, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 32694 + }, + { + "epoch": 0.32695, + "grad_norm": 0.8764780759811401, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 32695 + }, + { + "epoch": 0.32696, + "grad_norm": 0.8279525637626648, + "learning_rate": 0.003, + "loss": 3.986, + "step": 32696 + }, + { + "epoch": 0.32697, + "grad_norm": 0.8462624549865723, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 32697 + }, + { + "epoch": 0.32698, + "grad_norm": 0.8764651417732239, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 32698 + }, + { + "epoch": 0.32699, + "grad_norm": 0.9042829275131226, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 32699 + }, + { + "epoch": 0.327, + "grad_norm": 0.8409625887870789, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 32700 + }, + { + "epoch": 0.32701, + "grad_norm": 0.7230122089385986, + "learning_rate": 0.003, + "loss": 3.9329, + "step": 32701 + }, + { + "epoch": 0.32702, + "grad_norm": 0.6409006714820862, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 32702 + }, + { + "epoch": 0.32703, + "grad_norm": 0.5878980159759521, + "learning_rate": 0.003, + "loss": 3.966, + "step": 32703 + }, + { + "epoch": 0.32704, + "grad_norm": 0.5639416575431824, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 32704 + }, + { + "epoch": 0.32705, + "grad_norm": 0.6100924611091614, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 32705 + }, + { + "epoch": 0.32706, + "grad_norm": 0.6563337445259094, + "learning_rate": 0.003, + "loss": 3.984, + "step": 32706 + }, + { + "epoch": 0.32707, + "grad_norm": 0.7238401174545288, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 32707 + }, + { + "epoch": 0.32708, + "grad_norm": 0.8289658427238464, + "learning_rate": 0.003, + "loss": 3.959, + "step": 32708 + }, + { + "epoch": 0.32709, + "grad_norm": 0.973770022392273, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 32709 + }, + { + "epoch": 0.3271, + "grad_norm": 1.0694218873977661, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 32710 + }, + { + "epoch": 0.32711, + "grad_norm": 0.7564653158187866, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 32711 + }, + { + "epoch": 0.32712, + "grad_norm": 0.6172417998313904, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 32712 + }, + { + "epoch": 0.32713, + "grad_norm": 0.5990946292877197, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 32713 + }, + { + "epoch": 0.32714, + "grad_norm": 0.5760879516601562, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 32714 + }, + { + "epoch": 0.32715, + "grad_norm": 0.6629804968833923, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 32715 + }, + { + "epoch": 0.32716, + "grad_norm": 0.7397430539131165, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 32716 + }, + { + "epoch": 0.32717, + "grad_norm": 0.9611909985542297, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 32717 + }, + { + "epoch": 0.32718, + "grad_norm": 1.4628361463546753, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 32718 + }, + { + "epoch": 0.32719, + "grad_norm": 0.9141333699226379, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 32719 + }, + { + "epoch": 0.3272, + "grad_norm": 0.9466087818145752, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 32720 + }, + { + "epoch": 0.32721, + "grad_norm": 1.040136694908142, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 32721 + }, + { + "epoch": 0.32722, + "grad_norm": 1.162545084953308, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 32722 + }, + { + "epoch": 0.32723, + "grad_norm": 0.9019585847854614, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 32723 + }, + { + "epoch": 0.32724, + "grad_norm": 0.8515556454658508, + "learning_rate": 0.003, + "loss": 4.004, + "step": 32724 + }, + { + "epoch": 0.32725, + "grad_norm": 1.0446103811264038, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 32725 + }, + { + "epoch": 0.32726, + "grad_norm": 0.8921390175819397, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 32726 + }, + { + "epoch": 0.32727, + "grad_norm": 0.9149133563041687, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 32727 + }, + { + "epoch": 0.32728, + "grad_norm": 0.8773419857025146, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 32728 + }, + { + "epoch": 0.32729, + "grad_norm": 0.9936851859092712, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 32729 + }, + { + "epoch": 0.3273, + "grad_norm": 1.1183733940124512, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 32730 + }, + { + "epoch": 0.32731, + "grad_norm": 0.9370763897895813, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 32731 + }, + { + "epoch": 0.32732, + "grad_norm": 1.124130129814148, + "learning_rate": 0.003, + "loss": 3.98, + "step": 32732 + }, + { + "epoch": 0.32733, + "grad_norm": 0.9443601965904236, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 32733 + }, + { + "epoch": 0.32734, + "grad_norm": 1.0453137159347534, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 32734 + }, + { + "epoch": 0.32735, + "grad_norm": 0.8185724020004272, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 32735 + }, + { + "epoch": 0.32736, + "grad_norm": 0.6604447960853577, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 32736 + }, + { + "epoch": 0.32737, + "grad_norm": 0.7522901296615601, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 32737 + }, + { + "epoch": 0.32738, + "grad_norm": 0.8215345740318298, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 32738 + }, + { + "epoch": 0.32739, + "grad_norm": 0.8931112289428711, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 32739 + }, + { + "epoch": 0.3274, + "grad_norm": 0.9305903911590576, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 32740 + }, + { + "epoch": 0.32741, + "grad_norm": 0.8683371543884277, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 32741 + }, + { + "epoch": 0.32742, + "grad_norm": 0.7002133131027222, + "learning_rate": 0.003, + "loss": 3.974, + "step": 32742 + }, + { + "epoch": 0.32743, + "grad_norm": 0.795042872428894, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 32743 + }, + { + "epoch": 0.32744, + "grad_norm": 0.9347782731056213, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 32744 + }, + { + "epoch": 0.32745, + "grad_norm": 1.043495774269104, + "learning_rate": 0.003, + "loss": 3.97, + "step": 32745 + }, + { + "epoch": 0.32746, + "grad_norm": 0.8221429586410522, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 32746 + }, + { + "epoch": 0.32747, + "grad_norm": 0.7696993350982666, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 32747 + }, + { + "epoch": 0.32748, + "grad_norm": 0.7852022647857666, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 32748 + }, + { + "epoch": 0.32749, + "grad_norm": 0.7248318195343018, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 32749 + }, + { + "epoch": 0.3275, + "grad_norm": 0.8114910125732422, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 32750 + }, + { + "epoch": 0.32751, + "grad_norm": 0.8384526968002319, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 32751 + }, + { + "epoch": 0.32752, + "grad_norm": 1.114458680152893, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 32752 + }, + { + "epoch": 0.32753, + "grad_norm": 1.3822013139724731, + "learning_rate": 0.003, + "loss": 4.005, + "step": 32753 + }, + { + "epoch": 0.32754, + "grad_norm": 0.7284820079803467, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 32754 + }, + { + "epoch": 0.32755, + "grad_norm": 0.6565144062042236, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 32755 + }, + { + "epoch": 0.32756, + "grad_norm": 0.6362260580062866, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 32756 + }, + { + "epoch": 0.32757, + "grad_norm": 0.6384859681129456, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 32757 + }, + { + "epoch": 0.32758, + "grad_norm": 0.7152785062789917, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 32758 + }, + { + "epoch": 0.32759, + "grad_norm": 0.6770491600036621, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 32759 + }, + { + "epoch": 0.3276, + "grad_norm": 0.9011662602424622, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 32760 + }, + { + "epoch": 0.32761, + "grad_norm": 1.052607774734497, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 32761 + }, + { + "epoch": 0.32762, + "grad_norm": 1.0642908811569214, + "learning_rate": 0.003, + "loss": 3.994, + "step": 32762 + }, + { + "epoch": 0.32763, + "grad_norm": 1.0728535652160645, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 32763 + }, + { + "epoch": 0.32764, + "grad_norm": 0.9487437009811401, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 32764 + }, + { + "epoch": 0.32765, + "grad_norm": 0.8466985821723938, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 32765 + }, + { + "epoch": 0.32766, + "grad_norm": 0.7580529451370239, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 32766 + }, + { + "epoch": 0.32767, + "grad_norm": 0.8041210770606995, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 32767 + }, + { + "epoch": 0.32768, + "grad_norm": 0.9442958831787109, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 32768 + }, + { + "epoch": 0.32769, + "grad_norm": 1.0231817960739136, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 32769 + }, + { + "epoch": 0.3277, + "grad_norm": 1.0859543085098267, + "learning_rate": 0.003, + "loss": 3.989, + "step": 32770 + }, + { + "epoch": 0.32771, + "grad_norm": 0.9902774691581726, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 32771 + }, + { + "epoch": 0.32772, + "grad_norm": 0.9790458679199219, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 32772 + }, + { + "epoch": 0.32773, + "grad_norm": 0.8235830068588257, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 32773 + }, + { + "epoch": 0.32774, + "grad_norm": 0.7395147681236267, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 32774 + }, + { + "epoch": 0.32775, + "grad_norm": 0.8342896103858948, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 32775 + }, + { + "epoch": 0.32776, + "grad_norm": 0.9653837084770203, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 32776 + }, + { + "epoch": 0.32777, + "grad_norm": 1.0026121139526367, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 32777 + }, + { + "epoch": 0.32778, + "grad_norm": 0.952411949634552, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 32778 + }, + { + "epoch": 0.32779, + "grad_norm": 0.9597721099853516, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 32779 + }, + { + "epoch": 0.3278, + "grad_norm": 0.8170620203018188, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 32780 + }, + { + "epoch": 0.32781, + "grad_norm": 0.7967103719711304, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 32781 + }, + { + "epoch": 0.32782, + "grad_norm": 0.713830292224884, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 32782 + }, + { + "epoch": 0.32783, + "grad_norm": 0.7800357341766357, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 32783 + }, + { + "epoch": 0.32784, + "grad_norm": 0.8366248607635498, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 32784 + }, + { + "epoch": 0.32785, + "grad_norm": 0.6908177137374878, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 32785 + }, + { + "epoch": 0.32786, + "grad_norm": 0.6093934178352356, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 32786 + }, + { + "epoch": 0.32787, + "grad_norm": 0.6133504509925842, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 32787 + }, + { + "epoch": 0.32788, + "grad_norm": 0.6386643052101135, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 32788 + }, + { + "epoch": 0.32789, + "grad_norm": 0.7180176973342896, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 32789 + }, + { + "epoch": 0.3279, + "grad_norm": 0.844912052154541, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 32790 + }, + { + "epoch": 0.32791, + "grad_norm": 1.1447315216064453, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 32791 + }, + { + "epoch": 0.32792, + "grad_norm": 1.0001410245895386, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 32792 + }, + { + "epoch": 0.32793, + "grad_norm": 1.2277112007141113, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 32793 + }, + { + "epoch": 0.32794, + "grad_norm": 0.8762893676757812, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 32794 + }, + { + "epoch": 0.32795, + "grad_norm": 1.0750930309295654, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 32795 + }, + { + "epoch": 0.32796, + "grad_norm": 1.1810953617095947, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 32796 + }, + { + "epoch": 0.32797, + "grad_norm": 0.890525758266449, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 32797 + }, + { + "epoch": 0.32798, + "grad_norm": 0.9141677618026733, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 32798 + }, + { + "epoch": 0.32799, + "grad_norm": 0.9495511054992676, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 32799 + }, + { + "epoch": 0.328, + "grad_norm": 1.0360292196273804, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 32800 + }, + { + "epoch": 0.32801, + "grad_norm": 1.0130566358566284, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 32801 + }, + { + "epoch": 0.32802, + "grad_norm": 1.0566668510437012, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 32802 + }, + { + "epoch": 0.32803, + "grad_norm": 1.018157720565796, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 32803 + }, + { + "epoch": 0.32804, + "grad_norm": 1.238563060760498, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 32804 + }, + { + "epoch": 0.32805, + "grad_norm": 0.9053919315338135, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 32805 + }, + { + "epoch": 0.32806, + "grad_norm": 0.8527353405952454, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 32806 + }, + { + "epoch": 0.32807, + "grad_norm": 0.8408944606781006, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 32807 + }, + { + "epoch": 0.32808, + "grad_norm": 0.7700455188751221, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 32808 + }, + { + "epoch": 0.32809, + "grad_norm": 0.7224482297897339, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 32809 + }, + { + "epoch": 0.3281, + "grad_norm": 0.5925546884536743, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 32810 + }, + { + "epoch": 0.32811, + "grad_norm": 0.649031400680542, + "learning_rate": 0.003, + "loss": 3.997, + "step": 32811 + }, + { + "epoch": 0.32812, + "grad_norm": 0.6009167432785034, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 32812 + }, + { + "epoch": 0.32813, + "grad_norm": 0.5775517821311951, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 32813 + }, + { + "epoch": 0.32814, + "grad_norm": 0.6158239245414734, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 32814 + }, + { + "epoch": 0.32815, + "grad_norm": 0.6626220345497131, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 32815 + }, + { + "epoch": 0.32816, + "grad_norm": 0.6261868476867676, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 32816 + }, + { + "epoch": 0.32817, + "grad_norm": 0.7155467867851257, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 32817 + }, + { + "epoch": 0.32818, + "grad_norm": 0.8406774997711182, + "learning_rate": 0.003, + "loss": 3.9462, + "step": 32818 + }, + { + "epoch": 0.32819, + "grad_norm": 0.9371873736381531, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 32819 + }, + { + "epoch": 0.3282, + "grad_norm": 1.2663441896438599, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 32820 + }, + { + "epoch": 0.32821, + "grad_norm": 0.8887687921524048, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 32821 + }, + { + "epoch": 0.32822, + "grad_norm": 0.8181440234184265, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 32822 + }, + { + "epoch": 0.32823, + "grad_norm": 0.8527613282203674, + "learning_rate": 0.003, + "loss": 3.995, + "step": 32823 + }, + { + "epoch": 0.32824, + "grad_norm": 1.08590567111969, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 32824 + }, + { + "epoch": 0.32825, + "grad_norm": 1.0421407222747803, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 32825 + }, + { + "epoch": 0.32826, + "grad_norm": 1.1449872255325317, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 32826 + }, + { + "epoch": 0.32827, + "grad_norm": 0.8719208836555481, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 32827 + }, + { + "epoch": 0.32828, + "grad_norm": 0.8368306756019592, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 32828 + }, + { + "epoch": 0.32829, + "grad_norm": 0.9192578792572021, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 32829 + }, + { + "epoch": 0.3283, + "grad_norm": 1.0582367181777954, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 32830 + }, + { + "epoch": 0.32831, + "grad_norm": 1.1562347412109375, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 32831 + }, + { + "epoch": 0.32832, + "grad_norm": 0.8098315596580505, + "learning_rate": 0.003, + "loss": 3.98, + "step": 32832 + }, + { + "epoch": 0.32833, + "grad_norm": 0.6539410948753357, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 32833 + }, + { + "epoch": 0.32834, + "grad_norm": 0.811809241771698, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 32834 + }, + { + "epoch": 0.32835, + "grad_norm": 0.9752480983734131, + "learning_rate": 0.003, + "loss": 3.991, + "step": 32835 + }, + { + "epoch": 0.32836, + "grad_norm": 0.9762851595878601, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 32836 + }, + { + "epoch": 0.32837, + "grad_norm": 1.0349689722061157, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 32837 + }, + { + "epoch": 0.32838, + "grad_norm": 1.103812336921692, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 32838 + }, + { + "epoch": 0.32839, + "grad_norm": 0.7367599606513977, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 32839 + }, + { + "epoch": 0.3284, + "grad_norm": 0.6564750671386719, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 32840 + }, + { + "epoch": 0.32841, + "grad_norm": 0.7198147177696228, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 32841 + }, + { + "epoch": 0.32842, + "grad_norm": 0.9287322759628296, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 32842 + }, + { + "epoch": 0.32843, + "grad_norm": 1.2001981735229492, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 32843 + }, + { + "epoch": 0.32844, + "grad_norm": 0.7640447020530701, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 32844 + }, + { + "epoch": 0.32845, + "grad_norm": 0.6748675107955933, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 32845 + }, + { + "epoch": 0.32846, + "grad_norm": 0.6296458840370178, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 32846 + }, + { + "epoch": 0.32847, + "grad_norm": 0.6009144186973572, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 32847 + }, + { + "epoch": 0.32848, + "grad_norm": 0.638357400894165, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 32848 + }, + { + "epoch": 0.32849, + "grad_norm": 0.637729823589325, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 32849 + }, + { + "epoch": 0.3285, + "grad_norm": 0.6898089051246643, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 32850 + }, + { + "epoch": 0.32851, + "grad_norm": 0.8521791696548462, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 32851 + }, + { + "epoch": 0.32852, + "grad_norm": 1.0824999809265137, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 32852 + }, + { + "epoch": 0.32853, + "grad_norm": 1.1443674564361572, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 32853 + }, + { + "epoch": 0.32854, + "grad_norm": 0.852571427822113, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 32854 + }, + { + "epoch": 0.32855, + "grad_norm": 0.8061854839324951, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 32855 + }, + { + "epoch": 0.32856, + "grad_norm": 0.9532166719436646, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 32856 + }, + { + "epoch": 0.32857, + "grad_norm": 1.0930554866790771, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 32857 + }, + { + "epoch": 0.32858, + "grad_norm": 0.9275783896446228, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 32858 + }, + { + "epoch": 0.32859, + "grad_norm": 0.9672372341156006, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 32859 + }, + { + "epoch": 0.3286, + "grad_norm": 1.210656762123108, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 32860 + }, + { + "epoch": 0.32861, + "grad_norm": 0.845602810382843, + "learning_rate": 0.003, + "loss": 3.9376, + "step": 32861 + }, + { + "epoch": 0.32862, + "grad_norm": 0.9853912591934204, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 32862 + }, + { + "epoch": 0.32863, + "grad_norm": 1.2285250425338745, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 32863 + }, + { + "epoch": 0.32864, + "grad_norm": 1.068718671798706, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 32864 + }, + { + "epoch": 0.32865, + "grad_norm": 0.9282857179641724, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 32865 + }, + { + "epoch": 0.32866, + "grad_norm": 0.8707215189933777, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 32866 + }, + { + "epoch": 0.32867, + "grad_norm": 0.7034040093421936, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 32867 + }, + { + "epoch": 0.32868, + "grad_norm": 0.6550178527832031, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 32868 + }, + { + "epoch": 0.32869, + "grad_norm": 0.6666882038116455, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 32869 + }, + { + "epoch": 0.3287, + "grad_norm": 0.6081922054290771, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 32870 + }, + { + "epoch": 0.32871, + "grad_norm": 0.6227262020111084, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 32871 + }, + { + "epoch": 0.32872, + "grad_norm": 0.8549386858940125, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 32872 + }, + { + "epoch": 0.32873, + "grad_norm": 1.0605710744857788, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 32873 + }, + { + "epoch": 0.32874, + "grad_norm": 1.0644570589065552, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 32874 + }, + { + "epoch": 0.32875, + "grad_norm": 1.0382479429244995, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 32875 + }, + { + "epoch": 0.32876, + "grad_norm": 1.0426589250564575, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 32876 + }, + { + "epoch": 0.32877, + "grad_norm": 1.0340776443481445, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 32877 + }, + { + "epoch": 0.32878, + "grad_norm": 1.1006078720092773, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 32878 + }, + { + "epoch": 0.32879, + "grad_norm": 1.0017911195755005, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 32879 + }, + { + "epoch": 0.3288, + "grad_norm": 1.1220600605010986, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 32880 + }, + { + "epoch": 0.32881, + "grad_norm": 0.8964804410934448, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 32881 + }, + { + "epoch": 0.32882, + "grad_norm": 0.9893882870674133, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 32882 + }, + { + "epoch": 0.32883, + "grad_norm": 1.1505235433578491, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 32883 + }, + { + "epoch": 0.32884, + "grad_norm": 0.8474794626235962, + "learning_rate": 0.003, + "loss": 3.994, + "step": 32884 + }, + { + "epoch": 0.32885, + "grad_norm": 0.8068058490753174, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 32885 + }, + { + "epoch": 0.32886, + "grad_norm": 0.7477496862411499, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 32886 + }, + { + "epoch": 0.32887, + "grad_norm": 0.7489718198776245, + "learning_rate": 0.003, + "loss": 3.982, + "step": 32887 + }, + { + "epoch": 0.32888, + "grad_norm": 0.7481356263160706, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 32888 + }, + { + "epoch": 0.32889, + "grad_norm": 0.844550609588623, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 32889 + }, + { + "epoch": 0.3289, + "grad_norm": 1.1334326267242432, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 32890 + }, + { + "epoch": 0.32891, + "grad_norm": 0.9248389005661011, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 32891 + }, + { + "epoch": 0.32892, + "grad_norm": 0.6314566731452942, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 32892 + }, + { + "epoch": 0.32893, + "grad_norm": 0.7010231018066406, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 32893 + }, + { + "epoch": 0.32894, + "grad_norm": 0.72366863489151, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 32894 + }, + { + "epoch": 0.32895, + "grad_norm": 0.8347010016441345, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 32895 + }, + { + "epoch": 0.32896, + "grad_norm": 1.0302575826644897, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 32896 + }, + { + "epoch": 0.32897, + "grad_norm": 1.2061032056808472, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 32897 + }, + { + "epoch": 0.32898, + "grad_norm": 0.6878505349159241, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 32898 + }, + { + "epoch": 0.32899, + "grad_norm": 0.8613277077674866, + "learning_rate": 0.003, + "loss": 3.998, + "step": 32899 + }, + { + "epoch": 0.329, + "grad_norm": 1.2354670763015747, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 32900 + }, + { + "epoch": 0.32901, + "grad_norm": 0.7532533407211304, + "learning_rate": 0.003, + "loss": 4.011, + "step": 32901 + }, + { + "epoch": 0.32902, + "grad_norm": 0.6462257504463196, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 32902 + }, + { + "epoch": 0.32903, + "grad_norm": 0.6564210653305054, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 32903 + }, + { + "epoch": 0.32904, + "grad_norm": 0.7167737483978271, + "learning_rate": 0.003, + "loss": 3.9459, + "step": 32904 + }, + { + "epoch": 0.32905, + "grad_norm": 0.8238648772239685, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 32905 + }, + { + "epoch": 0.32906, + "grad_norm": 0.839614748954773, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 32906 + }, + { + "epoch": 0.32907, + "grad_norm": 0.8224316239356995, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 32907 + }, + { + "epoch": 0.32908, + "grad_norm": 0.8626474142074585, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 32908 + }, + { + "epoch": 0.32909, + "grad_norm": 0.8740398287773132, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 32909 + }, + { + "epoch": 0.3291, + "grad_norm": 0.8891167640686035, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 32910 + }, + { + "epoch": 0.32911, + "grad_norm": 0.8534160256385803, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 32911 + }, + { + "epoch": 0.32912, + "grad_norm": 0.8148693442344666, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 32912 + }, + { + "epoch": 0.32913, + "grad_norm": 0.8756507635116577, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 32913 + }, + { + "epoch": 0.32914, + "grad_norm": 1.057644009590149, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 32914 + }, + { + "epoch": 0.32915, + "grad_norm": 1.0693341493606567, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 32915 + }, + { + "epoch": 0.32916, + "grad_norm": 0.8992453813552856, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 32916 + }, + { + "epoch": 0.32917, + "grad_norm": 1.0667215585708618, + "learning_rate": 0.003, + "loss": 4.002, + "step": 32917 + }, + { + "epoch": 0.32918, + "grad_norm": 1.0586974620819092, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 32918 + }, + { + "epoch": 0.32919, + "grad_norm": 0.9238132238388062, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 32919 + }, + { + "epoch": 0.3292, + "grad_norm": 0.8602195978164673, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 32920 + }, + { + "epoch": 0.32921, + "grad_norm": 0.8724820017814636, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 32921 + }, + { + "epoch": 0.32922, + "grad_norm": 0.8883758187294006, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 32922 + }, + { + "epoch": 0.32923, + "grad_norm": 0.9703934192657471, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 32923 + }, + { + "epoch": 0.32924, + "grad_norm": 1.2227078676223755, + "learning_rate": 0.003, + "loss": 3.968, + "step": 32924 + }, + { + "epoch": 0.32925, + "grad_norm": 0.9266354441642761, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 32925 + }, + { + "epoch": 0.32926, + "grad_norm": 0.7785370945930481, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 32926 + }, + { + "epoch": 0.32927, + "grad_norm": 0.8030340075492859, + "learning_rate": 0.003, + "loss": 4.014, + "step": 32927 + }, + { + "epoch": 0.32928, + "grad_norm": 0.8877881765365601, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 32928 + }, + { + "epoch": 0.32929, + "grad_norm": 1.026104211807251, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 32929 + }, + { + "epoch": 0.3293, + "grad_norm": 1.2729439735412598, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 32930 + }, + { + "epoch": 0.32931, + "grad_norm": 1.2001529932022095, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 32931 + }, + { + "epoch": 0.32932, + "grad_norm": 1.023717999458313, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 32932 + }, + { + "epoch": 0.32933, + "grad_norm": 0.8832988739013672, + "learning_rate": 0.003, + "loss": 4.019, + "step": 32933 + }, + { + "epoch": 0.32934, + "grad_norm": 0.8862455487251282, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 32934 + }, + { + "epoch": 0.32935, + "grad_norm": 1.0115257501602173, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 32935 + }, + { + "epoch": 0.32936, + "grad_norm": 0.9591469168663025, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 32936 + }, + { + "epoch": 0.32937, + "grad_norm": 0.8839733004570007, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 32937 + }, + { + "epoch": 0.32938, + "grad_norm": 0.7704980969429016, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 32938 + }, + { + "epoch": 0.32939, + "grad_norm": 0.6692459583282471, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 32939 + }, + { + "epoch": 0.3294, + "grad_norm": 0.6956064105033875, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 32940 + }, + { + "epoch": 0.32941, + "grad_norm": 0.6961161494255066, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 32941 + }, + { + "epoch": 0.32942, + "grad_norm": 0.7158181071281433, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 32942 + }, + { + "epoch": 0.32943, + "grad_norm": 0.6957089304924011, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 32943 + }, + { + "epoch": 0.32944, + "grad_norm": 0.702486515045166, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 32944 + }, + { + "epoch": 0.32945, + "grad_norm": 0.7958844304084778, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 32945 + }, + { + "epoch": 0.32946, + "grad_norm": 0.9739514589309692, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 32946 + }, + { + "epoch": 0.32947, + "grad_norm": 1.1546005010604858, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 32947 + }, + { + "epoch": 0.32948, + "grad_norm": 0.9115421175956726, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 32948 + }, + { + "epoch": 0.32949, + "grad_norm": 0.9142782092094421, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 32949 + }, + { + "epoch": 0.3295, + "grad_norm": 0.9681139588356018, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 32950 + }, + { + "epoch": 0.32951, + "grad_norm": 1.112369418144226, + "learning_rate": 0.003, + "loss": 4.0, + "step": 32951 + }, + { + "epoch": 0.32952, + "grad_norm": 0.9164697527885437, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 32952 + }, + { + "epoch": 0.32953, + "grad_norm": 0.9662401676177979, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 32953 + }, + { + "epoch": 0.32954, + "grad_norm": 1.0259596109390259, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 32954 + }, + { + "epoch": 0.32955, + "grad_norm": 1.0117791891098022, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 32955 + }, + { + "epoch": 0.32956, + "grad_norm": 0.9963108897209167, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 32956 + }, + { + "epoch": 0.32957, + "grad_norm": 0.8257370591163635, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 32957 + }, + { + "epoch": 0.32958, + "grad_norm": 0.9883521199226379, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 32958 + }, + { + "epoch": 0.32959, + "grad_norm": 1.2157396078109741, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 32959 + }, + { + "epoch": 0.3296, + "grad_norm": 0.8566860556602478, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 32960 + }, + { + "epoch": 0.32961, + "grad_norm": 0.9295904636383057, + "learning_rate": 0.003, + "loss": 3.99, + "step": 32961 + }, + { + "epoch": 0.32962, + "grad_norm": 1.1822417974472046, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 32962 + }, + { + "epoch": 0.32963, + "grad_norm": 0.9353878498077393, + "learning_rate": 0.003, + "loss": 3.963, + "step": 32963 + }, + { + "epoch": 0.32964, + "grad_norm": 1.0581222772598267, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 32964 + }, + { + "epoch": 0.32965, + "grad_norm": 0.931227445602417, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 32965 + }, + { + "epoch": 0.32966, + "grad_norm": 0.8077581524848938, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 32966 + }, + { + "epoch": 0.32967, + "grad_norm": 0.7183004021644592, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 32967 + }, + { + "epoch": 0.32968, + "grad_norm": 0.7586396336555481, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 32968 + }, + { + "epoch": 0.32969, + "grad_norm": 0.6724128127098083, + "learning_rate": 0.003, + "loss": 4.002, + "step": 32969 + }, + { + "epoch": 0.3297, + "grad_norm": 0.7135902047157288, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 32970 + }, + { + "epoch": 0.32971, + "grad_norm": 0.9621875882148743, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 32971 + }, + { + "epoch": 0.32972, + "grad_norm": 1.1530883312225342, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 32972 + }, + { + "epoch": 0.32973, + "grad_norm": 0.7884253859519958, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 32973 + }, + { + "epoch": 0.32974, + "grad_norm": 0.7436503171920776, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 32974 + }, + { + "epoch": 0.32975, + "grad_norm": 0.8306210041046143, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 32975 + }, + { + "epoch": 0.32976, + "grad_norm": 0.9902637004852295, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 32976 + }, + { + "epoch": 0.32977, + "grad_norm": 1.059831142425537, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 32977 + }, + { + "epoch": 0.32978, + "grad_norm": 1.0130012035369873, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 32978 + }, + { + "epoch": 0.32979, + "grad_norm": 0.9761448502540588, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 32979 + }, + { + "epoch": 0.3298, + "grad_norm": 0.9118365049362183, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 32980 + }, + { + "epoch": 0.32981, + "grad_norm": 0.7579766511917114, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 32981 + }, + { + "epoch": 0.32982, + "grad_norm": 0.6672433614730835, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 32982 + }, + { + "epoch": 0.32983, + "grad_norm": 0.5904834270477295, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 32983 + }, + { + "epoch": 0.32984, + "grad_norm": 0.5945793390274048, + "learning_rate": 0.003, + "loss": 3.989, + "step": 32984 + }, + { + "epoch": 0.32985, + "grad_norm": 0.6358540654182434, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 32985 + }, + { + "epoch": 0.32986, + "grad_norm": 0.6277160048484802, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 32986 + }, + { + "epoch": 0.32987, + "grad_norm": 0.6975568532943726, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 32987 + }, + { + "epoch": 0.32988, + "grad_norm": 0.8194594383239746, + "learning_rate": 0.003, + "loss": 3.975, + "step": 32988 + }, + { + "epoch": 0.32989, + "grad_norm": 0.9466721415519714, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 32989 + }, + { + "epoch": 0.3299, + "grad_norm": 1.1281929016113281, + "learning_rate": 0.003, + "loss": 3.9388, + "step": 32990 + }, + { + "epoch": 0.32991, + "grad_norm": 1.1436468362808228, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 32991 + }, + { + "epoch": 0.32992, + "grad_norm": 0.8920917510986328, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 32992 + }, + { + "epoch": 0.32993, + "grad_norm": 0.9645689725875854, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 32993 + }, + { + "epoch": 0.32994, + "grad_norm": 1.001807689666748, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 32994 + }, + { + "epoch": 0.32995, + "grad_norm": 1.2743043899536133, + "learning_rate": 0.003, + "loss": 3.984, + "step": 32995 + }, + { + "epoch": 0.32996, + "grad_norm": 0.6770648956298828, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 32996 + }, + { + "epoch": 0.32997, + "grad_norm": 0.8179088234901428, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 32997 + }, + { + "epoch": 0.32998, + "grad_norm": 0.892920196056366, + "learning_rate": 0.003, + "loss": 3.985, + "step": 32998 + }, + { + "epoch": 0.32999, + "grad_norm": 0.8981190919876099, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 32999 + }, + { + "epoch": 0.33, + "grad_norm": 0.952021598815918, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 33000 + }, + { + "epoch": 0.33001, + "grad_norm": 1.0821113586425781, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 33001 + }, + { + "epoch": 0.33002, + "grad_norm": 1.1461098194122314, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 33002 + }, + { + "epoch": 0.33003, + "grad_norm": 0.7834789752960205, + "learning_rate": 0.003, + "loss": 4.005, + "step": 33003 + }, + { + "epoch": 0.33004, + "grad_norm": 0.6315568685531616, + "learning_rate": 0.003, + "loss": 3.988, + "step": 33004 + }, + { + "epoch": 0.33005, + "grad_norm": 1.0483967065811157, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 33005 + }, + { + "epoch": 0.33006, + "grad_norm": 1.0085155963897705, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 33006 + }, + { + "epoch": 0.33007, + "grad_norm": 0.8829138875007629, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 33007 + }, + { + "epoch": 0.33008, + "grad_norm": 0.9477053880691528, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 33008 + }, + { + "epoch": 0.33009, + "grad_norm": 1.0861448049545288, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 33009 + }, + { + "epoch": 0.3301, + "grad_norm": 0.8316453695297241, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 33010 + }, + { + "epoch": 0.33011, + "grad_norm": 0.8770897388458252, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 33011 + }, + { + "epoch": 0.33012, + "grad_norm": 0.9470415115356445, + "learning_rate": 0.003, + "loss": 3.989, + "step": 33012 + }, + { + "epoch": 0.33013, + "grad_norm": 1.0456143617630005, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 33013 + }, + { + "epoch": 0.33014, + "grad_norm": 0.9501416087150574, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 33014 + }, + { + "epoch": 0.33015, + "grad_norm": 1.0080584287643433, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 33015 + }, + { + "epoch": 0.33016, + "grad_norm": 1.0703943967819214, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 33016 + }, + { + "epoch": 0.33017, + "grad_norm": 0.9771105647087097, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 33017 + }, + { + "epoch": 0.33018, + "grad_norm": 0.9976072907447815, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 33018 + }, + { + "epoch": 0.33019, + "grad_norm": 0.9218783974647522, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 33019 + }, + { + "epoch": 0.3302, + "grad_norm": 0.8744208812713623, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 33020 + }, + { + "epoch": 0.33021, + "grad_norm": 0.7866498231887817, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 33021 + }, + { + "epoch": 0.33022, + "grad_norm": 0.9308577179908752, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 33022 + }, + { + "epoch": 0.33023, + "grad_norm": 1.0746090412139893, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 33023 + }, + { + "epoch": 0.33024, + "grad_norm": 1.0413289070129395, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 33024 + }, + { + "epoch": 0.33025, + "grad_norm": 1.138123631477356, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 33025 + }, + { + "epoch": 0.33026, + "grad_norm": 0.9232091903686523, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 33026 + }, + { + "epoch": 0.33027, + "grad_norm": 0.7590427994728088, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 33027 + }, + { + "epoch": 0.33028, + "grad_norm": 0.7859970927238464, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 33028 + }, + { + "epoch": 0.33029, + "grad_norm": 1.020012617111206, + "learning_rate": 0.003, + "loss": 3.99, + "step": 33029 + }, + { + "epoch": 0.3303, + "grad_norm": 1.2047462463378906, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 33030 + }, + { + "epoch": 0.33031, + "grad_norm": 0.8906852006912231, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 33031 + }, + { + "epoch": 0.33032, + "grad_norm": 1.0220338106155396, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 33032 + }, + { + "epoch": 0.33033, + "grad_norm": 1.0633891820907593, + "learning_rate": 0.003, + "loss": 3.983, + "step": 33033 + }, + { + "epoch": 0.33034, + "grad_norm": 0.9428620934486389, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 33034 + }, + { + "epoch": 0.33035, + "grad_norm": 0.7931355834007263, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 33035 + }, + { + "epoch": 0.33036, + "grad_norm": 0.6051740646362305, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 33036 + }, + { + "epoch": 0.33037, + "grad_norm": 0.5410848259925842, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 33037 + }, + { + "epoch": 0.33038, + "grad_norm": 0.5931380987167358, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 33038 + }, + { + "epoch": 0.33039, + "grad_norm": 0.6748098731040955, + "learning_rate": 0.003, + "loss": 3.998, + "step": 33039 + }, + { + "epoch": 0.3304, + "grad_norm": 0.6363991498947144, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 33040 + }, + { + "epoch": 0.33041, + "grad_norm": 0.6925687789916992, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 33041 + }, + { + "epoch": 0.33042, + "grad_norm": 0.8446574807167053, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 33042 + }, + { + "epoch": 0.33043, + "grad_norm": 1.1171568632125854, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 33043 + }, + { + "epoch": 0.33044, + "grad_norm": 0.9289556741714478, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 33044 + }, + { + "epoch": 0.33045, + "grad_norm": 0.7384862303733826, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 33045 + }, + { + "epoch": 0.33046, + "grad_norm": 0.6354304552078247, + "learning_rate": 0.003, + "loss": 3.9493, + "step": 33046 + }, + { + "epoch": 0.33047, + "grad_norm": 0.6146836280822754, + "learning_rate": 0.003, + "loss": 3.9457, + "step": 33047 + }, + { + "epoch": 0.33048, + "grad_norm": 0.768444836139679, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 33048 + }, + { + "epoch": 0.33049, + "grad_norm": 0.8296108841896057, + "learning_rate": 0.003, + "loss": 3.971, + "step": 33049 + }, + { + "epoch": 0.3305, + "grad_norm": 0.926270604133606, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 33050 + }, + { + "epoch": 0.33051, + "grad_norm": 1.0745229721069336, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 33051 + }, + { + "epoch": 0.33052, + "grad_norm": 0.7827836275100708, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 33052 + }, + { + "epoch": 0.33053, + "grad_norm": 0.829764187335968, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 33053 + }, + { + "epoch": 0.33054, + "grad_norm": 0.9513168931007385, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 33054 + }, + { + "epoch": 0.33055, + "grad_norm": 0.8059423565864563, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 33055 + }, + { + "epoch": 0.33056, + "grad_norm": 0.7293146252632141, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 33056 + }, + { + "epoch": 0.33057, + "grad_norm": 0.8139077425003052, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 33057 + }, + { + "epoch": 0.33058, + "grad_norm": 0.9606841802597046, + "learning_rate": 0.003, + "loss": 3.982, + "step": 33058 + }, + { + "epoch": 0.33059, + "grad_norm": 1.1062369346618652, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 33059 + }, + { + "epoch": 0.3306, + "grad_norm": 0.9449945688247681, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 33060 + }, + { + "epoch": 0.33061, + "grad_norm": 0.8899131417274475, + "learning_rate": 0.003, + "loss": 4.004, + "step": 33061 + }, + { + "epoch": 0.33062, + "grad_norm": 0.8781537413597107, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 33062 + }, + { + "epoch": 0.33063, + "grad_norm": 0.9465190768241882, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 33063 + }, + { + "epoch": 0.33064, + "grad_norm": 1.0850176811218262, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 33064 + }, + { + "epoch": 0.33065, + "grad_norm": 1.0586122274398804, + "learning_rate": 0.003, + "loss": 3.976, + "step": 33065 + }, + { + "epoch": 0.33066, + "grad_norm": 0.9603906869888306, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 33066 + }, + { + "epoch": 0.33067, + "grad_norm": 0.9823393821716309, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 33067 + }, + { + "epoch": 0.33068, + "grad_norm": 1.0456013679504395, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 33068 + }, + { + "epoch": 0.33069, + "grad_norm": 0.9306734204292297, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 33069 + }, + { + "epoch": 0.3307, + "grad_norm": 0.954736590385437, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 33070 + }, + { + "epoch": 0.33071, + "grad_norm": 0.8970369100570679, + "learning_rate": 0.003, + "loss": 3.994, + "step": 33071 + }, + { + "epoch": 0.33072, + "grad_norm": 0.8645350933074951, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 33072 + }, + { + "epoch": 0.33073, + "grad_norm": 1.007315754890442, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 33073 + }, + { + "epoch": 0.33074, + "grad_norm": 1.2315601110458374, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 33074 + }, + { + "epoch": 0.33075, + "grad_norm": 1.0237139463424683, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 33075 + }, + { + "epoch": 0.33076, + "grad_norm": 1.0007908344268799, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 33076 + }, + { + "epoch": 0.33077, + "grad_norm": 0.9715906977653503, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 33077 + }, + { + "epoch": 0.33078, + "grad_norm": 0.8430658578872681, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 33078 + }, + { + "epoch": 0.33079, + "grad_norm": 1.0490097999572754, + "learning_rate": 0.003, + "loss": 4.022, + "step": 33079 + }, + { + "epoch": 0.3308, + "grad_norm": 0.9883865118026733, + "learning_rate": 0.003, + "loss": 3.984, + "step": 33080 + }, + { + "epoch": 0.33081, + "grad_norm": 0.9512540698051453, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 33081 + }, + { + "epoch": 0.33082, + "grad_norm": 0.999781608581543, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 33082 + }, + { + "epoch": 0.33083, + "grad_norm": 0.8996179699897766, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 33083 + }, + { + "epoch": 0.33084, + "grad_norm": 1.0241618156433105, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 33084 + }, + { + "epoch": 0.33085, + "grad_norm": 1.4913982152938843, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 33085 + }, + { + "epoch": 0.33086, + "grad_norm": 0.5317667126655579, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 33086 + }, + { + "epoch": 0.33087, + "grad_norm": 1.0655726194381714, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 33087 + }, + { + "epoch": 0.33088, + "grad_norm": 1.2737767696380615, + "learning_rate": 0.003, + "loss": 3.998, + "step": 33088 + }, + { + "epoch": 0.33089, + "grad_norm": 0.634140133857727, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 33089 + }, + { + "epoch": 0.3309, + "grad_norm": 0.6462652087211609, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 33090 + }, + { + "epoch": 0.33091, + "grad_norm": 0.6081231236457825, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 33091 + }, + { + "epoch": 0.33092, + "grad_norm": 0.5670910477638245, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 33092 + }, + { + "epoch": 0.33093, + "grad_norm": 0.5159702301025391, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 33093 + }, + { + "epoch": 0.33094, + "grad_norm": 0.5795565843582153, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 33094 + }, + { + "epoch": 0.33095, + "grad_norm": 0.6519309878349304, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 33095 + }, + { + "epoch": 0.33096, + "grad_norm": 0.8648343086242676, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 33096 + }, + { + "epoch": 0.33097, + "grad_norm": 1.1239269971847534, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 33097 + }, + { + "epoch": 0.33098, + "grad_norm": 1.0143076181411743, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 33098 + }, + { + "epoch": 0.33099, + "grad_norm": 0.8235583305358887, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 33099 + }, + { + "epoch": 0.331, + "grad_norm": 0.6124919056892395, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 33100 + }, + { + "epoch": 0.33101, + "grad_norm": 0.8320285081863403, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 33101 + }, + { + "epoch": 0.33102, + "grad_norm": 0.9937469959259033, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 33102 + }, + { + "epoch": 0.33103, + "grad_norm": 1.1507375240325928, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 33103 + }, + { + "epoch": 0.33104, + "grad_norm": 0.7720727920532227, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 33104 + }, + { + "epoch": 0.33105, + "grad_norm": 0.7316970825195312, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 33105 + }, + { + "epoch": 0.33106, + "grad_norm": 0.7713351845741272, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 33106 + }, + { + "epoch": 0.33107, + "grad_norm": 0.8256540894508362, + "learning_rate": 0.003, + "loss": 3.9412, + "step": 33107 + }, + { + "epoch": 0.33108, + "grad_norm": 0.8923250436782837, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 33108 + }, + { + "epoch": 0.33109, + "grad_norm": 0.8498046398162842, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 33109 + }, + { + "epoch": 0.3311, + "grad_norm": 0.8339502215385437, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 33110 + }, + { + "epoch": 0.33111, + "grad_norm": 0.8221234083175659, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 33111 + }, + { + "epoch": 0.33112, + "grad_norm": 0.7585002779960632, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 33112 + }, + { + "epoch": 0.33113, + "grad_norm": 0.8178539276123047, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 33113 + }, + { + "epoch": 0.33114, + "grad_norm": 0.8420496582984924, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 33114 + }, + { + "epoch": 0.33115, + "grad_norm": 0.8548955917358398, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 33115 + }, + { + "epoch": 0.33116, + "grad_norm": 1.02829909324646, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 33116 + }, + { + "epoch": 0.33117, + "grad_norm": 1.3416224718093872, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 33117 + }, + { + "epoch": 0.33118, + "grad_norm": 0.85359787940979, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 33118 + }, + { + "epoch": 0.33119, + "grad_norm": 0.8090051412582397, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 33119 + }, + { + "epoch": 0.3312, + "grad_norm": 1.025249719619751, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 33120 + }, + { + "epoch": 0.33121, + "grad_norm": 1.1869909763336182, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 33121 + }, + { + "epoch": 0.33122, + "grad_norm": 0.6073952913284302, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 33122 + }, + { + "epoch": 0.33123, + "grad_norm": 0.6382406949996948, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 33123 + }, + { + "epoch": 0.33124, + "grad_norm": 0.6868864893913269, + "learning_rate": 0.003, + "loss": 3.954, + "step": 33124 + }, + { + "epoch": 0.33125, + "grad_norm": 0.6877424716949463, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 33125 + }, + { + "epoch": 0.33126, + "grad_norm": 0.7449150681495667, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 33126 + }, + { + "epoch": 0.33127, + "grad_norm": 0.8146490454673767, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 33127 + }, + { + "epoch": 0.33128, + "grad_norm": 0.9523589611053467, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 33128 + }, + { + "epoch": 0.33129, + "grad_norm": 1.0168359279632568, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 33129 + }, + { + "epoch": 0.3313, + "grad_norm": 1.0407766103744507, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 33130 + }, + { + "epoch": 0.33131, + "grad_norm": 1.133920431137085, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 33131 + }, + { + "epoch": 0.33132, + "grad_norm": 0.8493016362190247, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 33132 + }, + { + "epoch": 0.33133, + "grad_norm": 0.829299807548523, + "learning_rate": 0.003, + "loss": 3.988, + "step": 33133 + }, + { + "epoch": 0.33134, + "grad_norm": 0.9316754937171936, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 33134 + }, + { + "epoch": 0.33135, + "grad_norm": 1.0851393938064575, + "learning_rate": 0.003, + "loss": 3.983, + "step": 33135 + }, + { + "epoch": 0.33136, + "grad_norm": 0.9917900562286377, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 33136 + }, + { + "epoch": 0.33137, + "grad_norm": 0.7751608490943909, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 33137 + }, + { + "epoch": 0.33138, + "grad_norm": 0.63100266456604, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 33138 + }, + { + "epoch": 0.33139, + "grad_norm": 0.668402910232544, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 33139 + }, + { + "epoch": 0.3314, + "grad_norm": 0.6266415119171143, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 33140 + }, + { + "epoch": 0.33141, + "grad_norm": 0.7502108812332153, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 33141 + }, + { + "epoch": 0.33142, + "grad_norm": 0.8613996505737305, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 33142 + }, + { + "epoch": 0.33143, + "grad_norm": 0.961875319480896, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 33143 + }, + { + "epoch": 0.33144, + "grad_norm": 1.1261621713638306, + "learning_rate": 0.003, + "loss": 3.998, + "step": 33144 + }, + { + "epoch": 0.33145, + "grad_norm": 0.9034062027931213, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 33145 + }, + { + "epoch": 0.33146, + "grad_norm": 0.876934826374054, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 33146 + }, + { + "epoch": 0.33147, + "grad_norm": 1.0130311250686646, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 33147 + }, + { + "epoch": 0.33148, + "grad_norm": 1.0293993949890137, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 33148 + }, + { + "epoch": 0.33149, + "grad_norm": 0.912652850151062, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 33149 + }, + { + "epoch": 0.3315, + "grad_norm": 1.479037880897522, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 33150 + }, + { + "epoch": 0.33151, + "grad_norm": 0.5926223397254944, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 33151 + }, + { + "epoch": 0.33152, + "grad_norm": 0.7975197434425354, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 33152 + }, + { + "epoch": 0.33153, + "grad_norm": 0.7794082760810852, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 33153 + }, + { + "epoch": 0.33154, + "grad_norm": 0.7967977523803711, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 33154 + }, + { + "epoch": 0.33155, + "grad_norm": 1.02908456325531, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 33155 + }, + { + "epoch": 0.33156, + "grad_norm": 1.4279999732971191, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 33156 + }, + { + "epoch": 0.33157, + "grad_norm": 0.7074572443962097, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 33157 + }, + { + "epoch": 0.33158, + "grad_norm": 0.7862478494644165, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 33158 + }, + { + "epoch": 0.33159, + "grad_norm": 0.8397165536880493, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 33159 + }, + { + "epoch": 0.3316, + "grad_norm": 0.9184638857841492, + "learning_rate": 0.003, + "loss": 3.968, + "step": 33160 + }, + { + "epoch": 0.33161, + "grad_norm": 1.126503825187683, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 33161 + }, + { + "epoch": 0.33162, + "grad_norm": 0.8542036414146423, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 33162 + }, + { + "epoch": 0.33163, + "grad_norm": 0.5799782872200012, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 33163 + }, + { + "epoch": 0.33164, + "grad_norm": 0.5900892019271851, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 33164 + }, + { + "epoch": 0.33165, + "grad_norm": 0.6466618776321411, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 33165 + }, + { + "epoch": 0.33166, + "grad_norm": 0.7353314757347107, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 33166 + }, + { + "epoch": 0.33167, + "grad_norm": 0.821902334690094, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 33167 + }, + { + "epoch": 0.33168, + "grad_norm": 0.7868537902832031, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 33168 + }, + { + "epoch": 0.33169, + "grad_norm": 0.6821154356002808, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 33169 + }, + { + "epoch": 0.3317, + "grad_norm": 0.7216659188270569, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 33170 + }, + { + "epoch": 0.33171, + "grad_norm": 1.0509569644927979, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 33171 + }, + { + "epoch": 0.33172, + "grad_norm": 1.1423144340515137, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 33172 + }, + { + "epoch": 0.33173, + "grad_norm": 0.8707491159439087, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 33173 + }, + { + "epoch": 0.33174, + "grad_norm": 0.8111802935600281, + "learning_rate": 0.003, + "loss": 3.9388, + "step": 33174 + }, + { + "epoch": 0.33175, + "grad_norm": 0.8313079476356506, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 33175 + }, + { + "epoch": 0.33176, + "grad_norm": 0.8207656145095825, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 33176 + }, + { + "epoch": 0.33177, + "grad_norm": 0.9345130324363708, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 33177 + }, + { + "epoch": 0.33178, + "grad_norm": 0.9482846260070801, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 33178 + }, + { + "epoch": 0.33179, + "grad_norm": 0.961216926574707, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 33179 + }, + { + "epoch": 0.3318, + "grad_norm": 0.8687255382537842, + "learning_rate": 0.003, + "loss": 3.978, + "step": 33180 + }, + { + "epoch": 0.33181, + "grad_norm": 1.1127636432647705, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 33181 + }, + { + "epoch": 0.33182, + "grad_norm": 1.2107609510421753, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 33182 + }, + { + "epoch": 0.33183, + "grad_norm": 0.9308600425720215, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 33183 + }, + { + "epoch": 0.33184, + "grad_norm": 0.9172362089157104, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 33184 + }, + { + "epoch": 0.33185, + "grad_norm": 0.8937748670578003, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 33185 + }, + { + "epoch": 0.33186, + "grad_norm": 0.8431787490844727, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 33186 + }, + { + "epoch": 0.33187, + "grad_norm": 0.7839288711547852, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 33187 + }, + { + "epoch": 0.33188, + "grad_norm": 0.7879115343093872, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 33188 + }, + { + "epoch": 0.33189, + "grad_norm": 0.835395336151123, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 33189 + }, + { + "epoch": 0.3319, + "grad_norm": 0.9506126642227173, + "learning_rate": 0.003, + "loss": 4.013, + "step": 33190 + }, + { + "epoch": 0.33191, + "grad_norm": 0.9711897373199463, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 33191 + }, + { + "epoch": 0.33192, + "grad_norm": 1.0607006549835205, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 33192 + }, + { + "epoch": 0.33193, + "grad_norm": 1.1732622385025024, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 33193 + }, + { + "epoch": 0.33194, + "grad_norm": 0.9819129109382629, + "learning_rate": 0.003, + "loss": 3.999, + "step": 33194 + }, + { + "epoch": 0.33195, + "grad_norm": 0.9173434376716614, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 33195 + }, + { + "epoch": 0.33196, + "grad_norm": 0.9633626341819763, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 33196 + }, + { + "epoch": 0.33197, + "grad_norm": 1.3109687566757202, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 33197 + }, + { + "epoch": 0.33198, + "grad_norm": 0.7049962878227234, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 33198 + }, + { + "epoch": 0.33199, + "grad_norm": 0.7728586792945862, + "learning_rate": 0.003, + "loss": 3.9396, + "step": 33199 + }, + { + "epoch": 0.332, + "grad_norm": 0.8741982579231262, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 33200 + }, + { + "epoch": 0.33201, + "grad_norm": 0.9037245512008667, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 33201 + }, + { + "epoch": 0.33202, + "grad_norm": 0.8849324584007263, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 33202 + }, + { + "epoch": 0.33203, + "grad_norm": 0.7829990386962891, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 33203 + }, + { + "epoch": 0.33204, + "grad_norm": 0.837530791759491, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 33204 + }, + { + "epoch": 0.33205, + "grad_norm": 0.7735166549682617, + "learning_rate": 0.003, + "loss": 3.991, + "step": 33205 + }, + { + "epoch": 0.33206, + "grad_norm": 0.8456722497940063, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 33206 + }, + { + "epoch": 0.33207, + "grad_norm": 1.0326569080352783, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 33207 + }, + { + "epoch": 0.33208, + "grad_norm": 1.1855545043945312, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 33208 + }, + { + "epoch": 0.33209, + "grad_norm": 0.9344742298126221, + "learning_rate": 0.003, + "loss": 4.024, + "step": 33209 + }, + { + "epoch": 0.3321, + "grad_norm": 1.015474557876587, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 33210 + }, + { + "epoch": 0.33211, + "grad_norm": 1.0299005508422852, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 33211 + }, + { + "epoch": 0.33212, + "grad_norm": 0.9238229393959045, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 33212 + }, + { + "epoch": 0.33213, + "grad_norm": 0.9079163670539856, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 33213 + }, + { + "epoch": 0.33214, + "grad_norm": 0.9106621146202087, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 33214 + }, + { + "epoch": 0.33215, + "grad_norm": 0.9238945841789246, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 33215 + }, + { + "epoch": 0.33216, + "grad_norm": 0.8243985176086426, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 33216 + }, + { + "epoch": 0.33217, + "grad_norm": 0.6620566248893738, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 33217 + }, + { + "epoch": 0.33218, + "grad_norm": 0.6362363696098328, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 33218 + }, + { + "epoch": 0.33219, + "grad_norm": 0.5583633184432983, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 33219 + }, + { + "epoch": 0.3322, + "grad_norm": 0.4978553056716919, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 33220 + }, + { + "epoch": 0.33221, + "grad_norm": 0.6151140928268433, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 33221 + }, + { + "epoch": 0.33222, + "grad_norm": 0.7551579475402832, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 33222 + }, + { + "epoch": 0.33223, + "grad_norm": 0.820266842842102, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 33223 + }, + { + "epoch": 0.33224, + "grad_norm": 0.9310306906700134, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 33224 + }, + { + "epoch": 0.33225, + "grad_norm": 0.9961439371109009, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 33225 + }, + { + "epoch": 0.33226, + "grad_norm": 0.8620225787162781, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 33226 + }, + { + "epoch": 0.33227, + "grad_norm": 0.8452762365341187, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 33227 + }, + { + "epoch": 0.33228, + "grad_norm": 0.9209995865821838, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 33228 + }, + { + "epoch": 0.33229, + "grad_norm": 0.9014765620231628, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 33229 + }, + { + "epoch": 0.3323, + "grad_norm": 0.9414999485015869, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 33230 + }, + { + "epoch": 0.33231, + "grad_norm": 1.0563405752182007, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 33231 + }, + { + "epoch": 0.33232, + "grad_norm": 0.982158899307251, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 33232 + }, + { + "epoch": 0.33233, + "grad_norm": 0.956381618976593, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 33233 + }, + { + "epoch": 0.33234, + "grad_norm": 0.9472842216491699, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 33234 + }, + { + "epoch": 0.33235, + "grad_norm": 0.9529433250427246, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 33235 + }, + { + "epoch": 0.33236, + "grad_norm": 1.192480206489563, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 33236 + }, + { + "epoch": 0.33237, + "grad_norm": 0.9860550761222839, + "learning_rate": 0.003, + "loss": 3.988, + "step": 33237 + }, + { + "epoch": 0.33238, + "grad_norm": 1.0901579856872559, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 33238 + }, + { + "epoch": 0.33239, + "grad_norm": 0.8362670540809631, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 33239 + }, + { + "epoch": 0.3324, + "grad_norm": 0.7208173871040344, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 33240 + }, + { + "epoch": 0.33241, + "grad_norm": 0.8647270798683167, + "learning_rate": 0.003, + "loss": 3.971, + "step": 33241 + }, + { + "epoch": 0.33242, + "grad_norm": 1.129908561706543, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 33242 + }, + { + "epoch": 0.33243, + "grad_norm": 0.8608614802360535, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 33243 + }, + { + "epoch": 0.33244, + "grad_norm": 0.8194960951805115, + "learning_rate": 0.003, + "loss": 3.984, + "step": 33244 + }, + { + "epoch": 0.33245, + "grad_norm": 0.8871719837188721, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 33245 + }, + { + "epoch": 0.33246, + "grad_norm": 1.3762966394424438, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 33246 + }, + { + "epoch": 0.33247, + "grad_norm": 0.7749135494232178, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 33247 + }, + { + "epoch": 0.33248, + "grad_norm": 0.5680655241012573, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 33248 + }, + { + "epoch": 0.33249, + "grad_norm": 0.6048284769058228, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 33249 + }, + { + "epoch": 0.3325, + "grad_norm": 0.6708140969276428, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 33250 + }, + { + "epoch": 0.33251, + "grad_norm": 0.8828984498977661, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 33251 + }, + { + "epoch": 0.33252, + "grad_norm": 1.0671212673187256, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 33252 + }, + { + "epoch": 0.33253, + "grad_norm": 1.006119966506958, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 33253 + }, + { + "epoch": 0.33254, + "grad_norm": 0.9353200197219849, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 33254 + }, + { + "epoch": 0.33255, + "grad_norm": 0.7925071120262146, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 33255 + }, + { + "epoch": 0.33256, + "grad_norm": 0.7984153628349304, + "learning_rate": 0.003, + "loss": 3.964, + "step": 33256 + }, + { + "epoch": 0.33257, + "grad_norm": 1.045447826385498, + "learning_rate": 0.003, + "loss": 4.005, + "step": 33257 + }, + { + "epoch": 0.33258, + "grad_norm": 1.330660104751587, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 33258 + }, + { + "epoch": 0.33259, + "grad_norm": 0.6166536211967468, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 33259 + }, + { + "epoch": 0.3326, + "grad_norm": 0.7293148636817932, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 33260 + }, + { + "epoch": 0.33261, + "grad_norm": 0.8390899896621704, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 33261 + }, + { + "epoch": 0.33262, + "grad_norm": 0.8586038947105408, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 33262 + }, + { + "epoch": 0.33263, + "grad_norm": 0.9969210028648376, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 33263 + }, + { + "epoch": 0.33264, + "grad_norm": 0.8358091711997986, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 33264 + }, + { + "epoch": 0.33265, + "grad_norm": 0.6519429683685303, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 33265 + }, + { + "epoch": 0.33266, + "grad_norm": 0.6759891510009766, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 33266 + }, + { + "epoch": 0.33267, + "grad_norm": 0.609173595905304, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 33267 + }, + { + "epoch": 0.33268, + "grad_norm": 0.6088685393333435, + "learning_rate": 0.003, + "loss": 3.966, + "step": 33268 + }, + { + "epoch": 0.33269, + "grad_norm": 0.566277801990509, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 33269 + }, + { + "epoch": 0.3327, + "grad_norm": 0.5522683262825012, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 33270 + }, + { + "epoch": 0.33271, + "grad_norm": 0.5741465091705322, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 33271 + }, + { + "epoch": 0.33272, + "grad_norm": 0.6542105674743652, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 33272 + }, + { + "epoch": 0.33273, + "grad_norm": 0.7738354206085205, + "learning_rate": 0.003, + "loss": 3.957, + "step": 33273 + }, + { + "epoch": 0.33274, + "grad_norm": 0.8492299914360046, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 33274 + }, + { + "epoch": 0.33275, + "grad_norm": 0.8554628491401672, + "learning_rate": 0.003, + "loss": 3.968, + "step": 33275 + }, + { + "epoch": 0.33276, + "grad_norm": 1.1254754066467285, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 33276 + }, + { + "epoch": 0.33277, + "grad_norm": 1.2026385068893433, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 33277 + }, + { + "epoch": 0.33278, + "grad_norm": 0.9534925818443298, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 33278 + }, + { + "epoch": 0.33279, + "grad_norm": 0.967512845993042, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 33279 + }, + { + "epoch": 0.3328, + "grad_norm": 0.9742660522460938, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 33280 + }, + { + "epoch": 0.33281, + "grad_norm": 1.0403634309768677, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 33281 + }, + { + "epoch": 0.33282, + "grad_norm": 1.3075804710388184, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 33282 + }, + { + "epoch": 0.33283, + "grad_norm": 0.755254864692688, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 33283 + }, + { + "epoch": 0.33284, + "grad_norm": 0.744067370891571, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 33284 + }, + { + "epoch": 0.33285, + "grad_norm": 0.8115424513816833, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 33285 + }, + { + "epoch": 0.33286, + "grad_norm": 0.9662624597549438, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 33286 + }, + { + "epoch": 0.33287, + "grad_norm": 1.1473363637924194, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 33287 + }, + { + "epoch": 0.33288, + "grad_norm": 1.0931472778320312, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 33288 + }, + { + "epoch": 0.33289, + "grad_norm": 1.0673178434371948, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 33289 + }, + { + "epoch": 0.3329, + "grad_norm": 0.8902097940444946, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 33290 + }, + { + "epoch": 0.33291, + "grad_norm": 0.8278650045394897, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 33291 + }, + { + "epoch": 0.33292, + "grad_norm": 0.8583379983901978, + "learning_rate": 0.003, + "loss": 4.0, + "step": 33292 + }, + { + "epoch": 0.33293, + "grad_norm": 0.9049744606018066, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 33293 + }, + { + "epoch": 0.33294, + "grad_norm": 1.0081522464752197, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 33294 + }, + { + "epoch": 0.33295, + "grad_norm": 1.3703635931015015, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 33295 + }, + { + "epoch": 0.33296, + "grad_norm": 0.7925004959106445, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 33296 + }, + { + "epoch": 0.33297, + "grad_norm": 0.8004211187362671, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 33297 + }, + { + "epoch": 0.33298, + "grad_norm": 0.9166144132614136, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 33298 + }, + { + "epoch": 0.33299, + "grad_norm": 0.9513294100761414, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 33299 + }, + { + "epoch": 0.333, + "grad_norm": 0.9833652377128601, + "learning_rate": 0.003, + "loss": 3.976, + "step": 33300 + }, + { + "epoch": 0.33301, + "grad_norm": 0.882859468460083, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 33301 + }, + { + "epoch": 0.33302, + "grad_norm": 0.828890323638916, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 33302 + }, + { + "epoch": 0.33303, + "grad_norm": 0.8585402965545654, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 33303 + }, + { + "epoch": 0.33304, + "grad_norm": 1.0606616735458374, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 33304 + }, + { + "epoch": 0.33305, + "grad_norm": 1.108852505683899, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 33305 + }, + { + "epoch": 0.33306, + "grad_norm": 0.8474005460739136, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 33306 + }, + { + "epoch": 0.33307, + "grad_norm": 0.7722617983818054, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 33307 + }, + { + "epoch": 0.33308, + "grad_norm": 0.7805770039558411, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 33308 + }, + { + "epoch": 0.33309, + "grad_norm": 0.7499332427978516, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 33309 + }, + { + "epoch": 0.3331, + "grad_norm": 1.0448930263519287, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 33310 + }, + { + "epoch": 0.33311, + "grad_norm": 1.3111016750335693, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 33311 + }, + { + "epoch": 0.33312, + "grad_norm": 0.7219438552856445, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 33312 + }, + { + "epoch": 0.33313, + "grad_norm": 0.6498905420303345, + "learning_rate": 0.003, + "loss": 3.973, + "step": 33313 + }, + { + "epoch": 0.33314, + "grad_norm": 0.7142239212989807, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 33314 + }, + { + "epoch": 0.33315, + "grad_norm": 0.7118070721626282, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 33315 + }, + { + "epoch": 0.33316, + "grad_norm": 0.851085901260376, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 33316 + }, + { + "epoch": 0.33317, + "grad_norm": 0.8706521391868591, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 33317 + }, + { + "epoch": 0.33318, + "grad_norm": 0.8828452229499817, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 33318 + }, + { + "epoch": 0.33319, + "grad_norm": 1.0975732803344727, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 33319 + }, + { + "epoch": 0.3332, + "grad_norm": 1.1051511764526367, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 33320 + }, + { + "epoch": 0.33321, + "grad_norm": 0.8499352335929871, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 33321 + }, + { + "epoch": 0.33322, + "grad_norm": 0.7988424897193909, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 33322 + }, + { + "epoch": 0.33323, + "grad_norm": 0.8205262422561646, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 33323 + }, + { + "epoch": 0.33324, + "grad_norm": 0.8043299317359924, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 33324 + }, + { + "epoch": 0.33325, + "grad_norm": 0.8142148852348328, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 33325 + }, + { + "epoch": 0.33326, + "grad_norm": 0.9592801928520203, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 33326 + }, + { + "epoch": 0.33327, + "grad_norm": 1.3590786457061768, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 33327 + }, + { + "epoch": 0.33328, + "grad_norm": 0.660821259021759, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 33328 + }, + { + "epoch": 0.33329, + "grad_norm": 0.8131953477859497, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 33329 + }, + { + "epoch": 0.3333, + "grad_norm": 1.377326488494873, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 33330 + }, + { + "epoch": 0.33331, + "grad_norm": 0.7755995988845825, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 33331 + }, + { + "epoch": 0.33332, + "grad_norm": 0.7091944217681885, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 33332 + }, + { + "epoch": 0.33333, + "grad_norm": 0.7054687142372131, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 33333 + }, + { + "epoch": 0.33334, + "grad_norm": 0.8333284258842468, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 33334 + }, + { + "epoch": 0.33335, + "grad_norm": 0.7221879959106445, + "learning_rate": 0.003, + "loss": 3.976, + "step": 33335 + }, + { + "epoch": 0.33336, + "grad_norm": 0.7480661869049072, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 33336 + }, + { + "epoch": 0.33337, + "grad_norm": 0.7437936663627625, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 33337 + }, + { + "epoch": 0.33338, + "grad_norm": 0.9498668313026428, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 33338 + }, + { + "epoch": 0.33339, + "grad_norm": 1.0870710611343384, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 33339 + }, + { + "epoch": 0.3334, + "grad_norm": 0.9275280237197876, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 33340 + }, + { + "epoch": 0.33341, + "grad_norm": 1.1108360290527344, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 33341 + }, + { + "epoch": 0.33342, + "grad_norm": 1.0622470378875732, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 33342 + }, + { + "epoch": 0.33343, + "grad_norm": 0.9478028416633606, + "learning_rate": 0.003, + "loss": 4.009, + "step": 33343 + }, + { + "epoch": 0.33344, + "grad_norm": 0.9437253475189209, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 33344 + }, + { + "epoch": 0.33345, + "grad_norm": 1.0977834463119507, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 33345 + }, + { + "epoch": 0.33346, + "grad_norm": 0.8212341070175171, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 33346 + }, + { + "epoch": 0.33347, + "grad_norm": 0.7443996667861938, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 33347 + }, + { + "epoch": 0.33348, + "grad_norm": 0.8790258765220642, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 33348 + }, + { + "epoch": 0.33349, + "grad_norm": 1.2166688442230225, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 33349 + }, + { + "epoch": 0.3335, + "grad_norm": 0.9317178726196289, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 33350 + }, + { + "epoch": 0.33351, + "grad_norm": 1.2156940698623657, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 33351 + }, + { + "epoch": 0.33352, + "grad_norm": 0.9575168490409851, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 33352 + }, + { + "epoch": 0.33353, + "grad_norm": 0.845831573009491, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 33353 + }, + { + "epoch": 0.33354, + "grad_norm": 0.7999079823493958, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 33354 + }, + { + "epoch": 0.33355, + "grad_norm": 0.8455422520637512, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 33355 + }, + { + "epoch": 0.33356, + "grad_norm": 0.8438761234283447, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 33356 + }, + { + "epoch": 0.33357, + "grad_norm": 0.7142013907432556, + "learning_rate": 0.003, + "loss": 3.946, + "step": 33357 + }, + { + "epoch": 0.33358, + "grad_norm": 0.7987946271896362, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 33358 + }, + { + "epoch": 0.33359, + "grad_norm": 0.816582977771759, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 33359 + }, + { + "epoch": 0.3336, + "grad_norm": 0.8098604679107666, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 33360 + }, + { + "epoch": 0.33361, + "grad_norm": 0.770245373249054, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 33361 + }, + { + "epoch": 0.33362, + "grad_norm": 0.8171371817588806, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 33362 + }, + { + "epoch": 0.33363, + "grad_norm": 0.9816508889198303, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 33363 + }, + { + "epoch": 0.33364, + "grad_norm": 1.1301405429840088, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 33364 + }, + { + "epoch": 0.33365, + "grad_norm": 1.2181658744812012, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 33365 + }, + { + "epoch": 0.33366, + "grad_norm": 0.833936333656311, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 33366 + }, + { + "epoch": 0.33367, + "grad_norm": 0.7985481023788452, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 33367 + }, + { + "epoch": 0.33368, + "grad_norm": 0.8209153413772583, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 33368 + }, + { + "epoch": 0.33369, + "grad_norm": 0.7720400094985962, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 33369 + }, + { + "epoch": 0.3337, + "grad_norm": 0.7023248672485352, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 33370 + }, + { + "epoch": 0.33371, + "grad_norm": 0.7367078065872192, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 33371 + }, + { + "epoch": 0.33372, + "grad_norm": 0.8345584273338318, + "learning_rate": 0.003, + "loss": 3.973, + "step": 33372 + }, + { + "epoch": 0.33373, + "grad_norm": 0.8786050081253052, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 33373 + }, + { + "epoch": 0.33374, + "grad_norm": 1.1854000091552734, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 33374 + }, + { + "epoch": 0.33375, + "grad_norm": 1.0038615465164185, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 33375 + }, + { + "epoch": 0.33376, + "grad_norm": 1.0284919738769531, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 33376 + }, + { + "epoch": 0.33377, + "grad_norm": 0.9988184571266174, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 33377 + }, + { + "epoch": 0.33378, + "grad_norm": 1.0532200336456299, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 33378 + }, + { + "epoch": 0.33379, + "grad_norm": 0.8543786406517029, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 33379 + }, + { + "epoch": 0.3338, + "grad_norm": 0.7211754322052002, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 33380 + }, + { + "epoch": 0.33381, + "grad_norm": 0.6463448405265808, + "learning_rate": 0.003, + "loss": 3.982, + "step": 33381 + }, + { + "epoch": 0.33382, + "grad_norm": 0.6025155186653137, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 33382 + }, + { + "epoch": 0.33383, + "grad_norm": 0.6217491030693054, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 33383 + }, + { + "epoch": 0.33384, + "grad_norm": 0.7618924379348755, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 33384 + }, + { + "epoch": 0.33385, + "grad_norm": 1.0080649852752686, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 33385 + }, + { + "epoch": 0.33386, + "grad_norm": 1.2912092208862305, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 33386 + }, + { + "epoch": 0.33387, + "grad_norm": 0.7214446663856506, + "learning_rate": 0.003, + "loss": 3.9454, + "step": 33387 + }, + { + "epoch": 0.33388, + "grad_norm": 0.6232267022132874, + "learning_rate": 0.003, + "loss": 3.9469, + "step": 33388 + }, + { + "epoch": 0.33389, + "grad_norm": 0.7105653882026672, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 33389 + }, + { + "epoch": 0.3339, + "grad_norm": 0.8402934670448303, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 33390 + }, + { + "epoch": 0.33391, + "grad_norm": 0.9138760566711426, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 33391 + }, + { + "epoch": 0.33392, + "grad_norm": 1.0048214197158813, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 33392 + }, + { + "epoch": 0.33393, + "grad_norm": 1.0299277305603027, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 33393 + }, + { + "epoch": 0.33394, + "grad_norm": 1.1934070587158203, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 33394 + }, + { + "epoch": 0.33395, + "grad_norm": 0.8493682742118835, + "learning_rate": 0.003, + "loss": 3.997, + "step": 33395 + }, + { + "epoch": 0.33396, + "grad_norm": 0.8620051741600037, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 33396 + }, + { + "epoch": 0.33397, + "grad_norm": 1.0221940279006958, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 33397 + }, + { + "epoch": 0.33398, + "grad_norm": 1.0785620212554932, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 33398 + }, + { + "epoch": 0.33399, + "grad_norm": 0.936069130897522, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 33399 + }, + { + "epoch": 0.334, + "grad_norm": 1.0372314453125, + "learning_rate": 0.003, + "loss": 4.014, + "step": 33400 + }, + { + "epoch": 0.33401, + "grad_norm": 1.0963619947433472, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 33401 + }, + { + "epoch": 0.33402, + "grad_norm": 0.9582024812698364, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 33402 + }, + { + "epoch": 0.33403, + "grad_norm": 1.2753996849060059, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 33403 + }, + { + "epoch": 0.33404, + "grad_norm": 0.8568985462188721, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 33404 + }, + { + "epoch": 0.33405, + "grad_norm": 0.7931933403015137, + "learning_rate": 0.003, + "loss": 3.991, + "step": 33405 + }, + { + "epoch": 0.33406, + "grad_norm": 1.0176341533660889, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 33406 + }, + { + "epoch": 0.33407, + "grad_norm": 1.441214919090271, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 33407 + }, + { + "epoch": 0.33408, + "grad_norm": 0.953467071056366, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 33408 + }, + { + "epoch": 0.33409, + "grad_norm": 1.156485676765442, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 33409 + }, + { + "epoch": 0.3341, + "grad_norm": 1.2159500122070312, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 33410 + }, + { + "epoch": 0.33411, + "grad_norm": 0.8662699460983276, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 33411 + }, + { + "epoch": 0.33412, + "grad_norm": 0.6947513818740845, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 33412 + }, + { + "epoch": 0.33413, + "grad_norm": 0.8120648860931396, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 33413 + }, + { + "epoch": 0.33414, + "grad_norm": 1.008036494255066, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 33414 + }, + { + "epoch": 0.33415, + "grad_norm": 1.0716571807861328, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 33415 + }, + { + "epoch": 0.33416, + "grad_norm": 0.7767823338508606, + "learning_rate": 0.003, + "loss": 4.002, + "step": 33416 + }, + { + "epoch": 0.33417, + "grad_norm": 0.6856827735900879, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 33417 + }, + { + "epoch": 0.33418, + "grad_norm": 0.6960150599479675, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 33418 + }, + { + "epoch": 0.33419, + "grad_norm": 0.606023371219635, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 33419 + }, + { + "epoch": 0.3342, + "grad_norm": 0.6802657246589661, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 33420 + }, + { + "epoch": 0.33421, + "grad_norm": 0.6443341970443726, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 33421 + }, + { + "epoch": 0.33422, + "grad_norm": 0.5588348507881165, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 33422 + }, + { + "epoch": 0.33423, + "grad_norm": 0.5276538133621216, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 33423 + }, + { + "epoch": 0.33424, + "grad_norm": 0.6205162405967712, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 33424 + }, + { + "epoch": 0.33425, + "grad_norm": 0.747577965259552, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 33425 + }, + { + "epoch": 0.33426, + "grad_norm": 0.8004941940307617, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 33426 + }, + { + "epoch": 0.33427, + "grad_norm": 0.8493192791938782, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 33427 + }, + { + "epoch": 0.33428, + "grad_norm": 0.9747901558876038, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 33428 + }, + { + "epoch": 0.33429, + "grad_norm": 1.2574280500411987, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 33429 + }, + { + "epoch": 0.3343, + "grad_norm": 0.9456437230110168, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 33430 + }, + { + "epoch": 0.33431, + "grad_norm": 0.850011944770813, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 33431 + }, + { + "epoch": 0.33432, + "grad_norm": 0.9513918161392212, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 33432 + }, + { + "epoch": 0.33433, + "grad_norm": 1.1772000789642334, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 33433 + }, + { + "epoch": 0.33434, + "grad_norm": 0.772232174873352, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 33434 + }, + { + "epoch": 0.33435, + "grad_norm": 0.6969711780548096, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 33435 + }, + { + "epoch": 0.33436, + "grad_norm": 0.7776949405670166, + "learning_rate": 0.003, + "loss": 3.993, + "step": 33436 + }, + { + "epoch": 0.33437, + "grad_norm": 0.7758357524871826, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 33437 + }, + { + "epoch": 0.33438, + "grad_norm": 0.7675877213478088, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 33438 + }, + { + "epoch": 0.33439, + "grad_norm": 0.715604841709137, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 33439 + }, + { + "epoch": 0.3344, + "grad_norm": 0.8479792475700378, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 33440 + }, + { + "epoch": 0.33441, + "grad_norm": 0.9258811473846436, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 33441 + }, + { + "epoch": 0.33442, + "grad_norm": 1.0017751455307007, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 33442 + }, + { + "epoch": 0.33443, + "grad_norm": 0.998823344707489, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 33443 + }, + { + "epoch": 0.33444, + "grad_norm": 0.9316739439964294, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 33444 + }, + { + "epoch": 0.33445, + "grad_norm": 0.8726114630699158, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 33445 + }, + { + "epoch": 0.33446, + "grad_norm": 0.9129933714866638, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 33446 + }, + { + "epoch": 0.33447, + "grad_norm": 1.0942662954330444, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 33447 + }, + { + "epoch": 0.33448, + "grad_norm": 1.035989761352539, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 33448 + }, + { + "epoch": 0.33449, + "grad_norm": 0.9457242488861084, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 33449 + }, + { + "epoch": 0.3345, + "grad_norm": 0.8946390151977539, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 33450 + }, + { + "epoch": 0.33451, + "grad_norm": 0.9579622149467468, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 33451 + }, + { + "epoch": 0.33452, + "grad_norm": 1.2328007221221924, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 33452 + }, + { + "epoch": 0.33453, + "grad_norm": 0.8189416527748108, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 33453 + }, + { + "epoch": 0.33454, + "grad_norm": 0.8089562654495239, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 33454 + }, + { + "epoch": 0.33455, + "grad_norm": 0.8406727313995361, + "learning_rate": 0.003, + "loss": 3.993, + "step": 33455 + }, + { + "epoch": 0.33456, + "grad_norm": 1.0432311296463013, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 33456 + }, + { + "epoch": 0.33457, + "grad_norm": 0.9850112199783325, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 33457 + }, + { + "epoch": 0.33458, + "grad_norm": 1.0730507373809814, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 33458 + }, + { + "epoch": 0.33459, + "grad_norm": 0.9641826748847961, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 33459 + }, + { + "epoch": 0.3346, + "grad_norm": 0.9190773367881775, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 33460 + }, + { + "epoch": 0.33461, + "grad_norm": 0.9175162315368652, + "learning_rate": 0.003, + "loss": 3.996, + "step": 33461 + }, + { + "epoch": 0.33462, + "grad_norm": 0.9031410217285156, + "learning_rate": 0.003, + "loss": 4.021, + "step": 33462 + }, + { + "epoch": 0.33463, + "grad_norm": 0.8918395042419434, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 33463 + }, + { + "epoch": 0.33464, + "grad_norm": 0.9304821491241455, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 33464 + }, + { + "epoch": 0.33465, + "grad_norm": 1.0114444494247437, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 33465 + }, + { + "epoch": 0.33466, + "grad_norm": 1.1615746021270752, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 33466 + }, + { + "epoch": 0.33467, + "grad_norm": 1.0907446146011353, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 33467 + }, + { + "epoch": 0.33468, + "grad_norm": 1.1303415298461914, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 33468 + }, + { + "epoch": 0.33469, + "grad_norm": 1.1380993127822876, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 33469 + }, + { + "epoch": 0.3347, + "grad_norm": 0.7375786900520325, + "learning_rate": 0.003, + "loss": 4.041, + "step": 33470 + }, + { + "epoch": 0.33471, + "grad_norm": 0.5820444226264954, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 33471 + }, + { + "epoch": 0.33472, + "grad_norm": 0.610375702381134, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 33472 + }, + { + "epoch": 0.33473, + "grad_norm": 0.5466592907905579, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 33473 + }, + { + "epoch": 0.33474, + "grad_norm": 0.6636547446250916, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 33474 + }, + { + "epoch": 0.33475, + "grad_norm": 0.767010509967804, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 33475 + }, + { + "epoch": 0.33476, + "grad_norm": 0.8545445203781128, + "learning_rate": 0.003, + "loss": 3.991, + "step": 33476 + }, + { + "epoch": 0.33477, + "grad_norm": 0.9805036783218384, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 33477 + }, + { + "epoch": 0.33478, + "grad_norm": 1.1079145669937134, + "learning_rate": 0.003, + "loss": 3.995, + "step": 33478 + }, + { + "epoch": 0.33479, + "grad_norm": 0.8234482407569885, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 33479 + }, + { + "epoch": 0.3348, + "grad_norm": 0.6287835240364075, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 33480 + }, + { + "epoch": 0.33481, + "grad_norm": 0.648661732673645, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 33481 + }, + { + "epoch": 0.33482, + "grad_norm": 0.840247392654419, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 33482 + }, + { + "epoch": 0.33483, + "grad_norm": 0.9099528193473816, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 33483 + }, + { + "epoch": 0.33484, + "grad_norm": 0.9326075911521912, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 33484 + }, + { + "epoch": 0.33485, + "grad_norm": 1.0401535034179688, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 33485 + }, + { + "epoch": 0.33486, + "grad_norm": 1.0833977460861206, + "learning_rate": 0.003, + "loss": 3.984, + "step": 33486 + }, + { + "epoch": 0.33487, + "grad_norm": 0.8774539232254028, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 33487 + }, + { + "epoch": 0.33488, + "grad_norm": 0.8038880825042725, + "learning_rate": 0.003, + "loss": 3.976, + "step": 33488 + }, + { + "epoch": 0.33489, + "grad_norm": 0.838934600353241, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 33489 + }, + { + "epoch": 0.3349, + "grad_norm": 0.8459370136260986, + "learning_rate": 0.003, + "loss": 3.965, + "step": 33490 + }, + { + "epoch": 0.33491, + "grad_norm": 0.8882632255554199, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 33491 + }, + { + "epoch": 0.33492, + "grad_norm": 0.9251918196678162, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 33492 + }, + { + "epoch": 0.33493, + "grad_norm": 0.999549388885498, + "learning_rate": 0.003, + "loss": 3.967, + "step": 33493 + }, + { + "epoch": 0.33494, + "grad_norm": 1.1649808883666992, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 33494 + }, + { + "epoch": 0.33495, + "grad_norm": 0.8680398464202881, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 33495 + }, + { + "epoch": 0.33496, + "grad_norm": 0.851954460144043, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 33496 + }, + { + "epoch": 0.33497, + "grad_norm": 0.8512863516807556, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 33497 + }, + { + "epoch": 0.33498, + "grad_norm": 0.8211043477058411, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 33498 + }, + { + "epoch": 0.33499, + "grad_norm": 0.9344930648803711, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 33499 + }, + { + "epoch": 0.335, + "grad_norm": 1.251651644706726, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 33500 + }, + { + "epoch": 0.33501, + "grad_norm": 1.1440412998199463, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 33501 + }, + { + "epoch": 0.33502, + "grad_norm": 0.8811331987380981, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 33502 + }, + { + "epoch": 0.33503, + "grad_norm": 0.7443961501121521, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 33503 + }, + { + "epoch": 0.33504, + "grad_norm": 0.9332696199417114, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 33504 + }, + { + "epoch": 0.33505, + "grad_norm": 0.9936373829841614, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 33505 + }, + { + "epoch": 0.33506, + "grad_norm": 1.063723087310791, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 33506 + }, + { + "epoch": 0.33507, + "grad_norm": 0.9050944447517395, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 33507 + }, + { + "epoch": 0.33508, + "grad_norm": 0.8109772205352783, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 33508 + }, + { + "epoch": 0.33509, + "grad_norm": 0.8112784624099731, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 33509 + }, + { + "epoch": 0.3351, + "grad_norm": 0.8105851411819458, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 33510 + }, + { + "epoch": 0.33511, + "grad_norm": 0.7671585083007812, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 33511 + }, + { + "epoch": 0.33512, + "grad_norm": 0.8156996965408325, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 33512 + }, + { + "epoch": 0.33513, + "grad_norm": 0.9093772172927856, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 33513 + }, + { + "epoch": 0.33514, + "grad_norm": 1.0013943910598755, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 33514 + }, + { + "epoch": 0.33515, + "grad_norm": 1.1724348068237305, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 33515 + }, + { + "epoch": 0.33516, + "grad_norm": 0.8662506937980652, + "learning_rate": 0.003, + "loss": 3.962, + "step": 33516 + }, + { + "epoch": 0.33517, + "grad_norm": 0.6974263787269592, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 33517 + }, + { + "epoch": 0.33518, + "grad_norm": 0.8114238381385803, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 33518 + }, + { + "epoch": 0.33519, + "grad_norm": 0.8408022522926331, + "learning_rate": 0.003, + "loss": 3.954, + "step": 33519 + }, + { + "epoch": 0.3352, + "grad_norm": 0.9947931170463562, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 33520 + }, + { + "epoch": 0.33521, + "grad_norm": 1.1567775011062622, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 33521 + }, + { + "epoch": 0.33522, + "grad_norm": 1.0865352153778076, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 33522 + }, + { + "epoch": 0.33523, + "grad_norm": 1.0278961658477783, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 33523 + }, + { + "epoch": 0.33524, + "grad_norm": 0.7331315279006958, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 33524 + }, + { + "epoch": 0.33525, + "grad_norm": 0.8063281178474426, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 33525 + }, + { + "epoch": 0.33526, + "grad_norm": 0.7436653971672058, + "learning_rate": 0.003, + "loss": 3.991, + "step": 33526 + }, + { + "epoch": 0.33527, + "grad_norm": 0.668890118598938, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 33527 + }, + { + "epoch": 0.33528, + "grad_norm": 0.7427557110786438, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 33528 + }, + { + "epoch": 0.33529, + "grad_norm": 0.7644615173339844, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 33529 + }, + { + "epoch": 0.3353, + "grad_norm": 0.8325464725494385, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 33530 + }, + { + "epoch": 0.33531, + "grad_norm": 0.980459451675415, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 33531 + }, + { + "epoch": 0.33532, + "grad_norm": 1.164663314819336, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 33532 + }, + { + "epoch": 0.33533, + "grad_norm": 0.8390658497810364, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 33533 + }, + { + "epoch": 0.33534, + "grad_norm": 0.9296440482139587, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 33534 + }, + { + "epoch": 0.33535, + "grad_norm": 1.1352277994155884, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 33535 + }, + { + "epoch": 0.33536, + "grad_norm": 0.9863858819007874, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 33536 + }, + { + "epoch": 0.33537, + "grad_norm": 0.9238900542259216, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 33537 + }, + { + "epoch": 0.33538, + "grad_norm": 0.9819132685661316, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 33538 + }, + { + "epoch": 0.33539, + "grad_norm": 0.9724236130714417, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 33539 + }, + { + "epoch": 0.3354, + "grad_norm": 1.021095633506775, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 33540 + }, + { + "epoch": 0.33541, + "grad_norm": 1.0552995204925537, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 33541 + }, + { + "epoch": 0.33542, + "grad_norm": 1.0375081300735474, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 33542 + }, + { + "epoch": 0.33543, + "grad_norm": 0.907223105430603, + "learning_rate": 0.003, + "loss": 3.975, + "step": 33543 + }, + { + "epoch": 0.33544, + "grad_norm": 0.8704020977020264, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 33544 + }, + { + "epoch": 0.33545, + "grad_norm": 0.9500170350074768, + "learning_rate": 0.003, + "loss": 4.003, + "step": 33545 + }, + { + "epoch": 0.33546, + "grad_norm": 1.2564549446105957, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 33546 + }, + { + "epoch": 0.33547, + "grad_norm": 0.9980997443199158, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 33547 + }, + { + "epoch": 0.33548, + "grad_norm": 1.2219161987304688, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 33548 + }, + { + "epoch": 0.33549, + "grad_norm": 0.8421098589897156, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 33549 + }, + { + "epoch": 0.3355, + "grad_norm": 0.5826959013938904, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 33550 + }, + { + "epoch": 0.33551, + "grad_norm": 0.6010957360267639, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 33551 + }, + { + "epoch": 0.33552, + "grad_norm": 0.6287940740585327, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 33552 + }, + { + "epoch": 0.33553, + "grad_norm": 0.7221230864524841, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 33553 + }, + { + "epoch": 0.33554, + "grad_norm": 0.9462738633155823, + "learning_rate": 0.003, + "loss": 4.003, + "step": 33554 + }, + { + "epoch": 0.33555, + "grad_norm": 0.8161649107933044, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 33555 + }, + { + "epoch": 0.33556, + "grad_norm": 0.6805449724197388, + "learning_rate": 0.003, + "loss": 3.981, + "step": 33556 + }, + { + "epoch": 0.33557, + "grad_norm": 0.7076331973075867, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 33557 + }, + { + "epoch": 0.33558, + "grad_norm": 0.8369594216346741, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 33558 + }, + { + "epoch": 0.33559, + "grad_norm": 0.9339938163757324, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 33559 + }, + { + "epoch": 0.3356, + "grad_norm": 0.8130449652671814, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 33560 + }, + { + "epoch": 0.33561, + "grad_norm": 0.8924601674079895, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 33561 + }, + { + "epoch": 0.33562, + "grad_norm": 0.9184163212776184, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 33562 + }, + { + "epoch": 0.33563, + "grad_norm": 0.8405300378799438, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 33563 + }, + { + "epoch": 0.33564, + "grad_norm": 1.0590444803237915, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 33564 + }, + { + "epoch": 0.33565, + "grad_norm": 1.0376079082489014, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 33565 + }, + { + "epoch": 0.33566, + "grad_norm": 0.9512711763381958, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 33566 + }, + { + "epoch": 0.33567, + "grad_norm": 1.0857316255569458, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 33567 + }, + { + "epoch": 0.33568, + "grad_norm": 0.9859429001808167, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 33568 + }, + { + "epoch": 0.33569, + "grad_norm": 1.0960458517074585, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 33569 + }, + { + "epoch": 0.3357, + "grad_norm": 1.0337843894958496, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 33570 + }, + { + "epoch": 0.33571, + "grad_norm": 1.0510871410369873, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 33571 + }, + { + "epoch": 0.33572, + "grad_norm": 0.9007986187934875, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 33572 + }, + { + "epoch": 0.33573, + "grad_norm": 0.8171018362045288, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 33573 + }, + { + "epoch": 0.33574, + "grad_norm": 0.917945384979248, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 33574 + }, + { + "epoch": 0.33575, + "grad_norm": 1.1683341264724731, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 33575 + }, + { + "epoch": 0.33576, + "grad_norm": 1.1607773303985596, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 33576 + }, + { + "epoch": 0.33577, + "grad_norm": 0.7445974349975586, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 33577 + }, + { + "epoch": 0.33578, + "grad_norm": 0.6819556951522827, + "learning_rate": 0.003, + "loss": 4.035, + "step": 33578 + }, + { + "epoch": 0.33579, + "grad_norm": 0.6969217658042908, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 33579 + }, + { + "epoch": 0.3358, + "grad_norm": 0.6800373792648315, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 33580 + }, + { + "epoch": 0.33581, + "grad_norm": 0.679969310760498, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 33581 + }, + { + "epoch": 0.33582, + "grad_norm": 0.7166128158569336, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 33582 + }, + { + "epoch": 0.33583, + "grad_norm": 0.6580185890197754, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 33583 + }, + { + "epoch": 0.33584, + "grad_norm": 0.6035010814666748, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 33584 + }, + { + "epoch": 0.33585, + "grad_norm": 0.5487178564071655, + "learning_rate": 0.003, + "loss": 3.9365, + "step": 33585 + }, + { + "epoch": 0.33586, + "grad_norm": 0.679938018321991, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 33586 + }, + { + "epoch": 0.33587, + "grad_norm": 0.7655243873596191, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 33587 + }, + { + "epoch": 0.33588, + "grad_norm": 0.8114100694656372, + "learning_rate": 0.003, + "loss": 3.964, + "step": 33588 + }, + { + "epoch": 0.33589, + "grad_norm": 1.0778952836990356, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 33589 + }, + { + "epoch": 0.3359, + "grad_norm": 1.3079626560211182, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 33590 + }, + { + "epoch": 0.33591, + "grad_norm": 0.8965069055557251, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 33591 + }, + { + "epoch": 0.33592, + "grad_norm": 0.8267623782157898, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 33592 + }, + { + "epoch": 0.33593, + "grad_norm": 0.8057902455329895, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 33593 + }, + { + "epoch": 0.33594, + "grad_norm": 0.815340518951416, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 33594 + }, + { + "epoch": 0.33595, + "grad_norm": 1.0451809167861938, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 33595 + }, + { + "epoch": 0.33596, + "grad_norm": 1.1496281623840332, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 33596 + }, + { + "epoch": 0.33597, + "grad_norm": 0.8318554759025574, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 33597 + }, + { + "epoch": 0.33598, + "grad_norm": 0.6453879475593567, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 33598 + }, + { + "epoch": 0.33599, + "grad_norm": 0.8441749215126038, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 33599 + }, + { + "epoch": 0.336, + "grad_norm": 1.0035344362258911, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 33600 + }, + { + "epoch": 0.33601, + "grad_norm": 1.0106003284454346, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 33601 + }, + { + "epoch": 0.33602, + "grad_norm": 0.9595679640769958, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 33602 + }, + { + "epoch": 0.33603, + "grad_norm": 0.8077322244644165, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 33603 + }, + { + "epoch": 0.33604, + "grad_norm": 0.8053494095802307, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 33604 + }, + { + "epoch": 0.33605, + "grad_norm": 0.9412218332290649, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 33605 + }, + { + "epoch": 0.33606, + "grad_norm": 1.2090827226638794, + "learning_rate": 0.003, + "loss": 4.017, + "step": 33606 + }, + { + "epoch": 0.33607, + "grad_norm": 0.9975122213363647, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 33607 + }, + { + "epoch": 0.33608, + "grad_norm": 1.0623782873153687, + "learning_rate": 0.003, + "loss": 3.982, + "step": 33608 + }, + { + "epoch": 0.33609, + "grad_norm": 0.9522918462753296, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 33609 + }, + { + "epoch": 0.3361, + "grad_norm": 0.8914372324943542, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 33610 + }, + { + "epoch": 0.33611, + "grad_norm": 0.8775237798690796, + "learning_rate": 0.003, + "loss": 3.977, + "step": 33611 + }, + { + "epoch": 0.33612, + "grad_norm": 0.8337449431419373, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 33612 + }, + { + "epoch": 0.33613, + "grad_norm": 0.9217292070388794, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 33613 + }, + { + "epoch": 0.33614, + "grad_norm": 1.057450294494629, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 33614 + }, + { + "epoch": 0.33615, + "grad_norm": 0.9425153136253357, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 33615 + }, + { + "epoch": 0.33616, + "grad_norm": 0.9802325963973999, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 33616 + }, + { + "epoch": 0.33617, + "grad_norm": 0.9550592303276062, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 33617 + }, + { + "epoch": 0.33618, + "grad_norm": 0.9312248229980469, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 33618 + }, + { + "epoch": 0.33619, + "grad_norm": 0.9663880467414856, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 33619 + }, + { + "epoch": 0.3362, + "grad_norm": 1.0545326471328735, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 33620 + }, + { + "epoch": 0.33621, + "grad_norm": 1.050053596496582, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 33621 + }, + { + "epoch": 0.33622, + "grad_norm": 0.9501489996910095, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 33622 + }, + { + "epoch": 0.33623, + "grad_norm": 0.8686526417732239, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 33623 + }, + { + "epoch": 0.33624, + "grad_norm": 0.8675819635391235, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 33624 + }, + { + "epoch": 0.33625, + "grad_norm": 0.9976730942726135, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 33625 + }, + { + "epoch": 0.33626, + "grad_norm": 1.0665987730026245, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 33626 + }, + { + "epoch": 0.33627, + "grad_norm": 0.9387950897216797, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 33627 + }, + { + "epoch": 0.33628, + "grad_norm": 0.8999784588813782, + "learning_rate": 0.003, + "loss": 4.014, + "step": 33628 + }, + { + "epoch": 0.33629, + "grad_norm": 0.7707502245903015, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 33629 + }, + { + "epoch": 0.3363, + "grad_norm": 0.6791815757751465, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 33630 + }, + { + "epoch": 0.33631, + "grad_norm": 0.6181407570838928, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 33631 + }, + { + "epoch": 0.33632, + "grad_norm": 0.6419703364372253, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 33632 + }, + { + "epoch": 0.33633, + "grad_norm": 0.7402269840240479, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 33633 + }, + { + "epoch": 0.33634, + "grad_norm": 0.7546635270118713, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 33634 + }, + { + "epoch": 0.33635, + "grad_norm": 0.709947943687439, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 33635 + }, + { + "epoch": 0.33636, + "grad_norm": 0.8805006146430969, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 33636 + }, + { + "epoch": 0.33637, + "grad_norm": 1.0526204109191895, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 33637 + }, + { + "epoch": 0.33638, + "grad_norm": 0.9175156950950623, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 33638 + }, + { + "epoch": 0.33639, + "grad_norm": 0.7686423063278198, + "learning_rate": 0.003, + "loss": 3.972, + "step": 33639 + }, + { + "epoch": 0.3364, + "grad_norm": 0.7588375806808472, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 33640 + }, + { + "epoch": 0.33641, + "grad_norm": 0.8399437665939331, + "learning_rate": 0.003, + "loss": 3.968, + "step": 33641 + }, + { + "epoch": 0.33642, + "grad_norm": 0.9717065691947937, + "learning_rate": 0.003, + "loss": 3.981, + "step": 33642 + }, + { + "epoch": 0.33643, + "grad_norm": 1.2562906742095947, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 33643 + }, + { + "epoch": 0.33644, + "grad_norm": 0.889876663684845, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 33644 + }, + { + "epoch": 0.33645, + "grad_norm": 0.8152696490287781, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 33645 + }, + { + "epoch": 0.33646, + "grad_norm": 0.8308585286140442, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 33646 + }, + { + "epoch": 0.33647, + "grad_norm": 1.0005338191986084, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 33647 + }, + { + "epoch": 0.33648, + "grad_norm": 1.106770396232605, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 33648 + }, + { + "epoch": 0.33649, + "grad_norm": 0.8131284713745117, + "learning_rate": 0.003, + "loss": 3.976, + "step": 33649 + }, + { + "epoch": 0.3365, + "grad_norm": 0.8280332684516907, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 33650 + }, + { + "epoch": 0.33651, + "grad_norm": 0.8842000961303711, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 33651 + }, + { + "epoch": 0.33652, + "grad_norm": 0.8843600749969482, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 33652 + }, + { + "epoch": 0.33653, + "grad_norm": 0.9894416928291321, + "learning_rate": 0.003, + "loss": 3.982, + "step": 33653 + }, + { + "epoch": 0.33654, + "grad_norm": 1.184354305267334, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 33654 + }, + { + "epoch": 0.33655, + "grad_norm": 0.8221862316131592, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 33655 + }, + { + "epoch": 0.33656, + "grad_norm": 0.6749824285507202, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 33656 + }, + { + "epoch": 0.33657, + "grad_norm": 0.6451243162155151, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 33657 + }, + { + "epoch": 0.33658, + "grad_norm": 0.7339329719543457, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 33658 + }, + { + "epoch": 0.33659, + "grad_norm": 0.7401431798934937, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 33659 + }, + { + "epoch": 0.3366, + "grad_norm": 0.6929410696029663, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 33660 + }, + { + "epoch": 0.33661, + "grad_norm": 0.8561339974403381, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 33661 + }, + { + "epoch": 0.33662, + "grad_norm": 1.0753105878829956, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 33662 + }, + { + "epoch": 0.33663, + "grad_norm": 1.17317795753479, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 33663 + }, + { + "epoch": 0.33664, + "grad_norm": 0.7217439413070679, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 33664 + }, + { + "epoch": 0.33665, + "grad_norm": 0.7441021203994751, + "learning_rate": 0.003, + "loss": 3.993, + "step": 33665 + }, + { + "epoch": 0.33666, + "grad_norm": 0.758992075920105, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 33666 + }, + { + "epoch": 0.33667, + "grad_norm": 0.7200351357460022, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 33667 + }, + { + "epoch": 0.33668, + "grad_norm": 0.7788838744163513, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 33668 + }, + { + "epoch": 0.33669, + "grad_norm": 0.7966474294662476, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 33669 + }, + { + "epoch": 0.3367, + "grad_norm": 0.7380022406578064, + "learning_rate": 0.003, + "loss": 3.974, + "step": 33670 + }, + { + "epoch": 0.33671, + "grad_norm": 0.7487584352493286, + "learning_rate": 0.003, + "loss": 3.9436, + "step": 33671 + }, + { + "epoch": 0.33672, + "grad_norm": 0.9233666062355042, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 33672 + }, + { + "epoch": 0.33673, + "grad_norm": 1.5030585527420044, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 33673 + }, + { + "epoch": 0.33674, + "grad_norm": 0.6827166080474854, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 33674 + }, + { + "epoch": 0.33675, + "grad_norm": 0.6961929798126221, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 33675 + }, + { + "epoch": 0.33676, + "grad_norm": 1.0050983428955078, + "learning_rate": 0.003, + "loss": 4.004, + "step": 33676 + }, + { + "epoch": 0.33677, + "grad_norm": 1.10648512840271, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 33677 + }, + { + "epoch": 0.33678, + "grad_norm": 0.8543322682380676, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 33678 + }, + { + "epoch": 0.33679, + "grad_norm": 0.7996217012405396, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 33679 + }, + { + "epoch": 0.3368, + "grad_norm": 0.7183598875999451, + "learning_rate": 0.003, + "loss": 3.975, + "step": 33680 + }, + { + "epoch": 0.33681, + "grad_norm": 0.6327478289604187, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 33681 + }, + { + "epoch": 0.33682, + "grad_norm": 0.7723290324211121, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 33682 + }, + { + "epoch": 0.33683, + "grad_norm": 0.8407782316207886, + "learning_rate": 0.003, + "loss": 3.997, + "step": 33683 + }, + { + "epoch": 0.33684, + "grad_norm": 0.7828316688537598, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 33684 + }, + { + "epoch": 0.33685, + "grad_norm": 0.9148006439208984, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 33685 + }, + { + "epoch": 0.33686, + "grad_norm": 0.8771753907203674, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 33686 + }, + { + "epoch": 0.33687, + "grad_norm": 1.0254648923873901, + "learning_rate": 0.003, + "loss": 3.96, + "step": 33687 + }, + { + "epoch": 0.33688, + "grad_norm": 1.3511254787445068, + "learning_rate": 0.003, + "loss": 4.016, + "step": 33688 + }, + { + "epoch": 0.33689, + "grad_norm": 0.8339183926582336, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 33689 + }, + { + "epoch": 0.3369, + "grad_norm": 0.7304290533065796, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 33690 + }, + { + "epoch": 0.33691, + "grad_norm": 0.7652133703231812, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 33691 + }, + { + "epoch": 0.33692, + "grad_norm": 0.972414493560791, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 33692 + }, + { + "epoch": 0.33693, + "grad_norm": 1.0864495038986206, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 33693 + }, + { + "epoch": 0.33694, + "grad_norm": 0.8895688652992249, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 33694 + }, + { + "epoch": 0.33695, + "grad_norm": 0.81111741065979, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 33695 + }, + { + "epoch": 0.33696, + "grad_norm": 0.766028106212616, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 33696 + }, + { + "epoch": 0.33697, + "grad_norm": 0.8438752293586731, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 33697 + }, + { + "epoch": 0.33698, + "grad_norm": 0.8716627955436707, + "learning_rate": 0.003, + "loss": 4.008, + "step": 33698 + }, + { + "epoch": 0.33699, + "grad_norm": 0.9436254501342773, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 33699 + }, + { + "epoch": 0.337, + "grad_norm": 1.1309998035430908, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 33700 + }, + { + "epoch": 0.33701, + "grad_norm": 0.917843759059906, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 33701 + }, + { + "epoch": 0.33702, + "grad_norm": 0.8850109577178955, + "learning_rate": 0.003, + "loss": 3.981, + "step": 33702 + }, + { + "epoch": 0.33703, + "grad_norm": 0.797508716583252, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 33703 + }, + { + "epoch": 0.33704, + "grad_norm": 0.8503484129905701, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 33704 + }, + { + "epoch": 0.33705, + "grad_norm": 0.8965978622436523, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 33705 + }, + { + "epoch": 0.33706, + "grad_norm": 1.11007821559906, + "learning_rate": 0.003, + "loss": 4.013, + "step": 33706 + }, + { + "epoch": 0.33707, + "grad_norm": 1.0337505340576172, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 33707 + }, + { + "epoch": 0.33708, + "grad_norm": 1.1826599836349487, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 33708 + }, + { + "epoch": 0.33709, + "grad_norm": 0.9578114748001099, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 33709 + }, + { + "epoch": 0.3371, + "grad_norm": 1.1147881746292114, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 33710 + }, + { + "epoch": 0.33711, + "grad_norm": 0.9330757260322571, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 33711 + }, + { + "epoch": 0.33712, + "grad_norm": 0.8971151113510132, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 33712 + }, + { + "epoch": 0.33713, + "grad_norm": 1.0343003273010254, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 33713 + }, + { + "epoch": 0.33714, + "grad_norm": 1.079959750175476, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 33714 + }, + { + "epoch": 0.33715, + "grad_norm": 0.75529944896698, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 33715 + }, + { + "epoch": 0.33716, + "grad_norm": 0.6533664464950562, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 33716 + }, + { + "epoch": 0.33717, + "grad_norm": 0.7049921751022339, + "learning_rate": 0.003, + "loss": 3.972, + "step": 33717 + }, + { + "epoch": 0.33718, + "grad_norm": 0.6975719928741455, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 33718 + }, + { + "epoch": 0.33719, + "grad_norm": 0.7475070357322693, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 33719 + }, + { + "epoch": 0.3372, + "grad_norm": 0.8639022707939148, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 33720 + }, + { + "epoch": 0.33721, + "grad_norm": 0.8342199325561523, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 33721 + }, + { + "epoch": 0.33722, + "grad_norm": 0.7291203141212463, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 33722 + }, + { + "epoch": 0.33723, + "grad_norm": 0.6812360286712646, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 33723 + }, + { + "epoch": 0.33724, + "grad_norm": 0.7451435327529907, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 33724 + }, + { + "epoch": 0.33725, + "grad_norm": 0.7980358600616455, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 33725 + }, + { + "epoch": 0.33726, + "grad_norm": 0.8432754874229431, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 33726 + }, + { + "epoch": 0.33727, + "grad_norm": 1.1534661054611206, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 33727 + }, + { + "epoch": 0.33728, + "grad_norm": 1.1535627841949463, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 33728 + }, + { + "epoch": 0.33729, + "grad_norm": 0.7190380096435547, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 33729 + }, + { + "epoch": 0.3373, + "grad_norm": 0.6590969562530518, + "learning_rate": 0.003, + "loss": 3.979, + "step": 33730 + }, + { + "epoch": 0.33731, + "grad_norm": 0.6496301293373108, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 33731 + }, + { + "epoch": 0.33732, + "grad_norm": 0.7071911096572876, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 33732 + }, + { + "epoch": 0.33733, + "grad_norm": 0.9839214086532593, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 33733 + }, + { + "epoch": 0.33734, + "grad_norm": 1.2180966138839722, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 33734 + }, + { + "epoch": 0.33735, + "grad_norm": 0.789925754070282, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 33735 + }, + { + "epoch": 0.33736, + "grad_norm": 0.9044973850250244, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 33736 + }, + { + "epoch": 0.33737, + "grad_norm": 1.1590567827224731, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 33737 + }, + { + "epoch": 0.33738, + "grad_norm": 0.9227322340011597, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 33738 + }, + { + "epoch": 0.33739, + "grad_norm": 1.1094919443130493, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 33739 + }, + { + "epoch": 0.3374, + "grad_norm": 1.224934458732605, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 33740 + }, + { + "epoch": 0.33741, + "grad_norm": 0.7231660485267639, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 33741 + }, + { + "epoch": 0.33742, + "grad_norm": 0.5973290205001831, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 33742 + }, + { + "epoch": 0.33743, + "grad_norm": 0.5834373235702515, + "learning_rate": 0.003, + "loss": 3.9436, + "step": 33743 + }, + { + "epoch": 0.33744, + "grad_norm": 0.646532416343689, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 33744 + }, + { + "epoch": 0.33745, + "grad_norm": 0.8325283527374268, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 33745 + }, + { + "epoch": 0.33746, + "grad_norm": 0.9892037510871887, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 33746 + }, + { + "epoch": 0.33747, + "grad_norm": 0.9951537847518921, + "learning_rate": 0.003, + "loss": 3.973, + "step": 33747 + }, + { + "epoch": 0.33748, + "grad_norm": 0.9406344890594482, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 33748 + }, + { + "epoch": 0.33749, + "grad_norm": 0.9657214283943176, + "learning_rate": 0.003, + "loss": 3.9445, + "step": 33749 + }, + { + "epoch": 0.3375, + "grad_norm": 1.1225166320800781, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 33750 + }, + { + "epoch": 0.33751, + "grad_norm": 1.1318013668060303, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 33751 + }, + { + "epoch": 0.33752, + "grad_norm": 0.7645293474197388, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 33752 + }, + { + "epoch": 0.33753, + "grad_norm": 0.823349118232727, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 33753 + }, + { + "epoch": 0.33754, + "grad_norm": 0.8613853454589844, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 33754 + }, + { + "epoch": 0.33755, + "grad_norm": 0.9517074823379517, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 33755 + }, + { + "epoch": 0.33756, + "grad_norm": 0.9662336111068726, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 33756 + }, + { + "epoch": 0.33757, + "grad_norm": 0.8970470428466797, + "learning_rate": 0.003, + "loss": 3.984, + "step": 33757 + }, + { + "epoch": 0.33758, + "grad_norm": 0.8852062225341797, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 33758 + }, + { + "epoch": 0.33759, + "grad_norm": 0.9423020482063293, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 33759 + }, + { + "epoch": 0.3376, + "grad_norm": 0.9528287649154663, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 33760 + }, + { + "epoch": 0.33761, + "grad_norm": 1.0615360736846924, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 33761 + }, + { + "epoch": 0.33762, + "grad_norm": 1.0598050355911255, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 33762 + }, + { + "epoch": 0.33763, + "grad_norm": 0.9321108460426331, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 33763 + }, + { + "epoch": 0.33764, + "grad_norm": 0.9609721302986145, + "learning_rate": 0.003, + "loss": 3.981, + "step": 33764 + }, + { + "epoch": 0.33765, + "grad_norm": 0.957132875919342, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 33765 + }, + { + "epoch": 0.33766, + "grad_norm": 0.992692768573761, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 33766 + }, + { + "epoch": 0.33767, + "grad_norm": 1.247231125831604, + "learning_rate": 0.003, + "loss": 4.003, + "step": 33767 + }, + { + "epoch": 0.33768, + "grad_norm": 0.9802218675613403, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 33768 + }, + { + "epoch": 0.33769, + "grad_norm": 1.0977238416671753, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 33769 + }, + { + "epoch": 0.3377, + "grad_norm": 0.9514569044113159, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 33770 + }, + { + "epoch": 0.33771, + "grad_norm": 0.9212953448295593, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 33771 + }, + { + "epoch": 0.33772, + "grad_norm": 0.9547812938690186, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 33772 + }, + { + "epoch": 0.33773, + "grad_norm": 1.0169247388839722, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 33773 + }, + { + "epoch": 0.33774, + "grad_norm": 0.9714726209640503, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 33774 + }, + { + "epoch": 0.33775, + "grad_norm": 1.0394251346588135, + "learning_rate": 0.003, + "loss": 3.983, + "step": 33775 + }, + { + "epoch": 0.33776, + "grad_norm": 0.9340990781784058, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 33776 + }, + { + "epoch": 0.33777, + "grad_norm": 0.8588498830795288, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 33777 + }, + { + "epoch": 0.33778, + "grad_norm": 0.9710767269134521, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 33778 + }, + { + "epoch": 0.33779, + "grad_norm": 1.0770083665847778, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 33779 + }, + { + "epoch": 0.3378, + "grad_norm": 1.0930678844451904, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 33780 + }, + { + "epoch": 0.33781, + "grad_norm": 1.2925083637237549, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 33781 + }, + { + "epoch": 0.33782, + "grad_norm": 0.7415616512298584, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 33782 + }, + { + "epoch": 0.33783, + "grad_norm": 0.6861467957496643, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 33783 + }, + { + "epoch": 0.33784, + "grad_norm": 0.9711847305297852, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 33784 + }, + { + "epoch": 0.33785, + "grad_norm": 1.1311975717544556, + "learning_rate": 0.003, + "loss": 3.972, + "step": 33785 + }, + { + "epoch": 0.33786, + "grad_norm": 1.0164271593093872, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 33786 + }, + { + "epoch": 0.33787, + "grad_norm": 0.7187546491622925, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 33787 + }, + { + "epoch": 0.33788, + "grad_norm": 0.7138071656227112, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 33788 + }, + { + "epoch": 0.33789, + "grad_norm": 0.9803473949432373, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 33789 + }, + { + "epoch": 0.3379, + "grad_norm": 1.1419589519500732, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 33790 + }, + { + "epoch": 0.33791, + "grad_norm": 0.8231909871101379, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 33791 + }, + { + "epoch": 0.33792, + "grad_norm": 0.6870018243789673, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 33792 + }, + { + "epoch": 0.33793, + "grad_norm": 0.7267242670059204, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 33793 + }, + { + "epoch": 0.33794, + "grad_norm": 0.77128005027771, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 33794 + }, + { + "epoch": 0.33795, + "grad_norm": 0.8065897226333618, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 33795 + }, + { + "epoch": 0.33796, + "grad_norm": 0.7657335996627808, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 33796 + }, + { + "epoch": 0.33797, + "grad_norm": 0.8398439884185791, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 33797 + }, + { + "epoch": 0.33798, + "grad_norm": 0.9659525156021118, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 33798 + }, + { + "epoch": 0.33799, + "grad_norm": 1.041957139968872, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 33799 + }, + { + "epoch": 0.338, + "grad_norm": 0.8905584812164307, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 33800 + }, + { + "epoch": 0.33801, + "grad_norm": 0.7005693316459656, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 33801 + }, + { + "epoch": 0.33802, + "grad_norm": 0.6474481821060181, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 33802 + }, + { + "epoch": 0.33803, + "grad_norm": 0.7510114312171936, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 33803 + }, + { + "epoch": 0.33804, + "grad_norm": 0.871907114982605, + "learning_rate": 0.003, + "loss": 3.9267, + "step": 33804 + }, + { + "epoch": 0.33805, + "grad_norm": 1.09319269657135, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 33805 + }, + { + "epoch": 0.33806, + "grad_norm": 1.053057312965393, + "learning_rate": 0.003, + "loss": 3.99, + "step": 33806 + }, + { + "epoch": 0.33807, + "grad_norm": 0.8949390649795532, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 33807 + }, + { + "epoch": 0.33808, + "grad_norm": 0.8676890134811401, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 33808 + }, + { + "epoch": 0.33809, + "grad_norm": 0.8763558268547058, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 33809 + }, + { + "epoch": 0.3381, + "grad_norm": 0.862024188041687, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 33810 + }, + { + "epoch": 0.33811, + "grad_norm": 0.9216997623443604, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 33811 + }, + { + "epoch": 0.33812, + "grad_norm": 0.8464081883430481, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 33812 + }, + { + "epoch": 0.33813, + "grad_norm": 0.9098405241966248, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 33813 + }, + { + "epoch": 0.33814, + "grad_norm": 0.9463884830474854, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 33814 + }, + { + "epoch": 0.33815, + "grad_norm": 0.8926428556442261, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 33815 + }, + { + "epoch": 0.33816, + "grad_norm": 0.7735193967819214, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 33816 + }, + { + "epoch": 0.33817, + "grad_norm": 0.8192412853240967, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 33817 + }, + { + "epoch": 0.33818, + "grad_norm": 0.8659513592720032, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 33818 + }, + { + "epoch": 0.33819, + "grad_norm": 1.0059348344802856, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 33819 + }, + { + "epoch": 0.3382, + "grad_norm": 1.3523024320602417, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 33820 + }, + { + "epoch": 0.33821, + "grad_norm": 0.8539023399353027, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 33821 + }, + { + "epoch": 0.33822, + "grad_norm": 0.7331437468528748, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 33822 + }, + { + "epoch": 0.33823, + "grad_norm": 0.8529736399650574, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 33823 + }, + { + "epoch": 0.33824, + "grad_norm": 0.8353227376937866, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 33824 + }, + { + "epoch": 0.33825, + "grad_norm": 0.8833234310150146, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 33825 + }, + { + "epoch": 0.33826, + "grad_norm": 0.7585614919662476, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 33826 + }, + { + "epoch": 0.33827, + "grad_norm": 0.7811872959136963, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 33827 + }, + { + "epoch": 0.33828, + "grad_norm": 0.8733521699905396, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 33828 + }, + { + "epoch": 0.33829, + "grad_norm": 1.1227225065231323, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 33829 + }, + { + "epoch": 0.3383, + "grad_norm": 0.9650214910507202, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 33830 + }, + { + "epoch": 0.33831, + "grad_norm": 0.8450899720191956, + "learning_rate": 0.003, + "loss": 3.983, + "step": 33831 + }, + { + "epoch": 0.33832, + "grad_norm": 0.8685036897659302, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 33832 + }, + { + "epoch": 0.33833, + "grad_norm": 0.9433384537696838, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 33833 + }, + { + "epoch": 0.33834, + "grad_norm": 0.9880135655403137, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 33834 + }, + { + "epoch": 0.33835, + "grad_norm": 1.0064380168914795, + "learning_rate": 0.003, + "loss": 4.006, + "step": 33835 + }, + { + "epoch": 0.33836, + "grad_norm": 1.0948982238769531, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 33836 + }, + { + "epoch": 0.33837, + "grad_norm": 1.0549588203430176, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 33837 + }, + { + "epoch": 0.33838, + "grad_norm": 0.9929041862487793, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 33838 + }, + { + "epoch": 0.33839, + "grad_norm": 1.1393741369247437, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 33839 + }, + { + "epoch": 0.3384, + "grad_norm": 0.9517659544944763, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 33840 + }, + { + "epoch": 0.33841, + "grad_norm": 1.1493990421295166, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 33841 + }, + { + "epoch": 0.33842, + "grad_norm": 1.0585126876831055, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 33842 + }, + { + "epoch": 0.33843, + "grad_norm": 0.8741393685340881, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 33843 + }, + { + "epoch": 0.33844, + "grad_norm": 0.728969156742096, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 33844 + }, + { + "epoch": 0.33845, + "grad_norm": 0.735647976398468, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 33845 + }, + { + "epoch": 0.33846, + "grad_norm": 0.7893258333206177, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 33846 + }, + { + "epoch": 0.33847, + "grad_norm": 0.667068362236023, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 33847 + }, + { + "epoch": 0.33848, + "grad_norm": 0.8544107675552368, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 33848 + }, + { + "epoch": 0.33849, + "grad_norm": 0.9796803593635559, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 33849 + }, + { + "epoch": 0.3385, + "grad_norm": 0.9238835573196411, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 33850 + }, + { + "epoch": 0.33851, + "grad_norm": 0.8792781233787537, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 33851 + }, + { + "epoch": 0.33852, + "grad_norm": 0.8739051222801208, + "learning_rate": 0.003, + "loss": 3.96, + "step": 33852 + }, + { + "epoch": 0.33853, + "grad_norm": 0.8972802758216858, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 33853 + }, + { + "epoch": 0.33854, + "grad_norm": 0.9691957235336304, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 33854 + }, + { + "epoch": 0.33855, + "grad_norm": 0.9729677438735962, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 33855 + }, + { + "epoch": 0.33856, + "grad_norm": 0.7469257116317749, + "learning_rate": 0.003, + "loss": 3.958, + "step": 33856 + }, + { + "epoch": 0.33857, + "grad_norm": 0.6727645397186279, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 33857 + }, + { + "epoch": 0.33858, + "grad_norm": 0.6509019136428833, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 33858 + }, + { + "epoch": 0.33859, + "grad_norm": 0.717968761920929, + "learning_rate": 0.003, + "loss": 3.983, + "step": 33859 + }, + { + "epoch": 0.3386, + "grad_norm": 0.8108325004577637, + "learning_rate": 0.003, + "loss": 3.985, + "step": 33860 + }, + { + "epoch": 0.33861, + "grad_norm": 0.9632368087768555, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 33861 + }, + { + "epoch": 0.33862, + "grad_norm": 1.0850262641906738, + "learning_rate": 0.003, + "loss": 4.022, + "step": 33862 + }, + { + "epoch": 0.33863, + "grad_norm": 0.7313600182533264, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 33863 + }, + { + "epoch": 0.33864, + "grad_norm": 0.685490608215332, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 33864 + }, + { + "epoch": 0.33865, + "grad_norm": 0.8060835599899292, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 33865 + }, + { + "epoch": 0.33866, + "grad_norm": 0.8476738333702087, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 33866 + }, + { + "epoch": 0.33867, + "grad_norm": 0.8244439959526062, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 33867 + }, + { + "epoch": 0.33868, + "grad_norm": 0.9015598297119141, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 33868 + }, + { + "epoch": 0.33869, + "grad_norm": 0.9322713613510132, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 33869 + }, + { + "epoch": 0.3387, + "grad_norm": 0.9564865827560425, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 33870 + }, + { + "epoch": 0.33871, + "grad_norm": 0.9587309956550598, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 33871 + }, + { + "epoch": 0.33872, + "grad_norm": 0.9372667074203491, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 33872 + }, + { + "epoch": 0.33873, + "grad_norm": 1.1125426292419434, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 33873 + }, + { + "epoch": 0.33874, + "grad_norm": 0.9659783244132996, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 33874 + }, + { + "epoch": 0.33875, + "grad_norm": 0.8820645809173584, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 33875 + }, + { + "epoch": 0.33876, + "grad_norm": 0.8420048952102661, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 33876 + }, + { + "epoch": 0.33877, + "grad_norm": 0.8484020233154297, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 33877 + }, + { + "epoch": 0.33878, + "grad_norm": 0.8323063254356384, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 33878 + }, + { + "epoch": 0.33879, + "grad_norm": 0.9252663850784302, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 33879 + }, + { + "epoch": 0.3388, + "grad_norm": 1.2017310857772827, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 33880 + }, + { + "epoch": 0.33881, + "grad_norm": 1.0841165781021118, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 33881 + }, + { + "epoch": 0.33882, + "grad_norm": 0.873155415058136, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 33882 + }, + { + "epoch": 0.33883, + "grad_norm": 0.7629409432411194, + "learning_rate": 0.003, + "loss": 3.994, + "step": 33883 + }, + { + "epoch": 0.33884, + "grad_norm": 0.7062758207321167, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 33884 + }, + { + "epoch": 0.33885, + "grad_norm": 0.7552800178527832, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 33885 + }, + { + "epoch": 0.33886, + "grad_norm": 0.8383352756500244, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 33886 + }, + { + "epoch": 0.33887, + "grad_norm": 0.7554622292518616, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 33887 + }, + { + "epoch": 0.33888, + "grad_norm": 0.7007030844688416, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 33888 + }, + { + "epoch": 0.33889, + "grad_norm": 0.654003918170929, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 33889 + }, + { + "epoch": 0.3389, + "grad_norm": 0.7013734579086304, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 33890 + }, + { + "epoch": 0.33891, + "grad_norm": 0.7461391091346741, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 33891 + }, + { + "epoch": 0.33892, + "grad_norm": 0.8165204524993896, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 33892 + }, + { + "epoch": 0.33893, + "grad_norm": 0.7861067056655884, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 33893 + }, + { + "epoch": 0.33894, + "grad_norm": 0.9307798743247986, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 33894 + }, + { + "epoch": 0.33895, + "grad_norm": 1.377911925315857, + "learning_rate": 0.003, + "loss": 3.963, + "step": 33895 + }, + { + "epoch": 0.33896, + "grad_norm": 0.9062038064002991, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 33896 + }, + { + "epoch": 0.33897, + "grad_norm": 0.8758798241615295, + "learning_rate": 0.003, + "loss": 4.013, + "step": 33897 + }, + { + "epoch": 0.33898, + "grad_norm": 1.0893769264221191, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 33898 + }, + { + "epoch": 0.33899, + "grad_norm": 1.2158273458480835, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 33899 + }, + { + "epoch": 0.339, + "grad_norm": 0.6237785816192627, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 33900 + }, + { + "epoch": 0.33901, + "grad_norm": 0.6795762181282043, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 33901 + }, + { + "epoch": 0.33902, + "grad_norm": 0.8751939535140991, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 33902 + }, + { + "epoch": 0.33903, + "grad_norm": 0.9785947203636169, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 33903 + }, + { + "epoch": 0.33904, + "grad_norm": 1.122113585472107, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 33904 + }, + { + "epoch": 0.33905, + "grad_norm": 0.6784164309501648, + "learning_rate": 0.003, + "loss": 3.952, + "step": 33905 + }, + { + "epoch": 0.33906, + "grad_norm": 0.6387081742286682, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 33906 + }, + { + "epoch": 0.33907, + "grad_norm": 0.7752143740653992, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 33907 + }, + { + "epoch": 0.33908, + "grad_norm": 0.7896827459335327, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 33908 + }, + { + "epoch": 0.33909, + "grad_norm": 0.8328292369842529, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 33909 + }, + { + "epoch": 0.3391, + "grad_norm": 0.7768694162368774, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 33910 + }, + { + "epoch": 0.33911, + "grad_norm": 0.6580144166946411, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 33911 + }, + { + "epoch": 0.33912, + "grad_norm": 0.644828200340271, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 33912 + }, + { + "epoch": 0.33913, + "grad_norm": 0.7214594483375549, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 33913 + }, + { + "epoch": 0.33914, + "grad_norm": 0.7408504486083984, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 33914 + }, + { + "epoch": 0.33915, + "grad_norm": 0.7372466325759888, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 33915 + }, + { + "epoch": 0.33916, + "grad_norm": 0.9280502796173096, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 33916 + }, + { + "epoch": 0.33917, + "grad_norm": 1.2494479417800903, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 33917 + }, + { + "epoch": 0.33918, + "grad_norm": 1.1396915912628174, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 33918 + }, + { + "epoch": 0.33919, + "grad_norm": 0.8696605563163757, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 33919 + }, + { + "epoch": 0.3392, + "grad_norm": 0.7769641280174255, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 33920 + }, + { + "epoch": 0.33921, + "grad_norm": 0.7793871760368347, + "learning_rate": 0.003, + "loss": 3.975, + "step": 33921 + }, + { + "epoch": 0.33922, + "grad_norm": 1.0082019567489624, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 33922 + }, + { + "epoch": 0.33923, + "grad_norm": 1.0739202499389648, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 33923 + }, + { + "epoch": 0.33924, + "grad_norm": 0.8662359118461609, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 33924 + }, + { + "epoch": 0.33925, + "grad_norm": 0.8597113490104675, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 33925 + }, + { + "epoch": 0.33926, + "grad_norm": 0.963589072227478, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 33926 + }, + { + "epoch": 0.33927, + "grad_norm": 1.175758957862854, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 33927 + }, + { + "epoch": 0.33928, + "grad_norm": 1.018174171447754, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 33928 + }, + { + "epoch": 0.33929, + "grad_norm": 1.1977590322494507, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 33929 + }, + { + "epoch": 0.3393, + "grad_norm": 0.9117292165756226, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 33930 + }, + { + "epoch": 0.33931, + "grad_norm": 0.9122856855392456, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 33931 + }, + { + "epoch": 0.33932, + "grad_norm": 0.9673566222190857, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 33932 + }, + { + "epoch": 0.33933, + "grad_norm": 1.200395107269287, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 33933 + }, + { + "epoch": 0.33934, + "grad_norm": 0.7344029545783997, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 33934 + }, + { + "epoch": 0.33935, + "grad_norm": 0.7534857392311096, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 33935 + }, + { + "epoch": 0.33936, + "grad_norm": 0.787803590297699, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 33936 + }, + { + "epoch": 0.33937, + "grad_norm": 1.034033179283142, + "learning_rate": 0.003, + "loss": 3.983, + "step": 33937 + }, + { + "epoch": 0.33938, + "grad_norm": 1.1083773374557495, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 33938 + }, + { + "epoch": 0.33939, + "grad_norm": 0.9200851917266846, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 33939 + }, + { + "epoch": 0.3394, + "grad_norm": 0.8883707523345947, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 33940 + }, + { + "epoch": 0.33941, + "grad_norm": 0.9008223414421082, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 33941 + }, + { + "epoch": 0.33942, + "grad_norm": 1.0118054151535034, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 33942 + }, + { + "epoch": 0.33943, + "grad_norm": 1.093557357788086, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 33943 + }, + { + "epoch": 0.33944, + "grad_norm": 0.8127647638320923, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 33944 + }, + { + "epoch": 0.33945, + "grad_norm": 0.8315039873123169, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 33945 + }, + { + "epoch": 0.33946, + "grad_norm": 0.810836136341095, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 33946 + }, + { + "epoch": 0.33947, + "grad_norm": 0.8106192946434021, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 33947 + }, + { + "epoch": 0.33948, + "grad_norm": 0.8413642644882202, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 33948 + }, + { + "epoch": 0.33949, + "grad_norm": 0.8315308690071106, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 33949 + }, + { + "epoch": 0.3395, + "grad_norm": 0.7195736169815063, + "learning_rate": 0.003, + "loss": 3.99, + "step": 33950 + }, + { + "epoch": 0.33951, + "grad_norm": 0.8805912137031555, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 33951 + }, + { + "epoch": 0.33952, + "grad_norm": 0.7627028226852417, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 33952 + }, + { + "epoch": 0.33953, + "grad_norm": 0.7968209385871887, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 33953 + }, + { + "epoch": 0.33954, + "grad_norm": 0.98475581407547, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 33954 + }, + { + "epoch": 0.33955, + "grad_norm": 1.207260251045227, + "learning_rate": 0.003, + "loss": 4.028, + "step": 33955 + }, + { + "epoch": 0.33956, + "grad_norm": 1.1988497972488403, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 33956 + }, + { + "epoch": 0.33957, + "grad_norm": 0.8607190847396851, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 33957 + }, + { + "epoch": 0.33958, + "grad_norm": 0.7858001589775085, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 33958 + }, + { + "epoch": 0.33959, + "grad_norm": 0.8987900614738464, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 33959 + }, + { + "epoch": 0.3396, + "grad_norm": 1.063622236251831, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 33960 + }, + { + "epoch": 0.33961, + "grad_norm": 0.9053446054458618, + "learning_rate": 0.003, + "loss": 4.01, + "step": 33961 + }, + { + "epoch": 0.33962, + "grad_norm": 0.9685106873512268, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 33962 + }, + { + "epoch": 0.33963, + "grad_norm": 0.8694130182266235, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 33963 + }, + { + "epoch": 0.33964, + "grad_norm": 0.6743022203445435, + "learning_rate": 0.003, + "loss": 3.9383, + "step": 33964 + }, + { + "epoch": 0.33965, + "grad_norm": 0.7213859558105469, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 33965 + }, + { + "epoch": 0.33966, + "grad_norm": 0.843792200088501, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 33966 + }, + { + "epoch": 0.33967, + "grad_norm": 0.9752247333526611, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 33967 + }, + { + "epoch": 0.33968, + "grad_norm": 1.2402836084365845, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 33968 + }, + { + "epoch": 0.33969, + "grad_norm": 0.7877816557884216, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 33969 + }, + { + "epoch": 0.3397, + "grad_norm": 0.744081974029541, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 33970 + }, + { + "epoch": 0.33971, + "grad_norm": 0.8707138895988464, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 33971 + }, + { + "epoch": 0.33972, + "grad_norm": 1.0127248764038086, + "learning_rate": 0.003, + "loss": 4.003, + "step": 33972 + }, + { + "epoch": 0.33973, + "grad_norm": 1.1678484678268433, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 33973 + }, + { + "epoch": 0.33974, + "grad_norm": 1.0128246545791626, + "learning_rate": 0.003, + "loss": 3.974, + "step": 33974 + }, + { + "epoch": 0.33975, + "grad_norm": 1.1158918142318726, + "learning_rate": 0.003, + "loss": 3.9447, + "step": 33975 + }, + { + "epoch": 0.33976, + "grad_norm": 0.9902060627937317, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 33976 + }, + { + "epoch": 0.33977, + "grad_norm": 1.0819774866104126, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 33977 + }, + { + "epoch": 0.33978, + "grad_norm": 0.923795223236084, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 33978 + }, + { + "epoch": 0.33979, + "grad_norm": 0.8933278322219849, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 33979 + }, + { + "epoch": 0.3398, + "grad_norm": 0.8184022903442383, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 33980 + }, + { + "epoch": 0.33981, + "grad_norm": 0.93113112449646, + "learning_rate": 0.003, + "loss": 3.964, + "step": 33981 + }, + { + "epoch": 0.33982, + "grad_norm": 1.1346431970596313, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 33982 + }, + { + "epoch": 0.33983, + "grad_norm": 0.8947099447250366, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 33983 + }, + { + "epoch": 0.33984, + "grad_norm": 0.8062654733657837, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 33984 + }, + { + "epoch": 0.33985, + "grad_norm": 0.9632655382156372, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 33985 + }, + { + "epoch": 0.33986, + "grad_norm": 0.9771274328231812, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 33986 + }, + { + "epoch": 0.33987, + "grad_norm": 0.848258376121521, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 33987 + }, + { + "epoch": 0.33988, + "grad_norm": 1.1529754400253296, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 33988 + }, + { + "epoch": 0.33989, + "grad_norm": 1.031385898590088, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 33989 + }, + { + "epoch": 0.3399, + "grad_norm": 0.6749846935272217, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 33990 + }, + { + "epoch": 0.33991, + "grad_norm": 0.6770142316818237, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 33991 + }, + { + "epoch": 0.33992, + "grad_norm": 0.7330157160758972, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 33992 + }, + { + "epoch": 0.33993, + "grad_norm": 0.8147822618484497, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 33993 + }, + { + "epoch": 0.33994, + "grad_norm": 0.796364963054657, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 33994 + }, + { + "epoch": 0.33995, + "grad_norm": 0.8182478547096252, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 33995 + }, + { + "epoch": 0.33996, + "grad_norm": 0.8443918824195862, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 33996 + }, + { + "epoch": 0.33997, + "grad_norm": 0.9267309308052063, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 33997 + }, + { + "epoch": 0.33998, + "grad_norm": 1.1304476261138916, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 33998 + }, + { + "epoch": 0.33999, + "grad_norm": 1.258584976196289, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 33999 + }, + { + "epoch": 0.34, + "grad_norm": 0.776000440120697, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 34000 + }, + { + "epoch": 0.34001, + "grad_norm": 0.7325758934020996, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 34001 + }, + { + "epoch": 0.34002, + "grad_norm": 0.6396294236183167, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 34002 + }, + { + "epoch": 0.34003, + "grad_norm": 0.6327236890792847, + "learning_rate": 0.003, + "loss": 4.005, + "step": 34003 + }, + { + "epoch": 0.34004, + "grad_norm": 0.6584072709083557, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 34004 + }, + { + "epoch": 0.34005, + "grad_norm": 0.8114899396896362, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 34005 + }, + { + "epoch": 0.34006, + "grad_norm": 0.8856972455978394, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 34006 + }, + { + "epoch": 0.34007, + "grad_norm": 0.9366676211357117, + "learning_rate": 0.003, + "loss": 3.989, + "step": 34007 + }, + { + "epoch": 0.34008, + "grad_norm": 1.1337766647338867, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 34008 + }, + { + "epoch": 0.34009, + "grad_norm": 0.9504051208496094, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 34009 + }, + { + "epoch": 0.3401, + "grad_norm": 1.2198686599731445, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 34010 + }, + { + "epoch": 0.34011, + "grad_norm": 0.9347732067108154, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 34011 + }, + { + "epoch": 0.34012, + "grad_norm": 0.9205287098884583, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 34012 + }, + { + "epoch": 0.34013, + "grad_norm": 0.9286801815032959, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 34013 + }, + { + "epoch": 0.34014, + "grad_norm": 1.1542078256607056, + "learning_rate": 0.003, + "loss": 3.976, + "step": 34014 + }, + { + "epoch": 0.34015, + "grad_norm": 0.9787403345108032, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 34015 + }, + { + "epoch": 0.34016, + "grad_norm": 0.8848873376846313, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 34016 + }, + { + "epoch": 0.34017, + "grad_norm": 0.9125164151191711, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 34017 + }, + { + "epoch": 0.34018, + "grad_norm": 0.9844927191734314, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 34018 + }, + { + "epoch": 0.34019, + "grad_norm": 0.9562814235687256, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 34019 + }, + { + "epoch": 0.3402, + "grad_norm": 1.1552138328552246, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 34020 + }, + { + "epoch": 0.34021, + "grad_norm": 1.1595455408096313, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 34021 + }, + { + "epoch": 0.34022, + "grad_norm": 1.005078911781311, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 34022 + }, + { + "epoch": 0.34023, + "grad_norm": 1.1196856498718262, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 34023 + }, + { + "epoch": 0.34024, + "grad_norm": 0.8335196375846863, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 34024 + }, + { + "epoch": 0.34025, + "grad_norm": 0.6690870523452759, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 34025 + }, + { + "epoch": 0.34026, + "grad_norm": 0.644472062587738, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 34026 + }, + { + "epoch": 0.34027, + "grad_norm": 0.57020103931427, + "learning_rate": 0.003, + "loss": 3.99, + "step": 34027 + }, + { + "epoch": 0.34028, + "grad_norm": 0.5952790975570679, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 34028 + }, + { + "epoch": 0.34029, + "grad_norm": 0.613896906375885, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 34029 + }, + { + "epoch": 0.3403, + "grad_norm": 0.6873587965965271, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 34030 + }, + { + "epoch": 0.34031, + "grad_norm": 0.8640547394752502, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 34031 + }, + { + "epoch": 0.34032, + "grad_norm": 1.048680067062378, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 34032 + }, + { + "epoch": 0.34033, + "grad_norm": 1.1077991724014282, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 34033 + }, + { + "epoch": 0.34034, + "grad_norm": 0.8272826671600342, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 34034 + }, + { + "epoch": 0.34035, + "grad_norm": 0.641241729259491, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 34035 + }, + { + "epoch": 0.34036, + "grad_norm": 0.6289634704589844, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 34036 + }, + { + "epoch": 0.34037, + "grad_norm": 0.7384872436523438, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 34037 + }, + { + "epoch": 0.34038, + "grad_norm": 0.7593327760696411, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 34038 + }, + { + "epoch": 0.34039, + "grad_norm": 0.8093575239181519, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 34039 + }, + { + "epoch": 0.3404, + "grad_norm": 0.8557952642440796, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 34040 + }, + { + "epoch": 0.34041, + "grad_norm": 0.9042707681655884, + "learning_rate": 0.003, + "loss": 3.9107, + "step": 34041 + }, + { + "epoch": 0.34042, + "grad_norm": 0.8125980496406555, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 34042 + }, + { + "epoch": 0.34043, + "grad_norm": 0.8361361026763916, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 34043 + }, + { + "epoch": 0.34044, + "grad_norm": 0.9660698175430298, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 34044 + }, + { + "epoch": 0.34045, + "grad_norm": 1.0439250469207764, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 34045 + }, + { + "epoch": 0.34046, + "grad_norm": 0.8316056132316589, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 34046 + }, + { + "epoch": 0.34047, + "grad_norm": 0.7637447714805603, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 34047 + }, + { + "epoch": 0.34048, + "grad_norm": 0.7100105285644531, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 34048 + }, + { + "epoch": 0.34049, + "grad_norm": 0.7280190587043762, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 34049 + }, + { + "epoch": 0.3405, + "grad_norm": 0.7648276090621948, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 34050 + }, + { + "epoch": 0.34051, + "grad_norm": 0.8780141472816467, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 34051 + }, + { + "epoch": 0.34052, + "grad_norm": 0.9839565753936768, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 34052 + }, + { + "epoch": 0.34053, + "grad_norm": 1.183361530303955, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 34053 + }, + { + "epoch": 0.34054, + "grad_norm": 1.1474149227142334, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 34054 + }, + { + "epoch": 0.34055, + "grad_norm": 0.9899405241012573, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 34055 + }, + { + "epoch": 0.34056, + "grad_norm": 1.01242196559906, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 34056 + }, + { + "epoch": 0.34057, + "grad_norm": 1.0919415950775146, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 34057 + }, + { + "epoch": 0.34058, + "grad_norm": 0.8685453534126282, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 34058 + }, + { + "epoch": 0.34059, + "grad_norm": 0.729981005191803, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 34059 + }, + { + "epoch": 0.3406, + "grad_norm": 0.8546142578125, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 34060 + }, + { + "epoch": 0.34061, + "grad_norm": 0.8781875371932983, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 34061 + }, + { + "epoch": 0.34062, + "grad_norm": 0.8968924880027771, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 34062 + }, + { + "epoch": 0.34063, + "grad_norm": 1.1244256496429443, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 34063 + }, + { + "epoch": 0.34064, + "grad_norm": 0.8394032716751099, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 34064 + }, + { + "epoch": 0.34065, + "grad_norm": 0.9295394420623779, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 34065 + }, + { + "epoch": 0.34066, + "grad_norm": 1.1068427562713623, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 34066 + }, + { + "epoch": 0.34067, + "grad_norm": 1.4426127672195435, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 34067 + }, + { + "epoch": 0.34068, + "grad_norm": 0.7910992503166199, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 34068 + }, + { + "epoch": 0.34069, + "grad_norm": 0.760866105556488, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 34069 + }, + { + "epoch": 0.3407, + "grad_norm": 0.7134959697723389, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 34070 + }, + { + "epoch": 0.34071, + "grad_norm": 0.7017151713371277, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 34071 + }, + { + "epoch": 0.34072, + "grad_norm": 0.7090153098106384, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 34072 + }, + { + "epoch": 0.34073, + "grad_norm": 0.7455703020095825, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 34073 + }, + { + "epoch": 0.34074, + "grad_norm": 0.7267268300056458, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 34074 + }, + { + "epoch": 0.34075, + "grad_norm": 0.7904853820800781, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 34075 + }, + { + "epoch": 0.34076, + "grad_norm": 0.9018135666847229, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 34076 + }, + { + "epoch": 0.34077, + "grad_norm": 1.1719361543655396, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 34077 + }, + { + "epoch": 0.34078, + "grad_norm": 0.8538778424263, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 34078 + }, + { + "epoch": 0.34079, + "grad_norm": 0.7390376329421997, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 34079 + }, + { + "epoch": 0.3408, + "grad_norm": 0.813887894153595, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 34080 + }, + { + "epoch": 0.34081, + "grad_norm": 0.8795467615127563, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 34081 + }, + { + "epoch": 0.34082, + "grad_norm": 0.9331591129302979, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 34082 + }, + { + "epoch": 0.34083, + "grad_norm": 1.1047848463058472, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 34083 + }, + { + "epoch": 0.34084, + "grad_norm": 1.085903525352478, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 34084 + }, + { + "epoch": 0.34085, + "grad_norm": 0.9392151236534119, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 34085 + }, + { + "epoch": 0.34086, + "grad_norm": 0.9999030232429504, + "learning_rate": 0.003, + "loss": 3.976, + "step": 34086 + }, + { + "epoch": 0.34087, + "grad_norm": 1.2424204349517822, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 34087 + }, + { + "epoch": 0.34088, + "grad_norm": 0.8910965919494629, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 34088 + }, + { + "epoch": 0.34089, + "grad_norm": 0.8088398575782776, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 34089 + }, + { + "epoch": 0.3409, + "grad_norm": 0.7185824513435364, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 34090 + }, + { + "epoch": 0.34091, + "grad_norm": 0.6840623021125793, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 34091 + }, + { + "epoch": 0.34092, + "grad_norm": 0.8175874352455139, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 34092 + }, + { + "epoch": 0.34093, + "grad_norm": 0.841539204120636, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 34093 + }, + { + "epoch": 0.34094, + "grad_norm": 0.768546462059021, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 34094 + }, + { + "epoch": 0.34095, + "grad_norm": 0.7433310151100159, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 34095 + }, + { + "epoch": 0.34096, + "grad_norm": 0.7799928188323975, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 34096 + }, + { + "epoch": 0.34097, + "grad_norm": 0.9313113689422607, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 34097 + }, + { + "epoch": 0.34098, + "grad_norm": 0.8951665759086609, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 34098 + }, + { + "epoch": 0.34099, + "grad_norm": 0.8528792858123779, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 34099 + }, + { + "epoch": 0.341, + "grad_norm": 1.0524948835372925, + "learning_rate": 0.003, + "loss": 3.968, + "step": 34100 + }, + { + "epoch": 0.34101, + "grad_norm": 1.1299052238464355, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 34101 + }, + { + "epoch": 0.34102, + "grad_norm": 0.6345700621604919, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 34102 + }, + { + "epoch": 0.34103, + "grad_norm": 0.5289462208747864, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 34103 + }, + { + "epoch": 0.34104, + "grad_norm": 0.6133047342300415, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 34104 + }, + { + "epoch": 0.34105, + "grad_norm": 0.9024978876113892, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 34105 + }, + { + "epoch": 0.34106, + "grad_norm": 1.322149634361267, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 34106 + }, + { + "epoch": 0.34107, + "grad_norm": 0.6891055703163147, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 34107 + }, + { + "epoch": 0.34108, + "grad_norm": 0.5465244054794312, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 34108 + }, + { + "epoch": 0.34109, + "grad_norm": 0.6161264181137085, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 34109 + }, + { + "epoch": 0.3411, + "grad_norm": 0.6690406203269958, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 34110 + }, + { + "epoch": 0.34111, + "grad_norm": 0.7137323021888733, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 34111 + }, + { + "epoch": 0.34112, + "grad_norm": 0.7074917554855347, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 34112 + }, + { + "epoch": 0.34113, + "grad_norm": 0.7969538569450378, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 34113 + }, + { + "epoch": 0.34114, + "grad_norm": 0.9404736161231995, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 34114 + }, + { + "epoch": 0.34115, + "grad_norm": 1.112317442893982, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 34115 + }, + { + "epoch": 0.34116, + "grad_norm": 0.8709869384765625, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 34116 + }, + { + "epoch": 0.34117, + "grad_norm": 0.7640981674194336, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 34117 + }, + { + "epoch": 0.34118, + "grad_norm": 0.7774866819381714, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 34118 + }, + { + "epoch": 0.34119, + "grad_norm": 0.9631947875022888, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 34119 + }, + { + "epoch": 0.3412, + "grad_norm": 1.025522232055664, + "learning_rate": 0.003, + "loss": 3.9398, + "step": 34120 + }, + { + "epoch": 0.34121, + "grad_norm": 1.118577241897583, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 34121 + }, + { + "epoch": 0.34122, + "grad_norm": 0.786483108997345, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 34122 + }, + { + "epoch": 0.34123, + "grad_norm": 0.8778904676437378, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 34123 + }, + { + "epoch": 0.34124, + "grad_norm": 1.230366826057434, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 34124 + }, + { + "epoch": 0.34125, + "grad_norm": 0.8244633078575134, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 34125 + }, + { + "epoch": 0.34126, + "grad_norm": 0.9111184477806091, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 34126 + }, + { + "epoch": 0.34127, + "grad_norm": 0.9808758497238159, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 34127 + }, + { + "epoch": 0.34128, + "grad_norm": 1.0018277168273926, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 34128 + }, + { + "epoch": 0.34129, + "grad_norm": 1.0365277528762817, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 34129 + }, + { + "epoch": 0.3413, + "grad_norm": 1.0231162309646606, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 34130 + }, + { + "epoch": 0.34131, + "grad_norm": 1.285929799079895, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 34131 + }, + { + "epoch": 0.34132, + "grad_norm": 0.8231629133224487, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 34132 + }, + { + "epoch": 0.34133, + "grad_norm": 0.9179275035858154, + "learning_rate": 0.003, + "loss": 3.964, + "step": 34133 + }, + { + "epoch": 0.34134, + "grad_norm": 1.0124223232269287, + "learning_rate": 0.003, + "loss": 3.968, + "step": 34134 + }, + { + "epoch": 0.34135, + "grad_norm": 1.1179908514022827, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 34135 + }, + { + "epoch": 0.34136, + "grad_norm": 0.8766873478889465, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 34136 + }, + { + "epoch": 0.34137, + "grad_norm": 0.9456163048744202, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 34137 + }, + { + "epoch": 0.34138, + "grad_norm": 1.0688636302947998, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 34138 + }, + { + "epoch": 0.34139, + "grad_norm": 0.9708578586578369, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 34139 + }, + { + "epoch": 0.3414, + "grad_norm": 0.99611896276474, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 34140 + }, + { + "epoch": 0.34141, + "grad_norm": 0.9496821165084839, + "learning_rate": 0.003, + "loss": 3.984, + "step": 34141 + }, + { + "epoch": 0.34142, + "grad_norm": 0.9419203996658325, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 34142 + }, + { + "epoch": 0.34143, + "grad_norm": 0.9570938944816589, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 34143 + }, + { + "epoch": 0.34144, + "grad_norm": 0.9054812788963318, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 34144 + }, + { + "epoch": 0.34145, + "grad_norm": 0.8640575408935547, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 34145 + }, + { + "epoch": 0.34146, + "grad_norm": 0.9659289121627808, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 34146 + }, + { + "epoch": 0.34147, + "grad_norm": 0.8211184740066528, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 34147 + }, + { + "epoch": 0.34148, + "grad_norm": 0.6947760581970215, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 34148 + }, + { + "epoch": 0.34149, + "grad_norm": 0.7302140593528748, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 34149 + }, + { + "epoch": 0.3415, + "grad_norm": 0.7961023449897766, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 34150 + }, + { + "epoch": 0.34151, + "grad_norm": 0.9247610569000244, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 34151 + }, + { + "epoch": 0.34152, + "grad_norm": 1.1608754396438599, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 34152 + }, + { + "epoch": 0.34153, + "grad_norm": 0.8195092082023621, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 34153 + }, + { + "epoch": 0.34154, + "grad_norm": 0.8316060900688171, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 34154 + }, + { + "epoch": 0.34155, + "grad_norm": 1.1683719158172607, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 34155 + }, + { + "epoch": 0.34156, + "grad_norm": 1.1654911041259766, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 34156 + }, + { + "epoch": 0.34157, + "grad_norm": 0.9363168478012085, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 34157 + }, + { + "epoch": 0.34158, + "grad_norm": 1.0236397981643677, + "learning_rate": 0.003, + "loss": 3.992, + "step": 34158 + }, + { + "epoch": 0.34159, + "grad_norm": 1.171990990638733, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 34159 + }, + { + "epoch": 0.3416, + "grad_norm": 0.7788636684417725, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 34160 + }, + { + "epoch": 0.34161, + "grad_norm": 0.7630003690719604, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 34161 + }, + { + "epoch": 0.34162, + "grad_norm": 0.6785781383514404, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 34162 + }, + { + "epoch": 0.34163, + "grad_norm": 0.7210227251052856, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 34163 + }, + { + "epoch": 0.34164, + "grad_norm": 0.8148666024208069, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 34164 + }, + { + "epoch": 0.34165, + "grad_norm": 0.9465457797050476, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 34165 + }, + { + "epoch": 0.34166, + "grad_norm": 0.985747754573822, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 34166 + }, + { + "epoch": 0.34167, + "grad_norm": 0.8833534717559814, + "learning_rate": 0.003, + "loss": 3.9475, + "step": 34167 + }, + { + "epoch": 0.34168, + "grad_norm": 0.783381998538971, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 34168 + }, + { + "epoch": 0.34169, + "grad_norm": 0.8139287829399109, + "learning_rate": 0.003, + "loss": 3.988, + "step": 34169 + }, + { + "epoch": 0.3417, + "grad_norm": 0.9280197620391846, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 34170 + }, + { + "epoch": 0.34171, + "grad_norm": 0.9319996237754822, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 34171 + }, + { + "epoch": 0.34172, + "grad_norm": 0.8915484547615051, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 34172 + }, + { + "epoch": 0.34173, + "grad_norm": 0.9323473572731018, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 34173 + }, + { + "epoch": 0.34174, + "grad_norm": 0.8813246488571167, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 34174 + }, + { + "epoch": 0.34175, + "grad_norm": 0.9564193487167358, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 34175 + }, + { + "epoch": 0.34176, + "grad_norm": 1.1282193660736084, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 34176 + }, + { + "epoch": 0.34177, + "grad_norm": 0.8626007437705994, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 34177 + }, + { + "epoch": 0.34178, + "grad_norm": 0.8000825643539429, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 34178 + }, + { + "epoch": 0.34179, + "grad_norm": 0.7021636962890625, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 34179 + }, + { + "epoch": 0.3418, + "grad_norm": 0.8191959857940674, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 34180 + }, + { + "epoch": 0.34181, + "grad_norm": 0.9171397686004639, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 34181 + }, + { + "epoch": 0.34182, + "grad_norm": 0.9829855561256409, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 34182 + }, + { + "epoch": 0.34183, + "grad_norm": 1.119357943534851, + "learning_rate": 0.003, + "loss": 4.011, + "step": 34183 + }, + { + "epoch": 0.34184, + "grad_norm": 0.9077786803245544, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 34184 + }, + { + "epoch": 0.34185, + "grad_norm": 1.0690515041351318, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 34185 + }, + { + "epoch": 0.34186, + "grad_norm": 1.3328742980957031, + "learning_rate": 0.003, + "loss": 3.988, + "step": 34186 + }, + { + "epoch": 0.34187, + "grad_norm": 0.9734442234039307, + "learning_rate": 0.003, + "loss": 3.96, + "step": 34187 + }, + { + "epoch": 0.34188, + "grad_norm": 0.7185173034667969, + "learning_rate": 0.003, + "loss": 3.969, + "step": 34188 + }, + { + "epoch": 0.34189, + "grad_norm": 0.7175994515419006, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 34189 + }, + { + "epoch": 0.3419, + "grad_norm": 0.6788322925567627, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 34190 + }, + { + "epoch": 0.34191, + "grad_norm": 0.6265093088150024, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 34191 + }, + { + "epoch": 0.34192, + "grad_norm": 0.6863148212432861, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 34192 + }, + { + "epoch": 0.34193, + "grad_norm": 0.7594179511070251, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 34193 + }, + { + "epoch": 0.34194, + "grad_norm": 0.8980129957199097, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 34194 + }, + { + "epoch": 0.34195, + "grad_norm": 1.0402555465698242, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 34195 + }, + { + "epoch": 0.34196, + "grad_norm": 1.0952227115631104, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 34196 + }, + { + "epoch": 0.34197, + "grad_norm": 0.9061866998672485, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 34197 + }, + { + "epoch": 0.34198, + "grad_norm": 0.8177241086959839, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 34198 + }, + { + "epoch": 0.34199, + "grad_norm": 0.8574482798576355, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 34199 + }, + { + "epoch": 0.342, + "grad_norm": 0.8725963234901428, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 34200 + }, + { + "epoch": 0.34201, + "grad_norm": 0.8437882661819458, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 34201 + }, + { + "epoch": 0.34202, + "grad_norm": 0.790137767791748, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 34202 + }, + { + "epoch": 0.34203, + "grad_norm": 0.8350940346717834, + "learning_rate": 0.003, + "loss": 3.986, + "step": 34203 + }, + { + "epoch": 0.34204, + "grad_norm": 0.7655617594718933, + "learning_rate": 0.003, + "loss": 3.989, + "step": 34204 + }, + { + "epoch": 0.34205, + "grad_norm": 0.7159196138381958, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 34205 + }, + { + "epoch": 0.34206, + "grad_norm": 0.750983715057373, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 34206 + }, + { + "epoch": 0.34207, + "grad_norm": 1.0621203184127808, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 34207 + }, + { + "epoch": 0.34208, + "grad_norm": 1.328014850616455, + "learning_rate": 0.003, + "loss": 3.998, + "step": 34208 + }, + { + "epoch": 0.34209, + "grad_norm": 0.8295512199401855, + "learning_rate": 0.003, + "loss": 3.971, + "step": 34209 + }, + { + "epoch": 0.3421, + "grad_norm": 0.8635270595550537, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 34210 + }, + { + "epoch": 0.34211, + "grad_norm": 0.9793846011161804, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 34211 + }, + { + "epoch": 0.34212, + "grad_norm": 1.0982228517532349, + "learning_rate": 0.003, + "loss": 3.979, + "step": 34212 + }, + { + "epoch": 0.34213, + "grad_norm": 0.8845152258872986, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 34213 + }, + { + "epoch": 0.34214, + "grad_norm": 0.8574267029762268, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 34214 + }, + { + "epoch": 0.34215, + "grad_norm": 0.8532660603523254, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 34215 + }, + { + "epoch": 0.34216, + "grad_norm": 0.8328338265419006, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 34216 + }, + { + "epoch": 0.34217, + "grad_norm": 0.9375205039978027, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 34217 + }, + { + "epoch": 0.34218, + "grad_norm": 1.046411395072937, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 34218 + }, + { + "epoch": 0.34219, + "grad_norm": 1.04380202293396, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 34219 + }, + { + "epoch": 0.3422, + "grad_norm": 0.902998149394989, + "learning_rate": 0.003, + "loss": 3.994, + "step": 34220 + }, + { + "epoch": 0.34221, + "grad_norm": 0.8708569407463074, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 34221 + }, + { + "epoch": 0.34222, + "grad_norm": 0.8902411460876465, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 34222 + }, + { + "epoch": 0.34223, + "grad_norm": 0.8562315702438354, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 34223 + }, + { + "epoch": 0.34224, + "grad_norm": 0.9013319611549377, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 34224 + }, + { + "epoch": 0.34225, + "grad_norm": 1.230891466140747, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 34225 + }, + { + "epoch": 0.34226, + "grad_norm": 0.9944967031478882, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 34226 + }, + { + "epoch": 0.34227, + "grad_norm": 0.9619207382202148, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 34227 + }, + { + "epoch": 0.34228, + "grad_norm": 1.0050982236862183, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 34228 + }, + { + "epoch": 0.34229, + "grad_norm": 0.9742864370346069, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 34229 + }, + { + "epoch": 0.3423, + "grad_norm": 0.9092382788658142, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 34230 + }, + { + "epoch": 0.34231, + "grad_norm": 0.9373241066932678, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 34231 + }, + { + "epoch": 0.34232, + "grad_norm": 1.0337767601013184, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 34232 + }, + { + "epoch": 0.34233, + "grad_norm": 0.9450627565383911, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 34233 + }, + { + "epoch": 0.34234, + "grad_norm": 0.8511576056480408, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 34234 + }, + { + "epoch": 0.34235, + "grad_norm": 1.0138477087020874, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 34235 + }, + { + "epoch": 0.34236, + "grad_norm": 1.1718840599060059, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 34236 + }, + { + "epoch": 0.34237, + "grad_norm": 0.8172183036804199, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 34237 + }, + { + "epoch": 0.34238, + "grad_norm": 0.6715829372406006, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 34238 + }, + { + "epoch": 0.34239, + "grad_norm": 0.6590884923934937, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 34239 + }, + { + "epoch": 0.3424, + "grad_norm": 0.7275564074516296, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 34240 + }, + { + "epoch": 0.34241, + "grad_norm": 0.8427121639251709, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 34241 + }, + { + "epoch": 0.34242, + "grad_norm": 1.0139960050582886, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 34242 + }, + { + "epoch": 0.34243, + "grad_norm": 0.9779958724975586, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 34243 + }, + { + "epoch": 0.34244, + "grad_norm": 0.942302942276001, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 34244 + }, + { + "epoch": 0.34245, + "grad_norm": 0.867162823677063, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 34245 + }, + { + "epoch": 0.34246, + "grad_norm": 0.9119772911071777, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 34246 + }, + { + "epoch": 0.34247, + "grad_norm": 0.9011643528938293, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 34247 + }, + { + "epoch": 0.34248, + "grad_norm": 1.0230836868286133, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 34248 + }, + { + "epoch": 0.34249, + "grad_norm": 0.9045617580413818, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 34249 + }, + { + "epoch": 0.3425, + "grad_norm": 0.8421792387962341, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 34250 + }, + { + "epoch": 0.34251, + "grad_norm": 0.9263333082199097, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 34251 + }, + { + "epoch": 0.34252, + "grad_norm": 0.8313179612159729, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 34252 + }, + { + "epoch": 0.34253, + "grad_norm": 0.7095741629600525, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 34253 + }, + { + "epoch": 0.34254, + "grad_norm": 0.7322561740875244, + "learning_rate": 0.003, + "loss": 3.985, + "step": 34254 + }, + { + "epoch": 0.34255, + "grad_norm": 0.8417930603027344, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 34255 + }, + { + "epoch": 0.34256, + "grad_norm": 0.8689936995506287, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 34256 + }, + { + "epoch": 0.34257, + "grad_norm": 0.8632226586341858, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 34257 + }, + { + "epoch": 0.34258, + "grad_norm": 0.860907793045044, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 34258 + }, + { + "epoch": 0.34259, + "grad_norm": 0.8092222809791565, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 34259 + }, + { + "epoch": 0.3426, + "grad_norm": 0.804757297039032, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 34260 + }, + { + "epoch": 0.34261, + "grad_norm": 0.9871540069580078, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 34261 + }, + { + "epoch": 0.34262, + "grad_norm": 1.281238079071045, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 34262 + }, + { + "epoch": 0.34263, + "grad_norm": 0.7030850648880005, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 34263 + }, + { + "epoch": 0.34264, + "grad_norm": 0.6275056004524231, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 34264 + }, + { + "epoch": 0.34265, + "grad_norm": 0.6970253586769104, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 34265 + }, + { + "epoch": 0.34266, + "grad_norm": 0.8129541277885437, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 34266 + }, + { + "epoch": 0.34267, + "grad_norm": 0.9371607303619385, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 34267 + }, + { + "epoch": 0.34268, + "grad_norm": 0.9523378014564514, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 34268 + }, + { + "epoch": 0.34269, + "grad_norm": 1.0133740901947021, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 34269 + }, + { + "epoch": 0.3427, + "grad_norm": 1.160805583000183, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 34270 + }, + { + "epoch": 0.34271, + "grad_norm": 0.861980676651001, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 34271 + }, + { + "epoch": 0.34272, + "grad_norm": 0.7717623710632324, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 34272 + }, + { + "epoch": 0.34273, + "grad_norm": 0.8287400007247925, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 34273 + }, + { + "epoch": 0.34274, + "grad_norm": 0.8100370764732361, + "learning_rate": 0.003, + "loss": 3.984, + "step": 34274 + }, + { + "epoch": 0.34275, + "grad_norm": 0.844248354434967, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 34275 + }, + { + "epoch": 0.34276, + "grad_norm": 0.7871124148368835, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 34276 + }, + { + "epoch": 0.34277, + "grad_norm": 0.7978943586349487, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 34277 + }, + { + "epoch": 0.34278, + "grad_norm": 0.8138723373413086, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 34278 + }, + { + "epoch": 0.34279, + "grad_norm": 0.9614912867546082, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 34279 + }, + { + "epoch": 0.3428, + "grad_norm": 1.1488617658615112, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 34280 + }, + { + "epoch": 0.34281, + "grad_norm": 0.8698097467422485, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 34281 + }, + { + "epoch": 0.34282, + "grad_norm": 0.7636364698410034, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 34282 + }, + { + "epoch": 0.34283, + "grad_norm": 0.8037052750587463, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 34283 + }, + { + "epoch": 0.34284, + "grad_norm": 0.9579062461853027, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 34284 + }, + { + "epoch": 0.34285, + "grad_norm": 1.1118425130844116, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 34285 + }, + { + "epoch": 0.34286, + "grad_norm": 0.9773433208465576, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 34286 + }, + { + "epoch": 0.34287, + "grad_norm": 0.8363613486289978, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 34287 + }, + { + "epoch": 0.34288, + "grad_norm": 0.7689661383628845, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 34288 + }, + { + "epoch": 0.34289, + "grad_norm": 0.8211437463760376, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 34289 + }, + { + "epoch": 0.3429, + "grad_norm": 1.0476781129837036, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 34290 + }, + { + "epoch": 0.34291, + "grad_norm": 1.2165271043777466, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 34291 + }, + { + "epoch": 0.34292, + "grad_norm": 0.6685178875923157, + "learning_rate": 0.003, + "loss": 3.983, + "step": 34292 + }, + { + "epoch": 0.34293, + "grad_norm": 0.7015107870101929, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 34293 + }, + { + "epoch": 0.34294, + "grad_norm": 0.624569833278656, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 34294 + }, + { + "epoch": 0.34295, + "grad_norm": 0.7120619416236877, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 34295 + }, + { + "epoch": 0.34296, + "grad_norm": 0.8849573731422424, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 34296 + }, + { + "epoch": 0.34297, + "grad_norm": 0.9061974287033081, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 34297 + }, + { + "epoch": 0.34298, + "grad_norm": 0.8481357097625732, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 34298 + }, + { + "epoch": 0.34299, + "grad_norm": 0.7750709056854248, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 34299 + }, + { + "epoch": 0.343, + "grad_norm": 0.7226986289024353, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 34300 + }, + { + "epoch": 0.34301, + "grad_norm": 0.7868329882621765, + "learning_rate": 0.003, + "loss": 3.9411, + "step": 34301 + }, + { + "epoch": 0.34302, + "grad_norm": 1.1509714126586914, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 34302 + }, + { + "epoch": 0.34303, + "grad_norm": 1.1331701278686523, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 34303 + }, + { + "epoch": 0.34304, + "grad_norm": 0.7389916777610779, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 34304 + }, + { + "epoch": 0.34305, + "grad_norm": 0.6303780674934387, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 34305 + }, + { + "epoch": 0.34306, + "grad_norm": 0.6763666868209839, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 34306 + }, + { + "epoch": 0.34307, + "grad_norm": 0.6595414280891418, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 34307 + }, + { + "epoch": 0.34308, + "grad_norm": 0.6348878741264343, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 34308 + }, + { + "epoch": 0.34309, + "grad_norm": 0.662416398525238, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 34309 + }, + { + "epoch": 0.3431, + "grad_norm": 0.7793393731117249, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 34310 + }, + { + "epoch": 0.34311, + "grad_norm": 0.9522160291671753, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 34311 + }, + { + "epoch": 0.34312, + "grad_norm": 0.906379759311676, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 34312 + }, + { + "epoch": 0.34313, + "grad_norm": 1.0098494291305542, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 34313 + }, + { + "epoch": 0.34314, + "grad_norm": 1.186437964439392, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 34314 + }, + { + "epoch": 0.34315, + "grad_norm": 0.8889924883842468, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 34315 + }, + { + "epoch": 0.34316, + "grad_norm": 0.8927173614501953, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 34316 + }, + { + "epoch": 0.34317, + "grad_norm": 0.932362973690033, + "learning_rate": 0.003, + "loss": 3.975, + "step": 34317 + }, + { + "epoch": 0.34318, + "grad_norm": 0.986151397228241, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 34318 + }, + { + "epoch": 0.34319, + "grad_norm": 1.208954930305481, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 34319 + }, + { + "epoch": 0.3432, + "grad_norm": 0.7899855375289917, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 34320 + }, + { + "epoch": 0.34321, + "grad_norm": 0.7067559957504272, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 34321 + }, + { + "epoch": 0.34322, + "grad_norm": 0.8617085814476013, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 34322 + }, + { + "epoch": 0.34323, + "grad_norm": 0.9593177437782288, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 34323 + }, + { + "epoch": 0.34324, + "grad_norm": 1.1873856782913208, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 34324 + }, + { + "epoch": 0.34325, + "grad_norm": 0.9450768232345581, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 34325 + }, + { + "epoch": 0.34326, + "grad_norm": 0.8555667996406555, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 34326 + }, + { + "epoch": 0.34327, + "grad_norm": 0.8071604371070862, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 34327 + }, + { + "epoch": 0.34328, + "grad_norm": 0.831755518913269, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 34328 + }, + { + "epoch": 0.34329, + "grad_norm": 0.9878941774368286, + "learning_rate": 0.003, + "loss": 3.97, + "step": 34329 + }, + { + "epoch": 0.3433, + "grad_norm": 1.1253689527511597, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 34330 + }, + { + "epoch": 0.34331, + "grad_norm": 0.8893240690231323, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 34331 + }, + { + "epoch": 0.34332, + "grad_norm": 0.9150195121765137, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 34332 + }, + { + "epoch": 0.34333, + "grad_norm": 0.8569440841674805, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 34333 + }, + { + "epoch": 0.34334, + "grad_norm": 0.8696289658546448, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 34334 + }, + { + "epoch": 0.34335, + "grad_norm": 0.9248669147491455, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 34335 + }, + { + "epoch": 0.34336, + "grad_norm": 1.023245930671692, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 34336 + }, + { + "epoch": 0.34337, + "grad_norm": 1.1447917222976685, + "learning_rate": 0.003, + "loss": 4.014, + "step": 34337 + }, + { + "epoch": 0.34338, + "grad_norm": 0.9013797044754028, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 34338 + }, + { + "epoch": 0.34339, + "grad_norm": 0.9884989261627197, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 34339 + }, + { + "epoch": 0.3434, + "grad_norm": 1.1525120735168457, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 34340 + }, + { + "epoch": 0.34341, + "grad_norm": 1.0347691774368286, + "learning_rate": 0.003, + "loss": 3.971, + "step": 34341 + }, + { + "epoch": 0.34342, + "grad_norm": 0.9829759001731873, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 34342 + }, + { + "epoch": 0.34343, + "grad_norm": 1.0723460912704468, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 34343 + }, + { + "epoch": 0.34344, + "grad_norm": 1.014158010482788, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 34344 + }, + { + "epoch": 0.34345, + "grad_norm": 0.953138530254364, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 34345 + }, + { + "epoch": 0.34346, + "grad_norm": 0.900595486164093, + "learning_rate": 0.003, + "loss": 3.982, + "step": 34346 + }, + { + "epoch": 0.34347, + "grad_norm": 0.8455771803855896, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 34347 + }, + { + "epoch": 0.34348, + "grad_norm": 0.7463636994361877, + "learning_rate": 0.003, + "loss": 3.965, + "step": 34348 + }, + { + "epoch": 0.34349, + "grad_norm": 0.824893057346344, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 34349 + }, + { + "epoch": 0.3435, + "grad_norm": 0.9537352323532104, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 34350 + }, + { + "epoch": 0.34351, + "grad_norm": 0.8644811511039734, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 34351 + }, + { + "epoch": 0.34352, + "grad_norm": 0.7719224095344543, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 34352 + }, + { + "epoch": 0.34353, + "grad_norm": 0.7548385858535767, + "learning_rate": 0.003, + "loss": 3.985, + "step": 34353 + }, + { + "epoch": 0.34354, + "grad_norm": 0.916477382183075, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 34354 + }, + { + "epoch": 0.34355, + "grad_norm": 1.2151676416397095, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 34355 + }, + { + "epoch": 0.34356, + "grad_norm": 0.9826456904411316, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 34356 + }, + { + "epoch": 0.34357, + "grad_norm": 0.9372683167457581, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 34357 + }, + { + "epoch": 0.34358, + "grad_norm": 0.8799492716789246, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 34358 + }, + { + "epoch": 0.34359, + "grad_norm": 0.8712049126625061, + "learning_rate": 0.003, + "loss": 4.0, + "step": 34359 + }, + { + "epoch": 0.3436, + "grad_norm": 0.9115896821022034, + "learning_rate": 0.003, + "loss": 3.982, + "step": 34360 + }, + { + "epoch": 0.34361, + "grad_norm": 0.8985487818717957, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 34361 + }, + { + "epoch": 0.34362, + "grad_norm": 0.9438971281051636, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 34362 + }, + { + "epoch": 0.34363, + "grad_norm": 0.9662455916404724, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 34363 + }, + { + "epoch": 0.34364, + "grad_norm": 0.8251842260360718, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 34364 + }, + { + "epoch": 0.34365, + "grad_norm": 0.7538000345230103, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 34365 + }, + { + "epoch": 0.34366, + "grad_norm": 0.8044688105583191, + "learning_rate": 0.003, + "loss": 3.968, + "step": 34366 + }, + { + "epoch": 0.34367, + "grad_norm": 0.8307697772979736, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 34367 + }, + { + "epoch": 0.34368, + "grad_norm": 1.0279371738433838, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 34368 + }, + { + "epoch": 0.34369, + "grad_norm": 1.3319814205169678, + "learning_rate": 0.003, + "loss": 3.991, + "step": 34369 + }, + { + "epoch": 0.3437, + "grad_norm": 0.5956098437309265, + "learning_rate": 0.003, + "loss": 3.971, + "step": 34370 + }, + { + "epoch": 0.34371, + "grad_norm": 0.7911988496780396, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 34371 + }, + { + "epoch": 0.34372, + "grad_norm": 1.1669288873672485, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 34372 + }, + { + "epoch": 0.34373, + "grad_norm": 0.7752838730812073, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 34373 + }, + { + "epoch": 0.34374, + "grad_norm": 0.6304289698600769, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 34374 + }, + { + "epoch": 0.34375, + "grad_norm": 0.6412551403045654, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 34375 + }, + { + "epoch": 0.34376, + "grad_norm": 0.6860687136650085, + "learning_rate": 0.003, + "loss": 3.9455, + "step": 34376 + }, + { + "epoch": 0.34377, + "grad_norm": 0.8681167364120483, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 34377 + }, + { + "epoch": 0.34378, + "grad_norm": 1.1393414735794067, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 34378 + }, + { + "epoch": 0.34379, + "grad_norm": 1.0822395086288452, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 34379 + }, + { + "epoch": 0.3438, + "grad_norm": 0.8067560791969299, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 34380 + }, + { + "epoch": 0.34381, + "grad_norm": 0.7107057571411133, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 34381 + }, + { + "epoch": 0.34382, + "grad_norm": 0.7646576166152954, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 34382 + }, + { + "epoch": 0.34383, + "grad_norm": 0.8290819525718689, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 34383 + }, + { + "epoch": 0.34384, + "grad_norm": 0.7732831835746765, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 34384 + }, + { + "epoch": 0.34385, + "grad_norm": 0.7956564426422119, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 34385 + }, + { + "epoch": 0.34386, + "grad_norm": 0.8159328699111938, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 34386 + }, + { + "epoch": 0.34387, + "grad_norm": 0.9802791476249695, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 34387 + }, + { + "epoch": 0.34388, + "grad_norm": 1.3512184619903564, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 34388 + }, + { + "epoch": 0.34389, + "grad_norm": 0.8060258626937866, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 34389 + }, + { + "epoch": 0.3439, + "grad_norm": 0.7522668242454529, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 34390 + }, + { + "epoch": 0.34391, + "grad_norm": 0.8172879219055176, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 34391 + }, + { + "epoch": 0.34392, + "grad_norm": 1.1101844310760498, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 34392 + }, + { + "epoch": 0.34393, + "grad_norm": 1.1796367168426514, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 34393 + }, + { + "epoch": 0.34394, + "grad_norm": 0.625688374042511, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 34394 + }, + { + "epoch": 0.34395, + "grad_norm": 0.7301909923553467, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 34395 + }, + { + "epoch": 0.34396, + "grad_norm": 0.9407071471214294, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 34396 + }, + { + "epoch": 0.34397, + "grad_norm": 0.9482941031455994, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 34397 + }, + { + "epoch": 0.34398, + "grad_norm": 0.9659186005592346, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 34398 + }, + { + "epoch": 0.34399, + "grad_norm": 0.9854285717010498, + "learning_rate": 0.003, + "loss": 4.006, + "step": 34399 + }, + { + "epoch": 0.344, + "grad_norm": 0.9947935342788696, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 34400 + }, + { + "epoch": 0.34401, + "grad_norm": 1.07062828540802, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 34401 + }, + { + "epoch": 0.34402, + "grad_norm": 0.9249527454376221, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 34402 + }, + { + "epoch": 0.34403, + "grad_norm": 0.8886224627494812, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 34403 + }, + { + "epoch": 0.34404, + "grad_norm": 1.0243202447891235, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 34404 + }, + { + "epoch": 0.34405, + "grad_norm": 1.050980806350708, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 34405 + }, + { + "epoch": 0.34406, + "grad_norm": 0.9272882342338562, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 34406 + }, + { + "epoch": 0.34407, + "grad_norm": 0.8799100518226624, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 34407 + }, + { + "epoch": 0.34408, + "grad_norm": 0.9561105966567993, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 34408 + }, + { + "epoch": 0.34409, + "grad_norm": 0.8370293974876404, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 34409 + }, + { + "epoch": 0.3441, + "grad_norm": 0.6935255527496338, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 34410 + }, + { + "epoch": 0.34411, + "grad_norm": 0.7730186581611633, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 34411 + }, + { + "epoch": 0.34412, + "grad_norm": 0.8639588356018066, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 34412 + }, + { + "epoch": 0.34413, + "grad_norm": 0.9349215626716614, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 34413 + }, + { + "epoch": 0.34414, + "grad_norm": 0.8199653029441833, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 34414 + }, + { + "epoch": 0.34415, + "grad_norm": 0.7122595310211182, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 34415 + }, + { + "epoch": 0.34416, + "grad_norm": 0.7267784476280212, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 34416 + }, + { + "epoch": 0.34417, + "grad_norm": 0.6175834536552429, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 34417 + }, + { + "epoch": 0.34418, + "grad_norm": 0.6361780762672424, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 34418 + }, + { + "epoch": 0.34419, + "grad_norm": 0.6631248593330383, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 34419 + }, + { + "epoch": 0.3442, + "grad_norm": 0.8780115246772766, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 34420 + }, + { + "epoch": 0.34421, + "grad_norm": 1.3820009231567383, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 34421 + }, + { + "epoch": 0.34422, + "grad_norm": 0.7486633658409119, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 34422 + }, + { + "epoch": 0.34423, + "grad_norm": 0.6570396423339844, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 34423 + }, + { + "epoch": 0.34424, + "grad_norm": 0.7380676865577698, + "learning_rate": 0.003, + "loss": 3.972, + "step": 34424 + }, + { + "epoch": 0.34425, + "grad_norm": 0.6857942938804626, + "learning_rate": 0.003, + "loss": 3.972, + "step": 34425 + }, + { + "epoch": 0.34426, + "grad_norm": 0.7123207449913025, + "learning_rate": 0.003, + "loss": 3.991, + "step": 34426 + }, + { + "epoch": 0.34427, + "grad_norm": 0.8283118605613708, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 34427 + }, + { + "epoch": 0.34428, + "grad_norm": 1.163153886795044, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 34428 + }, + { + "epoch": 0.34429, + "grad_norm": 1.0862849950790405, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 34429 + }, + { + "epoch": 0.3443, + "grad_norm": 1.0124882459640503, + "learning_rate": 0.003, + "loss": 3.98, + "step": 34430 + }, + { + "epoch": 0.34431, + "grad_norm": 1.034604787826538, + "learning_rate": 0.003, + "loss": 3.958, + "step": 34431 + }, + { + "epoch": 0.34432, + "grad_norm": 0.8376259803771973, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 34432 + }, + { + "epoch": 0.34433, + "grad_norm": 0.7522556781768799, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 34433 + }, + { + "epoch": 0.34434, + "grad_norm": 0.7196252942085266, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 34434 + }, + { + "epoch": 0.34435, + "grad_norm": 0.8561823964118958, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 34435 + }, + { + "epoch": 0.34436, + "grad_norm": 0.9662783145904541, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 34436 + }, + { + "epoch": 0.34437, + "grad_norm": 1.0292900800704956, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 34437 + }, + { + "epoch": 0.34438, + "grad_norm": 1.0245946645736694, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 34438 + }, + { + "epoch": 0.34439, + "grad_norm": 0.7975664734840393, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 34439 + }, + { + "epoch": 0.3444, + "grad_norm": 0.669578492641449, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 34440 + }, + { + "epoch": 0.34441, + "grad_norm": 0.6514747142791748, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 34441 + }, + { + "epoch": 0.34442, + "grad_norm": 0.6387531757354736, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 34442 + }, + { + "epoch": 0.34443, + "grad_norm": 0.6593682169914246, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 34443 + }, + { + "epoch": 0.34444, + "grad_norm": 0.8254755735397339, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 34444 + }, + { + "epoch": 0.34445, + "grad_norm": 1.0665037631988525, + "learning_rate": 0.003, + "loss": 3.971, + "step": 34445 + }, + { + "epoch": 0.34446, + "grad_norm": 1.1502759456634521, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 34446 + }, + { + "epoch": 0.34447, + "grad_norm": 0.7132957577705383, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 34447 + }, + { + "epoch": 0.34448, + "grad_norm": 0.6325023770332336, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 34448 + }, + { + "epoch": 0.34449, + "grad_norm": 0.7525457143783569, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 34449 + }, + { + "epoch": 0.3445, + "grad_norm": 0.7644877433776855, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 34450 + }, + { + "epoch": 0.34451, + "grad_norm": 0.8735901713371277, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 34451 + }, + { + "epoch": 0.34452, + "grad_norm": 0.936516284942627, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 34452 + }, + { + "epoch": 0.34453, + "grad_norm": 0.8651028275489807, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 34453 + }, + { + "epoch": 0.34454, + "grad_norm": 0.9018081426620483, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 34454 + }, + { + "epoch": 0.34455, + "grad_norm": 0.9137197732925415, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 34455 + }, + { + "epoch": 0.34456, + "grad_norm": 0.8475934267044067, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 34456 + }, + { + "epoch": 0.34457, + "grad_norm": 0.7272558212280273, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 34457 + }, + { + "epoch": 0.34458, + "grad_norm": 0.8282168507575989, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 34458 + }, + { + "epoch": 0.34459, + "grad_norm": 1.0992540121078491, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 34459 + }, + { + "epoch": 0.3446, + "grad_norm": 1.0852106809616089, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 34460 + }, + { + "epoch": 0.34461, + "grad_norm": 1.0057660341262817, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 34461 + }, + { + "epoch": 0.34462, + "grad_norm": 0.9521392583847046, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 34462 + }, + { + "epoch": 0.34463, + "grad_norm": 1.2223420143127441, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 34463 + }, + { + "epoch": 0.34464, + "grad_norm": 0.8787345886230469, + "learning_rate": 0.003, + "loss": 4.001, + "step": 34464 + }, + { + "epoch": 0.34465, + "grad_norm": 0.9604967832565308, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 34465 + }, + { + "epoch": 0.34466, + "grad_norm": 1.3612874746322632, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 34466 + }, + { + "epoch": 0.34467, + "grad_norm": 0.7810909748077393, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 34467 + }, + { + "epoch": 0.34468, + "grad_norm": 0.7170125842094421, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 34468 + }, + { + "epoch": 0.34469, + "grad_norm": 0.6436113715171814, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 34469 + }, + { + "epoch": 0.3447, + "grad_norm": 0.8424472212791443, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 34470 + }, + { + "epoch": 0.34471, + "grad_norm": 1.0893906354904175, + "learning_rate": 0.003, + "loss": 3.983, + "step": 34471 + }, + { + "epoch": 0.34472, + "grad_norm": 0.9889529347419739, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 34472 + }, + { + "epoch": 0.34473, + "grad_norm": 1.0419955253601074, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 34473 + }, + { + "epoch": 0.34474, + "grad_norm": 0.9118795990943909, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 34474 + }, + { + "epoch": 0.34475, + "grad_norm": 0.8253859877586365, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 34475 + }, + { + "epoch": 0.34476, + "grad_norm": 0.7016973495483398, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 34476 + }, + { + "epoch": 0.34477, + "grad_norm": 0.8176684379577637, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 34477 + }, + { + "epoch": 0.34478, + "grad_norm": 1.07759428024292, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 34478 + }, + { + "epoch": 0.34479, + "grad_norm": 1.0707165002822876, + "learning_rate": 0.003, + "loss": 3.9414, + "step": 34479 + }, + { + "epoch": 0.3448, + "grad_norm": 0.8104775547981262, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 34480 + }, + { + "epoch": 0.34481, + "grad_norm": 0.7134526968002319, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 34481 + }, + { + "epoch": 0.34482, + "grad_norm": 0.7433558702468872, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 34482 + }, + { + "epoch": 0.34483, + "grad_norm": 0.7204609513282776, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 34483 + }, + { + "epoch": 0.34484, + "grad_norm": 0.8385145664215088, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 34484 + }, + { + "epoch": 0.34485, + "grad_norm": 0.9661549925804138, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 34485 + }, + { + "epoch": 0.34486, + "grad_norm": 1.0355523824691772, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 34486 + }, + { + "epoch": 0.34487, + "grad_norm": 1.0887491703033447, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 34487 + }, + { + "epoch": 0.34488, + "grad_norm": 1.0527056455612183, + "learning_rate": 0.003, + "loss": 3.9354, + "step": 34488 + }, + { + "epoch": 0.34489, + "grad_norm": 0.9629423022270203, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 34489 + }, + { + "epoch": 0.3449, + "grad_norm": 1.097220778465271, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 34490 + }, + { + "epoch": 0.34491, + "grad_norm": 1.113814353942871, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 34491 + }, + { + "epoch": 0.34492, + "grad_norm": 0.7578682899475098, + "learning_rate": 0.003, + "loss": 3.983, + "step": 34492 + }, + { + "epoch": 0.34493, + "grad_norm": 0.5520257949829102, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 34493 + }, + { + "epoch": 0.34494, + "grad_norm": 0.5587388277053833, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 34494 + }, + { + "epoch": 0.34495, + "grad_norm": 0.6209954023361206, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 34495 + }, + { + "epoch": 0.34496, + "grad_norm": 0.6858199238777161, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 34496 + }, + { + "epoch": 0.34497, + "grad_norm": 0.7349057197570801, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 34497 + }, + { + "epoch": 0.34498, + "grad_norm": 0.8230843544006348, + "learning_rate": 0.003, + "loss": 3.985, + "step": 34498 + }, + { + "epoch": 0.34499, + "grad_norm": 0.9732552766799927, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 34499 + }, + { + "epoch": 0.345, + "grad_norm": 1.0289571285247803, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 34500 + }, + { + "epoch": 0.34501, + "grad_norm": 1.0720300674438477, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 34501 + }, + { + "epoch": 0.34502, + "grad_norm": 0.9806328415870667, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 34502 + }, + { + "epoch": 0.34503, + "grad_norm": 0.9805777072906494, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 34503 + }, + { + "epoch": 0.34504, + "grad_norm": 1.1137213706970215, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 34504 + }, + { + "epoch": 0.34505, + "grad_norm": 0.8177998661994934, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 34505 + }, + { + "epoch": 0.34506, + "grad_norm": 0.8119290471076965, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 34506 + }, + { + "epoch": 0.34507, + "grad_norm": 0.8488447070121765, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 34507 + }, + { + "epoch": 0.34508, + "grad_norm": 0.9509136080741882, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 34508 + }, + { + "epoch": 0.34509, + "grad_norm": 0.983249843120575, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 34509 + }, + { + "epoch": 0.3451, + "grad_norm": 0.9573912620544434, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 34510 + }, + { + "epoch": 0.34511, + "grad_norm": 1.0759108066558838, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 34511 + }, + { + "epoch": 0.34512, + "grad_norm": 1.1517471075057983, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 34512 + }, + { + "epoch": 0.34513, + "grad_norm": 0.795722484588623, + "learning_rate": 0.003, + "loss": 3.996, + "step": 34513 + }, + { + "epoch": 0.34514, + "grad_norm": 0.8077188730239868, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 34514 + }, + { + "epoch": 0.34515, + "grad_norm": 0.8061200976371765, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 34515 + }, + { + "epoch": 0.34516, + "grad_norm": 0.8077476024627686, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 34516 + }, + { + "epoch": 0.34517, + "grad_norm": 0.8489881753921509, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 34517 + }, + { + "epoch": 0.34518, + "grad_norm": 1.0272445678710938, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 34518 + }, + { + "epoch": 0.34519, + "grad_norm": 1.1530386209487915, + "learning_rate": 0.003, + "loss": 3.972, + "step": 34519 + }, + { + "epoch": 0.3452, + "grad_norm": 0.8227440714836121, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 34520 + }, + { + "epoch": 0.34521, + "grad_norm": 0.8685489892959595, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 34521 + }, + { + "epoch": 0.34522, + "grad_norm": 0.8506094217300415, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 34522 + }, + { + "epoch": 0.34523, + "grad_norm": 0.9393786191940308, + "learning_rate": 0.003, + "loss": 4.003, + "step": 34523 + }, + { + "epoch": 0.34524, + "grad_norm": 0.9480283856391907, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 34524 + }, + { + "epoch": 0.34525, + "grad_norm": 1.0016570091247559, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 34525 + }, + { + "epoch": 0.34526, + "grad_norm": 1.0979641675949097, + "learning_rate": 0.003, + "loss": 3.984, + "step": 34526 + }, + { + "epoch": 0.34527, + "grad_norm": 0.8434511423110962, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 34527 + }, + { + "epoch": 0.34528, + "grad_norm": 0.7781137228012085, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 34528 + }, + { + "epoch": 0.34529, + "grad_norm": 0.9217750430107117, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 34529 + }, + { + "epoch": 0.3453, + "grad_norm": 1.2086138725280762, + "learning_rate": 0.003, + "loss": 3.988, + "step": 34530 + }, + { + "epoch": 0.34531, + "grad_norm": 0.9850289821624756, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 34531 + }, + { + "epoch": 0.34532, + "grad_norm": 0.9376546740531921, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 34532 + }, + { + "epoch": 0.34533, + "grad_norm": 1.0355430841445923, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 34533 + }, + { + "epoch": 0.34534, + "grad_norm": 0.924704372882843, + "learning_rate": 0.003, + "loss": 3.994, + "step": 34534 + }, + { + "epoch": 0.34535, + "grad_norm": 0.912506103515625, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 34535 + }, + { + "epoch": 0.34536, + "grad_norm": 0.9543155431747437, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 34536 + }, + { + "epoch": 0.34537, + "grad_norm": 0.8415480256080627, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 34537 + }, + { + "epoch": 0.34538, + "grad_norm": 0.9245783686637878, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 34538 + }, + { + "epoch": 0.34539, + "grad_norm": 1.0497558116912842, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 34539 + }, + { + "epoch": 0.3454, + "grad_norm": 1.0922006368637085, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 34540 + }, + { + "epoch": 0.34541, + "grad_norm": 1.0089988708496094, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 34541 + }, + { + "epoch": 0.34542, + "grad_norm": 1.0226999521255493, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 34542 + }, + { + "epoch": 0.34543, + "grad_norm": 1.0125421285629272, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 34543 + }, + { + "epoch": 0.34544, + "grad_norm": 0.98758864402771, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 34544 + }, + { + "epoch": 0.34545, + "grad_norm": 0.9669977426528931, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 34545 + }, + { + "epoch": 0.34546, + "grad_norm": 1.089102029800415, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 34546 + }, + { + "epoch": 0.34547, + "grad_norm": 0.8850140571594238, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 34547 + }, + { + "epoch": 0.34548, + "grad_norm": 1.143643856048584, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 34548 + }, + { + "epoch": 0.34549, + "grad_norm": 0.9094265103340149, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 34549 + }, + { + "epoch": 0.3455, + "grad_norm": 0.7876793146133423, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 34550 + }, + { + "epoch": 0.34551, + "grad_norm": 0.9195610284805298, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 34551 + }, + { + "epoch": 0.34552, + "grad_norm": 1.095802664756775, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 34552 + }, + { + "epoch": 0.34553, + "grad_norm": 0.8278201818466187, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 34553 + }, + { + "epoch": 0.34554, + "grad_norm": 0.7982791662216187, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 34554 + }, + { + "epoch": 0.34555, + "grad_norm": 0.7834436297416687, + "learning_rate": 0.003, + "loss": 4.007, + "step": 34555 + }, + { + "epoch": 0.34556, + "grad_norm": 0.7451352477073669, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 34556 + }, + { + "epoch": 0.34557, + "grad_norm": 0.8601987957954407, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 34557 + }, + { + "epoch": 0.34558, + "grad_norm": 0.8229082226753235, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 34558 + }, + { + "epoch": 0.34559, + "grad_norm": 0.7516164183616638, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 34559 + }, + { + "epoch": 0.3456, + "grad_norm": 0.7805548310279846, + "learning_rate": 0.003, + "loss": 3.973, + "step": 34560 + }, + { + "epoch": 0.34561, + "grad_norm": 0.9100666642189026, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 34561 + }, + { + "epoch": 0.34562, + "grad_norm": 0.8461819291114807, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 34562 + }, + { + "epoch": 0.34563, + "grad_norm": 0.7850950360298157, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 34563 + }, + { + "epoch": 0.34564, + "grad_norm": 0.8370240926742554, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 34564 + }, + { + "epoch": 0.34565, + "grad_norm": 0.9781057834625244, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 34565 + }, + { + "epoch": 0.34566, + "grad_norm": 1.0905003547668457, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 34566 + }, + { + "epoch": 0.34567, + "grad_norm": 0.8160979747772217, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 34567 + }, + { + "epoch": 0.34568, + "grad_norm": 0.8652650117874146, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 34568 + }, + { + "epoch": 0.34569, + "grad_norm": 1.0481053590774536, + "learning_rate": 0.003, + "loss": 3.982, + "step": 34569 + }, + { + "epoch": 0.3457, + "grad_norm": 1.0459922552108765, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 34570 + }, + { + "epoch": 0.34571, + "grad_norm": 0.8122027516365051, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 34571 + }, + { + "epoch": 0.34572, + "grad_norm": 0.5959889888763428, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 34572 + }, + { + "epoch": 0.34573, + "grad_norm": 0.5671717524528503, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 34573 + }, + { + "epoch": 0.34574, + "grad_norm": 0.5688528418540955, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 34574 + }, + { + "epoch": 0.34575, + "grad_norm": 0.5231848359107971, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 34575 + }, + { + "epoch": 0.34576, + "grad_norm": 0.49089986085891724, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 34576 + }, + { + "epoch": 0.34577, + "grad_norm": 0.6082242727279663, + "learning_rate": 0.003, + "loss": 3.962, + "step": 34577 + }, + { + "epoch": 0.34578, + "grad_norm": 0.6347395181655884, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 34578 + }, + { + "epoch": 0.34579, + "grad_norm": 0.6720582842826843, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 34579 + }, + { + "epoch": 0.3458, + "grad_norm": 0.9067041873931885, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 34580 + }, + { + "epoch": 0.34581, + "grad_norm": 1.3715734481811523, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 34581 + }, + { + "epoch": 0.34582, + "grad_norm": 0.7527507543563843, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 34582 + }, + { + "epoch": 0.34583, + "grad_norm": 0.6782777905464172, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 34583 + }, + { + "epoch": 0.34584, + "grad_norm": 0.7168471217155457, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 34584 + }, + { + "epoch": 0.34585, + "grad_norm": 0.8617438077926636, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 34585 + }, + { + "epoch": 0.34586, + "grad_norm": 1.1623811721801758, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 34586 + }, + { + "epoch": 0.34587, + "grad_norm": 0.9812150001525879, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 34587 + }, + { + "epoch": 0.34588, + "grad_norm": 0.8236289620399475, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 34588 + }, + { + "epoch": 0.34589, + "grad_norm": 0.8256896734237671, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 34589 + }, + { + "epoch": 0.3459, + "grad_norm": 0.8297581672668457, + "learning_rate": 0.003, + "loss": 3.987, + "step": 34590 + }, + { + "epoch": 0.34591, + "grad_norm": 1.0744749307632446, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 34591 + }, + { + "epoch": 0.34592, + "grad_norm": 1.0968865156173706, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 34592 + }, + { + "epoch": 0.34593, + "grad_norm": 0.8772927522659302, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 34593 + }, + { + "epoch": 0.34594, + "grad_norm": 0.7438201904296875, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 34594 + }, + { + "epoch": 0.34595, + "grad_norm": 0.9082021117210388, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 34595 + }, + { + "epoch": 0.34596, + "grad_norm": 1.023726463317871, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 34596 + }, + { + "epoch": 0.34597, + "grad_norm": 0.839100182056427, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 34597 + }, + { + "epoch": 0.34598, + "grad_norm": 0.8786401748657227, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 34598 + }, + { + "epoch": 0.34599, + "grad_norm": 1.012353539466858, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 34599 + }, + { + "epoch": 0.346, + "grad_norm": 0.9489909410476685, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 34600 + }, + { + "epoch": 0.34601, + "grad_norm": 1.152261734008789, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 34601 + }, + { + "epoch": 0.34602, + "grad_norm": 1.2007235288619995, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 34602 + }, + { + "epoch": 0.34603, + "grad_norm": 0.9272465705871582, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 34603 + }, + { + "epoch": 0.34604, + "grad_norm": 0.7605893015861511, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 34604 + }, + { + "epoch": 0.34605, + "grad_norm": 0.703569233417511, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 34605 + }, + { + "epoch": 0.34606, + "grad_norm": 0.6624396443367004, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 34606 + }, + { + "epoch": 0.34607, + "grad_norm": 0.8656072020530701, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 34607 + }, + { + "epoch": 0.34608, + "grad_norm": 0.9198071360588074, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 34608 + }, + { + "epoch": 0.34609, + "grad_norm": 1.1892788410186768, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 34609 + }, + { + "epoch": 0.3461, + "grad_norm": 1.054903507232666, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 34610 + }, + { + "epoch": 0.34611, + "grad_norm": 1.2648448944091797, + "learning_rate": 0.003, + "loss": 4.006, + "step": 34611 + }, + { + "epoch": 0.34612, + "grad_norm": 0.748464822769165, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 34612 + }, + { + "epoch": 0.34613, + "grad_norm": 0.8294523358345032, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 34613 + }, + { + "epoch": 0.34614, + "grad_norm": 0.867165207862854, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 34614 + }, + { + "epoch": 0.34615, + "grad_norm": 1.0204360485076904, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 34615 + }, + { + "epoch": 0.34616, + "grad_norm": 1.3317233324050903, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 34616 + }, + { + "epoch": 0.34617, + "grad_norm": 0.731116533279419, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 34617 + }, + { + "epoch": 0.34618, + "grad_norm": 0.6385423541069031, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 34618 + }, + { + "epoch": 0.34619, + "grad_norm": 0.7658708691596985, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 34619 + }, + { + "epoch": 0.3462, + "grad_norm": 0.940577507019043, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 34620 + }, + { + "epoch": 0.34621, + "grad_norm": 0.9888773560523987, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 34621 + }, + { + "epoch": 0.34622, + "grad_norm": 0.9622052907943726, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 34622 + }, + { + "epoch": 0.34623, + "grad_norm": 0.8053810596466064, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 34623 + }, + { + "epoch": 0.34624, + "grad_norm": 0.6432602405548096, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 34624 + }, + { + "epoch": 0.34625, + "grad_norm": 0.729764997959137, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 34625 + }, + { + "epoch": 0.34626, + "grad_norm": 0.8935821652412415, + "learning_rate": 0.003, + "loss": 4.01, + "step": 34626 + }, + { + "epoch": 0.34627, + "grad_norm": 0.9842808842658997, + "learning_rate": 0.003, + "loss": 3.9469, + "step": 34627 + }, + { + "epoch": 0.34628, + "grad_norm": 1.1456079483032227, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 34628 + }, + { + "epoch": 0.34629, + "grad_norm": 0.696544885635376, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 34629 + }, + { + "epoch": 0.3463, + "grad_norm": 0.6371253132820129, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 34630 + }, + { + "epoch": 0.34631, + "grad_norm": 0.863773763179779, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 34631 + }, + { + "epoch": 0.34632, + "grad_norm": 0.9388877749443054, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 34632 + }, + { + "epoch": 0.34633, + "grad_norm": 1.0254420042037964, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 34633 + }, + { + "epoch": 0.34634, + "grad_norm": 1.0144882202148438, + "learning_rate": 0.003, + "loss": 3.973, + "step": 34634 + }, + { + "epoch": 0.34635, + "grad_norm": 0.9281988143920898, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 34635 + }, + { + "epoch": 0.34636, + "grad_norm": 0.8840630054473877, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 34636 + }, + { + "epoch": 0.34637, + "grad_norm": 0.9304862022399902, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 34637 + }, + { + "epoch": 0.34638, + "grad_norm": 0.974524199962616, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 34638 + }, + { + "epoch": 0.34639, + "grad_norm": 1.037540316581726, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 34639 + }, + { + "epoch": 0.3464, + "grad_norm": 1.2848291397094727, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 34640 + }, + { + "epoch": 0.34641, + "grad_norm": 0.7880805134773254, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 34641 + }, + { + "epoch": 0.34642, + "grad_norm": 0.7850456833839417, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 34642 + }, + { + "epoch": 0.34643, + "grad_norm": 0.8672191500663757, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 34643 + }, + { + "epoch": 0.34644, + "grad_norm": 0.8677017688751221, + "learning_rate": 0.003, + "loss": 3.981, + "step": 34644 + }, + { + "epoch": 0.34645, + "grad_norm": 0.9300366640090942, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 34645 + }, + { + "epoch": 0.34646, + "grad_norm": 0.927908718585968, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 34646 + }, + { + "epoch": 0.34647, + "grad_norm": 0.7990273833274841, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 34647 + }, + { + "epoch": 0.34648, + "grad_norm": 0.917091965675354, + "learning_rate": 0.003, + "loss": 4.019, + "step": 34648 + }, + { + "epoch": 0.34649, + "grad_norm": 1.2287163734436035, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 34649 + }, + { + "epoch": 0.3465, + "grad_norm": 0.8632556200027466, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 34650 + }, + { + "epoch": 0.34651, + "grad_norm": 0.8783901929855347, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 34651 + }, + { + "epoch": 0.34652, + "grad_norm": 1.1072827577590942, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 34652 + }, + { + "epoch": 0.34653, + "grad_norm": 1.1332322359085083, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 34653 + }, + { + "epoch": 0.34654, + "grad_norm": 0.8074219226837158, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 34654 + }, + { + "epoch": 0.34655, + "grad_norm": 0.6618491411209106, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 34655 + }, + { + "epoch": 0.34656, + "grad_norm": 0.7383451461791992, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 34656 + }, + { + "epoch": 0.34657, + "grad_norm": 1.0867133140563965, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 34657 + }, + { + "epoch": 0.34658, + "grad_norm": 1.15659761428833, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 34658 + }, + { + "epoch": 0.34659, + "grad_norm": 0.7763815522193909, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 34659 + }, + { + "epoch": 0.3466, + "grad_norm": 0.6602960824966431, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 34660 + }, + { + "epoch": 0.34661, + "grad_norm": 0.6431627869606018, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 34661 + }, + { + "epoch": 0.34662, + "grad_norm": 0.6322453618049622, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 34662 + }, + { + "epoch": 0.34663, + "grad_norm": 0.6528359055519104, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 34663 + }, + { + "epoch": 0.34664, + "grad_norm": 0.6587511897087097, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 34664 + }, + { + "epoch": 0.34665, + "grad_norm": 0.7010579109191895, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 34665 + }, + { + "epoch": 0.34666, + "grad_norm": 0.6564729809761047, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 34666 + }, + { + "epoch": 0.34667, + "grad_norm": 0.7252622246742249, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 34667 + }, + { + "epoch": 0.34668, + "grad_norm": 0.7854233980178833, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 34668 + }, + { + "epoch": 0.34669, + "grad_norm": 1.042349934577942, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 34669 + }, + { + "epoch": 0.3467, + "grad_norm": 1.2474019527435303, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 34670 + }, + { + "epoch": 0.34671, + "grad_norm": 0.7257794737815857, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 34671 + }, + { + "epoch": 0.34672, + "grad_norm": 0.6484960317611694, + "learning_rate": 0.003, + "loss": 3.975, + "step": 34672 + }, + { + "epoch": 0.34673, + "grad_norm": 0.7397169470787048, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 34673 + }, + { + "epoch": 0.34674, + "grad_norm": 0.825655460357666, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 34674 + }, + { + "epoch": 0.34675, + "grad_norm": 1.1424901485443115, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 34675 + }, + { + "epoch": 0.34676, + "grad_norm": 1.0126203298568726, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 34676 + }, + { + "epoch": 0.34677, + "grad_norm": 0.9104232788085938, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 34677 + }, + { + "epoch": 0.34678, + "grad_norm": 0.7487277388572693, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 34678 + }, + { + "epoch": 0.34679, + "grad_norm": 0.7817580699920654, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 34679 + }, + { + "epoch": 0.3468, + "grad_norm": 0.7594075798988342, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 34680 + }, + { + "epoch": 0.34681, + "grad_norm": 0.8098450899124146, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 34681 + }, + { + "epoch": 0.34682, + "grad_norm": 1.0963878631591797, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 34682 + }, + { + "epoch": 0.34683, + "grad_norm": 0.9994804859161377, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 34683 + }, + { + "epoch": 0.34684, + "grad_norm": 0.8709931373596191, + "learning_rate": 0.003, + "loss": 3.987, + "step": 34684 + }, + { + "epoch": 0.34685, + "grad_norm": 0.7854844927787781, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 34685 + }, + { + "epoch": 0.34686, + "grad_norm": 0.8639016151428223, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 34686 + }, + { + "epoch": 0.34687, + "grad_norm": 1.0042988061904907, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 34687 + }, + { + "epoch": 0.34688, + "grad_norm": 1.1669775247573853, + "learning_rate": 0.003, + "loss": 4.0, + "step": 34688 + }, + { + "epoch": 0.34689, + "grad_norm": 0.8320513963699341, + "learning_rate": 0.003, + "loss": 3.96, + "step": 34689 + }, + { + "epoch": 0.3469, + "grad_norm": 0.8434905409812927, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 34690 + }, + { + "epoch": 0.34691, + "grad_norm": 0.9298394918441772, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 34691 + }, + { + "epoch": 0.34692, + "grad_norm": 1.0933945178985596, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 34692 + }, + { + "epoch": 0.34693, + "grad_norm": 0.9100098013877869, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 34693 + }, + { + "epoch": 0.34694, + "grad_norm": 0.8380478024482727, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 34694 + }, + { + "epoch": 0.34695, + "grad_norm": 0.838750422000885, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 34695 + }, + { + "epoch": 0.34696, + "grad_norm": 0.8780836462974548, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 34696 + }, + { + "epoch": 0.34697, + "grad_norm": 1.1322368383407593, + "learning_rate": 0.003, + "loss": 3.9367, + "step": 34697 + }, + { + "epoch": 0.34698, + "grad_norm": 0.9764858484268188, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 34698 + }, + { + "epoch": 0.34699, + "grad_norm": 0.8216850161552429, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 34699 + }, + { + "epoch": 0.347, + "grad_norm": 0.7840567231178284, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 34700 + }, + { + "epoch": 0.34701, + "grad_norm": 0.8956513404846191, + "learning_rate": 0.003, + "loss": 3.975, + "step": 34701 + }, + { + "epoch": 0.34702, + "grad_norm": 0.9380198121070862, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 34702 + }, + { + "epoch": 0.34703, + "grad_norm": 0.8902056217193604, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 34703 + }, + { + "epoch": 0.34704, + "grad_norm": 0.9058194160461426, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 34704 + }, + { + "epoch": 0.34705, + "grad_norm": 0.9675326943397522, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 34705 + }, + { + "epoch": 0.34706, + "grad_norm": 1.1063103675842285, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 34706 + }, + { + "epoch": 0.34707, + "grad_norm": 0.8269145488739014, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 34707 + }, + { + "epoch": 0.34708, + "grad_norm": 0.85235995054245, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 34708 + }, + { + "epoch": 0.34709, + "grad_norm": 0.9384817481040955, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 34709 + }, + { + "epoch": 0.3471, + "grad_norm": 0.822709321975708, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 34710 + }, + { + "epoch": 0.34711, + "grad_norm": 0.8667641878128052, + "learning_rate": 0.003, + "loss": 3.988, + "step": 34711 + }, + { + "epoch": 0.34712, + "grad_norm": 1.0996416807174683, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 34712 + }, + { + "epoch": 0.34713, + "grad_norm": 1.2838621139526367, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 34713 + }, + { + "epoch": 0.34714, + "grad_norm": 0.7770594358444214, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 34714 + }, + { + "epoch": 0.34715, + "grad_norm": 0.8351646065711975, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 34715 + }, + { + "epoch": 0.34716, + "grad_norm": 1.0842593908309937, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 34716 + }, + { + "epoch": 0.34717, + "grad_norm": 1.2006465196609497, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 34717 + }, + { + "epoch": 0.34718, + "grad_norm": 0.6740542054176331, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 34718 + }, + { + "epoch": 0.34719, + "grad_norm": 0.5738404989242554, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 34719 + }, + { + "epoch": 0.3472, + "grad_norm": 0.6162784695625305, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 34720 + }, + { + "epoch": 0.34721, + "grad_norm": 0.7348281741142273, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 34721 + }, + { + "epoch": 0.34722, + "grad_norm": 0.8137529492378235, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 34722 + }, + { + "epoch": 0.34723, + "grad_norm": 0.856346607208252, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 34723 + }, + { + "epoch": 0.34724, + "grad_norm": 0.8558158278465271, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 34724 + }, + { + "epoch": 0.34725, + "grad_norm": 0.846814751625061, + "learning_rate": 0.003, + "loss": 3.967, + "step": 34725 + }, + { + "epoch": 0.34726, + "grad_norm": 0.9409211277961731, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 34726 + }, + { + "epoch": 0.34727, + "grad_norm": 0.9364538788795471, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 34727 + }, + { + "epoch": 0.34728, + "grad_norm": 1.0614100694656372, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 34728 + }, + { + "epoch": 0.34729, + "grad_norm": 1.2086509466171265, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 34729 + }, + { + "epoch": 0.3473, + "grad_norm": 0.8142791390419006, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 34730 + }, + { + "epoch": 0.34731, + "grad_norm": 0.7079067230224609, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 34731 + }, + { + "epoch": 0.34732, + "grad_norm": 0.7296810150146484, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 34732 + }, + { + "epoch": 0.34733, + "grad_norm": 0.7131425142288208, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 34733 + }, + { + "epoch": 0.34734, + "grad_norm": 0.8055234551429749, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 34734 + }, + { + "epoch": 0.34735, + "grad_norm": 1.0911413431167603, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 34735 + }, + { + "epoch": 0.34736, + "grad_norm": 1.3018205165863037, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 34736 + }, + { + "epoch": 0.34737, + "grad_norm": 0.5739433169364929, + "learning_rate": 0.003, + "loss": 4.009, + "step": 34737 + }, + { + "epoch": 0.34738, + "grad_norm": 0.7016039490699768, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 34738 + }, + { + "epoch": 0.34739, + "grad_norm": 1.0233464241027832, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 34739 + }, + { + "epoch": 0.3474, + "grad_norm": 1.1433546543121338, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 34740 + }, + { + "epoch": 0.34741, + "grad_norm": 0.800154447555542, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 34741 + }, + { + "epoch": 0.34742, + "grad_norm": 0.7241703271865845, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 34742 + }, + { + "epoch": 0.34743, + "grad_norm": 0.6948196887969971, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 34743 + }, + { + "epoch": 0.34744, + "grad_norm": 0.6918114423751831, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 34744 + }, + { + "epoch": 0.34745, + "grad_norm": 0.7353536486625671, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 34745 + }, + { + "epoch": 0.34746, + "grad_norm": 0.7014166116714478, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 34746 + }, + { + "epoch": 0.34747, + "grad_norm": 0.7426641583442688, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 34747 + }, + { + "epoch": 0.34748, + "grad_norm": 0.8840955495834351, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 34748 + }, + { + "epoch": 0.34749, + "grad_norm": 0.9552009701728821, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 34749 + }, + { + "epoch": 0.3475, + "grad_norm": 1.2876235246658325, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 34750 + }, + { + "epoch": 0.34751, + "grad_norm": 0.9514825940132141, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 34751 + }, + { + "epoch": 0.34752, + "grad_norm": 0.9834404587745667, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 34752 + }, + { + "epoch": 0.34753, + "grad_norm": 1.2561836242675781, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 34753 + }, + { + "epoch": 0.34754, + "grad_norm": 0.8892229199409485, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 34754 + }, + { + "epoch": 0.34755, + "grad_norm": 0.8656673431396484, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 34755 + }, + { + "epoch": 0.34756, + "grad_norm": 1.0469276905059814, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 34756 + }, + { + "epoch": 0.34757, + "grad_norm": 1.1845834255218506, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 34757 + }, + { + "epoch": 0.34758, + "grad_norm": 0.7627046704292297, + "learning_rate": 0.003, + "loss": 3.986, + "step": 34758 + }, + { + "epoch": 0.34759, + "grad_norm": 0.848408043384552, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 34759 + }, + { + "epoch": 0.3476, + "grad_norm": 0.8333105444908142, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 34760 + }, + { + "epoch": 0.34761, + "grad_norm": 0.7982629537582397, + "learning_rate": 0.003, + "loss": 3.995, + "step": 34761 + }, + { + "epoch": 0.34762, + "grad_norm": 0.8013750314712524, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 34762 + }, + { + "epoch": 0.34763, + "grad_norm": 0.7888039350509644, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 34763 + }, + { + "epoch": 0.34764, + "grad_norm": 0.8951712250709534, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 34764 + }, + { + "epoch": 0.34765, + "grad_norm": 1.0870790481567383, + "learning_rate": 0.003, + "loss": 3.983, + "step": 34765 + }, + { + "epoch": 0.34766, + "grad_norm": 1.0653268098831177, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 34766 + }, + { + "epoch": 0.34767, + "grad_norm": 0.8961700201034546, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 34767 + }, + { + "epoch": 0.34768, + "grad_norm": 0.9291636943817139, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 34768 + }, + { + "epoch": 0.34769, + "grad_norm": 1.0952033996582031, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 34769 + }, + { + "epoch": 0.3477, + "grad_norm": 1.1418097019195557, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 34770 + }, + { + "epoch": 0.34771, + "grad_norm": 0.8101206421852112, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 34771 + }, + { + "epoch": 0.34772, + "grad_norm": 0.6874489784240723, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 34772 + }, + { + "epoch": 0.34773, + "grad_norm": 0.644173264503479, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 34773 + }, + { + "epoch": 0.34774, + "grad_norm": 0.6402318477630615, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 34774 + }, + { + "epoch": 0.34775, + "grad_norm": 0.7244625091552734, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 34775 + }, + { + "epoch": 0.34776, + "grad_norm": 1.048460602760315, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 34776 + }, + { + "epoch": 0.34777, + "grad_norm": 1.1294740438461304, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 34777 + }, + { + "epoch": 0.34778, + "grad_norm": 0.7519460320472717, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 34778 + }, + { + "epoch": 0.34779, + "grad_norm": 0.8226345777511597, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 34779 + }, + { + "epoch": 0.3478, + "grad_norm": 0.878525972366333, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 34780 + }, + { + "epoch": 0.34781, + "grad_norm": 0.9493208527565002, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 34781 + }, + { + "epoch": 0.34782, + "grad_norm": 0.9949884414672852, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 34782 + }, + { + "epoch": 0.34783, + "grad_norm": 0.9618869423866272, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 34783 + }, + { + "epoch": 0.34784, + "grad_norm": 0.9163301587104797, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 34784 + }, + { + "epoch": 0.34785, + "grad_norm": 0.8936095237731934, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 34785 + }, + { + "epoch": 0.34786, + "grad_norm": 0.932284951210022, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 34786 + }, + { + "epoch": 0.34787, + "grad_norm": 1.0109213590621948, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 34787 + }, + { + "epoch": 0.34788, + "grad_norm": 0.8810174465179443, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 34788 + }, + { + "epoch": 0.34789, + "grad_norm": 0.9280614256858826, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 34789 + }, + { + "epoch": 0.3479, + "grad_norm": 0.8751964569091797, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 34790 + }, + { + "epoch": 0.34791, + "grad_norm": 0.9215030074119568, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 34791 + }, + { + "epoch": 0.34792, + "grad_norm": 0.8103100657463074, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 34792 + }, + { + "epoch": 0.34793, + "grad_norm": 0.792326033115387, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 34793 + }, + { + "epoch": 0.34794, + "grad_norm": 0.7353072762489319, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 34794 + }, + { + "epoch": 0.34795, + "grad_norm": 0.8258078098297119, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 34795 + }, + { + "epoch": 0.34796, + "grad_norm": 0.9636012315750122, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 34796 + }, + { + "epoch": 0.34797, + "grad_norm": 1.08278489112854, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 34797 + }, + { + "epoch": 0.34798, + "grad_norm": 1.1088186502456665, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 34798 + }, + { + "epoch": 0.34799, + "grad_norm": 1.3497509956359863, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 34799 + }, + { + "epoch": 0.348, + "grad_norm": 0.7857226133346558, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 34800 + }, + { + "epoch": 0.34801, + "grad_norm": 0.6616933941841125, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 34801 + }, + { + "epoch": 0.34802, + "grad_norm": 0.798907995223999, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 34802 + }, + { + "epoch": 0.34803, + "grad_norm": 0.9145715236663818, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 34803 + }, + { + "epoch": 0.34804, + "grad_norm": 1.1668654680252075, + "learning_rate": 0.003, + "loss": 4.008, + "step": 34804 + }, + { + "epoch": 0.34805, + "grad_norm": 1.1007474660873413, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 34805 + }, + { + "epoch": 0.34806, + "grad_norm": 0.7713804244995117, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 34806 + }, + { + "epoch": 0.34807, + "grad_norm": 0.622166097164154, + "learning_rate": 0.003, + "loss": 3.965, + "step": 34807 + }, + { + "epoch": 0.34808, + "grad_norm": 0.564961314201355, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 34808 + }, + { + "epoch": 0.34809, + "grad_norm": 0.5380163788795471, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 34809 + }, + { + "epoch": 0.3481, + "grad_norm": 0.5325971841812134, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 34810 + }, + { + "epoch": 0.34811, + "grad_norm": 0.6121297478675842, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 34811 + }, + { + "epoch": 0.34812, + "grad_norm": 0.6414052248001099, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 34812 + }, + { + "epoch": 0.34813, + "grad_norm": 0.7827320694923401, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 34813 + }, + { + "epoch": 0.34814, + "grad_norm": 0.8462938666343689, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 34814 + }, + { + "epoch": 0.34815, + "grad_norm": 1.173732042312622, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 34815 + }, + { + "epoch": 0.34816, + "grad_norm": 0.986630916595459, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 34816 + }, + { + "epoch": 0.34817, + "grad_norm": 1.1207042932510376, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 34817 + }, + { + "epoch": 0.34818, + "grad_norm": 1.0247011184692383, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 34818 + }, + { + "epoch": 0.34819, + "grad_norm": 0.8651981353759766, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 34819 + }, + { + "epoch": 0.3482, + "grad_norm": 0.6851192116737366, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 34820 + }, + { + "epoch": 0.34821, + "grad_norm": 0.7184176445007324, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 34821 + }, + { + "epoch": 0.34822, + "grad_norm": 0.7041357755661011, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 34822 + }, + { + "epoch": 0.34823, + "grad_norm": 0.7485741376876831, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 34823 + }, + { + "epoch": 0.34824, + "grad_norm": 0.9552158117294312, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 34824 + }, + { + "epoch": 0.34825, + "grad_norm": 1.1504144668579102, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 34825 + }, + { + "epoch": 0.34826, + "grad_norm": 0.8796074986457825, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 34826 + }, + { + "epoch": 0.34827, + "grad_norm": 0.9005832672119141, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 34827 + }, + { + "epoch": 0.34828, + "grad_norm": 0.9139113426208496, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 34828 + }, + { + "epoch": 0.34829, + "grad_norm": 0.9258444905281067, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 34829 + }, + { + "epoch": 0.3483, + "grad_norm": 0.8089351058006287, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 34830 + }, + { + "epoch": 0.34831, + "grad_norm": 0.9317271709442139, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 34831 + }, + { + "epoch": 0.34832, + "grad_norm": 1.0832394361495972, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 34832 + }, + { + "epoch": 0.34833, + "grad_norm": 0.8268157839775085, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 34833 + }, + { + "epoch": 0.34834, + "grad_norm": 1.0571058988571167, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 34834 + }, + { + "epoch": 0.34835, + "grad_norm": 0.9783881902694702, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 34835 + }, + { + "epoch": 0.34836, + "grad_norm": 1.0294647216796875, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 34836 + }, + { + "epoch": 0.34837, + "grad_norm": 1.0914098024368286, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 34837 + }, + { + "epoch": 0.34838, + "grad_norm": 0.9013953804969788, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 34838 + }, + { + "epoch": 0.34839, + "grad_norm": 0.8968129754066467, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 34839 + }, + { + "epoch": 0.3484, + "grad_norm": 0.9868174195289612, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 34840 + }, + { + "epoch": 0.34841, + "grad_norm": 1.3272554874420166, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 34841 + }, + { + "epoch": 0.34842, + "grad_norm": 0.9063495397567749, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 34842 + }, + { + "epoch": 0.34843, + "grad_norm": 0.9280005097389221, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 34843 + }, + { + "epoch": 0.34844, + "grad_norm": 0.8903824687004089, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 34844 + }, + { + "epoch": 0.34845, + "grad_norm": 0.8700262904167175, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 34845 + }, + { + "epoch": 0.34846, + "grad_norm": 0.7956714630126953, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 34846 + }, + { + "epoch": 0.34847, + "grad_norm": 0.595596969127655, + "learning_rate": 0.003, + "loss": 3.972, + "step": 34847 + }, + { + "epoch": 0.34848, + "grad_norm": 0.6195836067199707, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 34848 + }, + { + "epoch": 0.34849, + "grad_norm": 0.6040182709693909, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 34849 + }, + { + "epoch": 0.3485, + "grad_norm": 0.6625202894210815, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 34850 + }, + { + "epoch": 0.34851, + "grad_norm": 0.8786371350288391, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 34851 + }, + { + "epoch": 0.34852, + "grad_norm": 1.1521086692810059, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 34852 + }, + { + "epoch": 0.34853, + "grad_norm": 1.153354525566101, + "learning_rate": 0.003, + "loss": 3.978, + "step": 34853 + }, + { + "epoch": 0.34854, + "grad_norm": 0.916576087474823, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 34854 + }, + { + "epoch": 0.34855, + "grad_norm": 0.7556350231170654, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 34855 + }, + { + "epoch": 0.34856, + "grad_norm": 0.8095990419387817, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 34856 + }, + { + "epoch": 0.34857, + "grad_norm": 0.6958017945289612, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 34857 + }, + { + "epoch": 0.34858, + "grad_norm": 0.6890007853507996, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 34858 + }, + { + "epoch": 0.34859, + "grad_norm": 0.9317807555198669, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 34859 + }, + { + "epoch": 0.3486, + "grad_norm": 1.362143635749817, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 34860 + }, + { + "epoch": 0.34861, + "grad_norm": 0.506596028804779, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 34861 + }, + { + "epoch": 0.34862, + "grad_norm": 0.7734477519989014, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 34862 + }, + { + "epoch": 0.34863, + "grad_norm": 1.0988731384277344, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 34863 + }, + { + "epoch": 0.34864, + "grad_norm": 0.9940036535263062, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 34864 + }, + { + "epoch": 0.34865, + "grad_norm": 0.9402612447738647, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 34865 + }, + { + "epoch": 0.34866, + "grad_norm": 0.9454014897346497, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 34866 + }, + { + "epoch": 0.34867, + "grad_norm": 1.0020196437835693, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 34867 + }, + { + "epoch": 0.34868, + "grad_norm": 1.0041428804397583, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 34868 + }, + { + "epoch": 0.34869, + "grad_norm": 1.078129529953003, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 34869 + }, + { + "epoch": 0.3487, + "grad_norm": 1.100785493850708, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 34870 + }, + { + "epoch": 0.34871, + "grad_norm": 0.9484154582023621, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 34871 + }, + { + "epoch": 0.34872, + "grad_norm": 0.9777306914329529, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 34872 + }, + { + "epoch": 0.34873, + "grad_norm": 1.029542326927185, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 34873 + }, + { + "epoch": 0.34874, + "grad_norm": 0.9138944149017334, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 34874 + }, + { + "epoch": 0.34875, + "grad_norm": 0.7632777094841003, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 34875 + }, + { + "epoch": 0.34876, + "grad_norm": 0.811041533946991, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 34876 + }, + { + "epoch": 0.34877, + "grad_norm": 0.7928783297538757, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 34877 + }, + { + "epoch": 0.34878, + "grad_norm": 0.8130239248275757, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 34878 + }, + { + "epoch": 0.34879, + "grad_norm": 0.8130394816398621, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 34879 + }, + { + "epoch": 0.3488, + "grad_norm": 0.7931410670280457, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 34880 + }, + { + "epoch": 0.34881, + "grad_norm": 0.8445945978164673, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 34881 + }, + { + "epoch": 0.34882, + "grad_norm": 0.9048891067504883, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 34882 + }, + { + "epoch": 0.34883, + "grad_norm": 0.9768567085266113, + "learning_rate": 0.003, + "loss": 3.97, + "step": 34883 + }, + { + "epoch": 0.34884, + "grad_norm": 0.9606625437736511, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 34884 + }, + { + "epoch": 0.34885, + "grad_norm": 0.8893373012542725, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 34885 + }, + { + "epoch": 0.34886, + "grad_norm": 0.9035897254943848, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 34886 + }, + { + "epoch": 0.34887, + "grad_norm": 0.8710338473320007, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 34887 + }, + { + "epoch": 0.34888, + "grad_norm": 0.7494446039199829, + "learning_rate": 0.003, + "loss": 3.991, + "step": 34888 + }, + { + "epoch": 0.34889, + "grad_norm": 0.7868604063987732, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 34889 + }, + { + "epoch": 0.3489, + "grad_norm": 0.937881588935852, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 34890 + }, + { + "epoch": 0.34891, + "grad_norm": 0.8463566899299622, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 34891 + }, + { + "epoch": 0.34892, + "grad_norm": 0.7503739595413208, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 34892 + }, + { + "epoch": 0.34893, + "grad_norm": 0.705805778503418, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 34893 + }, + { + "epoch": 0.34894, + "grad_norm": 0.8846867680549622, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 34894 + }, + { + "epoch": 0.34895, + "grad_norm": 0.9541426301002502, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 34895 + }, + { + "epoch": 0.34896, + "grad_norm": 1.1645580530166626, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 34896 + }, + { + "epoch": 0.34897, + "grad_norm": 0.9426249265670776, + "learning_rate": 0.003, + "loss": 3.969, + "step": 34897 + }, + { + "epoch": 0.34898, + "grad_norm": 0.8784621953964233, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 34898 + }, + { + "epoch": 0.34899, + "grad_norm": 0.8792023062705994, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 34899 + }, + { + "epoch": 0.349, + "grad_norm": 1.085510492324829, + "learning_rate": 0.003, + "loss": 4.013, + "step": 34900 + }, + { + "epoch": 0.34901, + "grad_norm": 1.0194541215896606, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 34901 + }, + { + "epoch": 0.34902, + "grad_norm": 0.9064772129058838, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 34902 + }, + { + "epoch": 0.34903, + "grad_norm": 1.052912950515747, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 34903 + }, + { + "epoch": 0.34904, + "grad_norm": 1.267911434173584, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 34904 + }, + { + "epoch": 0.34905, + "grad_norm": 0.8709411025047302, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 34905 + }, + { + "epoch": 0.34906, + "grad_norm": 0.8603829145431519, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 34906 + }, + { + "epoch": 0.34907, + "grad_norm": 0.8649342656135559, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 34907 + }, + { + "epoch": 0.34908, + "grad_norm": 0.9063426852226257, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 34908 + }, + { + "epoch": 0.34909, + "grad_norm": 0.9198972582817078, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 34909 + }, + { + "epoch": 0.3491, + "grad_norm": 0.9231833815574646, + "learning_rate": 0.003, + "loss": 4.013, + "step": 34910 + }, + { + "epoch": 0.34911, + "grad_norm": 0.9265281558036804, + "learning_rate": 0.003, + "loss": 3.99, + "step": 34911 + }, + { + "epoch": 0.34912, + "grad_norm": 1.0172382593154907, + "learning_rate": 0.003, + "loss": 3.994, + "step": 34912 + }, + { + "epoch": 0.34913, + "grad_norm": 1.089360237121582, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 34913 + }, + { + "epoch": 0.34914, + "grad_norm": 0.9205033183097839, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 34914 + }, + { + "epoch": 0.34915, + "grad_norm": 0.942065417766571, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 34915 + }, + { + "epoch": 0.34916, + "grad_norm": 0.8581379652023315, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 34916 + }, + { + "epoch": 0.34917, + "grad_norm": 0.8183024525642395, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 34917 + }, + { + "epoch": 0.34918, + "grad_norm": 0.8225533962249756, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 34918 + }, + { + "epoch": 0.34919, + "grad_norm": 0.9079436659812927, + "learning_rate": 0.003, + "loss": 3.993, + "step": 34919 + }, + { + "epoch": 0.3492, + "grad_norm": 1.189367651939392, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 34920 + }, + { + "epoch": 0.34921, + "grad_norm": 0.9997583031654358, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 34921 + }, + { + "epoch": 0.34922, + "grad_norm": 0.9835258722305298, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 34922 + }, + { + "epoch": 0.34923, + "grad_norm": 1.0776896476745605, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 34923 + }, + { + "epoch": 0.34924, + "grad_norm": 1.047839641571045, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 34924 + }, + { + "epoch": 0.34925, + "grad_norm": 0.8086878657341003, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 34925 + }, + { + "epoch": 0.34926, + "grad_norm": 0.7047544717788696, + "learning_rate": 0.003, + "loss": 3.9342, + "step": 34926 + }, + { + "epoch": 0.34927, + "grad_norm": 0.7808406352996826, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 34927 + }, + { + "epoch": 0.34928, + "grad_norm": 0.7102218866348267, + "learning_rate": 0.003, + "loss": 3.98, + "step": 34928 + }, + { + "epoch": 0.34929, + "grad_norm": 0.6989945769309998, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 34929 + }, + { + "epoch": 0.3493, + "grad_norm": 0.6490746736526489, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 34930 + }, + { + "epoch": 0.34931, + "grad_norm": 0.6586465835571289, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 34931 + }, + { + "epoch": 0.34932, + "grad_norm": 0.8422635197639465, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 34932 + }, + { + "epoch": 0.34933, + "grad_norm": 1.1775450706481934, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 34933 + }, + { + "epoch": 0.34934, + "grad_norm": 1.0730382204055786, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 34934 + }, + { + "epoch": 0.34935, + "grad_norm": 0.8203026056289673, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 34935 + }, + { + "epoch": 0.34936, + "grad_norm": 0.6983779072761536, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 34936 + }, + { + "epoch": 0.34937, + "grad_norm": 0.7247359752655029, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 34937 + }, + { + "epoch": 0.34938, + "grad_norm": 0.8479139804840088, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 34938 + }, + { + "epoch": 0.34939, + "grad_norm": 0.872444748878479, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 34939 + }, + { + "epoch": 0.3494, + "grad_norm": 0.8928108215332031, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 34940 + }, + { + "epoch": 0.34941, + "grad_norm": 1.0041228532791138, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 34941 + }, + { + "epoch": 0.34942, + "grad_norm": 1.0183358192443848, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 34942 + }, + { + "epoch": 0.34943, + "grad_norm": 1.0198121070861816, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 34943 + }, + { + "epoch": 0.34944, + "grad_norm": 1.0919241905212402, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 34944 + }, + { + "epoch": 0.34945, + "grad_norm": 0.9410253763198853, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 34945 + }, + { + "epoch": 0.34946, + "grad_norm": 0.8158665895462036, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 34946 + }, + { + "epoch": 0.34947, + "grad_norm": 0.8255749940872192, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 34947 + }, + { + "epoch": 0.34948, + "grad_norm": 0.9353623390197754, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 34948 + }, + { + "epoch": 0.34949, + "grad_norm": 0.984142541885376, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 34949 + }, + { + "epoch": 0.3495, + "grad_norm": 1.0135374069213867, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 34950 + }, + { + "epoch": 0.34951, + "grad_norm": 1.0653364658355713, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 34951 + }, + { + "epoch": 0.34952, + "grad_norm": 1.102640151977539, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 34952 + }, + { + "epoch": 0.34953, + "grad_norm": 0.9412261843681335, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 34953 + }, + { + "epoch": 0.34954, + "grad_norm": 0.8511659502983093, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 34954 + }, + { + "epoch": 0.34955, + "grad_norm": 0.7866503000259399, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 34955 + }, + { + "epoch": 0.34956, + "grad_norm": 0.8695470094680786, + "learning_rate": 0.003, + "loss": 3.983, + "step": 34956 + }, + { + "epoch": 0.34957, + "grad_norm": 1.064020037651062, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 34957 + }, + { + "epoch": 0.34958, + "grad_norm": 1.0037986040115356, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 34958 + }, + { + "epoch": 0.34959, + "grad_norm": 1.043962836265564, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 34959 + }, + { + "epoch": 0.3496, + "grad_norm": 0.8103291988372803, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 34960 + }, + { + "epoch": 0.34961, + "grad_norm": 0.7687110304832458, + "learning_rate": 0.003, + "loss": 4.012, + "step": 34961 + }, + { + "epoch": 0.34962, + "grad_norm": 0.6980137228965759, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 34962 + }, + { + "epoch": 0.34963, + "grad_norm": 0.653785228729248, + "learning_rate": 0.003, + "loss": 3.963, + "step": 34963 + }, + { + "epoch": 0.34964, + "grad_norm": 0.6922245025634766, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 34964 + }, + { + "epoch": 0.34965, + "grad_norm": 0.6760269999504089, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 34965 + }, + { + "epoch": 0.34966, + "grad_norm": 0.7349156141281128, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 34966 + }, + { + "epoch": 0.34967, + "grad_norm": 0.8876712918281555, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 34967 + }, + { + "epoch": 0.34968, + "grad_norm": 1.2314404249191284, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 34968 + }, + { + "epoch": 0.34969, + "grad_norm": 0.82976895570755, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 34969 + }, + { + "epoch": 0.3497, + "grad_norm": 0.5445650815963745, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 34970 + }, + { + "epoch": 0.34971, + "grad_norm": 0.5686646699905396, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 34971 + }, + { + "epoch": 0.34972, + "grad_norm": 0.5976795554161072, + "learning_rate": 0.003, + "loss": 3.951, + "step": 34972 + }, + { + "epoch": 0.34973, + "grad_norm": 0.768887460231781, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 34973 + }, + { + "epoch": 0.34974, + "grad_norm": 1.082108974456787, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 34974 + }, + { + "epoch": 0.34975, + "grad_norm": 1.1150554418563843, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 34975 + }, + { + "epoch": 0.34976, + "grad_norm": 0.7752797603607178, + "learning_rate": 0.003, + "loss": 3.9455, + "step": 34976 + }, + { + "epoch": 0.34977, + "grad_norm": 0.5908323526382446, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 34977 + }, + { + "epoch": 0.34978, + "grad_norm": 0.7128618359565735, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 34978 + }, + { + "epoch": 0.34979, + "grad_norm": 0.8403050303459167, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 34979 + }, + { + "epoch": 0.3498, + "grad_norm": 1.1003257036209106, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 34980 + }, + { + "epoch": 0.34981, + "grad_norm": 0.9432708024978638, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 34981 + }, + { + "epoch": 0.34982, + "grad_norm": 0.8003827333450317, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 34982 + }, + { + "epoch": 0.34983, + "grad_norm": 0.7272015810012817, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 34983 + }, + { + "epoch": 0.34984, + "grad_norm": 0.737602710723877, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 34984 + }, + { + "epoch": 0.34985, + "grad_norm": 0.6874889731407166, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 34985 + }, + { + "epoch": 0.34986, + "grad_norm": 0.6779497265815735, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 34986 + }, + { + "epoch": 0.34987, + "grad_norm": 0.8189074397087097, + "learning_rate": 0.003, + "loss": 3.9509, + "step": 34987 + }, + { + "epoch": 0.34988, + "grad_norm": 1.0418683290481567, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 34988 + }, + { + "epoch": 0.34989, + "grad_norm": 1.2260847091674805, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 34989 + }, + { + "epoch": 0.3499, + "grad_norm": 1.0767170190811157, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 34990 + }, + { + "epoch": 0.34991, + "grad_norm": 1.1040825843811035, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 34991 + }, + { + "epoch": 0.34992, + "grad_norm": 1.024254560470581, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 34992 + }, + { + "epoch": 0.34993, + "grad_norm": 0.9740331768989563, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 34993 + }, + { + "epoch": 0.34994, + "grad_norm": 1.0466117858886719, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 34994 + }, + { + "epoch": 0.34995, + "grad_norm": 1.071210503578186, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 34995 + }, + { + "epoch": 0.34996, + "grad_norm": 1.0831928253173828, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 34996 + }, + { + "epoch": 0.34997, + "grad_norm": 1.3470592498779297, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 34997 + }, + { + "epoch": 0.34998, + "grad_norm": 0.663561999797821, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 34998 + }, + { + "epoch": 0.34999, + "grad_norm": 0.8412469625473022, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 34999 + }, + { + "epoch": 0.35, + "grad_norm": 1.1952524185180664, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 35000 + }, + { + "epoch": 0.35001, + "grad_norm": 0.7031912803649902, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 35001 + }, + { + "epoch": 0.35002, + "grad_norm": 0.6286465525627136, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 35002 + }, + { + "epoch": 0.35003, + "grad_norm": 0.612640380859375, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 35003 + }, + { + "epoch": 0.35004, + "grad_norm": 0.6956764459609985, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 35004 + }, + { + "epoch": 0.35005, + "grad_norm": 0.9018158912658691, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 35005 + }, + { + "epoch": 0.35006, + "grad_norm": 0.9846453070640564, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 35006 + }, + { + "epoch": 0.35007, + "grad_norm": 0.9571338891983032, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 35007 + }, + { + "epoch": 0.35008, + "grad_norm": 1.022050142288208, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 35008 + }, + { + "epoch": 0.35009, + "grad_norm": 0.9813030362129211, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 35009 + }, + { + "epoch": 0.3501, + "grad_norm": 1.0102927684783936, + "learning_rate": 0.003, + "loss": 4.01, + "step": 35010 + }, + { + "epoch": 0.35011, + "grad_norm": 0.9553887844085693, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 35011 + }, + { + "epoch": 0.35012, + "grad_norm": 0.8977872133255005, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 35012 + }, + { + "epoch": 0.35013, + "grad_norm": 0.885191023349762, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 35013 + }, + { + "epoch": 0.35014, + "grad_norm": 0.7954698204994202, + "learning_rate": 0.003, + "loss": 3.979, + "step": 35014 + }, + { + "epoch": 0.35015, + "grad_norm": 0.8345710635185242, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 35015 + }, + { + "epoch": 0.35016, + "grad_norm": 1.2951594591140747, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 35016 + }, + { + "epoch": 0.35017, + "grad_norm": 0.8276659846305847, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 35017 + }, + { + "epoch": 0.35018, + "grad_norm": 0.9383276700973511, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 35018 + }, + { + "epoch": 0.35019, + "grad_norm": 1.1015781164169312, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 35019 + }, + { + "epoch": 0.3502, + "grad_norm": 1.020287036895752, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 35020 + }, + { + "epoch": 0.35021, + "grad_norm": 0.9264557957649231, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 35021 + }, + { + "epoch": 0.35022, + "grad_norm": 0.823933482170105, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 35022 + }, + { + "epoch": 0.35023, + "grad_norm": 0.7601601481437683, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 35023 + }, + { + "epoch": 0.35024, + "grad_norm": 0.8016993999481201, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 35024 + }, + { + "epoch": 0.35025, + "grad_norm": 0.8515185713768005, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 35025 + }, + { + "epoch": 0.35026, + "grad_norm": 0.8719425201416016, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 35026 + }, + { + "epoch": 0.35027, + "grad_norm": 0.8960185050964355, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 35027 + }, + { + "epoch": 0.35028, + "grad_norm": 1.105273962020874, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 35028 + }, + { + "epoch": 0.35029, + "grad_norm": 1.1858265399932861, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 35029 + }, + { + "epoch": 0.3503, + "grad_norm": 0.8595359325408936, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 35030 + }, + { + "epoch": 0.35031, + "grad_norm": 0.7312819361686707, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 35031 + }, + { + "epoch": 0.35032, + "grad_norm": 0.7358041405677795, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 35032 + }, + { + "epoch": 0.35033, + "grad_norm": 0.9308135509490967, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 35033 + }, + { + "epoch": 0.35034, + "grad_norm": 1.0586069822311401, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 35034 + }, + { + "epoch": 0.35035, + "grad_norm": 1.020595908164978, + "learning_rate": 0.003, + "loss": 4.002, + "step": 35035 + }, + { + "epoch": 0.35036, + "grad_norm": 1.059226155281067, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 35036 + }, + { + "epoch": 0.35037, + "grad_norm": 1.012649416923523, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 35037 + }, + { + "epoch": 0.35038, + "grad_norm": 1.068875789642334, + "learning_rate": 0.003, + "loss": 3.965, + "step": 35038 + }, + { + "epoch": 0.35039, + "grad_norm": 0.9178526401519775, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 35039 + }, + { + "epoch": 0.3504, + "grad_norm": 0.9196370840072632, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 35040 + }, + { + "epoch": 0.35041, + "grad_norm": 1.018831491470337, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 35041 + }, + { + "epoch": 0.35042, + "grad_norm": 0.9031400084495544, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 35042 + }, + { + "epoch": 0.35043, + "grad_norm": 0.8681761622428894, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 35043 + }, + { + "epoch": 0.35044, + "grad_norm": 0.928814172744751, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 35044 + }, + { + "epoch": 0.35045, + "grad_norm": 1.0555118322372437, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 35045 + }, + { + "epoch": 0.35046, + "grad_norm": 0.9919033050537109, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 35046 + }, + { + "epoch": 0.35047, + "grad_norm": 0.9874020218849182, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 35047 + }, + { + "epoch": 0.35048, + "grad_norm": 1.190258264541626, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 35048 + }, + { + "epoch": 0.35049, + "grad_norm": 0.7056639790534973, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 35049 + }, + { + "epoch": 0.3505, + "grad_norm": 0.8288295269012451, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 35050 + }, + { + "epoch": 0.35051, + "grad_norm": 1.1674422025680542, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 35051 + }, + { + "epoch": 0.35052, + "grad_norm": 0.7509384751319885, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 35052 + }, + { + "epoch": 0.35053, + "grad_norm": 0.73553466796875, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 35053 + }, + { + "epoch": 0.35054, + "grad_norm": 0.7028337121009827, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 35054 + }, + { + "epoch": 0.35055, + "grad_norm": 0.6415852308273315, + "learning_rate": 0.003, + "loss": 3.989, + "step": 35055 + }, + { + "epoch": 0.35056, + "grad_norm": 0.6551271080970764, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 35056 + }, + { + "epoch": 0.35057, + "grad_norm": 0.6425012946128845, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 35057 + }, + { + "epoch": 0.35058, + "grad_norm": 0.6476554274559021, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 35058 + }, + { + "epoch": 0.35059, + "grad_norm": 0.7646179795265198, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 35059 + }, + { + "epoch": 0.3506, + "grad_norm": 0.9268144965171814, + "learning_rate": 0.003, + "loss": 3.997, + "step": 35060 + }, + { + "epoch": 0.35061, + "grad_norm": 1.1257412433624268, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 35061 + }, + { + "epoch": 0.35062, + "grad_norm": 0.988994300365448, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 35062 + }, + { + "epoch": 0.35063, + "grad_norm": 1.1706640720367432, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 35063 + }, + { + "epoch": 0.35064, + "grad_norm": 1.0680975914001465, + "learning_rate": 0.003, + "loss": 3.974, + "step": 35064 + }, + { + "epoch": 0.35065, + "grad_norm": 1.0255517959594727, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 35065 + }, + { + "epoch": 0.35066, + "grad_norm": 1.1371744871139526, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 35066 + }, + { + "epoch": 0.35067, + "grad_norm": 1.1473214626312256, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 35067 + }, + { + "epoch": 0.35068, + "grad_norm": 0.8699187636375427, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 35068 + }, + { + "epoch": 0.35069, + "grad_norm": 0.7294921278953552, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 35069 + }, + { + "epoch": 0.3507, + "grad_norm": 0.6348940134048462, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 35070 + }, + { + "epoch": 0.35071, + "grad_norm": 0.6251999735832214, + "learning_rate": 0.003, + "loss": 3.975, + "step": 35071 + }, + { + "epoch": 0.35072, + "grad_norm": 0.7762973308563232, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 35072 + }, + { + "epoch": 0.35073, + "grad_norm": 0.8594342470169067, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 35073 + }, + { + "epoch": 0.35074, + "grad_norm": 0.9948176145553589, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 35074 + }, + { + "epoch": 0.35075, + "grad_norm": 1.1393338441848755, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 35075 + }, + { + "epoch": 0.35076, + "grad_norm": 1.0496242046356201, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 35076 + }, + { + "epoch": 0.35077, + "grad_norm": 1.1174542903900146, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 35077 + }, + { + "epoch": 0.35078, + "grad_norm": 0.9041407704353333, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 35078 + }, + { + "epoch": 0.35079, + "grad_norm": 0.7851302027702332, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 35079 + }, + { + "epoch": 0.3508, + "grad_norm": 0.7122425436973572, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 35080 + }, + { + "epoch": 0.35081, + "grad_norm": 0.6687360405921936, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 35081 + }, + { + "epoch": 0.35082, + "grad_norm": 0.7695944309234619, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 35082 + }, + { + "epoch": 0.35083, + "grad_norm": 1.0210317373275757, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 35083 + }, + { + "epoch": 0.35084, + "grad_norm": 1.2072614431381226, + "learning_rate": 0.003, + "loss": 3.972, + "step": 35084 + }, + { + "epoch": 0.35085, + "grad_norm": 0.8486311435699463, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 35085 + }, + { + "epoch": 0.35086, + "grad_norm": 0.8476829528808594, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 35086 + }, + { + "epoch": 0.35087, + "grad_norm": 0.8767332434654236, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 35087 + }, + { + "epoch": 0.35088, + "grad_norm": 0.9187741279602051, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 35088 + }, + { + "epoch": 0.35089, + "grad_norm": 0.8024381399154663, + "learning_rate": 0.003, + "loss": 3.942, + "step": 35089 + }, + { + "epoch": 0.3509, + "grad_norm": 0.7265815138816833, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 35090 + }, + { + "epoch": 0.35091, + "grad_norm": 0.7090408802032471, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 35091 + }, + { + "epoch": 0.35092, + "grad_norm": 0.6204546689987183, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 35092 + }, + { + "epoch": 0.35093, + "grad_norm": 0.6807122230529785, + "learning_rate": 0.003, + "loss": 3.9455, + "step": 35093 + }, + { + "epoch": 0.35094, + "grad_norm": 0.7953497767448425, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 35094 + }, + { + "epoch": 0.35095, + "grad_norm": 0.86823970079422, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 35095 + }, + { + "epoch": 0.35096, + "grad_norm": 0.842288613319397, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 35096 + }, + { + "epoch": 0.35097, + "grad_norm": 0.8364388346672058, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 35097 + }, + { + "epoch": 0.35098, + "grad_norm": 0.9934419989585876, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 35098 + }, + { + "epoch": 0.35099, + "grad_norm": 1.194867491722107, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 35099 + }, + { + "epoch": 0.351, + "grad_norm": 0.9956503510475159, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 35100 + }, + { + "epoch": 0.35101, + "grad_norm": 1.007959008216858, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 35101 + }, + { + "epoch": 0.35102, + "grad_norm": 0.875046968460083, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 35102 + }, + { + "epoch": 0.35103, + "grad_norm": 0.8242449164390564, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 35103 + }, + { + "epoch": 0.35104, + "grad_norm": 1.0009862184524536, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 35104 + }, + { + "epoch": 0.35105, + "grad_norm": 1.41642165184021, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 35105 + }, + { + "epoch": 0.35106, + "grad_norm": 0.7659485936164856, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 35106 + }, + { + "epoch": 0.35107, + "grad_norm": 0.7634344696998596, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 35107 + }, + { + "epoch": 0.35108, + "grad_norm": 0.8382549285888672, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 35108 + }, + { + "epoch": 0.35109, + "grad_norm": 0.8393747210502625, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 35109 + }, + { + "epoch": 0.3511, + "grad_norm": 0.8840888738632202, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 35110 + }, + { + "epoch": 0.35111, + "grad_norm": 1.108594298362732, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 35111 + }, + { + "epoch": 0.35112, + "grad_norm": 1.0432043075561523, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 35112 + }, + { + "epoch": 0.35113, + "grad_norm": 0.9119052290916443, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 35113 + }, + { + "epoch": 0.35114, + "grad_norm": 0.9330063462257385, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 35114 + }, + { + "epoch": 0.35115, + "grad_norm": 1.0180649757385254, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 35115 + }, + { + "epoch": 0.35116, + "grad_norm": 0.9399104714393616, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 35116 + }, + { + "epoch": 0.35117, + "grad_norm": 0.9574952125549316, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 35117 + }, + { + "epoch": 0.35118, + "grad_norm": 0.9284841418266296, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 35118 + }, + { + "epoch": 0.35119, + "grad_norm": 0.7081614136695862, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 35119 + }, + { + "epoch": 0.3512, + "grad_norm": 0.7244839072227478, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 35120 + }, + { + "epoch": 0.35121, + "grad_norm": 0.8002818822860718, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 35121 + }, + { + "epoch": 0.35122, + "grad_norm": 1.0736615657806396, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 35122 + }, + { + "epoch": 0.35123, + "grad_norm": 1.469853401184082, + "learning_rate": 0.003, + "loss": 3.988, + "step": 35123 + }, + { + "epoch": 0.35124, + "grad_norm": 0.6933988332748413, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 35124 + }, + { + "epoch": 0.35125, + "grad_norm": 0.723090410232544, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 35125 + }, + { + "epoch": 0.35126, + "grad_norm": 0.8122634291648865, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 35126 + }, + { + "epoch": 0.35127, + "grad_norm": 0.8668742775917053, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 35127 + }, + { + "epoch": 0.35128, + "grad_norm": 0.9956687092781067, + "learning_rate": 0.003, + "loss": 4.004, + "step": 35128 + }, + { + "epoch": 0.35129, + "grad_norm": 1.0769147872924805, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 35129 + }, + { + "epoch": 0.3513, + "grad_norm": 1.021628737449646, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 35130 + }, + { + "epoch": 0.35131, + "grad_norm": 0.9732219576835632, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 35131 + }, + { + "epoch": 0.35132, + "grad_norm": 0.8883460164070129, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 35132 + }, + { + "epoch": 0.35133, + "grad_norm": 0.7831789255142212, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 35133 + }, + { + "epoch": 0.35134, + "grad_norm": 0.9086045026779175, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 35134 + }, + { + "epoch": 0.35135, + "grad_norm": 1.0837527513504028, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 35135 + }, + { + "epoch": 0.35136, + "grad_norm": 0.9793734550476074, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 35136 + }, + { + "epoch": 0.35137, + "grad_norm": 0.9320245385169983, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 35137 + }, + { + "epoch": 0.35138, + "grad_norm": 0.9882290363311768, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 35138 + }, + { + "epoch": 0.35139, + "grad_norm": 0.9231003522872925, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 35139 + }, + { + "epoch": 0.3514, + "grad_norm": 0.941125214099884, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 35140 + }, + { + "epoch": 0.35141, + "grad_norm": 1.1173720359802246, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 35141 + }, + { + "epoch": 0.35142, + "grad_norm": 1.2203654050827026, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 35142 + }, + { + "epoch": 0.35143, + "grad_norm": 0.7492908835411072, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 35143 + }, + { + "epoch": 0.35144, + "grad_norm": 0.5648328065872192, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 35144 + }, + { + "epoch": 0.35145, + "grad_norm": 0.6632665395736694, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 35145 + }, + { + "epoch": 0.35146, + "grad_norm": 0.8014380931854248, + "learning_rate": 0.003, + "loss": 3.983, + "step": 35146 + }, + { + "epoch": 0.35147, + "grad_norm": 0.9414783716201782, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 35147 + }, + { + "epoch": 0.35148, + "grad_norm": 0.9714747667312622, + "learning_rate": 0.003, + "loss": 3.973, + "step": 35148 + }, + { + "epoch": 0.35149, + "grad_norm": 0.9024574756622314, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 35149 + }, + { + "epoch": 0.3515, + "grad_norm": 1.0694289207458496, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 35150 + }, + { + "epoch": 0.35151, + "grad_norm": 1.0373637676239014, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 35151 + }, + { + "epoch": 0.35152, + "grad_norm": 1.1986793279647827, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 35152 + }, + { + "epoch": 0.35153, + "grad_norm": 0.6808407306671143, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 35153 + }, + { + "epoch": 0.35154, + "grad_norm": 0.5226601958274841, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 35154 + }, + { + "epoch": 0.35155, + "grad_norm": 0.7449095845222473, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 35155 + }, + { + "epoch": 0.35156, + "grad_norm": 0.9592011570930481, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 35156 + }, + { + "epoch": 0.35157, + "grad_norm": 1.278747320175171, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 35157 + }, + { + "epoch": 0.35158, + "grad_norm": 0.6061877012252808, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 35158 + }, + { + "epoch": 0.35159, + "grad_norm": 0.63299959897995, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 35159 + }, + { + "epoch": 0.3516, + "grad_norm": 0.8306319713592529, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 35160 + }, + { + "epoch": 0.35161, + "grad_norm": 0.8474512100219727, + "learning_rate": 0.003, + "loss": 3.947, + "step": 35161 + }, + { + "epoch": 0.35162, + "grad_norm": 0.8328299522399902, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 35162 + }, + { + "epoch": 0.35163, + "grad_norm": 0.7433101534843445, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 35163 + }, + { + "epoch": 0.35164, + "grad_norm": 0.7776308059692383, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 35164 + }, + { + "epoch": 0.35165, + "grad_norm": 0.8759060502052307, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 35165 + }, + { + "epoch": 0.35166, + "grad_norm": 1.1544302701950073, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 35166 + }, + { + "epoch": 0.35167, + "grad_norm": 1.0819283723831177, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 35167 + }, + { + "epoch": 0.35168, + "grad_norm": 0.8826904296875, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 35168 + }, + { + "epoch": 0.35169, + "grad_norm": 0.9246611595153809, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 35169 + }, + { + "epoch": 0.3517, + "grad_norm": 0.9779358506202698, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 35170 + }, + { + "epoch": 0.35171, + "grad_norm": 0.9072208404541016, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 35171 + }, + { + "epoch": 0.35172, + "grad_norm": 0.8924888372421265, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 35172 + }, + { + "epoch": 0.35173, + "grad_norm": 1.004178524017334, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 35173 + }, + { + "epoch": 0.35174, + "grad_norm": 1.0130722522735596, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 35174 + }, + { + "epoch": 0.35175, + "grad_norm": 0.9056556820869446, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 35175 + }, + { + "epoch": 0.35176, + "grad_norm": 0.917845606803894, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 35176 + }, + { + "epoch": 0.35177, + "grad_norm": 0.9257668852806091, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 35177 + }, + { + "epoch": 0.35178, + "grad_norm": 0.9270830154418945, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 35178 + }, + { + "epoch": 0.35179, + "grad_norm": 0.8779510855674744, + "learning_rate": 0.003, + "loss": 3.976, + "step": 35179 + }, + { + "epoch": 0.3518, + "grad_norm": 0.8209325075149536, + "learning_rate": 0.003, + "loss": 3.9426, + "step": 35180 + }, + { + "epoch": 0.35181, + "grad_norm": 0.775204598903656, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 35181 + }, + { + "epoch": 0.35182, + "grad_norm": 0.8029803037643433, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 35182 + }, + { + "epoch": 0.35183, + "grad_norm": 0.929930567741394, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 35183 + }, + { + "epoch": 0.35184, + "grad_norm": 0.9424013495445251, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 35184 + }, + { + "epoch": 0.35185, + "grad_norm": 1.1469762325286865, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 35185 + }, + { + "epoch": 0.35186, + "grad_norm": 1.0501465797424316, + "learning_rate": 0.003, + "loss": 4.012, + "step": 35186 + }, + { + "epoch": 0.35187, + "grad_norm": 0.947636067867279, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 35187 + }, + { + "epoch": 0.35188, + "grad_norm": 1.0741206407546997, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 35188 + }, + { + "epoch": 0.35189, + "grad_norm": 1.1675184965133667, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 35189 + }, + { + "epoch": 0.3519, + "grad_norm": 0.7124341726303101, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 35190 + }, + { + "epoch": 0.35191, + "grad_norm": 0.5986812710762024, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 35191 + }, + { + "epoch": 0.35192, + "grad_norm": 0.6753842234611511, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 35192 + }, + { + "epoch": 0.35193, + "grad_norm": 0.8613893985748291, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 35193 + }, + { + "epoch": 0.35194, + "grad_norm": 1.0533875226974487, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 35194 + }, + { + "epoch": 0.35195, + "grad_norm": 1.0774996280670166, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 35195 + }, + { + "epoch": 0.35196, + "grad_norm": 0.7957283854484558, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 35196 + }, + { + "epoch": 0.35197, + "grad_norm": 0.8276906609535217, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 35197 + }, + { + "epoch": 0.35198, + "grad_norm": 0.8854073882102966, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 35198 + }, + { + "epoch": 0.35199, + "grad_norm": 0.8904858827590942, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 35199 + }, + { + "epoch": 0.352, + "grad_norm": 0.7825853824615479, + "learning_rate": 0.003, + "loss": 3.9432, + "step": 35200 + }, + { + "epoch": 0.35201, + "grad_norm": 0.7413291335105896, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 35201 + }, + { + "epoch": 0.35202, + "grad_norm": 0.8127537369728088, + "learning_rate": 0.003, + "loss": 3.961, + "step": 35202 + }, + { + "epoch": 0.35203, + "grad_norm": 0.8755364418029785, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 35203 + }, + { + "epoch": 0.35204, + "grad_norm": 0.8487808108329773, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 35204 + }, + { + "epoch": 0.35205, + "grad_norm": 0.9723891615867615, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 35205 + }, + { + "epoch": 0.35206, + "grad_norm": 1.1241954565048218, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 35206 + }, + { + "epoch": 0.35207, + "grad_norm": 1.0275077819824219, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 35207 + }, + { + "epoch": 0.35208, + "grad_norm": 0.980980396270752, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 35208 + }, + { + "epoch": 0.35209, + "grad_norm": 1.005293607711792, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 35209 + }, + { + "epoch": 0.3521, + "grad_norm": 0.9213323593139648, + "learning_rate": 0.003, + "loss": 3.97, + "step": 35210 + }, + { + "epoch": 0.35211, + "grad_norm": 0.8907598853111267, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 35211 + }, + { + "epoch": 0.35212, + "grad_norm": 0.9127060770988464, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 35212 + }, + { + "epoch": 0.35213, + "grad_norm": 1.0014716386795044, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 35213 + }, + { + "epoch": 0.35214, + "grad_norm": 1.4277255535125732, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 35214 + }, + { + "epoch": 0.35215, + "grad_norm": 0.8342350125312805, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 35215 + }, + { + "epoch": 0.35216, + "grad_norm": 0.8868569731712341, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 35216 + }, + { + "epoch": 0.35217, + "grad_norm": 0.859449565410614, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 35217 + }, + { + "epoch": 0.35218, + "grad_norm": 0.9108436703681946, + "learning_rate": 0.003, + "loss": 3.957, + "step": 35218 + }, + { + "epoch": 0.35219, + "grad_norm": 1.3101376295089722, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 35219 + }, + { + "epoch": 0.3522, + "grad_norm": 0.7896720170974731, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 35220 + }, + { + "epoch": 0.35221, + "grad_norm": 0.8086287379264832, + "learning_rate": 0.003, + "loss": 4.0, + "step": 35221 + }, + { + "epoch": 0.35222, + "grad_norm": 1.0601228475570679, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 35222 + }, + { + "epoch": 0.35223, + "grad_norm": 0.863907516002655, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 35223 + }, + { + "epoch": 0.35224, + "grad_norm": 0.7349860072135925, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 35224 + }, + { + "epoch": 0.35225, + "grad_norm": 0.7615607380867004, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 35225 + }, + { + "epoch": 0.35226, + "grad_norm": 0.9077712297439575, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 35226 + }, + { + "epoch": 0.35227, + "grad_norm": 1.0346013307571411, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 35227 + }, + { + "epoch": 0.35228, + "grad_norm": 1.0622833967208862, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 35228 + }, + { + "epoch": 0.35229, + "grad_norm": 0.9693940281867981, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 35229 + }, + { + "epoch": 0.3523, + "grad_norm": 0.7782523036003113, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 35230 + }, + { + "epoch": 0.35231, + "grad_norm": 0.6883679032325745, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 35231 + }, + { + "epoch": 0.35232, + "grad_norm": 0.718166708946228, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 35232 + }, + { + "epoch": 0.35233, + "grad_norm": 0.9844726920127869, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 35233 + }, + { + "epoch": 0.35234, + "grad_norm": 1.2133538722991943, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 35234 + }, + { + "epoch": 0.35235, + "grad_norm": 0.8402485847473145, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 35235 + }, + { + "epoch": 0.35236, + "grad_norm": 0.7784820795059204, + "learning_rate": 0.003, + "loss": 3.964, + "step": 35236 + }, + { + "epoch": 0.35237, + "grad_norm": 0.9642017483711243, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 35237 + }, + { + "epoch": 0.35238, + "grad_norm": 0.8903138637542725, + "learning_rate": 0.003, + "loss": 3.975, + "step": 35238 + }, + { + "epoch": 0.35239, + "grad_norm": 0.7265021800994873, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 35239 + }, + { + "epoch": 0.3524, + "grad_norm": 0.8606640100479126, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 35240 + }, + { + "epoch": 0.35241, + "grad_norm": 0.9261976480484009, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 35241 + }, + { + "epoch": 0.35242, + "grad_norm": 0.8834478855133057, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 35242 + }, + { + "epoch": 0.35243, + "grad_norm": 0.9512796401977539, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 35243 + }, + { + "epoch": 0.35244, + "grad_norm": 0.842113733291626, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 35244 + }, + { + "epoch": 0.35245, + "grad_norm": 0.8203734159469604, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 35245 + }, + { + "epoch": 0.35246, + "grad_norm": 0.8128701448440552, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 35246 + }, + { + "epoch": 0.35247, + "grad_norm": 0.7893966436386108, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 35247 + }, + { + "epoch": 0.35248, + "grad_norm": 0.8561751246452332, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 35248 + }, + { + "epoch": 0.35249, + "grad_norm": 0.9156550168991089, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 35249 + }, + { + "epoch": 0.3525, + "grad_norm": 1.0153067111968994, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 35250 + }, + { + "epoch": 0.35251, + "grad_norm": 1.1306862831115723, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 35251 + }, + { + "epoch": 0.35252, + "grad_norm": 0.7770462036132812, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 35252 + }, + { + "epoch": 0.35253, + "grad_norm": 0.8466492891311646, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 35253 + }, + { + "epoch": 0.35254, + "grad_norm": 0.9393610954284668, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 35254 + }, + { + "epoch": 0.35255, + "grad_norm": 1.0135191679000854, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 35255 + }, + { + "epoch": 0.35256, + "grad_norm": 1.1372640132904053, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 35256 + }, + { + "epoch": 0.35257, + "grad_norm": 0.9782729148864746, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 35257 + }, + { + "epoch": 0.35258, + "grad_norm": 0.994849443435669, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 35258 + }, + { + "epoch": 0.35259, + "grad_norm": 1.007148265838623, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 35259 + }, + { + "epoch": 0.3526, + "grad_norm": 0.9693145155906677, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 35260 + }, + { + "epoch": 0.35261, + "grad_norm": 0.9547803401947021, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 35261 + }, + { + "epoch": 0.35262, + "grad_norm": 0.9327895045280457, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 35262 + }, + { + "epoch": 0.35263, + "grad_norm": 0.9150780439376831, + "learning_rate": 0.003, + "loss": 4.016, + "step": 35263 + }, + { + "epoch": 0.35264, + "grad_norm": 0.9507908225059509, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 35264 + }, + { + "epoch": 0.35265, + "grad_norm": 1.366801142692566, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 35265 + }, + { + "epoch": 0.35266, + "grad_norm": 0.7135955095291138, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 35266 + }, + { + "epoch": 0.35267, + "grad_norm": 0.7145063281059265, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 35267 + }, + { + "epoch": 0.35268, + "grad_norm": 0.6695623397827148, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 35268 + }, + { + "epoch": 0.35269, + "grad_norm": 0.7032607197761536, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 35269 + }, + { + "epoch": 0.3527, + "grad_norm": 0.6700305938720703, + "learning_rate": 0.003, + "loss": 4.007, + "step": 35270 + }, + { + "epoch": 0.35271, + "grad_norm": 0.8377828598022461, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 35271 + }, + { + "epoch": 0.35272, + "grad_norm": 1.4803011417388916, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 35272 + }, + { + "epoch": 0.35273, + "grad_norm": 0.6571046113967896, + "learning_rate": 0.003, + "loss": 3.981, + "step": 35273 + }, + { + "epoch": 0.35274, + "grad_norm": 0.6335147023200989, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 35274 + }, + { + "epoch": 0.35275, + "grad_norm": 0.7145741581916809, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 35275 + }, + { + "epoch": 0.35276, + "grad_norm": 0.8031078577041626, + "learning_rate": 0.003, + "loss": 3.964, + "step": 35276 + }, + { + "epoch": 0.35277, + "grad_norm": 0.8516855239868164, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 35277 + }, + { + "epoch": 0.35278, + "grad_norm": 0.8731687664985657, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 35278 + }, + { + "epoch": 0.35279, + "grad_norm": 0.9233715534210205, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 35279 + }, + { + "epoch": 0.3528, + "grad_norm": 1.0796846151351929, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 35280 + }, + { + "epoch": 0.35281, + "grad_norm": 0.8988475203514099, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 35281 + }, + { + "epoch": 0.35282, + "grad_norm": 0.8588182926177979, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 35282 + }, + { + "epoch": 0.35283, + "grad_norm": 0.9472607970237732, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 35283 + }, + { + "epoch": 0.35284, + "grad_norm": 1.0846076011657715, + "learning_rate": 0.003, + "loss": 3.962, + "step": 35284 + }, + { + "epoch": 0.35285, + "grad_norm": 0.7394457459449768, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 35285 + }, + { + "epoch": 0.35286, + "grad_norm": 0.6155466437339783, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 35286 + }, + { + "epoch": 0.35287, + "grad_norm": 0.6536298394203186, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 35287 + }, + { + "epoch": 0.35288, + "grad_norm": 0.8173746466636658, + "learning_rate": 0.003, + "loss": 3.985, + "step": 35288 + }, + { + "epoch": 0.35289, + "grad_norm": 1.117172122001648, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 35289 + }, + { + "epoch": 0.3529, + "grad_norm": 1.29881751537323, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 35290 + }, + { + "epoch": 0.35291, + "grad_norm": 0.7601077556610107, + "learning_rate": 0.003, + "loss": 3.966, + "step": 35291 + }, + { + "epoch": 0.35292, + "grad_norm": 0.7477758526802063, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 35292 + }, + { + "epoch": 0.35293, + "grad_norm": 0.7160312533378601, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 35293 + }, + { + "epoch": 0.35294, + "grad_norm": 0.6312231421470642, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 35294 + }, + { + "epoch": 0.35295, + "grad_norm": 0.6320456862449646, + "learning_rate": 0.003, + "loss": 3.99, + "step": 35295 + }, + { + "epoch": 0.35296, + "grad_norm": 0.7646296620368958, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 35296 + }, + { + "epoch": 0.35297, + "grad_norm": 0.8417531251907349, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 35297 + }, + { + "epoch": 0.35298, + "grad_norm": 0.8785688281059265, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 35298 + }, + { + "epoch": 0.35299, + "grad_norm": 0.9916779398918152, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 35299 + }, + { + "epoch": 0.353, + "grad_norm": 1.3543047904968262, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 35300 + }, + { + "epoch": 0.35301, + "grad_norm": 0.7830851078033447, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 35301 + }, + { + "epoch": 0.35302, + "grad_norm": 0.7678887248039246, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 35302 + }, + { + "epoch": 0.35303, + "grad_norm": 0.7899602055549622, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 35303 + }, + { + "epoch": 0.35304, + "grad_norm": 0.899998128414154, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 35304 + }, + { + "epoch": 0.35305, + "grad_norm": 1.1335501670837402, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 35305 + }, + { + "epoch": 0.35306, + "grad_norm": 0.8936915993690491, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 35306 + }, + { + "epoch": 0.35307, + "grad_norm": 1.1022878885269165, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 35307 + }, + { + "epoch": 0.35308, + "grad_norm": 1.098985195159912, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 35308 + }, + { + "epoch": 0.35309, + "grad_norm": 0.8251752257347107, + "learning_rate": 0.003, + "loss": 3.9451, + "step": 35309 + }, + { + "epoch": 0.3531, + "grad_norm": 0.7433997392654419, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 35310 + }, + { + "epoch": 0.35311, + "grad_norm": 0.71710205078125, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 35311 + }, + { + "epoch": 0.35312, + "grad_norm": 0.8490052819252014, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 35312 + }, + { + "epoch": 0.35313, + "grad_norm": 0.9362645745277405, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 35313 + }, + { + "epoch": 0.35314, + "grad_norm": 1.2844951152801514, + "learning_rate": 0.003, + "loss": 3.9442, + "step": 35314 + }, + { + "epoch": 0.35315, + "grad_norm": 0.6811177134513855, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 35315 + }, + { + "epoch": 0.35316, + "grad_norm": 0.6853718161582947, + "learning_rate": 0.003, + "loss": 3.9299, + "step": 35316 + }, + { + "epoch": 0.35317, + "grad_norm": 0.6928508877754211, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 35317 + }, + { + "epoch": 0.35318, + "grad_norm": 0.6098600625991821, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 35318 + }, + { + "epoch": 0.35319, + "grad_norm": 0.6666731238365173, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 35319 + }, + { + "epoch": 0.3532, + "grad_norm": 0.9500723481178284, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 35320 + }, + { + "epoch": 0.35321, + "grad_norm": 1.4879956245422363, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 35321 + }, + { + "epoch": 0.35322, + "grad_norm": 0.5403308272361755, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 35322 + }, + { + "epoch": 0.35323, + "grad_norm": 0.8518301844596863, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 35323 + }, + { + "epoch": 0.35324, + "grad_norm": 0.9413281679153442, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 35324 + }, + { + "epoch": 0.35325, + "grad_norm": 0.9081847667694092, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 35325 + }, + { + "epoch": 0.35326, + "grad_norm": 1.0487157106399536, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 35326 + }, + { + "epoch": 0.35327, + "grad_norm": 1.1394994258880615, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 35327 + }, + { + "epoch": 0.35328, + "grad_norm": 0.8568704128265381, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 35328 + }, + { + "epoch": 0.35329, + "grad_norm": 0.9291459321975708, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 35329 + }, + { + "epoch": 0.3533, + "grad_norm": 0.8792229294776917, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 35330 + }, + { + "epoch": 0.35331, + "grad_norm": 1.1937721967697144, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 35331 + }, + { + "epoch": 0.35332, + "grad_norm": 1.0907506942749023, + "learning_rate": 0.003, + "loss": 4.006, + "step": 35332 + }, + { + "epoch": 0.35333, + "grad_norm": 0.8173921704292297, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 35333 + }, + { + "epoch": 0.35334, + "grad_norm": 0.8911704421043396, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 35334 + }, + { + "epoch": 0.35335, + "grad_norm": 0.9478369951248169, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 35335 + }, + { + "epoch": 0.35336, + "grad_norm": 0.8855272531509399, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 35336 + }, + { + "epoch": 0.35337, + "grad_norm": 0.7214560508728027, + "learning_rate": 0.003, + "loss": 3.956, + "step": 35337 + }, + { + "epoch": 0.35338, + "grad_norm": 0.8509708642959595, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 35338 + }, + { + "epoch": 0.35339, + "grad_norm": 0.782162606716156, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 35339 + }, + { + "epoch": 0.3534, + "grad_norm": 0.8214519619941711, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 35340 + }, + { + "epoch": 0.35341, + "grad_norm": 0.8036959767341614, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 35341 + }, + { + "epoch": 0.35342, + "grad_norm": 0.9328332543373108, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 35342 + }, + { + "epoch": 0.35343, + "grad_norm": 0.906527042388916, + "learning_rate": 0.003, + "loss": 3.988, + "step": 35343 + }, + { + "epoch": 0.35344, + "grad_norm": 0.9589279890060425, + "learning_rate": 0.003, + "loss": 3.976, + "step": 35344 + }, + { + "epoch": 0.35345, + "grad_norm": 1.4493154287338257, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 35345 + }, + { + "epoch": 0.35346, + "grad_norm": 0.7869017720222473, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 35346 + }, + { + "epoch": 0.35347, + "grad_norm": 0.8432360887527466, + "learning_rate": 0.003, + "loss": 3.999, + "step": 35347 + }, + { + "epoch": 0.35348, + "grad_norm": 0.9475392699241638, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 35348 + }, + { + "epoch": 0.35349, + "grad_norm": 0.8605174422264099, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 35349 + }, + { + "epoch": 0.3535, + "grad_norm": 0.9314536452293396, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 35350 + }, + { + "epoch": 0.35351, + "grad_norm": 1.0548399686813354, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 35351 + }, + { + "epoch": 0.35352, + "grad_norm": 1.0206551551818848, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 35352 + }, + { + "epoch": 0.35353, + "grad_norm": 1.2323381900787354, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 35353 + }, + { + "epoch": 0.35354, + "grad_norm": 1.0464237928390503, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 35354 + }, + { + "epoch": 0.35355, + "grad_norm": 1.1161460876464844, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 35355 + }, + { + "epoch": 0.35356, + "grad_norm": 0.7621923089027405, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 35356 + }, + { + "epoch": 0.35357, + "grad_norm": 0.7036759257316589, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 35357 + }, + { + "epoch": 0.35358, + "grad_norm": 0.8331112265586853, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 35358 + }, + { + "epoch": 0.35359, + "grad_norm": 1.1177122592926025, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 35359 + }, + { + "epoch": 0.3536, + "grad_norm": 0.9492641091346741, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 35360 + }, + { + "epoch": 0.35361, + "grad_norm": 0.7613819241523743, + "learning_rate": 0.003, + "loss": 3.963, + "step": 35361 + }, + { + "epoch": 0.35362, + "grad_norm": 0.8522034883499146, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 35362 + }, + { + "epoch": 0.35363, + "grad_norm": 0.8214003443717957, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 35363 + }, + { + "epoch": 0.35364, + "grad_norm": 0.7939426898956299, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 35364 + }, + { + "epoch": 0.35365, + "grad_norm": 0.7007430195808411, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 35365 + }, + { + "epoch": 0.35366, + "grad_norm": 0.6497138142585754, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 35366 + }, + { + "epoch": 0.35367, + "grad_norm": 0.6756111979484558, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 35367 + }, + { + "epoch": 0.35368, + "grad_norm": 0.857319176197052, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 35368 + }, + { + "epoch": 0.35369, + "grad_norm": 1.1820567846298218, + "learning_rate": 0.003, + "loss": 3.9386, + "step": 35369 + }, + { + "epoch": 0.3537, + "grad_norm": 0.9222732186317444, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 35370 + }, + { + "epoch": 0.35371, + "grad_norm": 0.9842205047607422, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 35371 + }, + { + "epoch": 0.35372, + "grad_norm": 1.0773873329162598, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 35372 + }, + { + "epoch": 0.35373, + "grad_norm": 0.8193306922912598, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 35373 + }, + { + "epoch": 0.35374, + "grad_norm": 0.5611608028411865, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 35374 + }, + { + "epoch": 0.35375, + "grad_norm": 0.7274740934371948, + "learning_rate": 0.003, + "loss": 3.971, + "step": 35375 + }, + { + "epoch": 0.35376, + "grad_norm": 0.8694683313369751, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 35376 + }, + { + "epoch": 0.35377, + "grad_norm": 0.9977729916572571, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 35377 + }, + { + "epoch": 0.35378, + "grad_norm": 1.2091748714447021, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 35378 + }, + { + "epoch": 0.35379, + "grad_norm": 0.8397759199142456, + "learning_rate": 0.003, + "loss": 3.961, + "step": 35379 + }, + { + "epoch": 0.3538, + "grad_norm": 1.1237396001815796, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 35380 + }, + { + "epoch": 0.35381, + "grad_norm": 1.2991292476654053, + "learning_rate": 0.003, + "loss": 3.988, + "step": 35381 + }, + { + "epoch": 0.35382, + "grad_norm": 0.8151851892471313, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 35382 + }, + { + "epoch": 0.35383, + "grad_norm": 0.8054969310760498, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 35383 + }, + { + "epoch": 0.35384, + "grad_norm": 0.7550609111785889, + "learning_rate": 0.003, + "loss": 3.98, + "step": 35384 + }, + { + "epoch": 0.35385, + "grad_norm": 0.8268976211547852, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 35385 + }, + { + "epoch": 0.35386, + "grad_norm": 0.9203519821166992, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 35386 + }, + { + "epoch": 0.35387, + "grad_norm": 1.034287929534912, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 35387 + }, + { + "epoch": 0.35388, + "grad_norm": 0.9447000026702881, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 35388 + }, + { + "epoch": 0.35389, + "grad_norm": 1.0739188194274902, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 35389 + }, + { + "epoch": 0.3539, + "grad_norm": 1.0430101156234741, + "learning_rate": 0.003, + "loss": 4.024, + "step": 35390 + }, + { + "epoch": 0.35391, + "grad_norm": 0.9160280823707581, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 35391 + }, + { + "epoch": 0.35392, + "grad_norm": 1.014533281326294, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 35392 + }, + { + "epoch": 0.35393, + "grad_norm": 1.0908145904541016, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 35393 + }, + { + "epoch": 0.35394, + "grad_norm": 0.8469796776771545, + "learning_rate": 0.003, + "loss": 4.01, + "step": 35394 + }, + { + "epoch": 0.35395, + "grad_norm": 0.7046872973442078, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 35395 + }, + { + "epoch": 0.35396, + "grad_norm": 0.7469837069511414, + "learning_rate": 0.003, + "loss": 3.993, + "step": 35396 + }, + { + "epoch": 0.35397, + "grad_norm": 0.8267151117324829, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 35397 + }, + { + "epoch": 0.35398, + "grad_norm": 1.0177562236785889, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 35398 + }, + { + "epoch": 0.35399, + "grad_norm": 1.0390369892120361, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 35399 + }, + { + "epoch": 0.354, + "grad_norm": 1.0481858253479004, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 35400 + }, + { + "epoch": 0.35401, + "grad_norm": 0.8463503122329712, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 35401 + }, + { + "epoch": 0.35402, + "grad_norm": 0.8192614316940308, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 35402 + }, + { + "epoch": 0.35403, + "grad_norm": 0.8754874467849731, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 35403 + }, + { + "epoch": 0.35404, + "grad_norm": 1.086499571800232, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 35404 + }, + { + "epoch": 0.35405, + "grad_norm": 1.2024633884429932, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 35405 + }, + { + "epoch": 0.35406, + "grad_norm": 0.8029123544692993, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 35406 + }, + { + "epoch": 0.35407, + "grad_norm": 0.6781187653541565, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 35407 + }, + { + "epoch": 0.35408, + "grad_norm": 0.6279599666595459, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 35408 + }, + { + "epoch": 0.35409, + "grad_norm": 0.618173360824585, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 35409 + }, + { + "epoch": 0.3541, + "grad_norm": 0.6953077912330627, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 35410 + }, + { + "epoch": 0.35411, + "grad_norm": 0.8103616237640381, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 35411 + }, + { + "epoch": 0.35412, + "grad_norm": 0.8788288831710815, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 35412 + }, + { + "epoch": 0.35413, + "grad_norm": 0.9157097935676575, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 35413 + }, + { + "epoch": 0.35414, + "grad_norm": 1.0306650400161743, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 35414 + }, + { + "epoch": 0.35415, + "grad_norm": 1.0000234842300415, + "learning_rate": 0.003, + "loss": 3.983, + "step": 35415 + }, + { + "epoch": 0.35416, + "grad_norm": 0.8667392134666443, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 35416 + }, + { + "epoch": 0.35417, + "grad_norm": 0.7301808595657349, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 35417 + }, + { + "epoch": 0.35418, + "grad_norm": 0.7449606657028198, + "learning_rate": 0.003, + "loss": 3.993, + "step": 35418 + }, + { + "epoch": 0.35419, + "grad_norm": 0.7600361704826355, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 35419 + }, + { + "epoch": 0.3542, + "grad_norm": 0.71144038438797, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 35420 + }, + { + "epoch": 0.35421, + "grad_norm": 0.6967000961303711, + "learning_rate": 0.003, + "loss": 3.98, + "step": 35421 + }, + { + "epoch": 0.35422, + "grad_norm": 0.7823388576507568, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 35422 + }, + { + "epoch": 0.35423, + "grad_norm": 0.9434642791748047, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 35423 + }, + { + "epoch": 0.35424, + "grad_norm": 1.095715045928955, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 35424 + }, + { + "epoch": 0.35425, + "grad_norm": 0.8532224297523499, + "learning_rate": 0.003, + "loss": 3.996, + "step": 35425 + }, + { + "epoch": 0.35426, + "grad_norm": 0.8605766296386719, + "learning_rate": 0.003, + "loss": 4.009, + "step": 35426 + }, + { + "epoch": 0.35427, + "grad_norm": 1.128010869026184, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 35427 + }, + { + "epoch": 0.35428, + "grad_norm": 1.2613928318023682, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 35428 + }, + { + "epoch": 0.35429, + "grad_norm": 0.8013793230056763, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 35429 + }, + { + "epoch": 0.3543, + "grad_norm": 0.8716922402381897, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 35430 + }, + { + "epoch": 0.35431, + "grad_norm": 1.1155626773834229, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 35431 + }, + { + "epoch": 0.35432, + "grad_norm": 1.1297112703323364, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 35432 + }, + { + "epoch": 0.35433, + "grad_norm": 0.7538715600967407, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 35433 + }, + { + "epoch": 0.35434, + "grad_norm": 0.8125194907188416, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 35434 + }, + { + "epoch": 0.35435, + "grad_norm": 0.8040494918823242, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 35435 + }, + { + "epoch": 0.35436, + "grad_norm": 0.9027365446090698, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 35436 + }, + { + "epoch": 0.35437, + "grad_norm": 1.013247013092041, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 35437 + }, + { + "epoch": 0.35438, + "grad_norm": 1.1300325393676758, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 35438 + }, + { + "epoch": 0.35439, + "grad_norm": 0.8648772239685059, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 35439 + }, + { + "epoch": 0.3544, + "grad_norm": 0.8757404088973999, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 35440 + }, + { + "epoch": 0.35441, + "grad_norm": 1.0758622884750366, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 35441 + }, + { + "epoch": 0.35442, + "grad_norm": 0.9672243595123291, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 35442 + }, + { + "epoch": 0.35443, + "grad_norm": 0.9462171196937561, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 35443 + }, + { + "epoch": 0.35444, + "grad_norm": 0.8410253524780273, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 35444 + }, + { + "epoch": 0.35445, + "grad_norm": 0.9107125401496887, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 35445 + }, + { + "epoch": 0.35446, + "grad_norm": 1.0851014852523804, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 35446 + }, + { + "epoch": 0.35447, + "grad_norm": 1.0668519735336304, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 35447 + }, + { + "epoch": 0.35448, + "grad_norm": 0.9362794160842896, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 35448 + }, + { + "epoch": 0.35449, + "grad_norm": 0.935173749923706, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 35449 + }, + { + "epoch": 0.3545, + "grad_norm": 0.9199188947677612, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 35450 + }, + { + "epoch": 0.35451, + "grad_norm": 1.0142021179199219, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 35451 + }, + { + "epoch": 0.35452, + "grad_norm": 1.0378820896148682, + "learning_rate": 0.003, + "loss": 3.984, + "step": 35452 + }, + { + "epoch": 0.35453, + "grad_norm": 0.9168224930763245, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 35453 + }, + { + "epoch": 0.35454, + "grad_norm": 0.8890273571014404, + "learning_rate": 0.003, + "loss": 3.985, + "step": 35454 + }, + { + "epoch": 0.35455, + "grad_norm": 0.8279982209205627, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 35455 + }, + { + "epoch": 0.35456, + "grad_norm": 1.0316218137741089, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 35456 + }, + { + "epoch": 0.35457, + "grad_norm": 0.9520313143730164, + "learning_rate": 0.003, + "loss": 4.003, + "step": 35457 + }, + { + "epoch": 0.35458, + "grad_norm": 0.9176859855651855, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 35458 + }, + { + "epoch": 0.35459, + "grad_norm": 1.2919073104858398, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 35459 + }, + { + "epoch": 0.3546, + "grad_norm": 0.9970171451568604, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 35460 + }, + { + "epoch": 0.35461, + "grad_norm": 0.9319819808006287, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 35461 + }, + { + "epoch": 0.35462, + "grad_norm": 1.0149905681610107, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 35462 + }, + { + "epoch": 0.35463, + "grad_norm": 1.2887245416641235, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 35463 + }, + { + "epoch": 0.35464, + "grad_norm": 0.703423798084259, + "learning_rate": 0.003, + "loss": 3.964, + "step": 35464 + }, + { + "epoch": 0.35465, + "grad_norm": 0.7487227916717529, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 35465 + }, + { + "epoch": 0.35466, + "grad_norm": 0.7941837906837463, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 35466 + }, + { + "epoch": 0.35467, + "grad_norm": 0.7907072901725769, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 35467 + }, + { + "epoch": 0.35468, + "grad_norm": 0.812412679195404, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 35468 + }, + { + "epoch": 0.35469, + "grad_norm": 0.9836657047271729, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 35469 + }, + { + "epoch": 0.3547, + "grad_norm": 1.0967901945114136, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 35470 + }, + { + "epoch": 0.35471, + "grad_norm": 0.8745702505111694, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 35471 + }, + { + "epoch": 0.35472, + "grad_norm": 0.7578063607215881, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 35472 + }, + { + "epoch": 0.35473, + "grad_norm": 0.6493932604789734, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 35473 + }, + { + "epoch": 0.35474, + "grad_norm": 0.733720600605011, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 35474 + }, + { + "epoch": 0.35475, + "grad_norm": 0.9625774025917053, + "learning_rate": 0.003, + "loss": 3.991, + "step": 35475 + }, + { + "epoch": 0.35476, + "grad_norm": 1.3611301183700562, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 35476 + }, + { + "epoch": 0.35477, + "grad_norm": 0.6345595717430115, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 35477 + }, + { + "epoch": 0.35478, + "grad_norm": 0.6860199570655823, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 35478 + }, + { + "epoch": 0.35479, + "grad_norm": 0.8111511468887329, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 35479 + }, + { + "epoch": 0.3548, + "grad_norm": 0.7453352808952332, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 35480 + }, + { + "epoch": 0.35481, + "grad_norm": 0.8871229290962219, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 35481 + }, + { + "epoch": 0.35482, + "grad_norm": 0.9878688454627991, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 35482 + }, + { + "epoch": 0.35483, + "grad_norm": 1.0040332078933716, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 35483 + }, + { + "epoch": 0.35484, + "grad_norm": 0.8749128580093384, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 35484 + }, + { + "epoch": 0.35485, + "grad_norm": 0.825527548789978, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 35485 + }, + { + "epoch": 0.35486, + "grad_norm": 0.8728793263435364, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 35486 + }, + { + "epoch": 0.35487, + "grad_norm": 0.9833179712295532, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 35487 + }, + { + "epoch": 0.35488, + "grad_norm": 1.089573621749878, + "learning_rate": 0.003, + "loss": 3.98, + "step": 35488 + }, + { + "epoch": 0.35489, + "grad_norm": 0.9619688391685486, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 35489 + }, + { + "epoch": 0.3549, + "grad_norm": 0.9527624249458313, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 35490 + }, + { + "epoch": 0.35491, + "grad_norm": 1.0251638889312744, + "learning_rate": 0.003, + "loss": 3.948, + "step": 35491 + }, + { + "epoch": 0.35492, + "grad_norm": 1.0619412660598755, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 35492 + }, + { + "epoch": 0.35493, + "grad_norm": 0.9820195436477661, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 35493 + }, + { + "epoch": 0.35494, + "grad_norm": 0.9392622113227844, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 35494 + }, + { + "epoch": 0.35495, + "grad_norm": 0.9515022039413452, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 35495 + }, + { + "epoch": 0.35496, + "grad_norm": 1.062843918800354, + "learning_rate": 0.003, + "loss": 3.9478, + "step": 35496 + }, + { + "epoch": 0.35497, + "grad_norm": 1.3063126802444458, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 35497 + }, + { + "epoch": 0.35498, + "grad_norm": 0.7803146243095398, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 35498 + }, + { + "epoch": 0.35499, + "grad_norm": 0.5793629884719849, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 35499 + }, + { + "epoch": 0.355, + "grad_norm": 0.6245570182800293, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 35500 + }, + { + "epoch": 0.35501, + "grad_norm": 0.7711560130119324, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 35501 + }, + { + "epoch": 0.35502, + "grad_norm": 1.0107042789459229, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 35502 + }, + { + "epoch": 0.35503, + "grad_norm": 1.1380282640457153, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 35503 + }, + { + "epoch": 0.35504, + "grad_norm": 0.8346614241600037, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 35504 + }, + { + "epoch": 0.35505, + "grad_norm": 0.8175058364868164, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 35505 + }, + { + "epoch": 0.35506, + "grad_norm": 0.7820791602134705, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 35506 + }, + { + "epoch": 0.35507, + "grad_norm": 0.999256432056427, + "learning_rate": 0.003, + "loss": 3.978, + "step": 35507 + }, + { + "epoch": 0.35508, + "grad_norm": 1.214651346206665, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 35508 + }, + { + "epoch": 0.35509, + "grad_norm": 0.7652456164360046, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 35509 + }, + { + "epoch": 0.3551, + "grad_norm": 0.6421256065368652, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 35510 + }, + { + "epoch": 0.35511, + "grad_norm": 0.6882007718086243, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 35511 + }, + { + "epoch": 0.35512, + "grad_norm": 0.6822824478149414, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 35512 + }, + { + "epoch": 0.35513, + "grad_norm": 0.6671320199966431, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 35513 + }, + { + "epoch": 0.35514, + "grad_norm": 0.7272623777389526, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 35514 + }, + { + "epoch": 0.35515, + "grad_norm": 0.700523853302002, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 35515 + }, + { + "epoch": 0.35516, + "grad_norm": 0.6863163113594055, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 35516 + }, + { + "epoch": 0.35517, + "grad_norm": 0.6981010437011719, + "learning_rate": 0.003, + "loss": 3.997, + "step": 35517 + }, + { + "epoch": 0.35518, + "grad_norm": 0.757818341255188, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 35518 + }, + { + "epoch": 0.35519, + "grad_norm": 1.0536009073257446, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 35519 + }, + { + "epoch": 0.3552, + "grad_norm": 1.3752481937408447, + "learning_rate": 0.003, + "loss": 3.98, + "step": 35520 + }, + { + "epoch": 0.35521, + "grad_norm": 0.6689863801002502, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 35521 + }, + { + "epoch": 0.35522, + "grad_norm": 0.6424327492713928, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 35522 + }, + { + "epoch": 0.35523, + "grad_norm": 0.9659270644187927, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 35523 + }, + { + "epoch": 0.35524, + "grad_norm": 1.1690627336502075, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 35524 + }, + { + "epoch": 0.35525, + "grad_norm": 0.9457510709762573, + "learning_rate": 0.003, + "loss": 3.994, + "step": 35525 + }, + { + "epoch": 0.35526, + "grad_norm": 0.916561484336853, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 35526 + }, + { + "epoch": 0.35527, + "grad_norm": 0.8006736636161804, + "learning_rate": 0.003, + "loss": 3.994, + "step": 35527 + }, + { + "epoch": 0.35528, + "grad_norm": 0.7330573201179504, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 35528 + }, + { + "epoch": 0.35529, + "grad_norm": 0.781795859336853, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 35529 + }, + { + "epoch": 0.3553, + "grad_norm": 0.7795097827911377, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 35530 + }, + { + "epoch": 0.35531, + "grad_norm": 0.7920577526092529, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 35531 + }, + { + "epoch": 0.35532, + "grad_norm": 0.8522545099258423, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 35532 + }, + { + "epoch": 0.35533, + "grad_norm": 1.124293327331543, + "learning_rate": 0.003, + "loss": 4.002, + "step": 35533 + }, + { + "epoch": 0.35534, + "grad_norm": 1.3547742366790771, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 35534 + }, + { + "epoch": 0.35535, + "grad_norm": 0.7757578492164612, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 35535 + }, + { + "epoch": 0.35536, + "grad_norm": 0.6745700240135193, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 35536 + }, + { + "epoch": 0.35537, + "grad_norm": 0.8500911593437195, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 35537 + }, + { + "epoch": 0.35538, + "grad_norm": 1.0200393199920654, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 35538 + }, + { + "epoch": 0.35539, + "grad_norm": 1.0315505266189575, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 35539 + }, + { + "epoch": 0.3554, + "grad_norm": 1.040452241897583, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 35540 + }, + { + "epoch": 0.35541, + "grad_norm": 0.9882364869117737, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 35541 + }, + { + "epoch": 0.35542, + "grad_norm": 0.9347600340843201, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 35542 + }, + { + "epoch": 0.35543, + "grad_norm": 0.7992668747901917, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 35543 + }, + { + "epoch": 0.35544, + "grad_norm": 0.8421900868415833, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 35544 + }, + { + "epoch": 0.35545, + "grad_norm": 0.8479865789413452, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 35545 + }, + { + "epoch": 0.35546, + "grad_norm": 0.9616513252258301, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 35546 + }, + { + "epoch": 0.35547, + "grad_norm": 1.0016796588897705, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 35547 + }, + { + "epoch": 0.35548, + "grad_norm": 1.0891304016113281, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 35548 + }, + { + "epoch": 0.35549, + "grad_norm": 1.1732654571533203, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 35549 + }, + { + "epoch": 0.3555, + "grad_norm": 0.8235320448875427, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 35550 + }, + { + "epoch": 0.35551, + "grad_norm": 0.734766960144043, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 35551 + }, + { + "epoch": 0.35552, + "grad_norm": 0.6129106879234314, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 35552 + }, + { + "epoch": 0.35553, + "grad_norm": 0.6646267771720886, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 35553 + }, + { + "epoch": 0.35554, + "grad_norm": 0.7705497145652771, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 35554 + }, + { + "epoch": 0.35555, + "grad_norm": 0.9253442287445068, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 35555 + }, + { + "epoch": 0.35556, + "grad_norm": 1.1804442405700684, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 35556 + }, + { + "epoch": 0.35557, + "grad_norm": 1.1431174278259277, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 35557 + }, + { + "epoch": 0.35558, + "grad_norm": 1.0447896718978882, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 35558 + }, + { + "epoch": 0.35559, + "grad_norm": 1.0627949237823486, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 35559 + }, + { + "epoch": 0.3556, + "grad_norm": 1.0467296838760376, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 35560 + }, + { + "epoch": 0.35561, + "grad_norm": 0.7669916152954102, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 35561 + }, + { + "epoch": 0.35562, + "grad_norm": 0.6258208155632019, + "learning_rate": 0.003, + "loss": 3.976, + "step": 35562 + }, + { + "epoch": 0.35563, + "grad_norm": 0.6409583687782288, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 35563 + }, + { + "epoch": 0.35564, + "grad_norm": 0.706049382686615, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 35564 + }, + { + "epoch": 0.35565, + "grad_norm": 0.7865723967552185, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 35565 + }, + { + "epoch": 0.35566, + "grad_norm": 0.9072043895721436, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 35566 + }, + { + "epoch": 0.35567, + "grad_norm": 0.952128529548645, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 35567 + }, + { + "epoch": 0.35568, + "grad_norm": 1.0123432874679565, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 35568 + }, + { + "epoch": 0.35569, + "grad_norm": 0.9036015868186951, + "learning_rate": 0.003, + "loss": 3.997, + "step": 35569 + }, + { + "epoch": 0.3557, + "grad_norm": 0.7306845188140869, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 35570 + }, + { + "epoch": 0.35571, + "grad_norm": 0.664493203163147, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 35571 + }, + { + "epoch": 0.35572, + "grad_norm": 0.7596365213394165, + "learning_rate": 0.003, + "loss": 3.969, + "step": 35572 + }, + { + "epoch": 0.35573, + "grad_norm": 0.8539884686470032, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 35573 + }, + { + "epoch": 0.35574, + "grad_norm": 0.949173629283905, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 35574 + }, + { + "epoch": 0.35575, + "grad_norm": 1.0024205446243286, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 35575 + }, + { + "epoch": 0.35576, + "grad_norm": 1.2514322996139526, + "learning_rate": 0.003, + "loss": 3.992, + "step": 35576 + }, + { + "epoch": 0.35577, + "grad_norm": 0.8491456508636475, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 35577 + }, + { + "epoch": 0.35578, + "grad_norm": 0.7267122864723206, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 35578 + }, + { + "epoch": 0.35579, + "grad_norm": 0.7332894206047058, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 35579 + }, + { + "epoch": 0.3558, + "grad_norm": 0.8950167894363403, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 35580 + }, + { + "epoch": 0.35581, + "grad_norm": 1.1209371089935303, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 35581 + }, + { + "epoch": 0.35582, + "grad_norm": 0.8186702132225037, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 35582 + }, + { + "epoch": 0.35583, + "grad_norm": 0.7833301424980164, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 35583 + }, + { + "epoch": 0.35584, + "grad_norm": 0.8680943846702576, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 35584 + }, + { + "epoch": 0.35585, + "grad_norm": 0.9332399368286133, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 35585 + }, + { + "epoch": 0.35586, + "grad_norm": 0.8722242712974548, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 35586 + }, + { + "epoch": 0.35587, + "grad_norm": 0.8519914746284485, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 35587 + }, + { + "epoch": 0.35588, + "grad_norm": 1.013642430305481, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 35588 + }, + { + "epoch": 0.35589, + "grad_norm": 0.9928181171417236, + "learning_rate": 0.003, + "loss": 3.997, + "step": 35589 + }, + { + "epoch": 0.3559, + "grad_norm": 0.9070024490356445, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 35590 + }, + { + "epoch": 0.35591, + "grad_norm": 1.028397798538208, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 35591 + }, + { + "epoch": 0.35592, + "grad_norm": 1.211435317993164, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 35592 + }, + { + "epoch": 0.35593, + "grad_norm": 1.1397463083267212, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 35593 + }, + { + "epoch": 0.35594, + "grad_norm": 0.9007595181465149, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 35594 + }, + { + "epoch": 0.35595, + "grad_norm": 0.8903830647468567, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 35595 + }, + { + "epoch": 0.35596, + "grad_norm": 1.0238497257232666, + "learning_rate": 0.003, + "loss": 3.972, + "step": 35596 + }, + { + "epoch": 0.35597, + "grad_norm": 1.2072514295578003, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 35597 + }, + { + "epoch": 0.35598, + "grad_norm": 0.9737263321876526, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 35598 + }, + { + "epoch": 0.35599, + "grad_norm": 0.8913156986236572, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 35599 + }, + { + "epoch": 0.356, + "grad_norm": 1.001407265663147, + "learning_rate": 0.003, + "loss": 3.976, + "step": 35600 + }, + { + "epoch": 0.35601, + "grad_norm": 1.0467878580093384, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 35601 + }, + { + "epoch": 0.35602, + "grad_norm": 0.882584810256958, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 35602 + }, + { + "epoch": 0.35603, + "grad_norm": 0.8704052567481995, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 35603 + }, + { + "epoch": 0.35604, + "grad_norm": 0.8422004580497742, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 35604 + }, + { + "epoch": 0.35605, + "grad_norm": 0.7861309051513672, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 35605 + }, + { + "epoch": 0.35606, + "grad_norm": 0.7343525290489197, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 35606 + }, + { + "epoch": 0.35607, + "grad_norm": 0.8212799429893494, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 35607 + }, + { + "epoch": 0.35608, + "grad_norm": 1.1553430557250977, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 35608 + }, + { + "epoch": 0.35609, + "grad_norm": 1.1166213750839233, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 35609 + }, + { + "epoch": 0.3561, + "grad_norm": 0.9144553542137146, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 35610 + }, + { + "epoch": 0.35611, + "grad_norm": 0.8108396530151367, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 35611 + }, + { + "epoch": 0.35612, + "grad_norm": 0.8781615495681763, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 35612 + }, + { + "epoch": 0.35613, + "grad_norm": 0.9354559183120728, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 35613 + }, + { + "epoch": 0.35614, + "grad_norm": 0.9056247472763062, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 35614 + }, + { + "epoch": 0.35615, + "grad_norm": 0.8079336881637573, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 35615 + }, + { + "epoch": 0.35616, + "grad_norm": 0.771009087562561, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 35616 + }, + { + "epoch": 0.35617, + "grad_norm": 0.8364607095718384, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 35617 + }, + { + "epoch": 0.35618, + "grad_norm": 0.9312424063682556, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 35618 + }, + { + "epoch": 0.35619, + "grad_norm": 0.9349567294120789, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 35619 + }, + { + "epoch": 0.3562, + "grad_norm": 0.9180462956428528, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 35620 + }, + { + "epoch": 0.35621, + "grad_norm": 1.0066604614257812, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 35621 + }, + { + "epoch": 0.35622, + "grad_norm": 1.1083803176879883, + "learning_rate": 0.003, + "loss": 3.995, + "step": 35622 + }, + { + "epoch": 0.35623, + "grad_norm": 1.04448664188385, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 35623 + }, + { + "epoch": 0.35624, + "grad_norm": 1.0001412630081177, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 35624 + }, + { + "epoch": 0.35625, + "grad_norm": 0.9622340202331543, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 35625 + }, + { + "epoch": 0.35626, + "grad_norm": 1.0000425577163696, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 35626 + }, + { + "epoch": 0.35627, + "grad_norm": 1.0855780839920044, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 35627 + }, + { + "epoch": 0.35628, + "grad_norm": 0.8923764824867249, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 35628 + }, + { + "epoch": 0.35629, + "grad_norm": 0.9693261981010437, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 35629 + }, + { + "epoch": 0.3563, + "grad_norm": 0.8878809213638306, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 35630 + }, + { + "epoch": 0.35631, + "grad_norm": 0.9051221609115601, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 35631 + }, + { + "epoch": 0.35632, + "grad_norm": 0.8494033813476562, + "learning_rate": 0.003, + "loss": 3.9376, + "step": 35632 + }, + { + "epoch": 0.35633, + "grad_norm": 0.7666134238243103, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 35633 + }, + { + "epoch": 0.35634, + "grad_norm": 0.8910103440284729, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 35634 + }, + { + "epoch": 0.35635, + "grad_norm": 1.0293086767196655, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 35635 + }, + { + "epoch": 0.35636, + "grad_norm": 1.1358373165130615, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 35636 + }, + { + "epoch": 0.35637, + "grad_norm": 0.971730649471283, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 35637 + }, + { + "epoch": 0.35638, + "grad_norm": 0.9876066446304321, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 35638 + }, + { + "epoch": 0.35639, + "grad_norm": 1.0357933044433594, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 35639 + }, + { + "epoch": 0.3564, + "grad_norm": 0.9787134528160095, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 35640 + }, + { + "epoch": 0.35641, + "grad_norm": 0.8981085419654846, + "learning_rate": 0.003, + "loss": 3.998, + "step": 35641 + }, + { + "epoch": 0.35642, + "grad_norm": 0.7431622743606567, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 35642 + }, + { + "epoch": 0.35643, + "grad_norm": 0.6509256362915039, + "learning_rate": 0.003, + "loss": 3.993, + "step": 35643 + }, + { + "epoch": 0.35644, + "grad_norm": 0.7099307775497437, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 35644 + }, + { + "epoch": 0.35645, + "grad_norm": 0.7534025311470032, + "learning_rate": 0.003, + "loss": 3.967, + "step": 35645 + }, + { + "epoch": 0.35646, + "grad_norm": 0.7039646506309509, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 35646 + }, + { + "epoch": 0.35647, + "grad_norm": 0.7836117148399353, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 35647 + }, + { + "epoch": 0.35648, + "grad_norm": 0.9516607522964478, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 35648 + }, + { + "epoch": 0.35649, + "grad_norm": 1.1478010416030884, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 35649 + }, + { + "epoch": 0.3565, + "grad_norm": 0.7536616921424866, + "learning_rate": 0.003, + "loss": 3.964, + "step": 35650 + }, + { + "epoch": 0.35651, + "grad_norm": 0.5946837663650513, + "learning_rate": 0.003, + "loss": 4.0, + "step": 35651 + }, + { + "epoch": 0.35652, + "grad_norm": 0.6282517313957214, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 35652 + }, + { + "epoch": 0.35653, + "grad_norm": 0.7353341579437256, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 35653 + }, + { + "epoch": 0.35654, + "grad_norm": 1.0119918584823608, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 35654 + }, + { + "epoch": 0.35655, + "grad_norm": 1.2333675622940063, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 35655 + }, + { + "epoch": 0.35656, + "grad_norm": 0.6801337599754333, + "learning_rate": 0.003, + "loss": 3.939, + "step": 35656 + }, + { + "epoch": 0.35657, + "grad_norm": 0.6416513919830322, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 35657 + }, + { + "epoch": 0.35658, + "grad_norm": 0.6246609687805176, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 35658 + }, + { + "epoch": 0.35659, + "grad_norm": 0.8469140529632568, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 35659 + }, + { + "epoch": 0.3566, + "grad_norm": 1.2028220891952515, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 35660 + }, + { + "epoch": 0.35661, + "grad_norm": 1.0547236204147339, + "learning_rate": 0.003, + "loss": 3.994, + "step": 35661 + }, + { + "epoch": 0.35662, + "grad_norm": 0.8504834175109863, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 35662 + }, + { + "epoch": 0.35663, + "grad_norm": 0.782042920589447, + "learning_rate": 0.003, + "loss": 3.952, + "step": 35663 + }, + { + "epoch": 0.35664, + "grad_norm": 0.819123387336731, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 35664 + }, + { + "epoch": 0.35665, + "grad_norm": 0.8761565685272217, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 35665 + }, + { + "epoch": 0.35666, + "grad_norm": 0.8420794606208801, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 35666 + }, + { + "epoch": 0.35667, + "grad_norm": 0.9461551308631897, + "learning_rate": 0.003, + "loss": 3.961, + "step": 35667 + }, + { + "epoch": 0.35668, + "grad_norm": 0.8943864703178406, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 35668 + }, + { + "epoch": 0.35669, + "grad_norm": 0.8864920139312744, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 35669 + }, + { + "epoch": 0.3567, + "grad_norm": 0.9456399083137512, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 35670 + }, + { + "epoch": 0.35671, + "grad_norm": 0.9368754029273987, + "learning_rate": 0.003, + "loss": 4.006, + "step": 35671 + }, + { + "epoch": 0.35672, + "grad_norm": 0.8562811613082886, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 35672 + }, + { + "epoch": 0.35673, + "grad_norm": 1.0987980365753174, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 35673 + }, + { + "epoch": 0.35674, + "grad_norm": 0.9994171857833862, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 35674 + }, + { + "epoch": 0.35675, + "grad_norm": 1.1946611404418945, + "learning_rate": 0.003, + "loss": 4.011, + "step": 35675 + }, + { + "epoch": 0.35676, + "grad_norm": 1.039693832397461, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 35676 + }, + { + "epoch": 0.35677, + "grad_norm": 0.99741131067276, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 35677 + }, + { + "epoch": 0.35678, + "grad_norm": 1.007856845855713, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 35678 + }, + { + "epoch": 0.35679, + "grad_norm": 1.217362403869629, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 35679 + }, + { + "epoch": 0.3568, + "grad_norm": 0.8786345720291138, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 35680 + }, + { + "epoch": 0.35681, + "grad_norm": 0.893526017665863, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 35681 + }, + { + "epoch": 0.35682, + "grad_norm": 1.038299798965454, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 35682 + }, + { + "epoch": 0.35683, + "grad_norm": 1.1521342992782593, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 35683 + }, + { + "epoch": 0.35684, + "grad_norm": 0.7412964701652527, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 35684 + }, + { + "epoch": 0.35685, + "grad_norm": 0.671424388885498, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 35685 + }, + { + "epoch": 0.35686, + "grad_norm": 0.6752375960350037, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 35686 + }, + { + "epoch": 0.35687, + "grad_norm": 0.7239757180213928, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 35687 + }, + { + "epoch": 0.35688, + "grad_norm": 0.9001651406288147, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 35688 + }, + { + "epoch": 0.35689, + "grad_norm": 0.8936821818351746, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 35689 + }, + { + "epoch": 0.3569, + "grad_norm": 0.7497587203979492, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 35690 + }, + { + "epoch": 0.35691, + "grad_norm": 0.86234050989151, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 35691 + }, + { + "epoch": 0.35692, + "grad_norm": 0.8933195471763611, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 35692 + }, + { + "epoch": 0.35693, + "grad_norm": 1.1251519918441772, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 35693 + }, + { + "epoch": 0.35694, + "grad_norm": 0.9490299820899963, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 35694 + }, + { + "epoch": 0.35695, + "grad_norm": 0.8371615409851074, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 35695 + }, + { + "epoch": 0.35696, + "grad_norm": 0.7908884286880493, + "learning_rate": 0.003, + "loss": 3.968, + "step": 35696 + }, + { + "epoch": 0.35697, + "grad_norm": 0.8403076529502869, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 35697 + }, + { + "epoch": 0.35698, + "grad_norm": 0.9020901918411255, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 35698 + }, + { + "epoch": 0.35699, + "grad_norm": 0.914879560470581, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 35699 + }, + { + "epoch": 0.357, + "grad_norm": 0.9642201662063599, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 35700 + }, + { + "epoch": 0.35701, + "grad_norm": 1.0146228075027466, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 35701 + }, + { + "epoch": 0.35702, + "grad_norm": 1.1677157878875732, + "learning_rate": 0.003, + "loss": 3.974, + "step": 35702 + }, + { + "epoch": 0.35703, + "grad_norm": 0.7307107448577881, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 35703 + }, + { + "epoch": 0.35704, + "grad_norm": 0.5982475876808167, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 35704 + }, + { + "epoch": 0.35705, + "grad_norm": 0.541718065738678, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 35705 + }, + { + "epoch": 0.35706, + "grad_norm": 0.6597070693969727, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 35706 + }, + { + "epoch": 0.35707, + "grad_norm": 0.5950496792793274, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 35707 + }, + { + "epoch": 0.35708, + "grad_norm": 0.6670334339141846, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 35708 + }, + { + "epoch": 0.35709, + "grad_norm": 0.8589178323745728, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 35709 + }, + { + "epoch": 0.3571, + "grad_norm": 1.1425284147262573, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 35710 + }, + { + "epoch": 0.35711, + "grad_norm": 0.7578059434890747, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 35711 + }, + { + "epoch": 0.35712, + "grad_norm": 0.6283740401268005, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 35712 + }, + { + "epoch": 0.35713, + "grad_norm": 0.8159963488578796, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 35713 + }, + { + "epoch": 0.35714, + "grad_norm": 0.7866744995117188, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 35714 + }, + { + "epoch": 0.35715, + "grad_norm": 0.7757623791694641, + "learning_rate": 0.003, + "loss": 3.959, + "step": 35715 + }, + { + "epoch": 0.35716, + "grad_norm": 0.8544276356697083, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 35716 + }, + { + "epoch": 0.35717, + "grad_norm": 0.9162160754203796, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 35717 + }, + { + "epoch": 0.35718, + "grad_norm": 1.009902834892273, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 35718 + }, + { + "epoch": 0.35719, + "grad_norm": 1.1716842651367188, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 35719 + }, + { + "epoch": 0.3572, + "grad_norm": 0.9219096899032593, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 35720 + }, + { + "epoch": 0.35721, + "grad_norm": 0.7858926653862, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 35721 + }, + { + "epoch": 0.35722, + "grad_norm": 0.6933205127716064, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 35722 + }, + { + "epoch": 0.35723, + "grad_norm": 0.6374828219413757, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 35723 + }, + { + "epoch": 0.35724, + "grad_norm": 0.6249655485153198, + "learning_rate": 0.003, + "loss": 3.9382, + "step": 35724 + }, + { + "epoch": 0.35725, + "grad_norm": 0.7250398397445679, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 35725 + }, + { + "epoch": 0.35726, + "grad_norm": 0.68715500831604, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 35726 + }, + { + "epoch": 0.35727, + "grad_norm": 0.6407122611999512, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 35727 + }, + { + "epoch": 0.35728, + "grad_norm": 0.7350702285766602, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 35728 + }, + { + "epoch": 0.35729, + "grad_norm": 0.8597891926765442, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 35729 + }, + { + "epoch": 0.3573, + "grad_norm": 1.2480647563934326, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 35730 + }, + { + "epoch": 0.35731, + "grad_norm": 1.1617238521575928, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 35731 + }, + { + "epoch": 0.35732, + "grad_norm": 0.9744846224784851, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 35732 + }, + { + "epoch": 0.35733, + "grad_norm": 0.93841153383255, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 35733 + }, + { + "epoch": 0.35734, + "grad_norm": 1.007804274559021, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 35734 + }, + { + "epoch": 0.35735, + "grad_norm": 1.1441636085510254, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 35735 + }, + { + "epoch": 0.35736, + "grad_norm": 0.9215797781944275, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 35736 + }, + { + "epoch": 0.35737, + "grad_norm": 1.1565535068511963, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 35737 + }, + { + "epoch": 0.35738, + "grad_norm": 0.7753909230232239, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 35738 + }, + { + "epoch": 0.35739, + "grad_norm": 0.9420831203460693, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 35739 + }, + { + "epoch": 0.3574, + "grad_norm": 0.9024673700332642, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 35740 + }, + { + "epoch": 0.35741, + "grad_norm": 0.8635263442993164, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 35741 + }, + { + "epoch": 0.35742, + "grad_norm": 1.0525723695755005, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 35742 + }, + { + "epoch": 0.35743, + "grad_norm": 1.1568148136138916, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 35743 + }, + { + "epoch": 0.35744, + "grad_norm": 1.2431195974349976, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 35744 + }, + { + "epoch": 0.35745, + "grad_norm": 1.1936769485473633, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 35745 + }, + { + "epoch": 0.35746, + "grad_norm": 0.8070017695426941, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 35746 + }, + { + "epoch": 0.35747, + "grad_norm": 0.8478720188140869, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 35747 + }, + { + "epoch": 0.35748, + "grad_norm": 0.9641151428222656, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 35748 + }, + { + "epoch": 0.35749, + "grad_norm": 1.1227909326553345, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 35749 + }, + { + "epoch": 0.3575, + "grad_norm": 0.9532807469367981, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 35750 + }, + { + "epoch": 0.35751, + "grad_norm": 0.8435110449790955, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 35751 + }, + { + "epoch": 0.35752, + "grad_norm": 0.7369087934494019, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 35752 + }, + { + "epoch": 0.35753, + "grad_norm": 0.7169075608253479, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 35753 + }, + { + "epoch": 0.35754, + "grad_norm": 0.7930052280426025, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 35754 + }, + { + "epoch": 0.35755, + "grad_norm": 0.7964222431182861, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 35755 + }, + { + "epoch": 0.35756, + "grad_norm": 0.7519850134849548, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 35756 + }, + { + "epoch": 0.35757, + "grad_norm": 0.8224310874938965, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 35757 + }, + { + "epoch": 0.35758, + "grad_norm": 0.8599464297294617, + "learning_rate": 0.003, + "loss": 4.001, + "step": 35758 + }, + { + "epoch": 0.35759, + "grad_norm": 0.9581980109214783, + "learning_rate": 0.003, + "loss": 4.002, + "step": 35759 + }, + { + "epoch": 0.3576, + "grad_norm": 1.109544277191162, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 35760 + }, + { + "epoch": 0.35761, + "grad_norm": 0.7679346203804016, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 35761 + }, + { + "epoch": 0.35762, + "grad_norm": 0.510012149810791, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 35762 + }, + { + "epoch": 0.35763, + "grad_norm": 0.6103097200393677, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 35763 + }, + { + "epoch": 0.35764, + "grad_norm": 0.781313419342041, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 35764 + }, + { + "epoch": 0.35765, + "grad_norm": 1.0471824407577515, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 35765 + }, + { + "epoch": 0.35766, + "grad_norm": 1.1710153818130493, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 35766 + }, + { + "epoch": 0.35767, + "grad_norm": 0.9650210738182068, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 35767 + }, + { + "epoch": 0.35768, + "grad_norm": 1.0730540752410889, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 35768 + }, + { + "epoch": 0.35769, + "grad_norm": 0.9904390573501587, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 35769 + }, + { + "epoch": 0.3577, + "grad_norm": 0.9198780655860901, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 35770 + }, + { + "epoch": 0.35771, + "grad_norm": 0.7377621531486511, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 35771 + }, + { + "epoch": 0.35772, + "grad_norm": 0.7176426649093628, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 35772 + }, + { + "epoch": 0.35773, + "grad_norm": 0.7993268966674805, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 35773 + }, + { + "epoch": 0.35774, + "grad_norm": 1.00766921043396, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 35774 + }, + { + "epoch": 0.35775, + "grad_norm": 1.0319072008132935, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 35775 + }, + { + "epoch": 0.35776, + "grad_norm": 0.8888317346572876, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 35776 + }, + { + "epoch": 0.35777, + "grad_norm": 0.9603275060653687, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 35777 + }, + { + "epoch": 0.35778, + "grad_norm": 1.234845757484436, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 35778 + }, + { + "epoch": 0.35779, + "grad_norm": 1.00291109085083, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 35779 + }, + { + "epoch": 0.3578, + "grad_norm": 1.0020543336868286, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 35780 + }, + { + "epoch": 0.35781, + "grad_norm": 0.9993874430656433, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 35781 + }, + { + "epoch": 0.35782, + "grad_norm": 1.0591504573822021, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 35782 + }, + { + "epoch": 0.35783, + "grad_norm": 0.8875681161880493, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 35783 + }, + { + "epoch": 0.35784, + "grad_norm": 0.9082662463188171, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 35784 + }, + { + "epoch": 0.35785, + "grad_norm": 0.9122529625892639, + "learning_rate": 0.003, + "loss": 3.962, + "step": 35785 + }, + { + "epoch": 0.35786, + "grad_norm": 0.9411823153495789, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 35786 + }, + { + "epoch": 0.35787, + "grad_norm": 1.1891937255859375, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 35787 + }, + { + "epoch": 0.35788, + "grad_norm": 0.9036888480186462, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 35788 + }, + { + "epoch": 0.35789, + "grad_norm": 1.0466694831848145, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 35789 + }, + { + "epoch": 0.3579, + "grad_norm": 0.998038649559021, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 35790 + }, + { + "epoch": 0.35791, + "grad_norm": 1.0026092529296875, + "learning_rate": 0.003, + "loss": 3.9454, + "step": 35791 + }, + { + "epoch": 0.35792, + "grad_norm": 1.017965316772461, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 35792 + }, + { + "epoch": 0.35793, + "grad_norm": 0.8393538594245911, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 35793 + }, + { + "epoch": 0.35794, + "grad_norm": 1.0245170593261719, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 35794 + }, + { + "epoch": 0.35795, + "grad_norm": 0.9860982298851013, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 35795 + }, + { + "epoch": 0.35796, + "grad_norm": 0.9712509512901306, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 35796 + }, + { + "epoch": 0.35797, + "grad_norm": 1.336400032043457, + "learning_rate": 0.003, + "loss": 4.0, + "step": 35797 + }, + { + "epoch": 0.35798, + "grad_norm": 0.8478973507881165, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 35798 + }, + { + "epoch": 0.35799, + "grad_norm": 0.6859142780303955, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 35799 + }, + { + "epoch": 0.358, + "grad_norm": 0.911845326423645, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 35800 + }, + { + "epoch": 0.35801, + "grad_norm": 0.9242119193077087, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 35801 + }, + { + "epoch": 0.35802, + "grad_norm": 0.9817672371864319, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 35802 + }, + { + "epoch": 0.35803, + "grad_norm": 1.093802571296692, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 35803 + }, + { + "epoch": 0.35804, + "grad_norm": 0.9403889775276184, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 35804 + }, + { + "epoch": 0.35805, + "grad_norm": 0.7183811068534851, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 35805 + }, + { + "epoch": 0.35806, + "grad_norm": 0.6368287801742554, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 35806 + }, + { + "epoch": 0.35807, + "grad_norm": 0.7030258178710938, + "learning_rate": 0.003, + "loss": 4.005, + "step": 35807 + }, + { + "epoch": 0.35808, + "grad_norm": 0.751969575881958, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 35808 + }, + { + "epoch": 0.35809, + "grad_norm": 0.7731482982635498, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 35809 + }, + { + "epoch": 0.3581, + "grad_norm": 0.7452133297920227, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 35810 + }, + { + "epoch": 0.35811, + "grad_norm": 0.8224204182624817, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 35811 + }, + { + "epoch": 0.35812, + "grad_norm": 0.839672863483429, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 35812 + }, + { + "epoch": 0.35813, + "grad_norm": 0.7662952542304993, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 35813 + }, + { + "epoch": 0.35814, + "grad_norm": 0.717024028301239, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 35814 + }, + { + "epoch": 0.35815, + "grad_norm": 0.6550969481468201, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 35815 + }, + { + "epoch": 0.35816, + "grad_norm": 0.6871244311332703, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 35816 + }, + { + "epoch": 0.35817, + "grad_norm": 0.7878357768058777, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 35817 + }, + { + "epoch": 0.35818, + "grad_norm": 1.0220104455947876, + "learning_rate": 0.003, + "loss": 3.997, + "step": 35818 + }, + { + "epoch": 0.35819, + "grad_norm": 1.0814399719238281, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 35819 + }, + { + "epoch": 0.3582, + "grad_norm": 0.7977302074432373, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 35820 + }, + { + "epoch": 0.35821, + "grad_norm": 0.6862973570823669, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 35821 + }, + { + "epoch": 0.35822, + "grad_norm": 0.7544977068901062, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 35822 + }, + { + "epoch": 0.35823, + "grad_norm": 0.7382851243019104, + "learning_rate": 0.003, + "loss": 3.991, + "step": 35823 + }, + { + "epoch": 0.35824, + "grad_norm": 0.8130156397819519, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 35824 + }, + { + "epoch": 0.35825, + "grad_norm": 0.9330124855041504, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 35825 + }, + { + "epoch": 0.35826, + "grad_norm": 0.950875461101532, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 35826 + }, + { + "epoch": 0.35827, + "grad_norm": 0.8297685384750366, + "learning_rate": 0.003, + "loss": 3.98, + "step": 35827 + }, + { + "epoch": 0.35828, + "grad_norm": 0.9817972779273987, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 35828 + }, + { + "epoch": 0.35829, + "grad_norm": 1.0352375507354736, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 35829 + }, + { + "epoch": 0.3583, + "grad_norm": 1.1275033950805664, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 35830 + }, + { + "epoch": 0.35831, + "grad_norm": 1.1451025009155273, + "learning_rate": 0.003, + "loss": 3.988, + "step": 35831 + }, + { + "epoch": 0.35832, + "grad_norm": 1.131698727607727, + "learning_rate": 0.003, + "loss": 3.99, + "step": 35832 + }, + { + "epoch": 0.35833, + "grad_norm": 1.0850584506988525, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 35833 + }, + { + "epoch": 0.35834, + "grad_norm": 0.8925266265869141, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 35834 + }, + { + "epoch": 0.35835, + "grad_norm": 0.8658837676048279, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 35835 + }, + { + "epoch": 0.35836, + "grad_norm": 0.9935423135757446, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 35836 + }, + { + "epoch": 0.35837, + "grad_norm": 1.195897102355957, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 35837 + }, + { + "epoch": 0.35838, + "grad_norm": 1.0504791736602783, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 35838 + }, + { + "epoch": 0.35839, + "grad_norm": 1.148151159286499, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 35839 + }, + { + "epoch": 0.3584, + "grad_norm": 0.9788755178451538, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 35840 + }, + { + "epoch": 0.35841, + "grad_norm": 0.839938223361969, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 35841 + }, + { + "epoch": 0.35842, + "grad_norm": 0.7193537354469299, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 35842 + }, + { + "epoch": 0.35843, + "grad_norm": 0.6998471617698669, + "learning_rate": 0.003, + "loss": 3.989, + "step": 35843 + }, + { + "epoch": 0.35844, + "grad_norm": 0.7320494055747986, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 35844 + }, + { + "epoch": 0.35845, + "grad_norm": 0.8336778283119202, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 35845 + }, + { + "epoch": 0.35846, + "grad_norm": 0.7922508716583252, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 35846 + }, + { + "epoch": 0.35847, + "grad_norm": 0.6758324503898621, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 35847 + }, + { + "epoch": 0.35848, + "grad_norm": 0.8549639582633972, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 35848 + }, + { + "epoch": 0.35849, + "grad_norm": 0.835518479347229, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 35849 + }, + { + "epoch": 0.3585, + "grad_norm": 0.7956916689872742, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 35850 + }, + { + "epoch": 0.35851, + "grad_norm": 0.747120201587677, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 35851 + }, + { + "epoch": 0.35852, + "grad_norm": 0.8704857230186462, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 35852 + }, + { + "epoch": 0.35853, + "grad_norm": 0.9009389281272888, + "learning_rate": 0.003, + "loss": 3.991, + "step": 35853 + }, + { + "epoch": 0.35854, + "grad_norm": 1.2021039724349976, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 35854 + }, + { + "epoch": 0.35855, + "grad_norm": 0.926023542881012, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 35855 + }, + { + "epoch": 0.35856, + "grad_norm": 1.0029102563858032, + "learning_rate": 0.003, + "loss": 3.97, + "step": 35856 + }, + { + "epoch": 0.35857, + "grad_norm": 0.9916598796844482, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 35857 + }, + { + "epoch": 0.35858, + "grad_norm": 1.2969274520874023, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 35858 + }, + { + "epoch": 0.35859, + "grad_norm": 0.6854649782180786, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 35859 + }, + { + "epoch": 0.3586, + "grad_norm": 0.6389893293380737, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 35860 + }, + { + "epoch": 0.35861, + "grad_norm": 0.7966915369033813, + "learning_rate": 0.003, + "loss": 3.974, + "step": 35861 + }, + { + "epoch": 0.35862, + "grad_norm": 0.842440128326416, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 35862 + }, + { + "epoch": 0.35863, + "grad_norm": 0.903033435344696, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 35863 + }, + { + "epoch": 0.35864, + "grad_norm": 0.8228189945220947, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 35864 + }, + { + "epoch": 0.35865, + "grad_norm": 0.8578748106956482, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 35865 + }, + { + "epoch": 0.35866, + "grad_norm": 0.9968085289001465, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 35866 + }, + { + "epoch": 0.35867, + "grad_norm": 1.230075478553772, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 35867 + }, + { + "epoch": 0.35868, + "grad_norm": 0.9667801260948181, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 35868 + }, + { + "epoch": 0.35869, + "grad_norm": 1.1388224363327026, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 35869 + }, + { + "epoch": 0.3587, + "grad_norm": 0.9234357476234436, + "learning_rate": 0.003, + "loss": 3.975, + "step": 35870 + }, + { + "epoch": 0.35871, + "grad_norm": 0.7822298407554626, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 35871 + }, + { + "epoch": 0.35872, + "grad_norm": 0.709493100643158, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 35872 + }, + { + "epoch": 0.35873, + "grad_norm": 0.8166400194168091, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 35873 + }, + { + "epoch": 0.35874, + "grad_norm": 0.9164003133773804, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 35874 + }, + { + "epoch": 0.35875, + "grad_norm": 1.125683307647705, + "learning_rate": 0.003, + "loss": 3.999, + "step": 35875 + }, + { + "epoch": 0.35876, + "grad_norm": 0.979030191898346, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 35876 + }, + { + "epoch": 0.35877, + "grad_norm": 0.9011402130126953, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 35877 + }, + { + "epoch": 0.35878, + "grad_norm": 0.926026463508606, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 35878 + }, + { + "epoch": 0.35879, + "grad_norm": 0.9921069145202637, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 35879 + }, + { + "epoch": 0.3588, + "grad_norm": 1.0419838428497314, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 35880 + }, + { + "epoch": 0.35881, + "grad_norm": 1.0513027906417847, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 35881 + }, + { + "epoch": 0.35882, + "grad_norm": 1.012803554534912, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 35882 + }, + { + "epoch": 0.35883, + "grad_norm": 0.9684647917747498, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 35883 + }, + { + "epoch": 0.35884, + "grad_norm": 0.8381258845329285, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 35884 + }, + { + "epoch": 0.35885, + "grad_norm": 0.7763985395431519, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 35885 + }, + { + "epoch": 0.35886, + "grad_norm": 0.9673192501068115, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 35886 + }, + { + "epoch": 0.35887, + "grad_norm": 1.1483309268951416, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 35887 + }, + { + "epoch": 0.35888, + "grad_norm": 0.8779354095458984, + "learning_rate": 0.003, + "loss": 4.003, + "step": 35888 + }, + { + "epoch": 0.35889, + "grad_norm": 0.8311321139335632, + "learning_rate": 0.003, + "loss": 3.981, + "step": 35889 + }, + { + "epoch": 0.3589, + "grad_norm": 0.906674325466156, + "learning_rate": 0.003, + "loss": 3.972, + "step": 35890 + }, + { + "epoch": 0.35891, + "grad_norm": 1.051655650138855, + "learning_rate": 0.003, + "loss": 3.985, + "step": 35891 + }, + { + "epoch": 0.35892, + "grad_norm": 1.1144577264785767, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 35892 + }, + { + "epoch": 0.35893, + "grad_norm": 1.0880658626556396, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 35893 + }, + { + "epoch": 0.35894, + "grad_norm": 1.0750641822814941, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 35894 + }, + { + "epoch": 0.35895, + "grad_norm": 0.6937015056610107, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 35895 + }, + { + "epoch": 0.35896, + "grad_norm": 0.7157841920852661, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 35896 + }, + { + "epoch": 0.35897, + "grad_norm": 0.8175703287124634, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 35897 + }, + { + "epoch": 0.35898, + "grad_norm": 1.070605993270874, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 35898 + }, + { + "epoch": 0.35899, + "grad_norm": 1.037493348121643, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 35899 + }, + { + "epoch": 0.359, + "grad_norm": 1.0401346683502197, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 35900 + }, + { + "epoch": 0.35901, + "grad_norm": 0.9095867276191711, + "learning_rate": 0.003, + "loss": 3.992, + "step": 35901 + }, + { + "epoch": 0.35902, + "grad_norm": 0.9175391793251038, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 35902 + }, + { + "epoch": 0.35903, + "grad_norm": 1.0215147733688354, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 35903 + }, + { + "epoch": 0.35904, + "grad_norm": 0.9699781537055969, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 35904 + }, + { + "epoch": 0.35905, + "grad_norm": 0.8955513834953308, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 35905 + }, + { + "epoch": 0.35906, + "grad_norm": 1.2295809984207153, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 35906 + }, + { + "epoch": 0.35907, + "grad_norm": 1.1101343631744385, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 35907 + }, + { + "epoch": 0.35908, + "grad_norm": 1.2462985515594482, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 35908 + }, + { + "epoch": 0.35909, + "grad_norm": 0.6751952767372131, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 35909 + }, + { + "epoch": 0.3591, + "grad_norm": 0.5592895746231079, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 35910 + }, + { + "epoch": 0.35911, + "grad_norm": 0.764862060546875, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 35911 + }, + { + "epoch": 0.35912, + "grad_norm": 1.0489306449890137, + "learning_rate": 0.003, + "loss": 3.983, + "step": 35912 + }, + { + "epoch": 0.35913, + "grad_norm": 1.1360548734664917, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 35913 + }, + { + "epoch": 0.35914, + "grad_norm": 0.5948429107666016, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 35914 + }, + { + "epoch": 0.35915, + "grad_norm": 0.661614716053009, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 35915 + }, + { + "epoch": 0.35916, + "grad_norm": 0.9942266941070557, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 35916 + }, + { + "epoch": 0.35917, + "grad_norm": 1.062700629234314, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 35917 + }, + { + "epoch": 0.35918, + "grad_norm": 0.7387388944625854, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 35918 + }, + { + "epoch": 0.35919, + "grad_norm": 0.6674962043762207, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 35919 + }, + { + "epoch": 0.3592, + "grad_norm": 0.8021997213363647, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 35920 + }, + { + "epoch": 0.35921, + "grad_norm": 0.8751733899116516, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 35921 + }, + { + "epoch": 0.35922, + "grad_norm": 0.7538155317306519, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 35922 + }, + { + "epoch": 0.35923, + "grad_norm": 0.6303317546844482, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 35923 + }, + { + "epoch": 0.35924, + "grad_norm": 0.6824501156806946, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 35924 + }, + { + "epoch": 0.35925, + "grad_norm": 0.7584856152534485, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 35925 + }, + { + "epoch": 0.35926, + "grad_norm": 0.7343990206718445, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 35926 + }, + { + "epoch": 0.35927, + "grad_norm": 0.8559405207633972, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 35927 + }, + { + "epoch": 0.35928, + "grad_norm": 0.9944117069244385, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 35928 + }, + { + "epoch": 0.35929, + "grad_norm": 1.2032301425933838, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 35929 + }, + { + "epoch": 0.3593, + "grad_norm": 0.8681849837303162, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 35930 + }, + { + "epoch": 0.35931, + "grad_norm": 0.8398818969726562, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 35931 + }, + { + "epoch": 0.35932, + "grad_norm": 1.0131254196166992, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 35932 + }, + { + "epoch": 0.35933, + "grad_norm": 1.5177055597305298, + "learning_rate": 0.003, + "loss": 3.998, + "step": 35933 + }, + { + "epoch": 0.35934, + "grad_norm": 0.6170707941055298, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 35934 + }, + { + "epoch": 0.35935, + "grad_norm": 0.865608274936676, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 35935 + }, + { + "epoch": 0.35936, + "grad_norm": 0.8202694058418274, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 35936 + }, + { + "epoch": 0.35937, + "grad_norm": 0.7651110291481018, + "learning_rate": 0.003, + "loss": 3.986, + "step": 35937 + }, + { + "epoch": 0.35938, + "grad_norm": 0.9603325724601746, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 35938 + }, + { + "epoch": 0.35939, + "grad_norm": 1.234480857849121, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 35939 + }, + { + "epoch": 0.3594, + "grad_norm": 1.0314077138900757, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 35940 + }, + { + "epoch": 0.35941, + "grad_norm": 0.9688170552253723, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 35941 + }, + { + "epoch": 0.35942, + "grad_norm": 1.1511880159378052, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 35942 + }, + { + "epoch": 0.35943, + "grad_norm": 0.9203600883483887, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 35943 + }, + { + "epoch": 0.35944, + "grad_norm": 0.9389334321022034, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 35944 + }, + { + "epoch": 0.35945, + "grad_norm": 1.0769332647323608, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 35945 + }, + { + "epoch": 0.35946, + "grad_norm": 1.190906286239624, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 35946 + }, + { + "epoch": 0.35947, + "grad_norm": 0.890359103679657, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 35947 + }, + { + "epoch": 0.35948, + "grad_norm": 0.6969262957572937, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 35948 + }, + { + "epoch": 0.35949, + "grad_norm": 0.6829444169998169, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 35949 + }, + { + "epoch": 0.3595, + "grad_norm": 0.8046309947967529, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 35950 + }, + { + "epoch": 0.35951, + "grad_norm": 0.9351231455802917, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 35951 + }, + { + "epoch": 0.35952, + "grad_norm": 0.896029531955719, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 35952 + }, + { + "epoch": 0.35953, + "grad_norm": 0.9520477056503296, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 35953 + }, + { + "epoch": 0.35954, + "grad_norm": 1.0313146114349365, + "learning_rate": 0.003, + "loss": 3.968, + "step": 35954 + }, + { + "epoch": 0.35955, + "grad_norm": 0.8672282099723816, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 35955 + }, + { + "epoch": 0.35956, + "grad_norm": 0.7425873875617981, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 35956 + }, + { + "epoch": 0.35957, + "grad_norm": 0.8226737976074219, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 35957 + }, + { + "epoch": 0.35958, + "grad_norm": 0.875385046005249, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 35958 + }, + { + "epoch": 0.35959, + "grad_norm": 0.829899251461029, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 35959 + }, + { + "epoch": 0.3596, + "grad_norm": 0.7243623733520508, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 35960 + }, + { + "epoch": 0.35961, + "grad_norm": 0.8079954981803894, + "learning_rate": 0.003, + "loss": 3.982, + "step": 35961 + }, + { + "epoch": 0.35962, + "grad_norm": 0.8936733603477478, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 35962 + }, + { + "epoch": 0.35963, + "grad_norm": 1.0772719383239746, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 35963 + }, + { + "epoch": 0.35964, + "grad_norm": 1.1656956672668457, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 35964 + }, + { + "epoch": 0.35965, + "grad_norm": 0.8024569749832153, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 35965 + }, + { + "epoch": 0.35966, + "grad_norm": 0.7771517038345337, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 35966 + }, + { + "epoch": 0.35967, + "grad_norm": 0.7866396903991699, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 35967 + }, + { + "epoch": 0.35968, + "grad_norm": 0.9164060950279236, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 35968 + }, + { + "epoch": 0.35969, + "grad_norm": 1.0331177711486816, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 35969 + }, + { + "epoch": 0.3597, + "grad_norm": 0.9641987085342407, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 35970 + }, + { + "epoch": 0.35971, + "grad_norm": 0.9659872651100159, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 35971 + }, + { + "epoch": 0.35972, + "grad_norm": 0.8965134024620056, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 35972 + }, + { + "epoch": 0.35973, + "grad_norm": 0.9037521481513977, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 35973 + }, + { + "epoch": 0.35974, + "grad_norm": 1.062214970588684, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 35974 + }, + { + "epoch": 0.35975, + "grad_norm": 1.0465350151062012, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 35975 + }, + { + "epoch": 0.35976, + "grad_norm": 0.9243862628936768, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 35976 + }, + { + "epoch": 0.35977, + "grad_norm": 0.9777161478996277, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 35977 + }, + { + "epoch": 0.35978, + "grad_norm": 1.0037314891815186, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 35978 + }, + { + "epoch": 0.35979, + "grad_norm": 1.1291903257369995, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 35979 + }, + { + "epoch": 0.3598, + "grad_norm": 1.1771321296691895, + "learning_rate": 0.003, + "loss": 3.963, + "step": 35980 + }, + { + "epoch": 0.35981, + "grad_norm": 0.8336952924728394, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 35981 + }, + { + "epoch": 0.35982, + "grad_norm": 0.7394837141036987, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 35982 + }, + { + "epoch": 0.35983, + "grad_norm": 0.8313787579536438, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 35983 + }, + { + "epoch": 0.35984, + "grad_norm": 0.9536665081977844, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 35984 + }, + { + "epoch": 0.35985, + "grad_norm": 0.9359556436538696, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 35985 + }, + { + "epoch": 0.35986, + "grad_norm": 0.7615119814872742, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 35986 + }, + { + "epoch": 0.35987, + "grad_norm": 0.7708956003189087, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 35987 + }, + { + "epoch": 0.35988, + "grad_norm": 0.7505605220794678, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 35988 + }, + { + "epoch": 0.35989, + "grad_norm": 0.8619018793106079, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 35989 + }, + { + "epoch": 0.3599, + "grad_norm": 0.9634534120559692, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 35990 + }, + { + "epoch": 0.35991, + "grad_norm": 1.1325513124465942, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 35991 + }, + { + "epoch": 0.35992, + "grad_norm": 1.0220074653625488, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 35992 + }, + { + "epoch": 0.35993, + "grad_norm": 0.800931453704834, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 35993 + }, + { + "epoch": 0.35994, + "grad_norm": 0.7297806143760681, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 35994 + }, + { + "epoch": 0.35995, + "grad_norm": 0.9403678774833679, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 35995 + }, + { + "epoch": 0.35996, + "grad_norm": 1.2563053369522095, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 35996 + }, + { + "epoch": 0.35997, + "grad_norm": 0.924223005771637, + "learning_rate": 0.003, + "loss": 3.968, + "step": 35997 + }, + { + "epoch": 0.35998, + "grad_norm": 0.9883263111114502, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 35998 + }, + { + "epoch": 0.35999, + "grad_norm": 1.2472913265228271, + "learning_rate": 0.003, + "loss": 4.042, + "step": 35999 + }, + { + "epoch": 0.36, + "grad_norm": 0.6396057605743408, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 36000 + }, + { + "epoch": 0.36001, + "grad_norm": 0.5558953881263733, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 36001 + }, + { + "epoch": 0.36002, + "grad_norm": 0.5801008939743042, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 36002 + }, + { + "epoch": 0.36003, + "grad_norm": 0.6998916268348694, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 36003 + }, + { + "epoch": 0.36004, + "grad_norm": 0.8157746195793152, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 36004 + }, + { + "epoch": 0.36005, + "grad_norm": 0.9416586756706238, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 36005 + }, + { + "epoch": 0.36006, + "grad_norm": 1.0191401243209839, + "learning_rate": 0.003, + "loss": 3.98, + "step": 36006 + }, + { + "epoch": 0.36007, + "grad_norm": 0.9482595920562744, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 36007 + }, + { + "epoch": 0.36008, + "grad_norm": 0.845923900604248, + "learning_rate": 0.003, + "loss": 3.971, + "step": 36008 + }, + { + "epoch": 0.36009, + "grad_norm": 0.7242968678474426, + "learning_rate": 0.003, + "loss": 3.9434, + "step": 36009 + }, + { + "epoch": 0.3601, + "grad_norm": 0.5625496506690979, + "learning_rate": 0.003, + "loss": 3.975, + "step": 36010 + }, + { + "epoch": 0.36011, + "grad_norm": 0.5165539979934692, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 36011 + }, + { + "epoch": 0.36012, + "grad_norm": 0.5468502044677734, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 36012 + }, + { + "epoch": 0.36013, + "grad_norm": 0.6396152377128601, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 36013 + }, + { + "epoch": 0.36014, + "grad_norm": 0.7471448183059692, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 36014 + }, + { + "epoch": 0.36015, + "grad_norm": 0.7801705002784729, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 36015 + }, + { + "epoch": 0.36016, + "grad_norm": 0.8921595215797424, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 36016 + }, + { + "epoch": 0.36017, + "grad_norm": 1.0146080255508423, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 36017 + }, + { + "epoch": 0.36018, + "grad_norm": 0.9635580778121948, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 36018 + }, + { + "epoch": 0.36019, + "grad_norm": 0.9110696315765381, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 36019 + }, + { + "epoch": 0.3602, + "grad_norm": 0.915355384349823, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 36020 + }, + { + "epoch": 0.36021, + "grad_norm": 0.928228497505188, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 36021 + }, + { + "epoch": 0.36022, + "grad_norm": 1.0338358879089355, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 36022 + }, + { + "epoch": 0.36023, + "grad_norm": 1.4639467000961304, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 36023 + }, + { + "epoch": 0.36024, + "grad_norm": 0.6893510222434998, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 36024 + }, + { + "epoch": 0.36025, + "grad_norm": 0.8432899713516235, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 36025 + }, + { + "epoch": 0.36026, + "grad_norm": 1.037713646888733, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 36026 + }, + { + "epoch": 0.36027, + "grad_norm": 1.3129411935806274, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 36027 + }, + { + "epoch": 0.36028, + "grad_norm": 1.055826187133789, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 36028 + }, + { + "epoch": 0.36029, + "grad_norm": 0.9104787111282349, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 36029 + }, + { + "epoch": 0.3603, + "grad_norm": 0.8008221387863159, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 36030 + }, + { + "epoch": 0.36031, + "grad_norm": 0.8245895504951477, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 36031 + }, + { + "epoch": 0.36032, + "grad_norm": 0.885284960269928, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 36032 + }, + { + "epoch": 0.36033, + "grad_norm": 0.9192755818367004, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 36033 + }, + { + "epoch": 0.36034, + "grad_norm": 1.0265111923217773, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 36034 + }, + { + "epoch": 0.36035, + "grad_norm": 0.9840852618217468, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 36035 + }, + { + "epoch": 0.36036, + "grad_norm": 0.9503747820854187, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 36036 + }, + { + "epoch": 0.36037, + "grad_norm": 0.9522866010665894, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 36037 + }, + { + "epoch": 0.36038, + "grad_norm": 1.0992175340652466, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 36038 + }, + { + "epoch": 0.36039, + "grad_norm": 1.055184245109558, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 36039 + }, + { + "epoch": 0.3604, + "grad_norm": 0.9938040375709534, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 36040 + }, + { + "epoch": 0.36041, + "grad_norm": 0.9130344986915588, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 36041 + }, + { + "epoch": 0.36042, + "grad_norm": 1.0381585359573364, + "learning_rate": 0.003, + "loss": 4.002, + "step": 36042 + }, + { + "epoch": 0.36043, + "grad_norm": 0.942233145236969, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 36043 + }, + { + "epoch": 0.36044, + "grad_norm": 0.9779905676841736, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 36044 + }, + { + "epoch": 0.36045, + "grad_norm": 0.9468114376068115, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 36045 + }, + { + "epoch": 0.36046, + "grad_norm": 0.8304805159568787, + "learning_rate": 0.003, + "loss": 3.997, + "step": 36046 + }, + { + "epoch": 0.36047, + "grad_norm": 0.791636049747467, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 36047 + }, + { + "epoch": 0.36048, + "grad_norm": 0.7606046199798584, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 36048 + }, + { + "epoch": 0.36049, + "grad_norm": 0.7033624649047852, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 36049 + }, + { + "epoch": 0.3605, + "grad_norm": 0.7739482522010803, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 36050 + }, + { + "epoch": 0.36051, + "grad_norm": 0.8185778260231018, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 36051 + }, + { + "epoch": 0.36052, + "grad_norm": 1.037488579750061, + "learning_rate": 0.003, + "loss": 4.0, + "step": 36052 + }, + { + "epoch": 0.36053, + "grad_norm": 1.31886625289917, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 36053 + }, + { + "epoch": 0.36054, + "grad_norm": 0.9552292227745056, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 36054 + }, + { + "epoch": 0.36055, + "grad_norm": 1.0684306621551514, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 36055 + }, + { + "epoch": 0.36056, + "grad_norm": 1.0631178617477417, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 36056 + }, + { + "epoch": 0.36057, + "grad_norm": 0.906614363193512, + "learning_rate": 0.003, + "loss": 3.996, + "step": 36057 + }, + { + "epoch": 0.36058, + "grad_norm": 0.8547419309616089, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 36058 + }, + { + "epoch": 0.36059, + "grad_norm": 0.8083584904670715, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 36059 + }, + { + "epoch": 0.3606, + "grad_norm": 0.7386074066162109, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 36060 + }, + { + "epoch": 0.36061, + "grad_norm": 0.7889652252197266, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 36061 + }, + { + "epoch": 0.36062, + "grad_norm": 0.8822491765022278, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 36062 + }, + { + "epoch": 0.36063, + "grad_norm": 1.015568733215332, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 36063 + }, + { + "epoch": 0.36064, + "grad_norm": 1.355622410774231, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 36064 + }, + { + "epoch": 0.36065, + "grad_norm": 0.5464690923690796, + "learning_rate": 0.003, + "loss": 3.9374, + "step": 36065 + }, + { + "epoch": 0.36066, + "grad_norm": 0.6410674452781677, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 36066 + }, + { + "epoch": 0.36067, + "grad_norm": 0.614759624004364, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 36067 + }, + { + "epoch": 0.36068, + "grad_norm": 0.703812837600708, + "learning_rate": 0.003, + "loss": 3.986, + "step": 36068 + }, + { + "epoch": 0.36069, + "grad_norm": 0.7660961151123047, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 36069 + }, + { + "epoch": 0.3607, + "grad_norm": 0.7905171513557434, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 36070 + }, + { + "epoch": 0.36071, + "grad_norm": 0.714677631855011, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 36071 + }, + { + "epoch": 0.36072, + "grad_norm": 0.7063193917274475, + "learning_rate": 0.003, + "loss": 3.944, + "step": 36072 + }, + { + "epoch": 0.36073, + "grad_norm": 0.7833319902420044, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 36073 + }, + { + "epoch": 0.36074, + "grad_norm": 0.9570627212524414, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 36074 + }, + { + "epoch": 0.36075, + "grad_norm": 1.2500591278076172, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 36075 + }, + { + "epoch": 0.36076, + "grad_norm": 0.8699328899383545, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 36076 + }, + { + "epoch": 0.36077, + "grad_norm": 0.8819198608398438, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 36077 + }, + { + "epoch": 0.36078, + "grad_norm": 0.8859657049179077, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 36078 + }, + { + "epoch": 0.36079, + "grad_norm": 0.8977394104003906, + "learning_rate": 0.003, + "loss": 3.99, + "step": 36079 + }, + { + "epoch": 0.3608, + "grad_norm": 1.0287827253341675, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 36080 + }, + { + "epoch": 0.36081, + "grad_norm": 1.009446382522583, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 36081 + }, + { + "epoch": 0.36082, + "grad_norm": 0.8519158363342285, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 36082 + }, + { + "epoch": 0.36083, + "grad_norm": 0.834469735622406, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 36083 + }, + { + "epoch": 0.36084, + "grad_norm": 0.9293833374977112, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 36084 + }, + { + "epoch": 0.36085, + "grad_norm": 1.0057878494262695, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 36085 + }, + { + "epoch": 0.36086, + "grad_norm": 1.1209962368011475, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 36086 + }, + { + "epoch": 0.36087, + "grad_norm": 0.985897958278656, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 36087 + }, + { + "epoch": 0.36088, + "grad_norm": 1.2145750522613525, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 36088 + }, + { + "epoch": 0.36089, + "grad_norm": 0.8708637952804565, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 36089 + }, + { + "epoch": 0.3609, + "grad_norm": 1.0160835981369019, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 36090 + }, + { + "epoch": 0.36091, + "grad_norm": 1.1447147130966187, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 36091 + }, + { + "epoch": 0.36092, + "grad_norm": 0.7998637557029724, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 36092 + }, + { + "epoch": 0.36093, + "grad_norm": 0.7195334434509277, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 36093 + }, + { + "epoch": 0.36094, + "grad_norm": 0.7087002992630005, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 36094 + }, + { + "epoch": 0.36095, + "grad_norm": 0.6620205044746399, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 36095 + }, + { + "epoch": 0.36096, + "grad_norm": 0.6801232695579529, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 36096 + }, + { + "epoch": 0.36097, + "grad_norm": 0.7694602012634277, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 36097 + }, + { + "epoch": 0.36098, + "grad_norm": 0.9584061503410339, + "learning_rate": 0.003, + "loss": 4.007, + "step": 36098 + }, + { + "epoch": 0.36099, + "grad_norm": 1.1654722690582275, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 36099 + }, + { + "epoch": 0.361, + "grad_norm": 1.2811646461486816, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 36100 + }, + { + "epoch": 0.36101, + "grad_norm": 0.8164727091789246, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 36101 + }, + { + "epoch": 0.36102, + "grad_norm": 0.7105634808540344, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 36102 + }, + { + "epoch": 0.36103, + "grad_norm": 0.7590923309326172, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 36103 + }, + { + "epoch": 0.36104, + "grad_norm": 0.8408402800559998, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 36104 + }, + { + "epoch": 0.36105, + "grad_norm": 0.777581512928009, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 36105 + }, + { + "epoch": 0.36106, + "grad_norm": 0.7868053913116455, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 36106 + }, + { + "epoch": 0.36107, + "grad_norm": 0.8721289038658142, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 36107 + }, + { + "epoch": 0.36108, + "grad_norm": 0.9231338500976562, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 36108 + }, + { + "epoch": 0.36109, + "grad_norm": 1.0343642234802246, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 36109 + }, + { + "epoch": 0.3611, + "grad_norm": 1.0183680057525635, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 36110 + }, + { + "epoch": 0.36111, + "grad_norm": 0.9627766609191895, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 36111 + }, + { + "epoch": 0.36112, + "grad_norm": 1.0250234603881836, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 36112 + }, + { + "epoch": 0.36113, + "grad_norm": 1.0419788360595703, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 36113 + }, + { + "epoch": 0.36114, + "grad_norm": 0.9010891318321228, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 36114 + }, + { + "epoch": 0.36115, + "grad_norm": 0.7915481925010681, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 36115 + }, + { + "epoch": 0.36116, + "grad_norm": 0.7199159860610962, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 36116 + }, + { + "epoch": 0.36117, + "grad_norm": 0.7463037967681885, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 36117 + }, + { + "epoch": 0.36118, + "grad_norm": 0.7533673644065857, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 36118 + }, + { + "epoch": 0.36119, + "grad_norm": 0.8684864640235901, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 36119 + }, + { + "epoch": 0.3612, + "grad_norm": 1.0478287935256958, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 36120 + }, + { + "epoch": 0.36121, + "grad_norm": 0.9129176735877991, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 36121 + }, + { + "epoch": 0.36122, + "grad_norm": 1.0058014392852783, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 36122 + }, + { + "epoch": 0.36123, + "grad_norm": 1.3202002048492432, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 36123 + }, + { + "epoch": 0.36124, + "grad_norm": 0.9454128742218018, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 36124 + }, + { + "epoch": 0.36125, + "grad_norm": 0.8785570859909058, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 36125 + }, + { + "epoch": 0.36126, + "grad_norm": 0.8785786628723145, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 36126 + }, + { + "epoch": 0.36127, + "grad_norm": 0.8109555840492249, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 36127 + }, + { + "epoch": 0.36128, + "grad_norm": 0.952059805393219, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 36128 + }, + { + "epoch": 0.36129, + "grad_norm": 1.0329437255859375, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 36129 + }, + { + "epoch": 0.3613, + "grad_norm": 0.9667741060256958, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 36130 + }, + { + "epoch": 0.36131, + "grad_norm": 1.105859637260437, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 36131 + }, + { + "epoch": 0.36132, + "grad_norm": 0.9825165867805481, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 36132 + }, + { + "epoch": 0.36133, + "grad_norm": 1.0063061714172363, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 36133 + }, + { + "epoch": 0.36134, + "grad_norm": 1.2911332845687866, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 36134 + }, + { + "epoch": 0.36135, + "grad_norm": 0.997704267501831, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 36135 + }, + { + "epoch": 0.36136, + "grad_norm": 0.9855402112007141, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 36136 + }, + { + "epoch": 0.36137, + "grad_norm": 1.0185012817382812, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 36137 + }, + { + "epoch": 0.36138, + "grad_norm": 1.0260400772094727, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 36138 + }, + { + "epoch": 0.36139, + "grad_norm": 0.927992045879364, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 36139 + }, + { + "epoch": 0.3614, + "grad_norm": 1.00204336643219, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 36140 + }, + { + "epoch": 0.36141, + "grad_norm": 0.8976835608482361, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 36141 + }, + { + "epoch": 0.36142, + "grad_norm": 0.909781813621521, + "learning_rate": 0.003, + "loss": 4.018, + "step": 36142 + }, + { + "epoch": 0.36143, + "grad_norm": 0.9012576937675476, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 36143 + }, + { + "epoch": 0.36144, + "grad_norm": 0.7568956613540649, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 36144 + }, + { + "epoch": 0.36145, + "grad_norm": 0.7671990394592285, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 36145 + }, + { + "epoch": 0.36146, + "grad_norm": 0.6010597348213196, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 36146 + }, + { + "epoch": 0.36147, + "grad_norm": 0.7353248596191406, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 36147 + }, + { + "epoch": 0.36148, + "grad_norm": 1.0101852416992188, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 36148 + }, + { + "epoch": 0.36149, + "grad_norm": 1.4396657943725586, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 36149 + }, + { + "epoch": 0.3615, + "grad_norm": 0.8442070484161377, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 36150 + }, + { + "epoch": 0.36151, + "grad_norm": 0.8749479055404663, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 36151 + }, + { + "epoch": 0.36152, + "grad_norm": 0.8969728350639343, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 36152 + }, + { + "epoch": 0.36153, + "grad_norm": 0.6311173439025879, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 36153 + }, + { + "epoch": 0.36154, + "grad_norm": 0.6073166131973267, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 36154 + }, + { + "epoch": 0.36155, + "grad_norm": 0.6456312537193298, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 36155 + }, + { + "epoch": 0.36156, + "grad_norm": 0.7543766498565674, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 36156 + }, + { + "epoch": 0.36157, + "grad_norm": 0.779406726360321, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 36157 + }, + { + "epoch": 0.36158, + "grad_norm": 0.7637103199958801, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 36158 + }, + { + "epoch": 0.36159, + "grad_norm": 0.9209310412406921, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 36159 + }, + { + "epoch": 0.3616, + "grad_norm": 1.2207363843917847, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 36160 + }, + { + "epoch": 0.36161, + "grad_norm": 0.9133251309394836, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 36161 + }, + { + "epoch": 0.36162, + "grad_norm": 0.7355939149856567, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 36162 + }, + { + "epoch": 0.36163, + "grad_norm": 0.7731209993362427, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 36163 + }, + { + "epoch": 0.36164, + "grad_norm": 0.8669020533561707, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 36164 + }, + { + "epoch": 0.36165, + "grad_norm": 0.8760286569595337, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 36165 + }, + { + "epoch": 0.36166, + "grad_norm": 0.7681072354316711, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 36166 + }, + { + "epoch": 0.36167, + "grad_norm": 0.6657683253288269, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 36167 + }, + { + "epoch": 0.36168, + "grad_norm": 0.6003568172454834, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 36168 + }, + { + "epoch": 0.36169, + "grad_norm": 0.6111166477203369, + "learning_rate": 0.003, + "loss": 3.9282, + "step": 36169 + }, + { + "epoch": 0.3617, + "grad_norm": 0.7832065224647522, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 36170 + }, + { + "epoch": 0.36171, + "grad_norm": 1.1268038749694824, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 36171 + }, + { + "epoch": 0.36172, + "grad_norm": 1.1043550968170166, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 36172 + }, + { + "epoch": 0.36173, + "grad_norm": 0.6707715392112732, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 36173 + }, + { + "epoch": 0.36174, + "grad_norm": 0.6252768635749817, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 36174 + }, + { + "epoch": 0.36175, + "grad_norm": 0.8477029204368591, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 36175 + }, + { + "epoch": 0.36176, + "grad_norm": 1.214210033416748, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 36176 + }, + { + "epoch": 0.36177, + "grad_norm": 0.677041232585907, + "learning_rate": 0.003, + "loss": 3.948, + "step": 36177 + }, + { + "epoch": 0.36178, + "grad_norm": 0.5506309866905212, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 36178 + }, + { + "epoch": 0.36179, + "grad_norm": 0.5961523056030273, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 36179 + }, + { + "epoch": 0.3618, + "grad_norm": 0.692940890789032, + "learning_rate": 0.003, + "loss": 3.9213, + "step": 36180 + }, + { + "epoch": 0.36181, + "grad_norm": 0.7532021403312683, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 36181 + }, + { + "epoch": 0.36182, + "grad_norm": 0.8387406468391418, + "learning_rate": 0.003, + "loss": 3.972, + "step": 36182 + }, + { + "epoch": 0.36183, + "grad_norm": 0.8542380928993225, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 36183 + }, + { + "epoch": 0.36184, + "grad_norm": 0.903258204460144, + "learning_rate": 0.003, + "loss": 3.9413, + "step": 36184 + }, + { + "epoch": 0.36185, + "grad_norm": 1.0361175537109375, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 36185 + }, + { + "epoch": 0.36186, + "grad_norm": 1.1229029893875122, + "learning_rate": 0.003, + "loss": 3.9434, + "step": 36186 + }, + { + "epoch": 0.36187, + "grad_norm": 0.8203414082527161, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 36187 + }, + { + "epoch": 0.36188, + "grad_norm": 0.7585783004760742, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 36188 + }, + { + "epoch": 0.36189, + "grad_norm": 0.9175700545310974, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 36189 + }, + { + "epoch": 0.3619, + "grad_norm": 1.1091065406799316, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 36190 + }, + { + "epoch": 0.36191, + "grad_norm": 1.1566972732543945, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 36191 + }, + { + "epoch": 0.36192, + "grad_norm": 0.916106104850769, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 36192 + }, + { + "epoch": 0.36193, + "grad_norm": 0.838434636592865, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 36193 + }, + { + "epoch": 0.36194, + "grad_norm": 0.7271400690078735, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 36194 + }, + { + "epoch": 0.36195, + "grad_norm": 0.7153394222259521, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 36195 + }, + { + "epoch": 0.36196, + "grad_norm": 0.6707715392112732, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 36196 + }, + { + "epoch": 0.36197, + "grad_norm": 0.7739907503128052, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 36197 + }, + { + "epoch": 0.36198, + "grad_norm": 0.8623161315917969, + "learning_rate": 0.003, + "loss": 3.958, + "step": 36198 + }, + { + "epoch": 0.36199, + "grad_norm": 0.8958646059036255, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 36199 + }, + { + "epoch": 0.362, + "grad_norm": 1.055206060409546, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 36200 + }, + { + "epoch": 0.36201, + "grad_norm": 1.0571995973587036, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 36201 + }, + { + "epoch": 0.36202, + "grad_norm": 1.0380141735076904, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 36202 + }, + { + "epoch": 0.36203, + "grad_norm": 1.2445530891418457, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 36203 + }, + { + "epoch": 0.36204, + "grad_norm": 0.7839359045028687, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 36204 + }, + { + "epoch": 0.36205, + "grad_norm": 0.6792985200881958, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 36205 + }, + { + "epoch": 0.36206, + "grad_norm": 0.7512255907058716, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 36206 + }, + { + "epoch": 0.36207, + "grad_norm": 0.8016072511672974, + "learning_rate": 0.003, + "loss": 3.954, + "step": 36207 + }, + { + "epoch": 0.36208, + "grad_norm": 0.8007936477661133, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 36208 + }, + { + "epoch": 0.36209, + "grad_norm": 0.7913188338279724, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 36209 + }, + { + "epoch": 0.3621, + "grad_norm": 0.863066554069519, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 36210 + }, + { + "epoch": 0.36211, + "grad_norm": 1.1091877222061157, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 36211 + }, + { + "epoch": 0.36212, + "grad_norm": 0.9973095655441284, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 36212 + }, + { + "epoch": 0.36213, + "grad_norm": 1.1891016960144043, + "learning_rate": 0.003, + "loss": 3.983, + "step": 36213 + }, + { + "epoch": 0.36214, + "grad_norm": 1.065670371055603, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 36214 + }, + { + "epoch": 0.36215, + "grad_norm": 0.8491978645324707, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 36215 + }, + { + "epoch": 0.36216, + "grad_norm": 0.7163688540458679, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 36216 + }, + { + "epoch": 0.36217, + "grad_norm": 0.8003910779953003, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 36217 + }, + { + "epoch": 0.36218, + "grad_norm": 0.9243523478507996, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 36218 + }, + { + "epoch": 0.36219, + "grad_norm": 1.1136255264282227, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 36219 + }, + { + "epoch": 0.3622, + "grad_norm": 0.7828724980354309, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 36220 + }, + { + "epoch": 0.36221, + "grad_norm": 0.8160341382026672, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 36221 + }, + { + "epoch": 0.36222, + "grad_norm": 0.9542453289031982, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 36222 + }, + { + "epoch": 0.36223, + "grad_norm": 1.0931925773620605, + "learning_rate": 0.003, + "loss": 3.999, + "step": 36223 + }, + { + "epoch": 0.36224, + "grad_norm": 0.877729594707489, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 36224 + }, + { + "epoch": 0.36225, + "grad_norm": 0.9705217480659485, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 36225 + }, + { + "epoch": 0.36226, + "grad_norm": 1.4272472858428955, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 36226 + }, + { + "epoch": 0.36227, + "grad_norm": 0.5872136950492859, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 36227 + }, + { + "epoch": 0.36228, + "grad_norm": 0.9057717323303223, + "learning_rate": 0.003, + "loss": 3.984, + "step": 36228 + }, + { + "epoch": 0.36229, + "grad_norm": 1.3698935508728027, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 36229 + }, + { + "epoch": 0.3623, + "grad_norm": 0.7867864370346069, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 36230 + }, + { + "epoch": 0.36231, + "grad_norm": 0.7584843635559082, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 36231 + }, + { + "epoch": 0.36232, + "grad_norm": 0.8367906808853149, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 36232 + }, + { + "epoch": 0.36233, + "grad_norm": 0.8033140897750854, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 36233 + }, + { + "epoch": 0.36234, + "grad_norm": 0.8295102119445801, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 36234 + }, + { + "epoch": 0.36235, + "grad_norm": 0.904066264629364, + "learning_rate": 0.003, + "loss": 4.01, + "step": 36235 + }, + { + "epoch": 0.36236, + "grad_norm": 0.985434889793396, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 36236 + }, + { + "epoch": 0.36237, + "grad_norm": 1.1136945486068726, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 36237 + }, + { + "epoch": 0.36238, + "grad_norm": 0.9173763990402222, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 36238 + }, + { + "epoch": 0.36239, + "grad_norm": 0.9217658042907715, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 36239 + }, + { + "epoch": 0.3624, + "grad_norm": 0.9798958897590637, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 36240 + }, + { + "epoch": 0.36241, + "grad_norm": 1.156337857246399, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 36241 + }, + { + "epoch": 0.36242, + "grad_norm": 0.8624242544174194, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 36242 + }, + { + "epoch": 0.36243, + "grad_norm": 0.8185104131698608, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 36243 + }, + { + "epoch": 0.36244, + "grad_norm": 1.0300647020339966, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 36244 + }, + { + "epoch": 0.36245, + "grad_norm": 1.1481122970581055, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 36245 + }, + { + "epoch": 0.36246, + "grad_norm": 0.8061193823814392, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 36246 + }, + { + "epoch": 0.36247, + "grad_norm": 0.8002191185951233, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 36247 + }, + { + "epoch": 0.36248, + "grad_norm": 1.074704647064209, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 36248 + }, + { + "epoch": 0.36249, + "grad_norm": 1.0171774625778198, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 36249 + }, + { + "epoch": 0.3625, + "grad_norm": 0.890031099319458, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 36250 + }, + { + "epoch": 0.36251, + "grad_norm": 1.021596074104309, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 36251 + }, + { + "epoch": 0.36252, + "grad_norm": 0.9680755138397217, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 36252 + }, + { + "epoch": 0.36253, + "grad_norm": 0.9907793998718262, + "learning_rate": 0.003, + "loss": 4.024, + "step": 36253 + }, + { + "epoch": 0.36254, + "grad_norm": 1.040075421333313, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 36254 + }, + { + "epoch": 0.36255, + "grad_norm": 1.2499139308929443, + "learning_rate": 0.003, + "loss": 3.984, + "step": 36255 + }, + { + "epoch": 0.36256, + "grad_norm": 0.7540169358253479, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 36256 + }, + { + "epoch": 0.36257, + "grad_norm": 0.6694825887680054, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 36257 + }, + { + "epoch": 0.36258, + "grad_norm": 0.8793380856513977, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 36258 + }, + { + "epoch": 0.36259, + "grad_norm": 0.9362257122993469, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 36259 + }, + { + "epoch": 0.3626, + "grad_norm": 0.9933204650878906, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 36260 + }, + { + "epoch": 0.36261, + "grad_norm": 0.9508855938911438, + "learning_rate": 0.003, + "loss": 3.988, + "step": 36261 + }, + { + "epoch": 0.36262, + "grad_norm": 0.8645362257957458, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 36262 + }, + { + "epoch": 0.36263, + "grad_norm": 1.0014911890029907, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 36263 + }, + { + "epoch": 0.36264, + "grad_norm": 1.114276647567749, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 36264 + }, + { + "epoch": 0.36265, + "grad_norm": 1.1073129177093506, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 36265 + }, + { + "epoch": 0.36266, + "grad_norm": 0.8058702349662781, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 36266 + }, + { + "epoch": 0.36267, + "grad_norm": 0.7549871206283569, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 36267 + }, + { + "epoch": 0.36268, + "grad_norm": 0.7445017695426941, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 36268 + }, + { + "epoch": 0.36269, + "grad_norm": 0.7724642157554626, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 36269 + }, + { + "epoch": 0.3627, + "grad_norm": 0.9927551746368408, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 36270 + }, + { + "epoch": 0.36271, + "grad_norm": 1.1760729551315308, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 36271 + }, + { + "epoch": 0.36272, + "grad_norm": 0.885515570640564, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 36272 + }, + { + "epoch": 0.36273, + "grad_norm": 0.7456080317497253, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 36273 + }, + { + "epoch": 0.36274, + "grad_norm": 0.7224969863891602, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 36274 + }, + { + "epoch": 0.36275, + "grad_norm": 0.8298744559288025, + "learning_rate": 0.003, + "loss": 3.982, + "step": 36275 + }, + { + "epoch": 0.36276, + "grad_norm": 0.8984490036964417, + "learning_rate": 0.003, + "loss": 3.981, + "step": 36276 + }, + { + "epoch": 0.36277, + "grad_norm": 0.870729923248291, + "learning_rate": 0.003, + "loss": 3.983, + "step": 36277 + }, + { + "epoch": 0.36278, + "grad_norm": 0.9153876304626465, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 36278 + }, + { + "epoch": 0.36279, + "grad_norm": 1.0102487802505493, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 36279 + }, + { + "epoch": 0.3628, + "grad_norm": 1.0286312103271484, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 36280 + }, + { + "epoch": 0.36281, + "grad_norm": 0.9859628081321716, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 36281 + }, + { + "epoch": 0.36282, + "grad_norm": 0.9221677184104919, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 36282 + }, + { + "epoch": 0.36283, + "grad_norm": 0.9292837381362915, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 36283 + }, + { + "epoch": 0.36284, + "grad_norm": 1.1213939189910889, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 36284 + }, + { + "epoch": 0.36285, + "grad_norm": 1.0840985774993896, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 36285 + }, + { + "epoch": 0.36286, + "grad_norm": 0.8049089908599854, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 36286 + }, + { + "epoch": 0.36287, + "grad_norm": 0.7010928988456726, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 36287 + }, + { + "epoch": 0.36288, + "grad_norm": 0.6597599387168884, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 36288 + }, + { + "epoch": 0.36289, + "grad_norm": 0.6496340036392212, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 36289 + }, + { + "epoch": 0.3629, + "grad_norm": 0.6656325459480286, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 36290 + }, + { + "epoch": 0.36291, + "grad_norm": 0.658548891544342, + "learning_rate": 0.003, + "loss": 3.966, + "step": 36291 + }, + { + "epoch": 0.36292, + "grad_norm": 0.6700641512870789, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 36292 + }, + { + "epoch": 0.36293, + "grad_norm": 0.7893662452697754, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 36293 + }, + { + "epoch": 0.36294, + "grad_norm": 0.9962747097015381, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 36294 + }, + { + "epoch": 0.36295, + "grad_norm": 1.1685795783996582, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 36295 + }, + { + "epoch": 0.36296, + "grad_norm": 1.2439337968826294, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 36296 + }, + { + "epoch": 0.36297, + "grad_norm": 0.710731029510498, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 36297 + }, + { + "epoch": 0.36298, + "grad_norm": 0.575681746006012, + "learning_rate": 0.003, + "loss": 3.9387, + "step": 36298 + }, + { + "epoch": 0.36299, + "grad_norm": 0.6573310494422913, + "learning_rate": 0.003, + "loss": 3.964, + "step": 36299 + }, + { + "epoch": 0.363, + "grad_norm": 0.7833206057548523, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 36300 + }, + { + "epoch": 0.36301, + "grad_norm": 0.9956979155540466, + "learning_rate": 0.003, + "loss": 3.991, + "step": 36301 + }, + { + "epoch": 0.36302, + "grad_norm": 1.1133769750595093, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 36302 + }, + { + "epoch": 0.36303, + "grad_norm": 0.8239755034446716, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 36303 + }, + { + "epoch": 0.36304, + "grad_norm": 0.7247451543807983, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 36304 + }, + { + "epoch": 0.36305, + "grad_norm": 0.6934221386909485, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 36305 + }, + { + "epoch": 0.36306, + "grad_norm": 0.6913109421730042, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 36306 + }, + { + "epoch": 0.36307, + "grad_norm": 0.8739315867424011, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 36307 + }, + { + "epoch": 0.36308, + "grad_norm": 1.1638317108154297, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 36308 + }, + { + "epoch": 0.36309, + "grad_norm": 0.9947952032089233, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 36309 + }, + { + "epoch": 0.3631, + "grad_norm": 1.0364195108413696, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 36310 + }, + { + "epoch": 0.36311, + "grad_norm": 0.991275429725647, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 36311 + }, + { + "epoch": 0.36312, + "grad_norm": 0.988101601600647, + "learning_rate": 0.003, + "loss": 3.969, + "step": 36312 + }, + { + "epoch": 0.36313, + "grad_norm": 1.0298163890838623, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 36313 + }, + { + "epoch": 0.36314, + "grad_norm": 1.0366355180740356, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 36314 + }, + { + "epoch": 0.36315, + "grad_norm": 0.9151508808135986, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 36315 + }, + { + "epoch": 0.36316, + "grad_norm": 1.0409982204437256, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 36316 + }, + { + "epoch": 0.36317, + "grad_norm": 0.922867476940155, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 36317 + }, + { + "epoch": 0.36318, + "grad_norm": 0.8365194201469421, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 36318 + }, + { + "epoch": 0.36319, + "grad_norm": 0.8591129779815674, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 36319 + }, + { + "epoch": 0.3632, + "grad_norm": 0.8713525533676147, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 36320 + }, + { + "epoch": 0.36321, + "grad_norm": 1.1457427740097046, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 36321 + }, + { + "epoch": 0.36322, + "grad_norm": 1.3817074298858643, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 36322 + }, + { + "epoch": 0.36323, + "grad_norm": 0.7288424372673035, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 36323 + }, + { + "epoch": 0.36324, + "grad_norm": 0.7171038389205933, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 36324 + }, + { + "epoch": 0.36325, + "grad_norm": 0.7784549593925476, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 36325 + }, + { + "epoch": 0.36326, + "grad_norm": 0.884393036365509, + "learning_rate": 0.003, + "loss": 3.985, + "step": 36326 + }, + { + "epoch": 0.36327, + "grad_norm": 1.036455750465393, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 36327 + }, + { + "epoch": 0.36328, + "grad_norm": 1.0116422176361084, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 36328 + }, + { + "epoch": 0.36329, + "grad_norm": 1.0165899991989136, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 36329 + }, + { + "epoch": 0.3633, + "grad_norm": 1.0928550958633423, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 36330 + }, + { + "epoch": 0.36331, + "grad_norm": 1.117455244064331, + "learning_rate": 0.003, + "loss": 4.001, + "step": 36331 + }, + { + "epoch": 0.36332, + "grad_norm": 0.7159485220909119, + "learning_rate": 0.003, + "loss": 3.993, + "step": 36332 + }, + { + "epoch": 0.36333, + "grad_norm": 0.5882036685943604, + "learning_rate": 0.003, + "loss": 3.972, + "step": 36333 + }, + { + "epoch": 0.36334, + "grad_norm": 0.7255657911300659, + "learning_rate": 0.003, + "loss": 4.012, + "step": 36334 + }, + { + "epoch": 0.36335, + "grad_norm": 0.8424400687217712, + "learning_rate": 0.003, + "loss": 3.965, + "step": 36335 + }, + { + "epoch": 0.36336, + "grad_norm": 0.9205389618873596, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 36336 + }, + { + "epoch": 0.36337, + "grad_norm": 0.8996075987815857, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 36337 + }, + { + "epoch": 0.36338, + "grad_norm": 0.7740461826324463, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 36338 + }, + { + "epoch": 0.36339, + "grad_norm": 0.7047325968742371, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 36339 + }, + { + "epoch": 0.3634, + "grad_norm": 0.7798894047737122, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 36340 + }, + { + "epoch": 0.36341, + "grad_norm": 1.0248438119888306, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 36341 + }, + { + "epoch": 0.36342, + "grad_norm": 1.0688751935958862, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 36342 + }, + { + "epoch": 0.36343, + "grad_norm": 1.053408145904541, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 36343 + }, + { + "epoch": 0.36344, + "grad_norm": 1.1461325883865356, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 36344 + }, + { + "epoch": 0.36345, + "grad_norm": 0.8875594735145569, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 36345 + }, + { + "epoch": 0.36346, + "grad_norm": 0.8949811458587646, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 36346 + }, + { + "epoch": 0.36347, + "grad_norm": 0.7587918043136597, + "learning_rate": 0.003, + "loss": 4.017, + "step": 36347 + }, + { + "epoch": 0.36348, + "grad_norm": 0.7307963371276855, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 36348 + }, + { + "epoch": 0.36349, + "grad_norm": 0.8973195552825928, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 36349 + }, + { + "epoch": 0.3635, + "grad_norm": 1.2527649402618408, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 36350 + }, + { + "epoch": 0.36351, + "grad_norm": 0.9519528746604919, + "learning_rate": 0.003, + "loss": 3.992, + "step": 36351 + }, + { + "epoch": 0.36352, + "grad_norm": 0.9696971774101257, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 36352 + }, + { + "epoch": 0.36353, + "grad_norm": 1.002807378768921, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 36353 + }, + { + "epoch": 0.36354, + "grad_norm": 1.3173396587371826, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 36354 + }, + { + "epoch": 0.36355, + "grad_norm": 0.7194259762763977, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 36355 + }, + { + "epoch": 0.36356, + "grad_norm": 0.7412980794906616, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 36356 + }, + { + "epoch": 0.36357, + "grad_norm": 0.6731964945793152, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 36357 + }, + { + "epoch": 0.36358, + "grad_norm": 0.7265130877494812, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 36358 + }, + { + "epoch": 0.36359, + "grad_norm": 0.7457084059715271, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 36359 + }, + { + "epoch": 0.3636, + "grad_norm": 0.8726292252540588, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 36360 + }, + { + "epoch": 0.36361, + "grad_norm": 1.293214201927185, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 36361 + }, + { + "epoch": 0.36362, + "grad_norm": 1.0155565738677979, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 36362 + }, + { + "epoch": 0.36363, + "grad_norm": 0.8449653387069702, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 36363 + }, + { + "epoch": 0.36364, + "grad_norm": 0.6972106099128723, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 36364 + }, + { + "epoch": 0.36365, + "grad_norm": 0.7546465992927551, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 36365 + }, + { + "epoch": 0.36366, + "grad_norm": 0.7669247984886169, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 36366 + }, + { + "epoch": 0.36367, + "grad_norm": 0.8615423440933228, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 36367 + }, + { + "epoch": 0.36368, + "grad_norm": 0.767210841178894, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 36368 + }, + { + "epoch": 0.36369, + "grad_norm": 0.683502197265625, + "learning_rate": 0.003, + "loss": 3.9364, + "step": 36369 + }, + { + "epoch": 0.3637, + "grad_norm": 0.6394029259681702, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 36370 + }, + { + "epoch": 0.36371, + "grad_norm": 0.7348243594169617, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 36371 + }, + { + "epoch": 0.36372, + "grad_norm": 0.7488658428192139, + "learning_rate": 0.003, + "loss": 3.994, + "step": 36372 + }, + { + "epoch": 0.36373, + "grad_norm": 0.7297083735466003, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 36373 + }, + { + "epoch": 0.36374, + "grad_norm": 0.755549967288971, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 36374 + }, + { + "epoch": 0.36375, + "grad_norm": 0.7910747528076172, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 36375 + }, + { + "epoch": 0.36376, + "grad_norm": 0.8537231087684631, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 36376 + }, + { + "epoch": 0.36377, + "grad_norm": 0.9387192130088806, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 36377 + }, + { + "epoch": 0.36378, + "grad_norm": 1.0476115942001343, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 36378 + }, + { + "epoch": 0.36379, + "grad_norm": 1.4507019519805908, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 36379 + }, + { + "epoch": 0.3638, + "grad_norm": 0.8133090138435364, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 36380 + }, + { + "epoch": 0.36381, + "grad_norm": 0.9471381306648254, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 36381 + }, + { + "epoch": 0.36382, + "grad_norm": 1.1807910203933716, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 36382 + }, + { + "epoch": 0.36383, + "grad_norm": 1.0554862022399902, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 36383 + }, + { + "epoch": 0.36384, + "grad_norm": 0.8208101391792297, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 36384 + }, + { + "epoch": 0.36385, + "grad_norm": 0.7356413006782532, + "learning_rate": 0.003, + "loss": 3.98, + "step": 36385 + }, + { + "epoch": 0.36386, + "grad_norm": 0.876285970211029, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 36386 + }, + { + "epoch": 0.36387, + "grad_norm": 1.16938316822052, + "learning_rate": 0.003, + "loss": 3.979, + "step": 36387 + }, + { + "epoch": 0.36388, + "grad_norm": 0.9649495482444763, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 36388 + }, + { + "epoch": 0.36389, + "grad_norm": 0.7775450348854065, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 36389 + }, + { + "epoch": 0.3639, + "grad_norm": 0.7216055393218994, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 36390 + }, + { + "epoch": 0.36391, + "grad_norm": 0.793452799320221, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 36391 + }, + { + "epoch": 0.36392, + "grad_norm": 0.940920352935791, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 36392 + }, + { + "epoch": 0.36393, + "grad_norm": 1.018301248550415, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 36393 + }, + { + "epoch": 0.36394, + "grad_norm": 0.9881923198699951, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 36394 + }, + { + "epoch": 0.36395, + "grad_norm": 1.0768961906433105, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 36395 + }, + { + "epoch": 0.36396, + "grad_norm": 0.9753762483596802, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 36396 + }, + { + "epoch": 0.36397, + "grad_norm": 1.2187345027923584, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 36397 + }, + { + "epoch": 0.36398, + "grad_norm": 1.0077468156814575, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 36398 + }, + { + "epoch": 0.36399, + "grad_norm": 1.0685802698135376, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 36399 + }, + { + "epoch": 0.364, + "grad_norm": 1.0109360218048096, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 36400 + }, + { + "epoch": 0.36401, + "grad_norm": 1.119766354560852, + "learning_rate": 0.003, + "loss": 4.009, + "step": 36401 + }, + { + "epoch": 0.36402, + "grad_norm": 0.9701829552650452, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 36402 + }, + { + "epoch": 0.36403, + "grad_norm": 0.9204695224761963, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 36403 + }, + { + "epoch": 0.36404, + "grad_norm": 0.886694073677063, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 36404 + }, + { + "epoch": 0.36405, + "grad_norm": 0.942781388759613, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 36405 + }, + { + "epoch": 0.36406, + "grad_norm": 0.8625243306159973, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 36406 + }, + { + "epoch": 0.36407, + "grad_norm": 0.832816481590271, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 36407 + }, + { + "epoch": 0.36408, + "grad_norm": 0.8359086513519287, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 36408 + }, + { + "epoch": 0.36409, + "grad_norm": 0.8852959275245667, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 36409 + }, + { + "epoch": 0.3641, + "grad_norm": 0.8981161713600159, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 36410 + }, + { + "epoch": 0.36411, + "grad_norm": 1.1053460836410522, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 36411 + }, + { + "epoch": 0.36412, + "grad_norm": 1.1185307502746582, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 36412 + }, + { + "epoch": 0.36413, + "grad_norm": 0.9807672500610352, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 36413 + }, + { + "epoch": 0.36414, + "grad_norm": 0.9224753975868225, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 36414 + }, + { + "epoch": 0.36415, + "grad_norm": 0.9681077599525452, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 36415 + }, + { + "epoch": 0.36416, + "grad_norm": 1.1922792196273804, + "learning_rate": 0.003, + "loss": 3.999, + "step": 36416 + }, + { + "epoch": 0.36417, + "grad_norm": 0.7806292772293091, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 36417 + }, + { + "epoch": 0.36418, + "grad_norm": 0.7315144538879395, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 36418 + }, + { + "epoch": 0.36419, + "grad_norm": 0.7857463955879211, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 36419 + }, + { + "epoch": 0.3642, + "grad_norm": 0.8618635535240173, + "learning_rate": 0.003, + "loss": 4.003, + "step": 36420 + }, + { + "epoch": 0.36421, + "grad_norm": 0.9373423457145691, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 36421 + }, + { + "epoch": 0.36422, + "grad_norm": 1.0050557851791382, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 36422 + }, + { + "epoch": 0.36423, + "grad_norm": 1.0043771266937256, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 36423 + }, + { + "epoch": 0.36424, + "grad_norm": 0.8928292989730835, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 36424 + }, + { + "epoch": 0.36425, + "grad_norm": 0.9458958506584167, + "learning_rate": 0.003, + "loss": 3.984, + "step": 36425 + }, + { + "epoch": 0.36426, + "grad_norm": 1.2605372667312622, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 36426 + }, + { + "epoch": 0.36427, + "grad_norm": 0.8849123120307922, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 36427 + }, + { + "epoch": 0.36428, + "grad_norm": 0.791580855846405, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 36428 + }, + { + "epoch": 0.36429, + "grad_norm": 0.9255066514015198, + "learning_rate": 0.003, + "loss": 3.995, + "step": 36429 + }, + { + "epoch": 0.3643, + "grad_norm": 1.0730067491531372, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 36430 + }, + { + "epoch": 0.36431, + "grad_norm": 0.9808925986289978, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 36431 + }, + { + "epoch": 0.36432, + "grad_norm": 0.8975308537483215, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 36432 + }, + { + "epoch": 0.36433, + "grad_norm": 0.753492534160614, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 36433 + }, + { + "epoch": 0.36434, + "grad_norm": 0.8011667728424072, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 36434 + }, + { + "epoch": 0.36435, + "grad_norm": 0.8572367429733276, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 36435 + }, + { + "epoch": 0.36436, + "grad_norm": 0.943211019039154, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 36436 + }, + { + "epoch": 0.36437, + "grad_norm": 0.8928398489952087, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 36437 + }, + { + "epoch": 0.36438, + "grad_norm": 0.9036343693733215, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 36438 + }, + { + "epoch": 0.36439, + "grad_norm": 0.8289978504180908, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 36439 + }, + { + "epoch": 0.3644, + "grad_norm": 0.8235885500907898, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 36440 + }, + { + "epoch": 0.36441, + "grad_norm": 0.8176711797714233, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 36441 + }, + { + "epoch": 0.36442, + "grad_norm": 0.901944100856781, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 36442 + }, + { + "epoch": 0.36443, + "grad_norm": 0.8311805129051208, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 36443 + }, + { + "epoch": 0.36444, + "grad_norm": 0.724233865737915, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 36444 + }, + { + "epoch": 0.36445, + "grad_norm": 0.7948705554008484, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 36445 + }, + { + "epoch": 0.36446, + "grad_norm": 1.1742877960205078, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 36446 + }, + { + "epoch": 0.36447, + "grad_norm": 0.9903634786605835, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 36447 + }, + { + "epoch": 0.36448, + "grad_norm": 0.8881961703300476, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 36448 + }, + { + "epoch": 0.36449, + "grad_norm": 0.8985059857368469, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 36449 + }, + { + "epoch": 0.3645, + "grad_norm": 0.8921286463737488, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 36450 + }, + { + "epoch": 0.36451, + "grad_norm": 0.8812556862831116, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 36451 + }, + { + "epoch": 0.36452, + "grad_norm": 0.8712377548217773, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 36452 + }, + { + "epoch": 0.36453, + "grad_norm": 1.0105340480804443, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 36453 + }, + { + "epoch": 0.36454, + "grad_norm": 1.0590894222259521, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 36454 + }, + { + "epoch": 0.36455, + "grad_norm": 0.9629542827606201, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 36455 + }, + { + "epoch": 0.36456, + "grad_norm": 1.1014587879180908, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 36456 + }, + { + "epoch": 0.36457, + "grad_norm": 0.8923196196556091, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 36457 + }, + { + "epoch": 0.36458, + "grad_norm": 1.0815839767456055, + "learning_rate": 0.003, + "loss": 4.033, + "step": 36458 + }, + { + "epoch": 0.36459, + "grad_norm": 1.1000267267227173, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 36459 + }, + { + "epoch": 0.3646, + "grad_norm": 0.9575393795967102, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 36460 + }, + { + "epoch": 0.36461, + "grad_norm": 1.0345470905303955, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 36461 + }, + { + "epoch": 0.36462, + "grad_norm": 0.9543490409851074, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 36462 + }, + { + "epoch": 0.36463, + "grad_norm": 0.9121909141540527, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 36463 + }, + { + "epoch": 0.36464, + "grad_norm": 0.9125641584396362, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 36464 + }, + { + "epoch": 0.36465, + "grad_norm": 0.9470889568328857, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 36465 + }, + { + "epoch": 0.36466, + "grad_norm": 1.0353909730911255, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 36466 + }, + { + "epoch": 0.36467, + "grad_norm": 0.9554426670074463, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 36467 + }, + { + "epoch": 0.36468, + "grad_norm": 0.9530978202819824, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 36468 + }, + { + "epoch": 0.36469, + "grad_norm": 0.9060129523277283, + "learning_rate": 0.003, + "loss": 3.975, + "step": 36469 + }, + { + "epoch": 0.3647, + "grad_norm": 0.9023591876029968, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 36470 + }, + { + "epoch": 0.36471, + "grad_norm": 0.7484405040740967, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 36471 + }, + { + "epoch": 0.36472, + "grad_norm": 0.6330456733703613, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 36472 + }, + { + "epoch": 0.36473, + "grad_norm": 0.7147196531295776, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 36473 + }, + { + "epoch": 0.36474, + "grad_norm": 0.7537676095962524, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 36474 + }, + { + "epoch": 0.36475, + "grad_norm": 0.7470688223838806, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 36475 + }, + { + "epoch": 0.36476, + "grad_norm": 0.751457154750824, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 36476 + }, + { + "epoch": 0.36477, + "grad_norm": 0.7615470290184021, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 36477 + }, + { + "epoch": 0.36478, + "grad_norm": 0.8934872150421143, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 36478 + }, + { + "epoch": 0.36479, + "grad_norm": 1.0744186639785767, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 36479 + }, + { + "epoch": 0.3648, + "grad_norm": 0.9923238754272461, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 36480 + }, + { + "epoch": 0.36481, + "grad_norm": 0.9436438083648682, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 36481 + }, + { + "epoch": 0.36482, + "grad_norm": 0.9163942337036133, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 36482 + }, + { + "epoch": 0.36483, + "grad_norm": 0.9636886119842529, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 36483 + }, + { + "epoch": 0.36484, + "grad_norm": 0.9511975646018982, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 36484 + }, + { + "epoch": 0.36485, + "grad_norm": 0.9472163915634155, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 36485 + }, + { + "epoch": 0.36486, + "grad_norm": 1.0805997848510742, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 36486 + }, + { + "epoch": 0.36487, + "grad_norm": 0.9662597179412842, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 36487 + }, + { + "epoch": 0.36488, + "grad_norm": 1.2002968788146973, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 36488 + }, + { + "epoch": 0.36489, + "grad_norm": 0.907866358757019, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 36489 + }, + { + "epoch": 0.3649, + "grad_norm": 0.7796376347541809, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 36490 + }, + { + "epoch": 0.36491, + "grad_norm": 0.819101095199585, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 36491 + }, + { + "epoch": 0.36492, + "grad_norm": 0.9898520112037659, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 36492 + }, + { + "epoch": 0.36493, + "grad_norm": 1.023210883140564, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 36493 + }, + { + "epoch": 0.36494, + "grad_norm": 1.093840479850769, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 36494 + }, + { + "epoch": 0.36495, + "grad_norm": 1.0642319917678833, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 36495 + }, + { + "epoch": 0.36496, + "grad_norm": 1.1162891387939453, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 36496 + }, + { + "epoch": 0.36497, + "grad_norm": 0.8573946952819824, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 36497 + }, + { + "epoch": 0.36498, + "grad_norm": 0.6961253881454468, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 36498 + }, + { + "epoch": 0.36499, + "grad_norm": 0.6521151661872864, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 36499 + }, + { + "epoch": 0.365, + "grad_norm": 0.623688817024231, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 36500 + }, + { + "epoch": 0.36501, + "grad_norm": 0.6567660570144653, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 36501 + }, + { + "epoch": 0.36502, + "grad_norm": 0.7379486560821533, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 36502 + }, + { + "epoch": 0.36503, + "grad_norm": 0.9357768297195435, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 36503 + }, + { + "epoch": 0.36504, + "grad_norm": 0.9912369251251221, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 36504 + }, + { + "epoch": 0.36505, + "grad_norm": 0.9972144365310669, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 36505 + }, + { + "epoch": 0.36506, + "grad_norm": 0.969127893447876, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 36506 + }, + { + "epoch": 0.36507, + "grad_norm": 0.8203965425491333, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 36507 + }, + { + "epoch": 0.36508, + "grad_norm": 0.7966840267181396, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 36508 + }, + { + "epoch": 0.36509, + "grad_norm": 0.924812912940979, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 36509 + }, + { + "epoch": 0.3651, + "grad_norm": 1.0204983949661255, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 36510 + }, + { + "epoch": 0.36511, + "grad_norm": 0.987029492855072, + "learning_rate": 0.003, + "loss": 4.001, + "step": 36511 + }, + { + "epoch": 0.36512, + "grad_norm": 1.035466194152832, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 36512 + }, + { + "epoch": 0.36513, + "grad_norm": 0.7858030200004578, + "learning_rate": 0.003, + "loss": 3.9463, + "step": 36513 + }, + { + "epoch": 0.36514, + "grad_norm": 0.7846704125404358, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 36514 + }, + { + "epoch": 0.36515, + "grad_norm": 0.7932121753692627, + "learning_rate": 0.003, + "loss": 3.941, + "step": 36515 + }, + { + "epoch": 0.36516, + "grad_norm": 0.9384203553199768, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 36516 + }, + { + "epoch": 0.36517, + "grad_norm": 1.1703604459762573, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 36517 + }, + { + "epoch": 0.36518, + "grad_norm": 0.9073598980903625, + "learning_rate": 0.003, + "loss": 3.964, + "step": 36518 + }, + { + "epoch": 0.36519, + "grad_norm": 0.8514218926429749, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 36519 + }, + { + "epoch": 0.3652, + "grad_norm": 0.9180405139923096, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 36520 + }, + { + "epoch": 0.36521, + "grad_norm": 1.1098990440368652, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 36521 + }, + { + "epoch": 0.36522, + "grad_norm": 0.8455879092216492, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 36522 + }, + { + "epoch": 0.36523, + "grad_norm": 0.6048752069473267, + "learning_rate": 0.003, + "loss": 3.983, + "step": 36523 + }, + { + "epoch": 0.36524, + "grad_norm": 0.6735074520111084, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 36524 + }, + { + "epoch": 0.36525, + "grad_norm": 0.8817867040634155, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 36525 + }, + { + "epoch": 0.36526, + "grad_norm": 1.0871502161026, + "learning_rate": 0.003, + "loss": 3.987, + "step": 36526 + }, + { + "epoch": 0.36527, + "grad_norm": 1.0123695135116577, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 36527 + }, + { + "epoch": 0.36528, + "grad_norm": 0.8868998885154724, + "learning_rate": 0.003, + "loss": 3.986, + "step": 36528 + }, + { + "epoch": 0.36529, + "grad_norm": 1.0132567882537842, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 36529 + }, + { + "epoch": 0.3653, + "grad_norm": 0.9926930665969849, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 36530 + }, + { + "epoch": 0.36531, + "grad_norm": 0.8791357278823853, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 36531 + }, + { + "epoch": 0.36532, + "grad_norm": 0.94716876745224, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 36532 + }, + { + "epoch": 0.36533, + "grad_norm": 1.0580391883850098, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 36533 + }, + { + "epoch": 0.36534, + "grad_norm": 1.1886273622512817, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 36534 + }, + { + "epoch": 0.36535, + "grad_norm": 0.9268249869346619, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 36535 + }, + { + "epoch": 0.36536, + "grad_norm": 1.1435670852661133, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 36536 + }, + { + "epoch": 0.36537, + "grad_norm": 0.8442835211753845, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 36537 + }, + { + "epoch": 0.36538, + "grad_norm": 0.6880837678909302, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 36538 + }, + { + "epoch": 0.36539, + "grad_norm": 0.7066935896873474, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 36539 + }, + { + "epoch": 0.3654, + "grad_norm": 0.7227423191070557, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 36540 + }, + { + "epoch": 0.36541, + "grad_norm": 0.6022458076477051, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 36541 + }, + { + "epoch": 0.36542, + "grad_norm": 0.6891031265258789, + "learning_rate": 0.003, + "loss": 3.969, + "step": 36542 + }, + { + "epoch": 0.36543, + "grad_norm": 0.59360671043396, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 36543 + }, + { + "epoch": 0.36544, + "grad_norm": 0.6468075513839722, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 36544 + }, + { + "epoch": 0.36545, + "grad_norm": 0.8233362436294556, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 36545 + }, + { + "epoch": 0.36546, + "grad_norm": 0.9054297208786011, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 36546 + }, + { + "epoch": 0.36547, + "grad_norm": 0.987572968006134, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 36547 + }, + { + "epoch": 0.36548, + "grad_norm": 1.066230297088623, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 36548 + }, + { + "epoch": 0.36549, + "grad_norm": 1.0928983688354492, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 36549 + }, + { + "epoch": 0.3655, + "grad_norm": 1.2146763801574707, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 36550 + }, + { + "epoch": 0.36551, + "grad_norm": 0.9368833899497986, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 36551 + }, + { + "epoch": 0.36552, + "grad_norm": 0.8545805811882019, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 36552 + }, + { + "epoch": 0.36553, + "grad_norm": 0.8148287534713745, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 36553 + }, + { + "epoch": 0.36554, + "grad_norm": 0.7561874985694885, + "learning_rate": 0.003, + "loss": 4.0, + "step": 36554 + }, + { + "epoch": 0.36555, + "grad_norm": 0.8098024725914001, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 36555 + }, + { + "epoch": 0.36556, + "grad_norm": 0.8967704772949219, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 36556 + }, + { + "epoch": 0.36557, + "grad_norm": 1.0049095153808594, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 36557 + }, + { + "epoch": 0.36558, + "grad_norm": 1.0851186513900757, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 36558 + }, + { + "epoch": 0.36559, + "grad_norm": 0.9673658609390259, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 36559 + }, + { + "epoch": 0.3656, + "grad_norm": 1.2417864799499512, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 36560 + }, + { + "epoch": 0.36561, + "grad_norm": 0.9619262218475342, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 36561 + }, + { + "epoch": 0.36562, + "grad_norm": 0.7473247051239014, + "learning_rate": 0.003, + "loss": 3.986, + "step": 36562 + }, + { + "epoch": 0.36563, + "grad_norm": 0.7230894565582275, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 36563 + }, + { + "epoch": 0.36564, + "grad_norm": 0.674808144569397, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 36564 + }, + { + "epoch": 0.36565, + "grad_norm": 0.7427915930747986, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 36565 + }, + { + "epoch": 0.36566, + "grad_norm": 0.9411801695823669, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 36566 + }, + { + "epoch": 0.36567, + "grad_norm": 1.081712007522583, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 36567 + }, + { + "epoch": 0.36568, + "grad_norm": 1.1009010076522827, + "learning_rate": 0.003, + "loss": 3.995, + "step": 36568 + }, + { + "epoch": 0.36569, + "grad_norm": 0.9842122197151184, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 36569 + }, + { + "epoch": 0.3657, + "grad_norm": 0.9344068169593811, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 36570 + }, + { + "epoch": 0.36571, + "grad_norm": 0.9442040920257568, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 36571 + }, + { + "epoch": 0.36572, + "grad_norm": 0.9971546530723572, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 36572 + }, + { + "epoch": 0.36573, + "grad_norm": 1.2698410749435425, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 36573 + }, + { + "epoch": 0.36574, + "grad_norm": 0.8494325280189514, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 36574 + }, + { + "epoch": 0.36575, + "grad_norm": 0.6607924103736877, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 36575 + }, + { + "epoch": 0.36576, + "grad_norm": 0.7316749691963196, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 36576 + }, + { + "epoch": 0.36577, + "grad_norm": 0.8060632944107056, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 36577 + }, + { + "epoch": 0.36578, + "grad_norm": 0.8364569544792175, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 36578 + }, + { + "epoch": 0.36579, + "grad_norm": 0.8915988206863403, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 36579 + }, + { + "epoch": 0.3658, + "grad_norm": 1.017247200012207, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 36580 + }, + { + "epoch": 0.36581, + "grad_norm": 1.0705325603485107, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 36581 + }, + { + "epoch": 0.36582, + "grad_norm": 0.9992272853851318, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 36582 + }, + { + "epoch": 0.36583, + "grad_norm": 1.028091311454773, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 36583 + }, + { + "epoch": 0.36584, + "grad_norm": 1.0560439825057983, + "learning_rate": 0.003, + "loss": 3.98, + "step": 36584 + }, + { + "epoch": 0.36585, + "grad_norm": 1.013310432434082, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 36585 + }, + { + "epoch": 0.36586, + "grad_norm": 0.9821938872337341, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 36586 + }, + { + "epoch": 0.36587, + "grad_norm": 0.9407991170883179, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 36587 + }, + { + "epoch": 0.36588, + "grad_norm": 0.9067440032958984, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 36588 + }, + { + "epoch": 0.36589, + "grad_norm": 0.7784480452537537, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 36589 + }, + { + "epoch": 0.3659, + "grad_norm": 0.7666623592376709, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 36590 + }, + { + "epoch": 0.36591, + "grad_norm": 0.8642516136169434, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 36591 + }, + { + "epoch": 0.36592, + "grad_norm": 1.0034244060516357, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 36592 + }, + { + "epoch": 0.36593, + "grad_norm": 1.230804681777954, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 36593 + }, + { + "epoch": 0.36594, + "grad_norm": 0.9748013019561768, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 36594 + }, + { + "epoch": 0.36595, + "grad_norm": 1.2551188468933105, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 36595 + }, + { + "epoch": 0.36596, + "grad_norm": 0.8009209036827087, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 36596 + }, + { + "epoch": 0.36597, + "grad_norm": 0.6337122917175293, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 36597 + }, + { + "epoch": 0.36598, + "grad_norm": 0.5818371772766113, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 36598 + }, + { + "epoch": 0.36599, + "grad_norm": 0.6790701746940613, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 36599 + }, + { + "epoch": 0.366, + "grad_norm": 0.7954164147377014, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 36600 + }, + { + "epoch": 0.36601, + "grad_norm": 0.8530604839324951, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 36601 + }, + { + "epoch": 0.36602, + "grad_norm": 1.1437451839447021, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 36602 + }, + { + "epoch": 0.36603, + "grad_norm": 1.260772943496704, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 36603 + }, + { + "epoch": 0.36604, + "grad_norm": 0.7268264293670654, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 36604 + }, + { + "epoch": 0.36605, + "grad_norm": 0.7584230899810791, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 36605 + }, + { + "epoch": 0.36606, + "grad_norm": 0.8291401267051697, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 36606 + }, + { + "epoch": 0.36607, + "grad_norm": 0.8860329389572144, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 36607 + }, + { + "epoch": 0.36608, + "grad_norm": 0.9720618724822998, + "learning_rate": 0.003, + "loss": 3.992, + "step": 36608 + }, + { + "epoch": 0.36609, + "grad_norm": 1.0051233768463135, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 36609 + }, + { + "epoch": 0.3661, + "grad_norm": 0.7121849656105042, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 36610 + }, + { + "epoch": 0.36611, + "grad_norm": 0.5211007595062256, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 36611 + }, + { + "epoch": 0.36612, + "grad_norm": 0.5986360311508179, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 36612 + }, + { + "epoch": 0.36613, + "grad_norm": 0.5702938437461853, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 36613 + }, + { + "epoch": 0.36614, + "grad_norm": 0.48001137375831604, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 36614 + }, + { + "epoch": 0.36615, + "grad_norm": 0.5320805907249451, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 36615 + }, + { + "epoch": 0.36616, + "grad_norm": 0.5996074676513672, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 36616 + }, + { + "epoch": 0.36617, + "grad_norm": 0.6742087006568909, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 36617 + }, + { + "epoch": 0.36618, + "grad_norm": 0.8260096907615662, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 36618 + }, + { + "epoch": 0.36619, + "grad_norm": 0.8091428279876709, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 36619 + }, + { + "epoch": 0.3662, + "grad_norm": 0.807845413684845, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 36620 + }, + { + "epoch": 0.36621, + "grad_norm": 0.9221017956733704, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 36621 + }, + { + "epoch": 0.36622, + "grad_norm": 1.4189685583114624, + "learning_rate": 0.003, + "loss": 3.985, + "step": 36622 + }, + { + "epoch": 0.36623, + "grad_norm": 0.7459923028945923, + "learning_rate": 0.003, + "loss": 3.965, + "step": 36623 + }, + { + "epoch": 0.36624, + "grad_norm": 0.7090433239936829, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 36624 + }, + { + "epoch": 0.36625, + "grad_norm": 0.8111494779586792, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 36625 + }, + { + "epoch": 0.36626, + "grad_norm": 0.9971103668212891, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 36626 + }, + { + "epoch": 0.36627, + "grad_norm": 1.126158356666565, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 36627 + }, + { + "epoch": 0.36628, + "grad_norm": 0.889037549495697, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 36628 + }, + { + "epoch": 0.36629, + "grad_norm": 0.9359325766563416, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 36629 + }, + { + "epoch": 0.3663, + "grad_norm": 1.029882788658142, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 36630 + }, + { + "epoch": 0.36631, + "grad_norm": 1.0854626893997192, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 36631 + }, + { + "epoch": 0.36632, + "grad_norm": 1.0018056631088257, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 36632 + }, + { + "epoch": 0.36633, + "grad_norm": 1.296768307685852, + "learning_rate": 0.003, + "loss": 3.972, + "step": 36633 + }, + { + "epoch": 0.36634, + "grad_norm": 0.8651896715164185, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 36634 + }, + { + "epoch": 0.36635, + "grad_norm": 0.8053105473518372, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 36635 + }, + { + "epoch": 0.36636, + "grad_norm": 0.8931826949119568, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 36636 + }, + { + "epoch": 0.36637, + "grad_norm": 1.2151808738708496, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 36637 + }, + { + "epoch": 0.36638, + "grad_norm": 0.8973534107208252, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 36638 + }, + { + "epoch": 0.36639, + "grad_norm": 0.7042010426521301, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 36639 + }, + { + "epoch": 0.3664, + "grad_norm": 0.6336286067962646, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 36640 + }, + { + "epoch": 0.36641, + "grad_norm": 0.5291603207588196, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 36641 + }, + { + "epoch": 0.36642, + "grad_norm": 0.6818662285804749, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 36642 + }, + { + "epoch": 0.36643, + "grad_norm": 0.9502124190330505, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 36643 + }, + { + "epoch": 0.36644, + "grad_norm": 1.1563657522201538, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 36644 + }, + { + "epoch": 0.36645, + "grad_norm": 0.795744776725769, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 36645 + }, + { + "epoch": 0.36646, + "grad_norm": 0.5728160738945007, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 36646 + }, + { + "epoch": 0.36647, + "grad_norm": 0.5353488922119141, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 36647 + }, + { + "epoch": 0.36648, + "grad_norm": 0.6028926968574524, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 36648 + }, + { + "epoch": 0.36649, + "grad_norm": 0.7161435484886169, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 36649 + }, + { + "epoch": 0.3665, + "grad_norm": 0.9114841222763062, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 36650 + }, + { + "epoch": 0.36651, + "grad_norm": 1.2197816371917725, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 36651 + }, + { + "epoch": 0.36652, + "grad_norm": 0.9649457931518555, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 36652 + }, + { + "epoch": 0.36653, + "grad_norm": 1.008151888847351, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 36653 + }, + { + "epoch": 0.36654, + "grad_norm": 0.9827769994735718, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 36654 + }, + { + "epoch": 0.36655, + "grad_norm": 0.9621031880378723, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 36655 + }, + { + "epoch": 0.36656, + "grad_norm": 1.034130334854126, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 36656 + }, + { + "epoch": 0.36657, + "grad_norm": 1.207200527191162, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 36657 + }, + { + "epoch": 0.36658, + "grad_norm": 0.9901199340820312, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 36658 + }, + { + "epoch": 0.36659, + "grad_norm": 1.024147868156433, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 36659 + }, + { + "epoch": 0.3666, + "grad_norm": 1.138702630996704, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 36660 + }, + { + "epoch": 0.36661, + "grad_norm": 0.9660364389419556, + "learning_rate": 0.003, + "loss": 3.974, + "step": 36661 + }, + { + "epoch": 0.36662, + "grad_norm": 0.8212222456932068, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 36662 + }, + { + "epoch": 0.36663, + "grad_norm": 0.7811955809593201, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 36663 + }, + { + "epoch": 0.36664, + "grad_norm": 0.8021105527877808, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 36664 + }, + { + "epoch": 0.36665, + "grad_norm": 0.9525015950202942, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 36665 + }, + { + "epoch": 0.36666, + "grad_norm": 1.1665397882461548, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 36666 + }, + { + "epoch": 0.36667, + "grad_norm": 0.84259033203125, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 36667 + }, + { + "epoch": 0.36668, + "grad_norm": 0.6610124111175537, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 36668 + }, + { + "epoch": 0.36669, + "grad_norm": 0.6017925143241882, + "learning_rate": 0.003, + "loss": 3.999, + "step": 36669 + }, + { + "epoch": 0.3667, + "grad_norm": 0.7434792518615723, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 36670 + }, + { + "epoch": 0.36671, + "grad_norm": 0.6776030659675598, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 36671 + }, + { + "epoch": 0.36672, + "grad_norm": 0.6676545143127441, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 36672 + }, + { + "epoch": 0.36673, + "grad_norm": 0.6864649057388306, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 36673 + }, + { + "epoch": 0.36674, + "grad_norm": 0.8249058127403259, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 36674 + }, + { + "epoch": 0.36675, + "grad_norm": 0.856673002243042, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 36675 + }, + { + "epoch": 0.36676, + "grad_norm": 0.952380359172821, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 36676 + }, + { + "epoch": 0.36677, + "grad_norm": 1.1301895380020142, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 36677 + }, + { + "epoch": 0.36678, + "grad_norm": 0.9533231854438782, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 36678 + }, + { + "epoch": 0.36679, + "grad_norm": 0.9195442199707031, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 36679 + }, + { + "epoch": 0.3668, + "grad_norm": 1.0273544788360596, + "learning_rate": 0.003, + "loss": 3.981, + "step": 36680 + }, + { + "epoch": 0.36681, + "grad_norm": 1.6104003190994263, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 36681 + }, + { + "epoch": 0.36682, + "grad_norm": 0.6850157380104065, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 36682 + }, + { + "epoch": 0.36683, + "grad_norm": 0.8503463268280029, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 36683 + }, + { + "epoch": 0.36684, + "grad_norm": 1.2129074335098267, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 36684 + }, + { + "epoch": 0.36685, + "grad_norm": 0.9360294938087463, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 36685 + }, + { + "epoch": 0.36686, + "grad_norm": 1.1029667854309082, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 36686 + }, + { + "epoch": 0.36687, + "grad_norm": 0.841696560382843, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 36687 + }, + { + "epoch": 0.36688, + "grad_norm": 0.7209200859069824, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 36688 + }, + { + "epoch": 0.36689, + "grad_norm": 0.5707867741584778, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 36689 + }, + { + "epoch": 0.3669, + "grad_norm": 0.6804700493812561, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 36690 + }, + { + "epoch": 0.36691, + "grad_norm": 0.7309439182281494, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 36691 + }, + { + "epoch": 0.36692, + "grad_norm": 0.8265941143035889, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 36692 + }, + { + "epoch": 0.36693, + "grad_norm": 0.9699952006340027, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 36693 + }, + { + "epoch": 0.36694, + "grad_norm": 0.9875581860542297, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 36694 + }, + { + "epoch": 0.36695, + "grad_norm": 0.9985460042953491, + "learning_rate": 0.003, + "loss": 3.998, + "step": 36695 + }, + { + "epoch": 0.36696, + "grad_norm": 0.9798828959465027, + "learning_rate": 0.003, + "loss": 4.012, + "step": 36696 + }, + { + "epoch": 0.36697, + "grad_norm": 1.0599421262741089, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 36697 + }, + { + "epoch": 0.36698, + "grad_norm": 1.2262952327728271, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 36698 + }, + { + "epoch": 0.36699, + "grad_norm": 0.9201129078865051, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 36699 + }, + { + "epoch": 0.367, + "grad_norm": 0.8620260953903198, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 36700 + }, + { + "epoch": 0.36701, + "grad_norm": 1.1183762550354004, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 36701 + }, + { + "epoch": 0.36702, + "grad_norm": 1.0520997047424316, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 36702 + }, + { + "epoch": 0.36703, + "grad_norm": 0.85958331823349, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 36703 + }, + { + "epoch": 0.36704, + "grad_norm": 0.8415271043777466, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 36704 + }, + { + "epoch": 0.36705, + "grad_norm": 0.9263831377029419, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 36705 + }, + { + "epoch": 0.36706, + "grad_norm": 0.9275064468383789, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 36706 + }, + { + "epoch": 0.36707, + "grad_norm": 0.8760188221931458, + "learning_rate": 0.003, + "loss": 3.997, + "step": 36707 + }, + { + "epoch": 0.36708, + "grad_norm": 0.8263141512870789, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 36708 + }, + { + "epoch": 0.36709, + "grad_norm": 0.8781777024269104, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 36709 + }, + { + "epoch": 0.3671, + "grad_norm": 0.9988304972648621, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 36710 + }, + { + "epoch": 0.36711, + "grad_norm": 1.141880750656128, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 36711 + }, + { + "epoch": 0.36712, + "grad_norm": 0.8005127906799316, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 36712 + }, + { + "epoch": 0.36713, + "grad_norm": 0.8394144773483276, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 36713 + }, + { + "epoch": 0.36714, + "grad_norm": 0.9206026792526245, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 36714 + }, + { + "epoch": 0.36715, + "grad_norm": 1.1535261869430542, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 36715 + }, + { + "epoch": 0.36716, + "grad_norm": 1.1306525468826294, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 36716 + }, + { + "epoch": 0.36717, + "grad_norm": 0.934327244758606, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 36717 + }, + { + "epoch": 0.36718, + "grad_norm": 0.9303128719329834, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 36718 + }, + { + "epoch": 0.36719, + "grad_norm": 0.8807304501533508, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 36719 + }, + { + "epoch": 0.3672, + "grad_norm": 0.8612138628959656, + "learning_rate": 0.003, + "loss": 3.994, + "step": 36720 + }, + { + "epoch": 0.36721, + "grad_norm": 0.8324612975120544, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 36721 + }, + { + "epoch": 0.36722, + "grad_norm": 0.7133486866950989, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 36722 + }, + { + "epoch": 0.36723, + "grad_norm": 0.6202331185340881, + "learning_rate": 0.003, + "loss": 3.953, + "step": 36723 + }, + { + "epoch": 0.36724, + "grad_norm": 0.726762056350708, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 36724 + }, + { + "epoch": 0.36725, + "grad_norm": 0.8485122919082642, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 36725 + }, + { + "epoch": 0.36726, + "grad_norm": 0.9574657678604126, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 36726 + }, + { + "epoch": 0.36727, + "grad_norm": 0.885899543762207, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 36727 + }, + { + "epoch": 0.36728, + "grad_norm": 0.9483583569526672, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 36728 + }, + { + "epoch": 0.36729, + "grad_norm": 0.8867186307907104, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 36729 + }, + { + "epoch": 0.3673, + "grad_norm": 0.8679955005645752, + "learning_rate": 0.003, + "loss": 4.017, + "step": 36730 + }, + { + "epoch": 0.36731, + "grad_norm": 1.1410762071609497, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 36731 + }, + { + "epoch": 0.36732, + "grad_norm": 0.9260802865028381, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 36732 + }, + { + "epoch": 0.36733, + "grad_norm": 0.9783821105957031, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 36733 + }, + { + "epoch": 0.36734, + "grad_norm": 1.5047847032546997, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 36734 + }, + { + "epoch": 0.36735, + "grad_norm": 0.6685523390769958, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 36735 + }, + { + "epoch": 0.36736, + "grad_norm": 0.8818989396095276, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 36736 + }, + { + "epoch": 0.36737, + "grad_norm": 1.4582626819610596, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 36737 + }, + { + "epoch": 0.36738, + "grad_norm": 0.5236583948135376, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 36738 + }, + { + "epoch": 0.36739, + "grad_norm": 0.8235771059989929, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 36739 + }, + { + "epoch": 0.3674, + "grad_norm": 0.9264410138130188, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 36740 + }, + { + "epoch": 0.36741, + "grad_norm": 1.0418505668640137, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 36741 + }, + { + "epoch": 0.36742, + "grad_norm": 1.0127958059310913, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 36742 + }, + { + "epoch": 0.36743, + "grad_norm": 0.8217326402664185, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 36743 + }, + { + "epoch": 0.36744, + "grad_norm": 0.6953132152557373, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 36744 + }, + { + "epoch": 0.36745, + "grad_norm": 0.7004735469818115, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 36745 + }, + { + "epoch": 0.36746, + "grad_norm": 0.7192983031272888, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 36746 + }, + { + "epoch": 0.36747, + "grad_norm": 0.7549105286598206, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 36747 + }, + { + "epoch": 0.36748, + "grad_norm": 0.7867814898490906, + "learning_rate": 0.003, + "loss": 3.9303, + "step": 36748 + }, + { + "epoch": 0.36749, + "grad_norm": 0.8269910216331482, + "learning_rate": 0.003, + "loss": 3.982, + "step": 36749 + }, + { + "epoch": 0.3675, + "grad_norm": 0.7609298825263977, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 36750 + }, + { + "epoch": 0.36751, + "grad_norm": 0.8315577507019043, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 36751 + }, + { + "epoch": 0.36752, + "grad_norm": 0.8879876732826233, + "learning_rate": 0.003, + "loss": 3.998, + "step": 36752 + }, + { + "epoch": 0.36753, + "grad_norm": 0.6984572410583496, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 36753 + }, + { + "epoch": 0.36754, + "grad_norm": 0.6816555857658386, + "learning_rate": 0.003, + "loss": 4.003, + "step": 36754 + }, + { + "epoch": 0.36755, + "grad_norm": 0.6841795444488525, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 36755 + }, + { + "epoch": 0.36756, + "grad_norm": 0.8977696895599365, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 36756 + }, + { + "epoch": 0.36757, + "grad_norm": 1.086033582687378, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 36757 + }, + { + "epoch": 0.36758, + "grad_norm": 1.185468077659607, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 36758 + }, + { + "epoch": 0.36759, + "grad_norm": 0.7561252117156982, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 36759 + }, + { + "epoch": 0.3676, + "grad_norm": 0.6758939623832703, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 36760 + }, + { + "epoch": 0.36761, + "grad_norm": 0.7163647413253784, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 36761 + }, + { + "epoch": 0.36762, + "grad_norm": 0.8967345952987671, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 36762 + }, + { + "epoch": 0.36763, + "grad_norm": 0.8960419297218323, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 36763 + }, + { + "epoch": 0.36764, + "grad_norm": 0.8954609632492065, + "learning_rate": 0.003, + "loss": 3.956, + "step": 36764 + }, + { + "epoch": 0.36765, + "grad_norm": 1.0486537218093872, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 36765 + }, + { + "epoch": 0.36766, + "grad_norm": 1.1532073020935059, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 36766 + }, + { + "epoch": 0.36767, + "grad_norm": 0.8481074571609497, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 36767 + }, + { + "epoch": 0.36768, + "grad_norm": 0.8019057512283325, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 36768 + }, + { + "epoch": 0.36769, + "grad_norm": 0.8877273201942444, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 36769 + }, + { + "epoch": 0.3677, + "grad_norm": 0.8760189414024353, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 36770 + }, + { + "epoch": 0.36771, + "grad_norm": 0.9683693051338196, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 36771 + }, + { + "epoch": 0.36772, + "grad_norm": 1.3025563955307007, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 36772 + }, + { + "epoch": 0.36773, + "grad_norm": 0.7881420850753784, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 36773 + }, + { + "epoch": 0.36774, + "grad_norm": 0.8236131072044373, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 36774 + }, + { + "epoch": 0.36775, + "grad_norm": 1.0459176301956177, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 36775 + }, + { + "epoch": 0.36776, + "grad_norm": 1.1279562711715698, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 36776 + }, + { + "epoch": 0.36777, + "grad_norm": 0.6839395761489868, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 36777 + }, + { + "epoch": 0.36778, + "grad_norm": 0.6421745419502258, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 36778 + }, + { + "epoch": 0.36779, + "grad_norm": 0.7653096914291382, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 36779 + }, + { + "epoch": 0.3678, + "grad_norm": 1.0020655393600464, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 36780 + }, + { + "epoch": 0.36781, + "grad_norm": 1.1273399591445923, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 36781 + }, + { + "epoch": 0.36782, + "grad_norm": 1.0574703216552734, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 36782 + }, + { + "epoch": 0.36783, + "grad_norm": 0.912652313709259, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 36783 + }, + { + "epoch": 0.36784, + "grad_norm": 0.9142255187034607, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 36784 + }, + { + "epoch": 0.36785, + "grad_norm": 0.8665688633918762, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 36785 + }, + { + "epoch": 0.36786, + "grad_norm": 0.8954645991325378, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 36786 + }, + { + "epoch": 0.36787, + "grad_norm": 0.9474170804023743, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 36787 + }, + { + "epoch": 0.36788, + "grad_norm": 1.152955412864685, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 36788 + }, + { + "epoch": 0.36789, + "grad_norm": 0.9868454933166504, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 36789 + }, + { + "epoch": 0.3679, + "grad_norm": 1.042512059211731, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 36790 + }, + { + "epoch": 0.36791, + "grad_norm": 1.1932748556137085, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 36791 + }, + { + "epoch": 0.36792, + "grad_norm": 1.1313639879226685, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 36792 + }, + { + "epoch": 0.36793, + "grad_norm": 0.9373892545700073, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 36793 + }, + { + "epoch": 0.36794, + "grad_norm": 0.8538760542869568, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 36794 + }, + { + "epoch": 0.36795, + "grad_norm": 1.0636487007141113, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 36795 + }, + { + "epoch": 0.36796, + "grad_norm": 1.2534620761871338, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 36796 + }, + { + "epoch": 0.36797, + "grad_norm": 0.6324447989463806, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 36797 + }, + { + "epoch": 0.36798, + "grad_norm": 0.6577920913696289, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 36798 + }, + { + "epoch": 0.36799, + "grad_norm": 0.8200725317001343, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 36799 + }, + { + "epoch": 0.368, + "grad_norm": 0.9487665295600891, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 36800 + }, + { + "epoch": 0.36801, + "grad_norm": 1.1013203859329224, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 36801 + }, + { + "epoch": 0.36802, + "grad_norm": 0.913017213344574, + "learning_rate": 0.003, + "loss": 3.972, + "step": 36802 + }, + { + "epoch": 0.36803, + "grad_norm": 0.7866990566253662, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 36803 + }, + { + "epoch": 0.36804, + "grad_norm": 0.7406496405601501, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 36804 + }, + { + "epoch": 0.36805, + "grad_norm": 0.7884745001792908, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 36805 + }, + { + "epoch": 0.36806, + "grad_norm": 0.8133347630500793, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 36806 + }, + { + "epoch": 0.36807, + "grad_norm": 0.9206989407539368, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 36807 + }, + { + "epoch": 0.36808, + "grad_norm": 1.3070764541625977, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 36808 + }, + { + "epoch": 0.36809, + "grad_norm": 0.6838752031326294, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 36809 + }, + { + "epoch": 0.3681, + "grad_norm": 0.6213175654411316, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 36810 + }, + { + "epoch": 0.36811, + "grad_norm": 0.8167229890823364, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 36811 + }, + { + "epoch": 0.36812, + "grad_norm": 0.8632939457893372, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 36812 + }, + { + "epoch": 0.36813, + "grad_norm": 1.0131876468658447, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 36813 + }, + { + "epoch": 0.36814, + "grad_norm": 1.0957767963409424, + "learning_rate": 0.003, + "loss": 3.9283, + "step": 36814 + }, + { + "epoch": 0.36815, + "grad_norm": 0.9240907430648804, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 36815 + }, + { + "epoch": 0.36816, + "grad_norm": 0.8226408958435059, + "learning_rate": 0.003, + "loss": 3.997, + "step": 36816 + }, + { + "epoch": 0.36817, + "grad_norm": 0.8094374537467957, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 36817 + }, + { + "epoch": 0.36818, + "grad_norm": 0.8810203075408936, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 36818 + }, + { + "epoch": 0.36819, + "grad_norm": 0.9073318839073181, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 36819 + }, + { + "epoch": 0.3682, + "grad_norm": 0.8995482921600342, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 36820 + }, + { + "epoch": 0.36821, + "grad_norm": 0.8279207944869995, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 36821 + }, + { + "epoch": 0.36822, + "grad_norm": 0.784768283367157, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 36822 + }, + { + "epoch": 0.36823, + "grad_norm": 0.9786350131034851, + "learning_rate": 0.003, + "loss": 3.989, + "step": 36823 + }, + { + "epoch": 0.36824, + "grad_norm": 1.0801595449447632, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 36824 + }, + { + "epoch": 0.36825, + "grad_norm": 1.2236857414245605, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 36825 + }, + { + "epoch": 0.36826, + "grad_norm": 0.7670919895172119, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 36826 + }, + { + "epoch": 0.36827, + "grad_norm": 0.6768054366111755, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 36827 + }, + { + "epoch": 0.36828, + "grad_norm": 0.828389048576355, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 36828 + }, + { + "epoch": 0.36829, + "grad_norm": 0.954582929611206, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 36829 + }, + { + "epoch": 0.3683, + "grad_norm": 1.0765739679336548, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 36830 + }, + { + "epoch": 0.36831, + "grad_norm": 1.009067177772522, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 36831 + }, + { + "epoch": 0.36832, + "grad_norm": 0.8807523250579834, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 36832 + }, + { + "epoch": 0.36833, + "grad_norm": 0.9006160497665405, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 36833 + }, + { + "epoch": 0.36834, + "grad_norm": 1.2342474460601807, + "learning_rate": 0.003, + "loss": 4.003, + "step": 36834 + }, + { + "epoch": 0.36835, + "grad_norm": 0.9846972227096558, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 36835 + }, + { + "epoch": 0.36836, + "grad_norm": 0.843177080154419, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 36836 + }, + { + "epoch": 0.36837, + "grad_norm": 0.8281829953193665, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 36837 + }, + { + "epoch": 0.36838, + "grad_norm": 0.8627418279647827, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 36838 + }, + { + "epoch": 0.36839, + "grad_norm": 0.8281732201576233, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 36839 + }, + { + "epoch": 0.3684, + "grad_norm": 1.0788183212280273, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 36840 + }, + { + "epoch": 0.36841, + "grad_norm": 1.1931825876235962, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 36841 + }, + { + "epoch": 0.36842, + "grad_norm": 0.778333306312561, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 36842 + }, + { + "epoch": 0.36843, + "grad_norm": 0.6462261080741882, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 36843 + }, + { + "epoch": 0.36844, + "grad_norm": 0.6752718091011047, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 36844 + }, + { + "epoch": 0.36845, + "grad_norm": 0.746918797492981, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 36845 + }, + { + "epoch": 0.36846, + "grad_norm": 0.8183074593544006, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 36846 + }, + { + "epoch": 0.36847, + "grad_norm": 0.8802846670150757, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 36847 + }, + { + "epoch": 0.36848, + "grad_norm": 0.7715796232223511, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 36848 + }, + { + "epoch": 0.36849, + "grad_norm": 0.6885365843772888, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 36849 + }, + { + "epoch": 0.3685, + "grad_norm": 0.728550136089325, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 36850 + }, + { + "epoch": 0.36851, + "grad_norm": 0.8406098484992981, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 36851 + }, + { + "epoch": 0.36852, + "grad_norm": 1.070440411567688, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 36852 + }, + { + "epoch": 0.36853, + "grad_norm": 1.0858591794967651, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 36853 + }, + { + "epoch": 0.36854, + "grad_norm": 1.094638705253601, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 36854 + }, + { + "epoch": 0.36855, + "grad_norm": 1.0600056648254395, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 36855 + }, + { + "epoch": 0.36856, + "grad_norm": 0.8091448545455933, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 36856 + }, + { + "epoch": 0.36857, + "grad_norm": 0.7704145908355713, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 36857 + }, + { + "epoch": 0.36858, + "grad_norm": 0.7746199369430542, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 36858 + }, + { + "epoch": 0.36859, + "grad_norm": 0.8046976923942566, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 36859 + }, + { + "epoch": 0.3686, + "grad_norm": 0.7696680426597595, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 36860 + }, + { + "epoch": 0.36861, + "grad_norm": 0.9175959825515747, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 36861 + }, + { + "epoch": 0.36862, + "grad_norm": 1.0066388845443726, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 36862 + }, + { + "epoch": 0.36863, + "grad_norm": 1.0799908638000488, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 36863 + }, + { + "epoch": 0.36864, + "grad_norm": 0.8758724927902222, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 36864 + }, + { + "epoch": 0.36865, + "grad_norm": 0.9778779149055481, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 36865 + }, + { + "epoch": 0.36866, + "grad_norm": 1.077376365661621, + "learning_rate": 0.003, + "loss": 4.004, + "step": 36866 + }, + { + "epoch": 0.36867, + "grad_norm": 1.0480772256851196, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 36867 + }, + { + "epoch": 0.36868, + "grad_norm": 1.1466807126998901, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 36868 + }, + { + "epoch": 0.36869, + "grad_norm": 0.8408313989639282, + "learning_rate": 0.003, + "loss": 4.0, + "step": 36869 + }, + { + "epoch": 0.3687, + "grad_norm": 0.7419973015785217, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 36870 + }, + { + "epoch": 0.36871, + "grad_norm": 0.8429141044616699, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 36871 + }, + { + "epoch": 0.36872, + "grad_norm": 1.1605995893478394, + "learning_rate": 0.003, + "loss": 3.938, + "step": 36872 + }, + { + "epoch": 0.36873, + "grad_norm": 1.0317045450210571, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 36873 + }, + { + "epoch": 0.36874, + "grad_norm": 0.9718703627586365, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 36874 + }, + { + "epoch": 0.36875, + "grad_norm": 1.1055803298950195, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 36875 + }, + { + "epoch": 0.36876, + "grad_norm": 1.0637595653533936, + "learning_rate": 0.003, + "loss": 3.984, + "step": 36876 + }, + { + "epoch": 0.36877, + "grad_norm": 0.9594369530677795, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 36877 + }, + { + "epoch": 0.36878, + "grad_norm": 0.8528957962989807, + "learning_rate": 0.003, + "loss": 3.968, + "step": 36878 + }, + { + "epoch": 0.36879, + "grad_norm": 0.8614755272865295, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 36879 + }, + { + "epoch": 0.3688, + "grad_norm": 0.7349494695663452, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 36880 + }, + { + "epoch": 0.36881, + "grad_norm": 0.7932853698730469, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 36881 + }, + { + "epoch": 0.36882, + "grad_norm": 0.9866887927055359, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 36882 + }, + { + "epoch": 0.36883, + "grad_norm": 1.2814221382141113, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 36883 + }, + { + "epoch": 0.36884, + "grad_norm": 0.8188600540161133, + "learning_rate": 0.003, + "loss": 4.003, + "step": 36884 + }, + { + "epoch": 0.36885, + "grad_norm": 0.846968948841095, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 36885 + }, + { + "epoch": 0.36886, + "grad_norm": 1.0167385339736938, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 36886 + }, + { + "epoch": 0.36887, + "grad_norm": 1.0470051765441895, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 36887 + }, + { + "epoch": 0.36888, + "grad_norm": 1.1046339273452759, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 36888 + }, + { + "epoch": 0.36889, + "grad_norm": 0.8202678561210632, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 36889 + }, + { + "epoch": 0.3689, + "grad_norm": 0.603931725025177, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 36890 + }, + { + "epoch": 0.36891, + "grad_norm": 0.7001110911369324, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 36891 + }, + { + "epoch": 0.36892, + "grad_norm": 0.7952529788017273, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 36892 + }, + { + "epoch": 0.36893, + "grad_norm": 0.9307608008384705, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 36893 + }, + { + "epoch": 0.36894, + "grad_norm": 0.989386260509491, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 36894 + }, + { + "epoch": 0.36895, + "grad_norm": 0.9734946489334106, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 36895 + }, + { + "epoch": 0.36896, + "grad_norm": 0.9069214463233948, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 36896 + }, + { + "epoch": 0.36897, + "grad_norm": 0.857344388961792, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 36897 + }, + { + "epoch": 0.36898, + "grad_norm": 0.6876819133758545, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 36898 + }, + { + "epoch": 0.36899, + "grad_norm": 0.5803753137588501, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 36899 + }, + { + "epoch": 0.369, + "grad_norm": 0.561255693435669, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 36900 + }, + { + "epoch": 0.36901, + "grad_norm": 0.5453017950057983, + "learning_rate": 0.003, + "loss": 3.985, + "step": 36901 + }, + { + "epoch": 0.36902, + "grad_norm": 0.6560730934143066, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 36902 + }, + { + "epoch": 0.36903, + "grad_norm": 1.1876846551895142, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 36903 + }, + { + "epoch": 0.36904, + "grad_norm": 1.0278468132019043, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 36904 + }, + { + "epoch": 0.36905, + "grad_norm": 0.8908202052116394, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 36905 + }, + { + "epoch": 0.36906, + "grad_norm": 1.1046534776687622, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 36906 + }, + { + "epoch": 0.36907, + "grad_norm": 1.042088270187378, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 36907 + }, + { + "epoch": 0.36908, + "grad_norm": 0.8452457785606384, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 36908 + }, + { + "epoch": 0.36909, + "grad_norm": 0.7279841899871826, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 36909 + }, + { + "epoch": 0.3691, + "grad_norm": 0.7153027653694153, + "learning_rate": 0.003, + "loss": 3.946, + "step": 36910 + }, + { + "epoch": 0.36911, + "grad_norm": 0.8676890730857849, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 36911 + }, + { + "epoch": 0.36912, + "grad_norm": 0.9789462685585022, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 36912 + }, + { + "epoch": 0.36913, + "grad_norm": 1.073143720626831, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 36913 + }, + { + "epoch": 0.36914, + "grad_norm": 0.835477888584137, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 36914 + }, + { + "epoch": 0.36915, + "grad_norm": 0.8543207049369812, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 36915 + }, + { + "epoch": 0.36916, + "grad_norm": 0.9910640716552734, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 36916 + }, + { + "epoch": 0.36917, + "grad_norm": 0.967138409614563, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 36917 + }, + { + "epoch": 0.36918, + "grad_norm": 1.0604959726333618, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 36918 + }, + { + "epoch": 0.36919, + "grad_norm": 0.8414232134819031, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 36919 + }, + { + "epoch": 0.3692, + "grad_norm": 0.7143886089324951, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 36920 + }, + { + "epoch": 0.36921, + "grad_norm": 0.7949140071868896, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 36921 + }, + { + "epoch": 0.36922, + "grad_norm": 0.8351624608039856, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 36922 + }, + { + "epoch": 0.36923, + "grad_norm": 0.9090775847434998, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 36923 + }, + { + "epoch": 0.36924, + "grad_norm": 1.1173261404037476, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 36924 + }, + { + "epoch": 0.36925, + "grad_norm": 1.1520864963531494, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 36925 + }, + { + "epoch": 0.36926, + "grad_norm": 0.9910605549812317, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 36926 + }, + { + "epoch": 0.36927, + "grad_norm": 1.1326085329055786, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 36927 + }, + { + "epoch": 0.36928, + "grad_norm": 1.0646898746490479, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 36928 + }, + { + "epoch": 0.36929, + "grad_norm": 0.8442399501800537, + "learning_rate": 0.003, + "loss": 4.004, + "step": 36929 + }, + { + "epoch": 0.3693, + "grad_norm": 0.7857387065887451, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 36930 + }, + { + "epoch": 0.36931, + "grad_norm": 0.8487773537635803, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 36931 + }, + { + "epoch": 0.36932, + "grad_norm": 0.7830185294151306, + "learning_rate": 0.003, + "loss": 3.987, + "step": 36932 + }, + { + "epoch": 0.36933, + "grad_norm": 0.924168050289154, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 36933 + }, + { + "epoch": 0.36934, + "grad_norm": 1.0767024755477905, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 36934 + }, + { + "epoch": 0.36935, + "grad_norm": 1.0253673791885376, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 36935 + }, + { + "epoch": 0.36936, + "grad_norm": 1.199730634689331, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 36936 + }, + { + "epoch": 0.36937, + "grad_norm": 0.8364176154136658, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 36937 + }, + { + "epoch": 0.36938, + "grad_norm": 0.998901903629303, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 36938 + }, + { + "epoch": 0.36939, + "grad_norm": 1.3598605394363403, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 36939 + }, + { + "epoch": 0.3694, + "grad_norm": 0.6351978778839111, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 36940 + }, + { + "epoch": 0.36941, + "grad_norm": 0.6180613040924072, + "learning_rate": 0.003, + "loss": 3.967, + "step": 36941 + }, + { + "epoch": 0.36942, + "grad_norm": 0.7786617875099182, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 36942 + }, + { + "epoch": 0.36943, + "grad_norm": 0.9112129807472229, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 36943 + }, + { + "epoch": 0.36944, + "grad_norm": 1.1130257844924927, + "learning_rate": 0.003, + "loss": 3.995, + "step": 36944 + }, + { + "epoch": 0.36945, + "grad_norm": 0.8280860185623169, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 36945 + }, + { + "epoch": 0.36946, + "grad_norm": 0.7635908722877502, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 36946 + }, + { + "epoch": 0.36947, + "grad_norm": 0.7539331912994385, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 36947 + }, + { + "epoch": 0.36948, + "grad_norm": 0.7010751366615295, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 36948 + }, + { + "epoch": 0.36949, + "grad_norm": 0.7750644087791443, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 36949 + }, + { + "epoch": 0.3695, + "grad_norm": 0.7289663553237915, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 36950 + }, + { + "epoch": 0.36951, + "grad_norm": 0.6862159967422485, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 36951 + }, + { + "epoch": 0.36952, + "grad_norm": 0.6470468044281006, + "learning_rate": 0.003, + "loss": 3.995, + "step": 36952 + }, + { + "epoch": 0.36953, + "grad_norm": 0.7541556358337402, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 36953 + }, + { + "epoch": 0.36954, + "grad_norm": 0.9560497403144836, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 36954 + }, + { + "epoch": 0.36955, + "grad_norm": 1.1424953937530518, + "learning_rate": 0.003, + "loss": 3.968, + "step": 36955 + }, + { + "epoch": 0.36956, + "grad_norm": 0.7828877568244934, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 36956 + }, + { + "epoch": 0.36957, + "grad_norm": 0.8884869813919067, + "learning_rate": 0.003, + "loss": 4.016, + "step": 36957 + }, + { + "epoch": 0.36958, + "grad_norm": 1.3107149600982666, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 36958 + }, + { + "epoch": 0.36959, + "grad_norm": 0.9476369619369507, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 36959 + }, + { + "epoch": 0.3696, + "grad_norm": 0.8511696457862854, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 36960 + }, + { + "epoch": 0.36961, + "grad_norm": 0.8853407502174377, + "learning_rate": 0.003, + "loss": 3.975, + "step": 36961 + }, + { + "epoch": 0.36962, + "grad_norm": 1.0284334421157837, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 36962 + }, + { + "epoch": 0.36963, + "grad_norm": 1.1281284093856812, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 36963 + }, + { + "epoch": 0.36964, + "grad_norm": 0.7441313862800598, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 36964 + }, + { + "epoch": 0.36965, + "grad_norm": 0.6908243298530579, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 36965 + }, + { + "epoch": 0.36966, + "grad_norm": 0.7065862417221069, + "learning_rate": 0.003, + "loss": 3.982, + "step": 36966 + }, + { + "epoch": 0.36967, + "grad_norm": 0.7939404249191284, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 36967 + }, + { + "epoch": 0.36968, + "grad_norm": 0.9594210386276245, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 36968 + }, + { + "epoch": 0.36969, + "grad_norm": 1.2309565544128418, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 36969 + }, + { + "epoch": 0.3697, + "grad_norm": 0.7427654266357422, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 36970 + }, + { + "epoch": 0.36971, + "grad_norm": 0.6743876338005066, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 36971 + }, + { + "epoch": 0.36972, + "grad_norm": 0.6988741159439087, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 36972 + }, + { + "epoch": 0.36973, + "grad_norm": 0.6893583536148071, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 36973 + }, + { + "epoch": 0.36974, + "grad_norm": 0.8494730591773987, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 36974 + }, + { + "epoch": 0.36975, + "grad_norm": 0.8981438279151917, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 36975 + }, + { + "epoch": 0.36976, + "grad_norm": 0.995896577835083, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 36976 + }, + { + "epoch": 0.36977, + "grad_norm": 1.2650444507598877, + "learning_rate": 0.003, + "loss": 3.993, + "step": 36977 + }, + { + "epoch": 0.36978, + "grad_norm": 0.7352771759033203, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 36978 + }, + { + "epoch": 0.36979, + "grad_norm": 0.6930034756660461, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 36979 + }, + { + "epoch": 0.3698, + "grad_norm": 0.7071332335472107, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 36980 + }, + { + "epoch": 0.36981, + "grad_norm": 0.7644148468971252, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 36981 + }, + { + "epoch": 0.36982, + "grad_norm": 0.8089757561683655, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 36982 + }, + { + "epoch": 0.36983, + "grad_norm": 0.8036205768585205, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 36983 + }, + { + "epoch": 0.36984, + "grad_norm": 0.6939908862113953, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 36984 + }, + { + "epoch": 0.36985, + "grad_norm": 0.848086416721344, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 36985 + }, + { + "epoch": 0.36986, + "grad_norm": 1.2061856985092163, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 36986 + }, + { + "epoch": 0.36987, + "grad_norm": 1.1320945024490356, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 36987 + }, + { + "epoch": 0.36988, + "grad_norm": 0.8255122900009155, + "learning_rate": 0.003, + "loss": 3.977, + "step": 36988 + }, + { + "epoch": 0.36989, + "grad_norm": 0.6896318793296814, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 36989 + }, + { + "epoch": 0.3699, + "grad_norm": 0.8098825812339783, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 36990 + }, + { + "epoch": 0.36991, + "grad_norm": 0.9862973093986511, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 36991 + }, + { + "epoch": 0.36992, + "grad_norm": 0.9059943556785583, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 36992 + }, + { + "epoch": 0.36993, + "grad_norm": 1.0350013971328735, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 36993 + }, + { + "epoch": 0.36994, + "grad_norm": 1.1068389415740967, + "learning_rate": 0.003, + "loss": 3.967, + "step": 36994 + }, + { + "epoch": 0.36995, + "grad_norm": 0.6577797532081604, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 36995 + }, + { + "epoch": 0.36996, + "grad_norm": 0.6378918290138245, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 36996 + }, + { + "epoch": 0.36997, + "grad_norm": 0.7987729907035828, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 36997 + }, + { + "epoch": 0.36998, + "grad_norm": 0.9951341152191162, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 36998 + }, + { + "epoch": 0.36999, + "grad_norm": 1.155949354171753, + "learning_rate": 0.003, + "loss": 3.987, + "step": 36999 + }, + { + "epoch": 0.37, + "grad_norm": 0.8480376601219177, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 37000 + }, + { + "epoch": 0.37001, + "grad_norm": 0.9650610089302063, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 37001 + }, + { + "epoch": 0.37002, + "grad_norm": 1.0396307706832886, + "learning_rate": 0.003, + "loss": 3.955, + "step": 37002 + }, + { + "epoch": 0.37003, + "grad_norm": 1.2078237533569336, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 37003 + }, + { + "epoch": 0.37004, + "grad_norm": 1.0226410627365112, + "learning_rate": 0.003, + "loss": 3.965, + "step": 37004 + }, + { + "epoch": 0.37005, + "grad_norm": 0.9424207210540771, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 37005 + }, + { + "epoch": 0.37006, + "grad_norm": 1.0875686407089233, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 37006 + }, + { + "epoch": 0.37007, + "grad_norm": 1.2267954349517822, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 37007 + }, + { + "epoch": 0.37008, + "grad_norm": 0.7341787815093994, + "learning_rate": 0.003, + "loss": 3.9509, + "step": 37008 + }, + { + "epoch": 0.37009, + "grad_norm": 0.8532608151435852, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 37009 + }, + { + "epoch": 0.3701, + "grad_norm": 1.0819170475006104, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 37010 + }, + { + "epoch": 0.37011, + "grad_norm": 1.0629777908325195, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 37011 + }, + { + "epoch": 0.37012, + "grad_norm": 0.8998352885246277, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 37012 + }, + { + "epoch": 0.37013, + "grad_norm": 0.9881258010864258, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 37013 + }, + { + "epoch": 0.37014, + "grad_norm": 1.0586068630218506, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 37014 + }, + { + "epoch": 0.37015, + "grad_norm": 0.9900878667831421, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 37015 + }, + { + "epoch": 0.37016, + "grad_norm": 0.867002546787262, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 37016 + }, + { + "epoch": 0.37017, + "grad_norm": 0.9403209686279297, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 37017 + }, + { + "epoch": 0.37018, + "grad_norm": 1.1596912145614624, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 37018 + }, + { + "epoch": 0.37019, + "grad_norm": 0.9733276963233948, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 37019 + }, + { + "epoch": 0.3702, + "grad_norm": 0.9283165335655212, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 37020 + }, + { + "epoch": 0.37021, + "grad_norm": 1.0823544263839722, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 37021 + }, + { + "epoch": 0.37022, + "grad_norm": 0.900855541229248, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 37022 + }, + { + "epoch": 0.37023, + "grad_norm": 0.8960193395614624, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 37023 + }, + { + "epoch": 0.37024, + "grad_norm": 0.90533047914505, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 37024 + }, + { + "epoch": 0.37025, + "grad_norm": 0.9168426990509033, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 37025 + }, + { + "epoch": 0.37026, + "grad_norm": 1.0016593933105469, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 37026 + }, + { + "epoch": 0.37027, + "grad_norm": 0.9358053803443909, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 37027 + }, + { + "epoch": 0.37028, + "grad_norm": 1.0025672912597656, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 37028 + }, + { + "epoch": 0.37029, + "grad_norm": 1.0815958976745605, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 37029 + }, + { + "epoch": 0.3703, + "grad_norm": 1.1881353855133057, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 37030 + }, + { + "epoch": 0.37031, + "grad_norm": 0.9777210354804993, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 37031 + }, + { + "epoch": 0.37032, + "grad_norm": 0.8781248331069946, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 37032 + }, + { + "epoch": 0.37033, + "grad_norm": 0.7451189756393433, + "learning_rate": 0.003, + "loss": 4.022, + "step": 37033 + }, + { + "epoch": 0.37034, + "grad_norm": 0.6234452128410339, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 37034 + }, + { + "epoch": 0.37035, + "grad_norm": 0.7151374816894531, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 37035 + }, + { + "epoch": 0.37036, + "grad_norm": 0.7758432030677795, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 37036 + }, + { + "epoch": 0.37037, + "grad_norm": 0.8766210675239563, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 37037 + }, + { + "epoch": 0.37038, + "grad_norm": 1.198695182800293, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 37038 + }, + { + "epoch": 0.37039, + "grad_norm": 0.8132272362709045, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 37039 + }, + { + "epoch": 0.3704, + "grad_norm": 0.5544961094856262, + "learning_rate": 0.003, + "loss": 3.9401, + "step": 37040 + }, + { + "epoch": 0.37041, + "grad_norm": 0.7843204736709595, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 37041 + }, + { + "epoch": 0.37042, + "grad_norm": 1.0159064531326294, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 37042 + }, + { + "epoch": 0.37043, + "grad_norm": 1.0880342721939087, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 37043 + }, + { + "epoch": 0.37044, + "grad_norm": 0.8131694793701172, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 37044 + }, + { + "epoch": 0.37045, + "grad_norm": 0.7694173455238342, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 37045 + }, + { + "epoch": 0.37046, + "grad_norm": 0.6950709223747253, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 37046 + }, + { + "epoch": 0.37047, + "grad_norm": 0.7095657587051392, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 37047 + }, + { + "epoch": 0.37048, + "grad_norm": 0.6756837368011475, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 37048 + }, + { + "epoch": 0.37049, + "grad_norm": 0.7321543097496033, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 37049 + }, + { + "epoch": 0.3705, + "grad_norm": 0.8550647497177124, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 37050 + }, + { + "epoch": 0.37051, + "grad_norm": 1.0084127187728882, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 37051 + }, + { + "epoch": 0.37052, + "grad_norm": 0.9869688749313354, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 37052 + }, + { + "epoch": 0.37053, + "grad_norm": 0.8719035387039185, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 37053 + }, + { + "epoch": 0.37054, + "grad_norm": 1.0108656883239746, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 37054 + }, + { + "epoch": 0.37055, + "grad_norm": 1.3648319244384766, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 37055 + }, + { + "epoch": 0.37056, + "grad_norm": 0.6175218820571899, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 37056 + }, + { + "epoch": 0.37057, + "grad_norm": 0.7099364995956421, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 37057 + }, + { + "epoch": 0.37058, + "grad_norm": 0.7294310927391052, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 37058 + }, + { + "epoch": 0.37059, + "grad_norm": 0.793144166469574, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 37059 + }, + { + "epoch": 0.3706, + "grad_norm": 1.1395446062088013, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 37060 + }, + { + "epoch": 0.37061, + "grad_norm": 1.0670396089553833, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 37061 + }, + { + "epoch": 0.37062, + "grad_norm": 1.0250731706619263, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 37062 + }, + { + "epoch": 0.37063, + "grad_norm": 1.144058346748352, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 37063 + }, + { + "epoch": 0.37064, + "grad_norm": 0.8915567398071289, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 37064 + }, + { + "epoch": 0.37065, + "grad_norm": 0.6544596552848816, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 37065 + }, + { + "epoch": 0.37066, + "grad_norm": 0.7657761573791504, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 37066 + }, + { + "epoch": 0.37067, + "grad_norm": 0.7533538341522217, + "learning_rate": 0.003, + "loss": 3.988, + "step": 37067 + }, + { + "epoch": 0.37068, + "grad_norm": 0.8080588579177856, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 37068 + }, + { + "epoch": 0.37069, + "grad_norm": 0.838137686252594, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 37069 + }, + { + "epoch": 0.3707, + "grad_norm": 0.9734652042388916, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 37070 + }, + { + "epoch": 0.37071, + "grad_norm": 0.9610617756843567, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 37071 + }, + { + "epoch": 0.37072, + "grad_norm": 0.8409895300865173, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 37072 + }, + { + "epoch": 0.37073, + "grad_norm": 0.8951683044433594, + "learning_rate": 0.003, + "loss": 3.987, + "step": 37073 + }, + { + "epoch": 0.37074, + "grad_norm": 0.9730215668678284, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 37074 + }, + { + "epoch": 0.37075, + "grad_norm": 0.9812310338020325, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 37075 + }, + { + "epoch": 0.37076, + "grad_norm": 1.2820229530334473, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 37076 + }, + { + "epoch": 0.37077, + "grad_norm": 0.904712975025177, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 37077 + }, + { + "epoch": 0.37078, + "grad_norm": 1.0265789031982422, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 37078 + }, + { + "epoch": 0.37079, + "grad_norm": 1.1804418563842773, + "learning_rate": 0.003, + "loss": 3.991, + "step": 37079 + }, + { + "epoch": 0.3708, + "grad_norm": 0.8717527389526367, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 37080 + }, + { + "epoch": 0.37081, + "grad_norm": 0.9045213460922241, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 37081 + }, + { + "epoch": 0.37082, + "grad_norm": 0.9367362260818481, + "learning_rate": 0.003, + "loss": 3.981, + "step": 37082 + }, + { + "epoch": 0.37083, + "grad_norm": 0.9407592415809631, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 37083 + }, + { + "epoch": 0.37084, + "grad_norm": 1.0585895776748657, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 37084 + }, + { + "epoch": 0.37085, + "grad_norm": 1.0259891748428345, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 37085 + }, + { + "epoch": 0.37086, + "grad_norm": 0.9714613556861877, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 37086 + }, + { + "epoch": 0.37087, + "grad_norm": 0.9432087540626526, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 37087 + }, + { + "epoch": 0.37088, + "grad_norm": 1.155334234237671, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 37088 + }, + { + "epoch": 0.37089, + "grad_norm": 0.8993036150932312, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 37089 + }, + { + "epoch": 0.3709, + "grad_norm": 0.8614706993103027, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 37090 + }, + { + "epoch": 0.37091, + "grad_norm": 1.1699390411376953, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 37091 + }, + { + "epoch": 0.37092, + "grad_norm": 1.2750556468963623, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 37092 + }, + { + "epoch": 0.37093, + "grad_norm": 0.8425441384315491, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 37093 + }, + { + "epoch": 0.37094, + "grad_norm": 0.7334075570106506, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 37094 + }, + { + "epoch": 0.37095, + "grad_norm": 0.7200164198875427, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 37095 + }, + { + "epoch": 0.37096, + "grad_norm": 0.6630911231040955, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 37096 + }, + { + "epoch": 0.37097, + "grad_norm": 0.6556231379508972, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 37097 + }, + { + "epoch": 0.37098, + "grad_norm": 0.6324580311775208, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 37098 + }, + { + "epoch": 0.37099, + "grad_norm": 0.5696724057197571, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 37099 + }, + { + "epoch": 0.371, + "grad_norm": 0.5970602035522461, + "learning_rate": 0.003, + "loss": 3.9365, + "step": 37100 + }, + { + "epoch": 0.37101, + "grad_norm": 0.7263101935386658, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 37101 + }, + { + "epoch": 0.37102, + "grad_norm": 0.9218031764030457, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 37102 + }, + { + "epoch": 0.37103, + "grad_norm": 1.21561861038208, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 37103 + }, + { + "epoch": 0.37104, + "grad_norm": 0.9003336429595947, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 37104 + }, + { + "epoch": 0.37105, + "grad_norm": 0.8950603008270264, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 37105 + }, + { + "epoch": 0.37106, + "grad_norm": 0.8666157722473145, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 37106 + }, + { + "epoch": 0.37107, + "grad_norm": 1.058619737625122, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 37107 + }, + { + "epoch": 0.37108, + "grad_norm": 1.0460821390151978, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 37108 + }, + { + "epoch": 0.37109, + "grad_norm": 1.0205835103988647, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 37109 + }, + { + "epoch": 0.3711, + "grad_norm": 0.8670299649238586, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 37110 + }, + { + "epoch": 0.37111, + "grad_norm": 0.6874682903289795, + "learning_rate": 0.003, + "loss": 3.956, + "step": 37111 + }, + { + "epoch": 0.37112, + "grad_norm": 0.751916766166687, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 37112 + }, + { + "epoch": 0.37113, + "grad_norm": 1.0011732578277588, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 37113 + }, + { + "epoch": 0.37114, + "grad_norm": 1.2032546997070312, + "learning_rate": 0.003, + "loss": 3.997, + "step": 37114 + }, + { + "epoch": 0.37115, + "grad_norm": 0.9026545882225037, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 37115 + }, + { + "epoch": 0.37116, + "grad_norm": 0.7948858737945557, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 37116 + }, + { + "epoch": 0.37117, + "grad_norm": 0.8753494620323181, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 37117 + }, + { + "epoch": 0.37118, + "grad_norm": 1.1389180421829224, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 37118 + }, + { + "epoch": 0.37119, + "grad_norm": 0.9926146268844604, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 37119 + }, + { + "epoch": 0.3712, + "grad_norm": 0.8344734907150269, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 37120 + }, + { + "epoch": 0.37121, + "grad_norm": 0.9314437508583069, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 37121 + }, + { + "epoch": 0.37122, + "grad_norm": 1.1512407064437866, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 37122 + }, + { + "epoch": 0.37123, + "grad_norm": 0.7222261428833008, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 37123 + }, + { + "epoch": 0.37124, + "grad_norm": 0.5921404957771301, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 37124 + }, + { + "epoch": 0.37125, + "grad_norm": 0.6597774028778076, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 37125 + }, + { + "epoch": 0.37126, + "grad_norm": 0.9163988828659058, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 37126 + }, + { + "epoch": 0.37127, + "grad_norm": 1.0719921588897705, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 37127 + }, + { + "epoch": 0.37128, + "grad_norm": 1.1076430082321167, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 37128 + }, + { + "epoch": 0.37129, + "grad_norm": 0.8148869276046753, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 37129 + }, + { + "epoch": 0.3713, + "grad_norm": 0.6592761278152466, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 37130 + }, + { + "epoch": 0.37131, + "grad_norm": 0.6607849597930908, + "learning_rate": 0.003, + "loss": 3.996, + "step": 37131 + }, + { + "epoch": 0.37132, + "grad_norm": 0.6573082208633423, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 37132 + }, + { + "epoch": 0.37133, + "grad_norm": 0.6657180786132812, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 37133 + }, + { + "epoch": 0.37134, + "grad_norm": 0.7460392713546753, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 37134 + }, + { + "epoch": 0.37135, + "grad_norm": 0.7617326974868774, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 37135 + }, + { + "epoch": 0.37136, + "grad_norm": 0.7959724068641663, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 37136 + }, + { + "epoch": 0.37137, + "grad_norm": 0.7426033616065979, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 37137 + }, + { + "epoch": 0.37138, + "grad_norm": 0.802419900894165, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 37138 + }, + { + "epoch": 0.37139, + "grad_norm": 0.8818849921226501, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 37139 + }, + { + "epoch": 0.3714, + "grad_norm": 1.1160365343093872, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 37140 + }, + { + "epoch": 0.37141, + "grad_norm": 1.2113488912582397, + "learning_rate": 0.003, + "loss": 3.949, + "step": 37141 + }, + { + "epoch": 0.37142, + "grad_norm": 1.0666513442993164, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 37142 + }, + { + "epoch": 0.37143, + "grad_norm": 1.054213285446167, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 37143 + }, + { + "epoch": 0.37144, + "grad_norm": 0.8875797986984253, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 37144 + }, + { + "epoch": 0.37145, + "grad_norm": 0.8301931023597717, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 37145 + }, + { + "epoch": 0.37146, + "grad_norm": 0.7412774562835693, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 37146 + }, + { + "epoch": 0.37147, + "grad_norm": 1.1275625228881836, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 37147 + }, + { + "epoch": 0.37148, + "grad_norm": 1.1206583976745605, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 37148 + }, + { + "epoch": 0.37149, + "grad_norm": 0.7979243993759155, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 37149 + }, + { + "epoch": 0.3715, + "grad_norm": 0.7775332927703857, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 37150 + }, + { + "epoch": 0.37151, + "grad_norm": 1.0015536546707153, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 37151 + }, + { + "epoch": 0.37152, + "grad_norm": 1.1434203386306763, + "learning_rate": 0.003, + "loss": 3.997, + "step": 37152 + }, + { + "epoch": 0.37153, + "grad_norm": 0.8195577263832092, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 37153 + }, + { + "epoch": 0.37154, + "grad_norm": 0.7339721918106079, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 37154 + }, + { + "epoch": 0.37155, + "grad_norm": 0.839361310005188, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 37155 + }, + { + "epoch": 0.37156, + "grad_norm": 0.8569461703300476, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 37156 + }, + { + "epoch": 0.37157, + "grad_norm": 0.8099280595779419, + "learning_rate": 0.003, + "loss": 3.974, + "step": 37157 + }, + { + "epoch": 0.37158, + "grad_norm": 0.9442126750946045, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 37158 + }, + { + "epoch": 0.37159, + "grad_norm": 1.087133765220642, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 37159 + }, + { + "epoch": 0.3716, + "grad_norm": 0.9793602824211121, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 37160 + }, + { + "epoch": 0.37161, + "grad_norm": 0.9285367131233215, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 37161 + }, + { + "epoch": 0.37162, + "grad_norm": 1.0501604080200195, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 37162 + }, + { + "epoch": 0.37163, + "grad_norm": 1.1089855432510376, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 37163 + }, + { + "epoch": 0.37164, + "grad_norm": 0.9021442532539368, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 37164 + }, + { + "epoch": 0.37165, + "grad_norm": 0.8500725626945496, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 37165 + }, + { + "epoch": 0.37166, + "grad_norm": 0.9334098696708679, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 37166 + }, + { + "epoch": 0.37167, + "grad_norm": 1.095099925994873, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 37167 + }, + { + "epoch": 0.37168, + "grad_norm": 0.9401664137840271, + "learning_rate": 0.003, + "loss": 3.959, + "step": 37168 + }, + { + "epoch": 0.37169, + "grad_norm": 0.9560405015945435, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 37169 + }, + { + "epoch": 0.3717, + "grad_norm": 1.3423986434936523, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 37170 + }, + { + "epoch": 0.37171, + "grad_norm": 0.6721263527870178, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 37171 + }, + { + "epoch": 0.37172, + "grad_norm": 0.7563371658325195, + "learning_rate": 0.003, + "loss": 3.998, + "step": 37172 + }, + { + "epoch": 0.37173, + "grad_norm": 0.8149079084396362, + "learning_rate": 0.003, + "loss": 3.963, + "step": 37173 + }, + { + "epoch": 0.37174, + "grad_norm": 0.8288574814796448, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 37174 + }, + { + "epoch": 0.37175, + "grad_norm": 0.9780520796775818, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 37175 + }, + { + "epoch": 0.37176, + "grad_norm": 1.165347933769226, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 37176 + }, + { + "epoch": 0.37177, + "grad_norm": 0.7574217915534973, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 37177 + }, + { + "epoch": 0.37178, + "grad_norm": 0.6748964190483093, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 37178 + }, + { + "epoch": 0.37179, + "grad_norm": 0.8289321660995483, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 37179 + }, + { + "epoch": 0.3718, + "grad_norm": 1.1459097862243652, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 37180 + }, + { + "epoch": 0.37181, + "grad_norm": 1.023309588432312, + "learning_rate": 0.003, + "loss": 3.9478, + "step": 37181 + }, + { + "epoch": 0.37182, + "grad_norm": 1.147483229637146, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 37182 + }, + { + "epoch": 0.37183, + "grad_norm": 0.6732507348060608, + "learning_rate": 0.003, + "loss": 3.9411, + "step": 37183 + }, + { + "epoch": 0.37184, + "grad_norm": 0.6857383847236633, + "learning_rate": 0.003, + "loss": 3.9465, + "step": 37184 + }, + { + "epoch": 0.37185, + "grad_norm": 0.8449041247367859, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 37185 + }, + { + "epoch": 0.37186, + "grad_norm": 0.9662480354309082, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 37186 + }, + { + "epoch": 0.37187, + "grad_norm": 1.1194751262664795, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 37187 + }, + { + "epoch": 0.37188, + "grad_norm": 0.8374146819114685, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 37188 + }, + { + "epoch": 0.37189, + "grad_norm": 0.8380528092384338, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 37189 + }, + { + "epoch": 0.3719, + "grad_norm": 1.0108075141906738, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 37190 + }, + { + "epoch": 0.37191, + "grad_norm": 1.3672782182693481, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 37191 + }, + { + "epoch": 0.37192, + "grad_norm": 0.7304951548576355, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 37192 + }, + { + "epoch": 0.37193, + "grad_norm": 0.6709659695625305, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 37193 + }, + { + "epoch": 0.37194, + "grad_norm": 0.760771632194519, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 37194 + }, + { + "epoch": 0.37195, + "grad_norm": 0.879575252532959, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 37195 + }, + { + "epoch": 0.37196, + "grad_norm": 0.8802063465118408, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 37196 + }, + { + "epoch": 0.37197, + "grad_norm": 0.7830325961112976, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 37197 + }, + { + "epoch": 0.37198, + "grad_norm": 0.8271722793579102, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 37198 + }, + { + "epoch": 0.37199, + "grad_norm": 0.9876917600631714, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 37199 + }, + { + "epoch": 0.372, + "grad_norm": 0.9978885650634766, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 37200 + }, + { + "epoch": 0.37201, + "grad_norm": 1.3937963247299194, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 37201 + }, + { + "epoch": 0.37202, + "grad_norm": 0.7049383521080017, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 37202 + }, + { + "epoch": 0.37203, + "grad_norm": 0.6365799307823181, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 37203 + }, + { + "epoch": 0.37204, + "grad_norm": 0.7951638698577881, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 37204 + }, + { + "epoch": 0.37205, + "grad_norm": 0.9576162695884705, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 37205 + }, + { + "epoch": 0.37206, + "grad_norm": 1.1558071374893188, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 37206 + }, + { + "epoch": 0.37207, + "grad_norm": 0.6973257064819336, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 37207 + }, + { + "epoch": 0.37208, + "grad_norm": 1.0271668434143066, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 37208 + }, + { + "epoch": 0.37209, + "grad_norm": 1.0437511205673218, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 37209 + }, + { + "epoch": 0.3721, + "grad_norm": 1.0492814779281616, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 37210 + }, + { + "epoch": 0.37211, + "grad_norm": 1.0131875276565552, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 37211 + }, + { + "epoch": 0.37212, + "grad_norm": 0.9167686104774475, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 37212 + }, + { + "epoch": 0.37213, + "grad_norm": 0.9262983798980713, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 37213 + }, + { + "epoch": 0.37214, + "grad_norm": 0.9110230803489685, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 37214 + }, + { + "epoch": 0.37215, + "grad_norm": 1.0077701807022095, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 37215 + }, + { + "epoch": 0.37216, + "grad_norm": 1.1596015691757202, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 37216 + }, + { + "epoch": 0.37217, + "grad_norm": 0.918013334274292, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 37217 + }, + { + "epoch": 0.37218, + "grad_norm": 0.8578289151191711, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 37218 + }, + { + "epoch": 0.37219, + "grad_norm": 0.9064183235168457, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 37219 + }, + { + "epoch": 0.3722, + "grad_norm": 0.9828143119812012, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 37220 + }, + { + "epoch": 0.37221, + "grad_norm": 1.0691726207733154, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 37221 + }, + { + "epoch": 0.37222, + "grad_norm": 1.0419119596481323, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 37222 + }, + { + "epoch": 0.37223, + "grad_norm": 0.8782952427864075, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 37223 + }, + { + "epoch": 0.37224, + "grad_norm": 0.9029203653335571, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 37224 + }, + { + "epoch": 0.37225, + "grad_norm": 1.1069079637527466, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 37225 + }, + { + "epoch": 0.37226, + "grad_norm": 0.7867429256439209, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 37226 + }, + { + "epoch": 0.37227, + "grad_norm": 0.7790812253952026, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 37227 + }, + { + "epoch": 0.37228, + "grad_norm": 0.7767866253852844, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 37228 + }, + { + "epoch": 0.37229, + "grad_norm": 0.8947355151176453, + "learning_rate": 0.003, + "loss": 3.972, + "step": 37229 + }, + { + "epoch": 0.3723, + "grad_norm": 1.0100175142288208, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 37230 + }, + { + "epoch": 0.37231, + "grad_norm": 1.2303526401519775, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 37231 + }, + { + "epoch": 0.37232, + "grad_norm": 0.659388542175293, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 37232 + }, + { + "epoch": 0.37233, + "grad_norm": 0.5932953357696533, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 37233 + }, + { + "epoch": 0.37234, + "grad_norm": 0.7821550965309143, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 37234 + }, + { + "epoch": 0.37235, + "grad_norm": 0.9049550890922546, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 37235 + }, + { + "epoch": 0.37236, + "grad_norm": 1.1531623601913452, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 37236 + }, + { + "epoch": 0.37237, + "grad_norm": 0.8387067317962646, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 37237 + }, + { + "epoch": 0.37238, + "grad_norm": 0.8507299423217773, + "learning_rate": 0.003, + "loss": 3.986, + "step": 37238 + }, + { + "epoch": 0.37239, + "grad_norm": 0.9016196727752686, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 37239 + }, + { + "epoch": 0.3724, + "grad_norm": 0.7951169013977051, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 37240 + }, + { + "epoch": 0.37241, + "grad_norm": 0.8169518709182739, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 37241 + }, + { + "epoch": 0.37242, + "grad_norm": 0.9298339486122131, + "learning_rate": 0.003, + "loss": 4.021, + "step": 37242 + }, + { + "epoch": 0.37243, + "grad_norm": 0.7873278856277466, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 37243 + }, + { + "epoch": 0.37244, + "grad_norm": 0.7046746611595154, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 37244 + }, + { + "epoch": 0.37245, + "grad_norm": 0.7148481011390686, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 37245 + }, + { + "epoch": 0.37246, + "grad_norm": 0.7474754452705383, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 37246 + }, + { + "epoch": 0.37247, + "grad_norm": 0.7656787037849426, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 37247 + }, + { + "epoch": 0.37248, + "grad_norm": 0.7734026908874512, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 37248 + }, + { + "epoch": 0.37249, + "grad_norm": 0.7438535094261169, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 37249 + }, + { + "epoch": 0.3725, + "grad_norm": 0.8334265947341919, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 37250 + }, + { + "epoch": 0.37251, + "grad_norm": 0.8754532933235168, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 37251 + }, + { + "epoch": 0.37252, + "grad_norm": 0.7644717693328857, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 37252 + }, + { + "epoch": 0.37253, + "grad_norm": 0.7552086710929871, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 37253 + }, + { + "epoch": 0.37254, + "grad_norm": 0.7193397283554077, + "learning_rate": 0.003, + "loss": 3.9358, + "step": 37254 + }, + { + "epoch": 0.37255, + "grad_norm": 0.7339372038841248, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 37255 + }, + { + "epoch": 0.37256, + "grad_norm": 0.9036675691604614, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 37256 + }, + { + "epoch": 0.37257, + "grad_norm": 1.1719911098480225, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 37257 + }, + { + "epoch": 0.37258, + "grad_norm": 1.1707621812820435, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 37258 + }, + { + "epoch": 0.37259, + "grad_norm": 1.1385705471038818, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 37259 + }, + { + "epoch": 0.3726, + "grad_norm": 0.9568917751312256, + "learning_rate": 0.003, + "loss": 3.975, + "step": 37260 + }, + { + "epoch": 0.37261, + "grad_norm": 0.9926353693008423, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 37261 + }, + { + "epoch": 0.37262, + "grad_norm": 1.1636451482772827, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 37262 + }, + { + "epoch": 0.37263, + "grad_norm": 1.1583460569381714, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 37263 + }, + { + "epoch": 0.37264, + "grad_norm": 0.8393154740333557, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 37264 + }, + { + "epoch": 0.37265, + "grad_norm": 0.7835245132446289, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 37265 + }, + { + "epoch": 0.37266, + "grad_norm": 0.9497269988059998, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 37266 + }, + { + "epoch": 0.37267, + "grad_norm": 1.0879652500152588, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 37267 + }, + { + "epoch": 0.37268, + "grad_norm": 0.8629530668258667, + "learning_rate": 0.003, + "loss": 3.974, + "step": 37268 + }, + { + "epoch": 0.37269, + "grad_norm": 0.8297415971755981, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 37269 + }, + { + "epoch": 0.3727, + "grad_norm": 0.8757036328315735, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 37270 + }, + { + "epoch": 0.37271, + "grad_norm": 0.9682446122169495, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 37271 + }, + { + "epoch": 0.37272, + "grad_norm": 1.2792710065841675, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 37272 + }, + { + "epoch": 0.37273, + "grad_norm": 0.735395610332489, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 37273 + }, + { + "epoch": 0.37274, + "grad_norm": 0.7381970882415771, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 37274 + }, + { + "epoch": 0.37275, + "grad_norm": 1.0856236219406128, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 37275 + }, + { + "epoch": 0.37276, + "grad_norm": 1.2367810010910034, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 37276 + }, + { + "epoch": 0.37277, + "grad_norm": 0.8891799449920654, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 37277 + }, + { + "epoch": 0.37278, + "grad_norm": 0.7994269728660583, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 37278 + }, + { + "epoch": 0.37279, + "grad_norm": 0.7658758163452148, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 37279 + }, + { + "epoch": 0.3728, + "grad_norm": 0.812741219997406, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 37280 + }, + { + "epoch": 0.37281, + "grad_norm": 0.8600624203681946, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 37281 + }, + { + "epoch": 0.37282, + "grad_norm": 0.8046388626098633, + "learning_rate": 0.003, + "loss": 3.968, + "step": 37282 + }, + { + "epoch": 0.37283, + "grad_norm": 0.8632801175117493, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 37283 + }, + { + "epoch": 0.37284, + "grad_norm": 0.9114347696304321, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 37284 + }, + { + "epoch": 0.37285, + "grad_norm": 1.0515540838241577, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 37285 + }, + { + "epoch": 0.37286, + "grad_norm": 1.118456482887268, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 37286 + }, + { + "epoch": 0.37287, + "grad_norm": 1.0894936323165894, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 37287 + }, + { + "epoch": 0.37288, + "grad_norm": 0.8131495714187622, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 37288 + }, + { + "epoch": 0.37289, + "grad_norm": 0.8759163618087769, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 37289 + }, + { + "epoch": 0.3729, + "grad_norm": 1.120607614517212, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 37290 + }, + { + "epoch": 0.37291, + "grad_norm": 0.9766330122947693, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 37291 + }, + { + "epoch": 0.37292, + "grad_norm": 0.873801589012146, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 37292 + }, + { + "epoch": 0.37293, + "grad_norm": 1.0253219604492188, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 37293 + }, + { + "epoch": 0.37294, + "grad_norm": 1.1546287536621094, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 37294 + }, + { + "epoch": 0.37295, + "grad_norm": 0.7256019711494446, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 37295 + }, + { + "epoch": 0.37296, + "grad_norm": 0.7245150804519653, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 37296 + }, + { + "epoch": 0.37297, + "grad_norm": 0.7026397585868835, + "learning_rate": 0.003, + "loss": 3.96, + "step": 37297 + }, + { + "epoch": 0.37298, + "grad_norm": 0.771270751953125, + "learning_rate": 0.003, + "loss": 3.991, + "step": 37298 + }, + { + "epoch": 0.37299, + "grad_norm": 0.8777480125427246, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 37299 + }, + { + "epoch": 0.373, + "grad_norm": 0.8068657517433167, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 37300 + }, + { + "epoch": 0.37301, + "grad_norm": 0.7142436504364014, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 37301 + }, + { + "epoch": 0.37302, + "grad_norm": 0.6684564352035522, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 37302 + }, + { + "epoch": 0.37303, + "grad_norm": 0.6652061343193054, + "learning_rate": 0.003, + "loss": 3.9333, + "step": 37303 + }, + { + "epoch": 0.37304, + "grad_norm": 0.6426818370819092, + "learning_rate": 0.003, + "loss": 3.9428, + "step": 37304 + }, + { + "epoch": 0.37305, + "grad_norm": 0.7791720032691956, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 37305 + }, + { + "epoch": 0.37306, + "grad_norm": 0.9628361463546753, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 37306 + }, + { + "epoch": 0.37307, + "grad_norm": 1.2786030769348145, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 37307 + }, + { + "epoch": 0.37308, + "grad_norm": 0.9001626372337341, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 37308 + }, + { + "epoch": 0.37309, + "grad_norm": 0.7805397510528564, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 37309 + }, + { + "epoch": 0.3731, + "grad_norm": 0.7432160973548889, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 37310 + }, + { + "epoch": 0.37311, + "grad_norm": 0.7090791463851929, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 37311 + }, + { + "epoch": 0.37312, + "grad_norm": 0.7876272797584534, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 37312 + }, + { + "epoch": 0.37313, + "grad_norm": 0.9443870186805725, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 37313 + }, + { + "epoch": 0.37314, + "grad_norm": 1.075921654701233, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 37314 + }, + { + "epoch": 0.37315, + "grad_norm": 1.0085991621017456, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 37315 + }, + { + "epoch": 0.37316, + "grad_norm": 0.9799695611000061, + "learning_rate": 0.003, + "loss": 3.993, + "step": 37316 + }, + { + "epoch": 0.37317, + "grad_norm": 1.0005269050598145, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 37317 + }, + { + "epoch": 0.37318, + "grad_norm": 1.0694594383239746, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 37318 + }, + { + "epoch": 0.37319, + "grad_norm": 1.068430781364441, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 37319 + }, + { + "epoch": 0.3732, + "grad_norm": 1.0198016166687012, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 37320 + }, + { + "epoch": 0.37321, + "grad_norm": 0.8592920303344727, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 37321 + }, + { + "epoch": 0.37322, + "grad_norm": 0.8390135765075684, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 37322 + }, + { + "epoch": 0.37323, + "grad_norm": 0.8031332492828369, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 37323 + }, + { + "epoch": 0.37324, + "grad_norm": 0.7545059323310852, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 37324 + }, + { + "epoch": 0.37325, + "grad_norm": 0.6974469423294067, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 37325 + }, + { + "epoch": 0.37326, + "grad_norm": 0.6277225613594055, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 37326 + }, + { + "epoch": 0.37327, + "grad_norm": 0.9196205735206604, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 37327 + }, + { + "epoch": 0.37328, + "grad_norm": 1.686104416847229, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 37328 + }, + { + "epoch": 0.37329, + "grad_norm": 0.5648958086967468, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 37329 + }, + { + "epoch": 0.3733, + "grad_norm": 0.9967575073242188, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 37330 + }, + { + "epoch": 0.37331, + "grad_norm": 1.3605620861053467, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 37331 + }, + { + "epoch": 0.37332, + "grad_norm": 0.7730592489242554, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 37332 + }, + { + "epoch": 0.37333, + "grad_norm": 0.8204920887947083, + "learning_rate": 0.003, + "loss": 3.96, + "step": 37333 + }, + { + "epoch": 0.37334, + "grad_norm": 0.8347780704498291, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 37334 + }, + { + "epoch": 0.37335, + "grad_norm": 0.9820380210876465, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 37335 + }, + { + "epoch": 0.37336, + "grad_norm": 1.2095749378204346, + "learning_rate": 0.003, + "loss": 4.029, + "step": 37336 + }, + { + "epoch": 0.37337, + "grad_norm": 0.9229565858840942, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 37337 + }, + { + "epoch": 0.37338, + "grad_norm": 0.9271348714828491, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 37338 + }, + { + "epoch": 0.37339, + "grad_norm": 0.9389652013778687, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 37339 + }, + { + "epoch": 0.3734, + "grad_norm": 0.9101753830909729, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 37340 + }, + { + "epoch": 0.37341, + "grad_norm": 0.8272475004196167, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 37341 + }, + { + "epoch": 0.37342, + "grad_norm": 0.9930359721183777, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 37342 + }, + { + "epoch": 0.37343, + "grad_norm": 1.0682258605957031, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 37343 + }, + { + "epoch": 0.37344, + "grad_norm": 0.9129781126976013, + "learning_rate": 0.003, + "loss": 3.979, + "step": 37344 + }, + { + "epoch": 0.37345, + "grad_norm": 0.8073350787162781, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 37345 + }, + { + "epoch": 0.37346, + "grad_norm": 0.8166892528533936, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 37346 + }, + { + "epoch": 0.37347, + "grad_norm": 1.0291740894317627, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 37347 + }, + { + "epoch": 0.37348, + "grad_norm": 1.1282069683074951, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 37348 + }, + { + "epoch": 0.37349, + "grad_norm": 0.9856895804405212, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 37349 + }, + { + "epoch": 0.3735, + "grad_norm": 0.9748697876930237, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 37350 + }, + { + "epoch": 0.37351, + "grad_norm": 0.8971541523933411, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 37351 + }, + { + "epoch": 0.37352, + "grad_norm": 0.9031245112419128, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 37352 + }, + { + "epoch": 0.37353, + "grad_norm": 1.0660446882247925, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 37353 + }, + { + "epoch": 0.37354, + "grad_norm": 1.213388442993164, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 37354 + }, + { + "epoch": 0.37355, + "grad_norm": 0.8600258827209473, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 37355 + }, + { + "epoch": 0.37356, + "grad_norm": 0.6558176875114441, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 37356 + }, + { + "epoch": 0.37357, + "grad_norm": 0.5856007933616638, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 37357 + }, + { + "epoch": 0.37358, + "grad_norm": 0.5732006430625916, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 37358 + }, + { + "epoch": 0.37359, + "grad_norm": 0.6980565786361694, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 37359 + }, + { + "epoch": 0.3736, + "grad_norm": 0.8322768807411194, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 37360 + }, + { + "epoch": 0.37361, + "grad_norm": 0.8865121006965637, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 37361 + }, + { + "epoch": 0.37362, + "grad_norm": 0.9546343684196472, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 37362 + }, + { + "epoch": 0.37363, + "grad_norm": 0.8906335234642029, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 37363 + }, + { + "epoch": 0.37364, + "grad_norm": 0.7337836623191833, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 37364 + }, + { + "epoch": 0.37365, + "grad_norm": 0.75478196144104, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 37365 + }, + { + "epoch": 0.37366, + "grad_norm": 0.8041022419929504, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 37366 + }, + { + "epoch": 0.37367, + "grad_norm": 1.0549877882003784, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 37367 + }, + { + "epoch": 0.37368, + "grad_norm": 1.2569702863693237, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 37368 + }, + { + "epoch": 0.37369, + "grad_norm": 0.7820621728897095, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 37369 + }, + { + "epoch": 0.3737, + "grad_norm": 0.770089328289032, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 37370 + }, + { + "epoch": 0.37371, + "grad_norm": 0.8568240404129028, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 37371 + }, + { + "epoch": 0.37372, + "grad_norm": 0.9214556217193604, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 37372 + }, + { + "epoch": 0.37373, + "grad_norm": 0.9520178437232971, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 37373 + }, + { + "epoch": 0.37374, + "grad_norm": 0.8669770359992981, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 37374 + }, + { + "epoch": 0.37375, + "grad_norm": 0.8697465658187866, + "learning_rate": 0.003, + "loss": 4.002, + "step": 37375 + }, + { + "epoch": 0.37376, + "grad_norm": 0.8472004532814026, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 37376 + }, + { + "epoch": 0.37377, + "grad_norm": 0.9549016356468201, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 37377 + }, + { + "epoch": 0.37378, + "grad_norm": 1.0846900939941406, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 37378 + }, + { + "epoch": 0.37379, + "grad_norm": 0.9752529859542847, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 37379 + }, + { + "epoch": 0.3738, + "grad_norm": 1.0073683261871338, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 37380 + }, + { + "epoch": 0.37381, + "grad_norm": 1.0393390655517578, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 37381 + }, + { + "epoch": 0.37382, + "grad_norm": 1.0035136938095093, + "learning_rate": 0.003, + "loss": 4.013, + "step": 37382 + }, + { + "epoch": 0.37383, + "grad_norm": 0.9364776611328125, + "learning_rate": 0.003, + "loss": 3.985, + "step": 37383 + }, + { + "epoch": 0.37384, + "grad_norm": 0.8626548647880554, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 37384 + }, + { + "epoch": 0.37385, + "grad_norm": 1.1882567405700684, + "learning_rate": 0.003, + "loss": 3.976, + "step": 37385 + }, + { + "epoch": 0.37386, + "grad_norm": 1.211451768875122, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 37386 + }, + { + "epoch": 0.37387, + "grad_norm": 0.8003457188606262, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 37387 + }, + { + "epoch": 0.37388, + "grad_norm": 0.6270265579223633, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 37388 + }, + { + "epoch": 0.37389, + "grad_norm": 0.558799684047699, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 37389 + }, + { + "epoch": 0.3739, + "grad_norm": 0.5747310519218445, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 37390 + }, + { + "epoch": 0.37391, + "grad_norm": 0.5637305974960327, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 37391 + }, + { + "epoch": 0.37392, + "grad_norm": 0.5806949734687805, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 37392 + }, + { + "epoch": 0.37393, + "grad_norm": 0.5989385843276978, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 37393 + }, + { + "epoch": 0.37394, + "grad_norm": 0.6168765425682068, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 37394 + }, + { + "epoch": 0.37395, + "grad_norm": 0.5219328999519348, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 37395 + }, + { + "epoch": 0.37396, + "grad_norm": 0.5264004468917847, + "learning_rate": 0.003, + "loss": 3.984, + "step": 37396 + }, + { + "epoch": 0.37397, + "grad_norm": 0.6022440195083618, + "learning_rate": 0.003, + "loss": 3.969, + "step": 37397 + }, + { + "epoch": 0.37398, + "grad_norm": 0.6215300559997559, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 37398 + }, + { + "epoch": 0.37399, + "grad_norm": 0.8352498412132263, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 37399 + }, + { + "epoch": 0.374, + "grad_norm": 1.3675092458724976, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 37400 + }, + { + "epoch": 0.37401, + "grad_norm": 0.9243541359901428, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 37401 + }, + { + "epoch": 0.37402, + "grad_norm": 0.800549328327179, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 37402 + }, + { + "epoch": 0.37403, + "grad_norm": 0.8196024298667908, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 37403 + }, + { + "epoch": 0.37404, + "grad_norm": 0.9793288707733154, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 37404 + }, + { + "epoch": 0.37405, + "grad_norm": 1.3447322845458984, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 37405 + }, + { + "epoch": 0.37406, + "grad_norm": 0.7448618412017822, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 37406 + }, + { + "epoch": 0.37407, + "grad_norm": 0.7304264903068542, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 37407 + }, + { + "epoch": 0.37408, + "grad_norm": 0.7365788817405701, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 37408 + }, + { + "epoch": 0.37409, + "grad_norm": 0.9527022838592529, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 37409 + }, + { + "epoch": 0.3741, + "grad_norm": 1.2216615676879883, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 37410 + }, + { + "epoch": 0.37411, + "grad_norm": 0.7383777499198914, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 37411 + }, + { + "epoch": 0.37412, + "grad_norm": 0.8362016081809998, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 37412 + }, + { + "epoch": 0.37413, + "grad_norm": 1.0048776865005493, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 37413 + }, + { + "epoch": 0.37414, + "grad_norm": 1.3359872102737427, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 37414 + }, + { + "epoch": 0.37415, + "grad_norm": 0.8244677782058716, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 37415 + }, + { + "epoch": 0.37416, + "grad_norm": 0.8155438899993896, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 37416 + }, + { + "epoch": 0.37417, + "grad_norm": 0.8643932342529297, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 37417 + }, + { + "epoch": 0.37418, + "grad_norm": 0.9721914529800415, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 37418 + }, + { + "epoch": 0.37419, + "grad_norm": 1.1411436796188354, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 37419 + }, + { + "epoch": 0.3742, + "grad_norm": 0.8987215161323547, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 37420 + }, + { + "epoch": 0.37421, + "grad_norm": 0.7920986413955688, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 37421 + }, + { + "epoch": 0.37422, + "grad_norm": 0.8885725140571594, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 37422 + }, + { + "epoch": 0.37423, + "grad_norm": 1.04700767993927, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 37423 + }, + { + "epoch": 0.37424, + "grad_norm": 1.0358688831329346, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 37424 + }, + { + "epoch": 0.37425, + "grad_norm": 1.0642428398132324, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 37425 + }, + { + "epoch": 0.37426, + "grad_norm": 0.9424983859062195, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 37426 + }, + { + "epoch": 0.37427, + "grad_norm": 0.9800695776939392, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 37427 + }, + { + "epoch": 0.37428, + "grad_norm": 1.0488382577896118, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 37428 + }, + { + "epoch": 0.37429, + "grad_norm": 0.8998373746871948, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 37429 + }, + { + "epoch": 0.3743, + "grad_norm": 1.0811294317245483, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 37430 + }, + { + "epoch": 0.37431, + "grad_norm": 1.2393522262573242, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 37431 + }, + { + "epoch": 0.37432, + "grad_norm": 1.243838906288147, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 37432 + }, + { + "epoch": 0.37433, + "grad_norm": 0.8223748207092285, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 37433 + }, + { + "epoch": 0.37434, + "grad_norm": 0.8192471861839294, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 37434 + }, + { + "epoch": 0.37435, + "grad_norm": 1.089129090309143, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 37435 + }, + { + "epoch": 0.37436, + "grad_norm": 1.0435535907745361, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 37436 + }, + { + "epoch": 0.37437, + "grad_norm": 0.9975606203079224, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 37437 + }, + { + "epoch": 0.37438, + "grad_norm": 1.0850545167922974, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 37438 + }, + { + "epoch": 0.37439, + "grad_norm": 1.0872284173965454, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 37439 + }, + { + "epoch": 0.3744, + "grad_norm": 0.8244867324829102, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 37440 + }, + { + "epoch": 0.37441, + "grad_norm": 0.6780752539634705, + "learning_rate": 0.003, + "loss": 3.973, + "step": 37441 + }, + { + "epoch": 0.37442, + "grad_norm": 0.8225213885307312, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 37442 + }, + { + "epoch": 0.37443, + "grad_norm": 0.9670777916908264, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 37443 + }, + { + "epoch": 0.37444, + "grad_norm": 1.102015733718872, + "learning_rate": 0.003, + "loss": 3.9381, + "step": 37444 + }, + { + "epoch": 0.37445, + "grad_norm": 0.98354572057724, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 37445 + }, + { + "epoch": 0.37446, + "grad_norm": 0.9816216826438904, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 37446 + }, + { + "epoch": 0.37447, + "grad_norm": 0.9274555444717407, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 37447 + }, + { + "epoch": 0.37448, + "grad_norm": 0.9073542952537537, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 37448 + }, + { + "epoch": 0.37449, + "grad_norm": 0.9805872440338135, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 37449 + }, + { + "epoch": 0.3745, + "grad_norm": 0.856300413608551, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 37450 + }, + { + "epoch": 0.37451, + "grad_norm": 0.6811423301696777, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 37451 + }, + { + "epoch": 0.37452, + "grad_norm": 0.7547053098678589, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 37452 + }, + { + "epoch": 0.37453, + "grad_norm": 0.7894959449768066, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 37453 + }, + { + "epoch": 0.37454, + "grad_norm": 0.7593517303466797, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 37454 + }, + { + "epoch": 0.37455, + "grad_norm": 0.7611907124519348, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 37455 + }, + { + "epoch": 0.37456, + "grad_norm": 0.8421993255615234, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 37456 + }, + { + "epoch": 0.37457, + "grad_norm": 1.052202582359314, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 37457 + }, + { + "epoch": 0.37458, + "grad_norm": 1.1089606285095215, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 37458 + }, + { + "epoch": 0.37459, + "grad_norm": 0.8959953188896179, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 37459 + }, + { + "epoch": 0.3746, + "grad_norm": 0.9083190560340881, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 37460 + }, + { + "epoch": 0.37461, + "grad_norm": 0.9496435523033142, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 37461 + }, + { + "epoch": 0.37462, + "grad_norm": 0.9835770726203918, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 37462 + }, + { + "epoch": 0.37463, + "grad_norm": 0.986220121383667, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 37463 + }, + { + "epoch": 0.37464, + "grad_norm": 1.032047152519226, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 37464 + }, + { + "epoch": 0.37465, + "grad_norm": 0.9366419315338135, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 37465 + }, + { + "epoch": 0.37466, + "grad_norm": 0.9349289536476135, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 37466 + }, + { + "epoch": 0.37467, + "grad_norm": 1.1357851028442383, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 37467 + }, + { + "epoch": 0.37468, + "grad_norm": 1.117041826248169, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 37468 + }, + { + "epoch": 0.37469, + "grad_norm": 0.9923757910728455, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 37469 + }, + { + "epoch": 0.3747, + "grad_norm": 1.0925613641738892, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 37470 + }, + { + "epoch": 0.37471, + "grad_norm": 0.9507389664649963, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 37471 + }, + { + "epoch": 0.37472, + "grad_norm": 0.9758026599884033, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 37472 + }, + { + "epoch": 0.37473, + "grad_norm": 0.899285078048706, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 37473 + }, + { + "epoch": 0.37474, + "grad_norm": 0.9656832218170166, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 37474 + }, + { + "epoch": 0.37475, + "grad_norm": 1.1104059219360352, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 37475 + }, + { + "epoch": 0.37476, + "grad_norm": 1.0533696413040161, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 37476 + }, + { + "epoch": 0.37477, + "grad_norm": 0.8001932501792908, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 37477 + }, + { + "epoch": 0.37478, + "grad_norm": 0.6282013058662415, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 37478 + }, + { + "epoch": 0.37479, + "grad_norm": 0.9011516571044922, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 37479 + }, + { + "epoch": 0.3748, + "grad_norm": 1.1094073057174683, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 37480 + }, + { + "epoch": 0.37481, + "grad_norm": 0.7937501072883606, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 37481 + }, + { + "epoch": 0.37482, + "grad_norm": 0.748267412185669, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 37482 + }, + { + "epoch": 0.37483, + "grad_norm": 0.6949784159660339, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 37483 + }, + { + "epoch": 0.37484, + "grad_norm": 0.7025122046470642, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 37484 + }, + { + "epoch": 0.37485, + "grad_norm": 0.6665582656860352, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 37485 + }, + { + "epoch": 0.37486, + "grad_norm": 0.6180562973022461, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 37486 + }, + { + "epoch": 0.37487, + "grad_norm": 0.7463956475257874, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 37487 + }, + { + "epoch": 0.37488, + "grad_norm": 0.7595266103744507, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 37488 + }, + { + "epoch": 0.37489, + "grad_norm": 0.7904514074325562, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 37489 + }, + { + "epoch": 0.3749, + "grad_norm": 0.7725411653518677, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 37490 + }, + { + "epoch": 0.37491, + "grad_norm": 0.7965022921562195, + "learning_rate": 0.003, + "loss": 3.981, + "step": 37491 + }, + { + "epoch": 0.37492, + "grad_norm": 0.8607519865036011, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 37492 + }, + { + "epoch": 0.37493, + "grad_norm": 0.801102340221405, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 37493 + }, + { + "epoch": 0.37494, + "grad_norm": 0.7450896501541138, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 37494 + }, + { + "epoch": 0.37495, + "grad_norm": 0.9744552969932556, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 37495 + }, + { + "epoch": 0.37496, + "grad_norm": 1.5754023790359497, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 37496 + }, + { + "epoch": 0.37497, + "grad_norm": 0.6783145666122437, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 37497 + }, + { + "epoch": 0.37498, + "grad_norm": 0.6941813230514526, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 37498 + }, + { + "epoch": 0.37499, + "grad_norm": 0.7367433309555054, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 37499 + }, + { + "epoch": 0.375, + "grad_norm": 0.8430785536766052, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 37500 + }, + { + "epoch": 0.37501, + "grad_norm": 0.8760191798210144, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 37501 + }, + { + "epoch": 0.37502, + "grad_norm": 0.9606825113296509, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 37502 + }, + { + "epoch": 0.37503, + "grad_norm": 0.926093339920044, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 37503 + }, + { + "epoch": 0.37504, + "grad_norm": 0.9389569759368896, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 37504 + }, + { + "epoch": 0.37505, + "grad_norm": 0.9832819700241089, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 37505 + }, + { + "epoch": 0.37506, + "grad_norm": 1.0324383974075317, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 37506 + }, + { + "epoch": 0.37507, + "grad_norm": 0.984309196472168, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 37507 + }, + { + "epoch": 0.37508, + "grad_norm": 1.3738471269607544, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 37508 + }, + { + "epoch": 0.37509, + "grad_norm": 0.8461614847183228, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 37509 + }, + { + "epoch": 0.3751, + "grad_norm": 0.8205536603927612, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 37510 + }, + { + "epoch": 0.37511, + "grad_norm": 0.9615084528923035, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 37511 + }, + { + "epoch": 0.37512, + "grad_norm": 1.4070919752120972, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 37512 + }, + { + "epoch": 0.37513, + "grad_norm": 0.7658939957618713, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 37513 + }, + { + "epoch": 0.37514, + "grad_norm": 0.7785660028457642, + "learning_rate": 0.003, + "loss": 3.993, + "step": 37514 + }, + { + "epoch": 0.37515, + "grad_norm": 1.0231767892837524, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 37515 + }, + { + "epoch": 0.37516, + "grad_norm": 1.688747763633728, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 37516 + }, + { + "epoch": 0.37517, + "grad_norm": 0.6101855039596558, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 37517 + }, + { + "epoch": 0.37518, + "grad_norm": 0.8404138684272766, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 37518 + }, + { + "epoch": 0.37519, + "grad_norm": 0.7681109309196472, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 37519 + }, + { + "epoch": 0.3752, + "grad_norm": 0.8900719881057739, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 37520 + }, + { + "epoch": 0.37521, + "grad_norm": 1.20862877368927, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 37521 + }, + { + "epoch": 0.37522, + "grad_norm": 0.969092071056366, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 37522 + }, + { + "epoch": 0.37523, + "grad_norm": 1.1347569227218628, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 37523 + }, + { + "epoch": 0.37524, + "grad_norm": 0.9258004426956177, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 37524 + }, + { + "epoch": 0.37525, + "grad_norm": 0.822664201259613, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 37525 + }, + { + "epoch": 0.37526, + "grad_norm": 0.7932360172271729, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 37526 + }, + { + "epoch": 0.37527, + "grad_norm": 0.8466471433639526, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 37527 + }, + { + "epoch": 0.37528, + "grad_norm": 0.9858060479164124, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 37528 + }, + { + "epoch": 0.37529, + "grad_norm": 1.3854050636291504, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 37529 + }, + { + "epoch": 0.3753, + "grad_norm": 0.7769443988800049, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 37530 + }, + { + "epoch": 0.37531, + "grad_norm": 0.6981359720230103, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 37531 + }, + { + "epoch": 0.37532, + "grad_norm": 0.6555920839309692, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 37532 + }, + { + "epoch": 0.37533, + "grad_norm": 0.6878018975257874, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 37533 + }, + { + "epoch": 0.37534, + "grad_norm": 0.7868050932884216, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 37534 + }, + { + "epoch": 0.37535, + "grad_norm": 0.9991856217384338, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 37535 + }, + { + "epoch": 0.37536, + "grad_norm": 1.0454307794570923, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 37536 + }, + { + "epoch": 0.37537, + "grad_norm": 0.8835763931274414, + "learning_rate": 0.003, + "loss": 3.99, + "step": 37537 + }, + { + "epoch": 0.37538, + "grad_norm": 1.1159263849258423, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 37538 + }, + { + "epoch": 0.37539, + "grad_norm": 1.071101188659668, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 37539 + }, + { + "epoch": 0.3754, + "grad_norm": 1.0877571105957031, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 37540 + }, + { + "epoch": 0.37541, + "grad_norm": 0.8774957656860352, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 37541 + }, + { + "epoch": 0.37542, + "grad_norm": 0.6259041428565979, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 37542 + }, + { + "epoch": 0.37543, + "grad_norm": 0.6597819328308105, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 37543 + }, + { + "epoch": 0.37544, + "grad_norm": 0.7778456807136536, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 37544 + }, + { + "epoch": 0.37545, + "grad_norm": 0.862392008304596, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 37545 + }, + { + "epoch": 0.37546, + "grad_norm": 0.9366529583930969, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 37546 + }, + { + "epoch": 0.37547, + "grad_norm": 1.237081527709961, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 37547 + }, + { + "epoch": 0.37548, + "grad_norm": 0.9561769366264343, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 37548 + }, + { + "epoch": 0.37549, + "grad_norm": 0.8495546579360962, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 37549 + }, + { + "epoch": 0.3755, + "grad_norm": 0.8580581545829773, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 37550 + }, + { + "epoch": 0.37551, + "grad_norm": 0.9557216167449951, + "learning_rate": 0.003, + "loss": 3.983, + "step": 37551 + }, + { + "epoch": 0.37552, + "grad_norm": 0.9302075505256653, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 37552 + }, + { + "epoch": 0.37553, + "grad_norm": 0.8221277594566345, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 37553 + }, + { + "epoch": 0.37554, + "grad_norm": 0.7349765300750732, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 37554 + }, + { + "epoch": 0.37555, + "grad_norm": 0.8540971279144287, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 37555 + }, + { + "epoch": 0.37556, + "grad_norm": 1.0125510692596436, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 37556 + }, + { + "epoch": 0.37557, + "grad_norm": 0.9029754996299744, + "learning_rate": 0.003, + "loss": 3.965, + "step": 37557 + }, + { + "epoch": 0.37558, + "grad_norm": 0.7449281215667725, + "learning_rate": 0.003, + "loss": 3.979, + "step": 37558 + }, + { + "epoch": 0.37559, + "grad_norm": 0.670850396156311, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 37559 + }, + { + "epoch": 0.3756, + "grad_norm": 0.660923421382904, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 37560 + }, + { + "epoch": 0.37561, + "grad_norm": 0.7694377303123474, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 37561 + }, + { + "epoch": 0.37562, + "grad_norm": 1.1394778490066528, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 37562 + }, + { + "epoch": 0.37563, + "grad_norm": 1.1435519456863403, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 37563 + }, + { + "epoch": 0.37564, + "grad_norm": 0.907808780670166, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 37564 + }, + { + "epoch": 0.37565, + "grad_norm": 0.9638621211051941, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 37565 + }, + { + "epoch": 0.37566, + "grad_norm": 1.064929723739624, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 37566 + }, + { + "epoch": 0.37567, + "grad_norm": 0.905001163482666, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 37567 + }, + { + "epoch": 0.37568, + "grad_norm": 0.8209678530693054, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 37568 + }, + { + "epoch": 0.37569, + "grad_norm": 0.7959985733032227, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 37569 + }, + { + "epoch": 0.3757, + "grad_norm": 0.794310450553894, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 37570 + }, + { + "epoch": 0.37571, + "grad_norm": 0.8471026420593262, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 37571 + }, + { + "epoch": 0.37572, + "grad_norm": 0.7901346683502197, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 37572 + }, + { + "epoch": 0.37573, + "grad_norm": 0.7642295956611633, + "learning_rate": 0.003, + "loss": 3.978, + "step": 37573 + }, + { + "epoch": 0.37574, + "grad_norm": 0.7260158061981201, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 37574 + }, + { + "epoch": 0.37575, + "grad_norm": 0.8484503030776978, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 37575 + }, + { + "epoch": 0.37576, + "grad_norm": 1.2627131938934326, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 37576 + }, + { + "epoch": 0.37577, + "grad_norm": 0.9237498641014099, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 37577 + }, + { + "epoch": 0.37578, + "grad_norm": 0.9443372488021851, + "learning_rate": 0.003, + "loss": 3.99, + "step": 37578 + }, + { + "epoch": 0.37579, + "grad_norm": 1.1377174854278564, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 37579 + }, + { + "epoch": 0.3758, + "grad_norm": 0.9088180661201477, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 37580 + }, + { + "epoch": 0.37581, + "grad_norm": 0.8206301331520081, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 37581 + }, + { + "epoch": 0.37582, + "grad_norm": 0.8452156782150269, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 37582 + }, + { + "epoch": 0.37583, + "grad_norm": 0.8370166420936584, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 37583 + }, + { + "epoch": 0.37584, + "grad_norm": 0.818200409412384, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 37584 + }, + { + "epoch": 0.37585, + "grad_norm": 0.8063117265701294, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 37585 + }, + { + "epoch": 0.37586, + "grad_norm": 0.9775764346122742, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 37586 + }, + { + "epoch": 0.37587, + "grad_norm": 1.133955955505371, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 37587 + }, + { + "epoch": 0.37588, + "grad_norm": 0.9646317958831787, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 37588 + }, + { + "epoch": 0.37589, + "grad_norm": 0.8230567574501038, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 37589 + }, + { + "epoch": 0.3759, + "grad_norm": 0.8768205642700195, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 37590 + }, + { + "epoch": 0.37591, + "grad_norm": 1.1569238901138306, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 37591 + }, + { + "epoch": 0.37592, + "grad_norm": 1.0452015399932861, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 37592 + }, + { + "epoch": 0.37593, + "grad_norm": 0.851327657699585, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 37593 + }, + { + "epoch": 0.37594, + "grad_norm": 0.6873839497566223, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 37594 + }, + { + "epoch": 0.37595, + "grad_norm": 0.7755624055862427, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 37595 + }, + { + "epoch": 0.37596, + "grad_norm": 0.9845823049545288, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 37596 + }, + { + "epoch": 0.37597, + "grad_norm": 0.9429706931114197, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 37597 + }, + { + "epoch": 0.37598, + "grad_norm": 0.956299364566803, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 37598 + }, + { + "epoch": 0.37599, + "grad_norm": 0.9145247936248779, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 37599 + }, + { + "epoch": 0.376, + "grad_norm": 0.7707775831222534, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 37600 + }, + { + "epoch": 0.37601, + "grad_norm": 0.7101722359657288, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 37601 + }, + { + "epoch": 0.37602, + "grad_norm": 0.7581698298454285, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 37602 + }, + { + "epoch": 0.37603, + "grad_norm": 0.8764510154724121, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 37603 + }, + { + "epoch": 0.37604, + "grad_norm": 0.8996073007583618, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 37604 + }, + { + "epoch": 0.37605, + "grad_norm": 0.9428448677062988, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 37605 + }, + { + "epoch": 0.37606, + "grad_norm": 1.131577491760254, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 37606 + }, + { + "epoch": 0.37607, + "grad_norm": 0.8678311705589294, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 37607 + }, + { + "epoch": 0.37608, + "grad_norm": 0.7549024224281311, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 37608 + }, + { + "epoch": 0.37609, + "grad_norm": 0.8617907762527466, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 37609 + }, + { + "epoch": 0.3761, + "grad_norm": 0.8960815072059631, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 37610 + }, + { + "epoch": 0.37611, + "grad_norm": 0.8849450349807739, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 37611 + }, + { + "epoch": 0.37612, + "grad_norm": 0.9541747570037842, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 37612 + }, + { + "epoch": 0.37613, + "grad_norm": 1.1463629007339478, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 37613 + }, + { + "epoch": 0.37614, + "grad_norm": 1.095234751701355, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 37614 + }, + { + "epoch": 0.37615, + "grad_norm": 0.9141149520874023, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 37615 + }, + { + "epoch": 0.37616, + "grad_norm": 0.8038759827613831, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 37616 + }, + { + "epoch": 0.37617, + "grad_norm": 0.9021785259246826, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 37617 + }, + { + "epoch": 0.37618, + "grad_norm": 1.2830554246902466, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 37618 + }, + { + "epoch": 0.37619, + "grad_norm": 0.9640439748764038, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 37619 + }, + { + "epoch": 0.3762, + "grad_norm": 0.7380743622779846, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 37620 + }, + { + "epoch": 0.37621, + "grad_norm": 0.8011260032653809, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 37621 + }, + { + "epoch": 0.37622, + "grad_norm": 1.0875385999679565, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 37622 + }, + { + "epoch": 0.37623, + "grad_norm": 1.1626557111740112, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 37623 + }, + { + "epoch": 0.37624, + "grad_norm": 0.7633092999458313, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 37624 + }, + { + "epoch": 0.37625, + "grad_norm": 0.9208902716636658, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 37625 + }, + { + "epoch": 0.37626, + "grad_norm": 0.9060560464859009, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 37626 + }, + { + "epoch": 0.37627, + "grad_norm": 0.8252425789833069, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 37627 + }, + { + "epoch": 0.37628, + "grad_norm": 0.8232213258743286, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 37628 + }, + { + "epoch": 0.37629, + "grad_norm": 0.9696314334869385, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 37629 + }, + { + "epoch": 0.3763, + "grad_norm": 1.0703785419464111, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 37630 + }, + { + "epoch": 0.37631, + "grad_norm": 0.9145122766494751, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 37631 + }, + { + "epoch": 0.37632, + "grad_norm": 1.0080674886703491, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 37632 + }, + { + "epoch": 0.37633, + "grad_norm": 1.2950806617736816, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 37633 + }, + { + "epoch": 0.37634, + "grad_norm": 0.9130512475967407, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 37634 + }, + { + "epoch": 0.37635, + "grad_norm": 0.8694928884506226, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 37635 + }, + { + "epoch": 0.37636, + "grad_norm": 1.0594449043273926, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 37636 + }, + { + "epoch": 0.37637, + "grad_norm": 0.9504410028457642, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 37637 + }, + { + "epoch": 0.37638, + "grad_norm": 0.9365146160125732, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 37638 + }, + { + "epoch": 0.37639, + "grad_norm": 1.0811723470687866, + "learning_rate": 0.003, + "loss": 3.966, + "step": 37639 + }, + { + "epoch": 0.3764, + "grad_norm": 1.0241540670394897, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 37640 + }, + { + "epoch": 0.37641, + "grad_norm": 0.8403652906417847, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 37641 + }, + { + "epoch": 0.37642, + "grad_norm": 0.861466109752655, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 37642 + }, + { + "epoch": 0.37643, + "grad_norm": 0.7803069949150085, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 37643 + }, + { + "epoch": 0.37644, + "grad_norm": 0.7973328828811646, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 37644 + }, + { + "epoch": 0.37645, + "grad_norm": 0.9210329055786133, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 37645 + }, + { + "epoch": 0.37646, + "grad_norm": 1.3243821859359741, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 37646 + }, + { + "epoch": 0.37647, + "grad_norm": 0.8832153677940369, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 37647 + }, + { + "epoch": 0.37648, + "grad_norm": 0.7719971537590027, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 37648 + }, + { + "epoch": 0.37649, + "grad_norm": 0.7206680774688721, + "learning_rate": 0.003, + "loss": 3.998, + "step": 37649 + }, + { + "epoch": 0.3765, + "grad_norm": 0.6430049538612366, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 37650 + }, + { + "epoch": 0.37651, + "grad_norm": 0.7180471420288086, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 37651 + }, + { + "epoch": 0.37652, + "grad_norm": 0.7679041028022766, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 37652 + }, + { + "epoch": 0.37653, + "grad_norm": 0.8148255944252014, + "learning_rate": 0.003, + "loss": 3.962, + "step": 37653 + }, + { + "epoch": 0.37654, + "grad_norm": 0.873351514339447, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 37654 + }, + { + "epoch": 0.37655, + "grad_norm": 0.8900653719902039, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 37655 + }, + { + "epoch": 0.37656, + "grad_norm": 1.0414491891860962, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 37656 + }, + { + "epoch": 0.37657, + "grad_norm": 1.001498818397522, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 37657 + }, + { + "epoch": 0.37658, + "grad_norm": 1.1219011545181274, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 37658 + }, + { + "epoch": 0.37659, + "grad_norm": 0.708319902420044, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 37659 + }, + { + "epoch": 0.3766, + "grad_norm": 0.5320887565612793, + "learning_rate": 0.003, + "loss": 3.9388, + "step": 37660 + }, + { + "epoch": 0.37661, + "grad_norm": 0.5280128717422485, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 37661 + }, + { + "epoch": 0.37662, + "grad_norm": 0.49779897928237915, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 37662 + }, + { + "epoch": 0.37663, + "grad_norm": 0.5884284377098083, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 37663 + }, + { + "epoch": 0.37664, + "grad_norm": 0.7632103562355042, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 37664 + }, + { + "epoch": 0.37665, + "grad_norm": 1.0072052478790283, + "learning_rate": 0.003, + "loss": 3.959, + "step": 37665 + }, + { + "epoch": 0.37666, + "grad_norm": 1.1176990270614624, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 37666 + }, + { + "epoch": 0.37667, + "grad_norm": 0.7455126643180847, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 37667 + }, + { + "epoch": 0.37668, + "grad_norm": 0.5999515056610107, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 37668 + }, + { + "epoch": 0.37669, + "grad_norm": 0.5956615209579468, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 37669 + }, + { + "epoch": 0.3767, + "grad_norm": 0.613898754119873, + "learning_rate": 0.003, + "loss": 3.9454, + "step": 37670 + }, + { + "epoch": 0.37671, + "grad_norm": 0.6640515923500061, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 37671 + }, + { + "epoch": 0.37672, + "grad_norm": 0.6646856069564819, + "learning_rate": 0.003, + "loss": 3.959, + "step": 37672 + }, + { + "epoch": 0.37673, + "grad_norm": 0.7006389498710632, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 37673 + }, + { + "epoch": 0.37674, + "grad_norm": 0.7886336445808411, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 37674 + }, + { + "epoch": 0.37675, + "grad_norm": 0.8202678561210632, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 37675 + }, + { + "epoch": 0.37676, + "grad_norm": 1.0471762418746948, + "learning_rate": 0.003, + "loss": 3.9459, + "step": 37676 + }, + { + "epoch": 0.37677, + "grad_norm": 1.368546485900879, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 37677 + }, + { + "epoch": 0.37678, + "grad_norm": 0.7511842250823975, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 37678 + }, + { + "epoch": 0.37679, + "grad_norm": 0.8158900141716003, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 37679 + }, + { + "epoch": 0.3768, + "grad_norm": 1.0191045999526978, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 37680 + }, + { + "epoch": 0.37681, + "grad_norm": 1.4057226181030273, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 37681 + }, + { + "epoch": 0.37682, + "grad_norm": 0.6877594590187073, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 37682 + }, + { + "epoch": 0.37683, + "grad_norm": 0.8662709593772888, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 37683 + }, + { + "epoch": 0.37684, + "grad_norm": 0.9685482382774353, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 37684 + }, + { + "epoch": 0.37685, + "grad_norm": 1.091866374015808, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 37685 + }, + { + "epoch": 0.37686, + "grad_norm": 0.9332873225212097, + "learning_rate": 0.003, + "loss": 3.979, + "step": 37686 + }, + { + "epoch": 0.37687, + "grad_norm": 0.9363253116607666, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 37687 + }, + { + "epoch": 0.37688, + "grad_norm": 0.8647321462631226, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 37688 + }, + { + "epoch": 0.37689, + "grad_norm": 0.8590725064277649, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 37689 + }, + { + "epoch": 0.3769, + "grad_norm": 0.9224189519882202, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 37690 + }, + { + "epoch": 0.37691, + "grad_norm": 1.0359742641448975, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 37691 + }, + { + "epoch": 0.37692, + "grad_norm": 1.3030500411987305, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 37692 + }, + { + "epoch": 0.37693, + "grad_norm": 1.1023926734924316, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 37693 + }, + { + "epoch": 0.37694, + "grad_norm": 0.9175246953964233, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 37694 + }, + { + "epoch": 0.37695, + "grad_norm": 0.860657274723053, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 37695 + }, + { + "epoch": 0.37696, + "grad_norm": 0.8402391076087952, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 37696 + }, + { + "epoch": 0.37697, + "grad_norm": 0.8212873935699463, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 37697 + }, + { + "epoch": 0.37698, + "grad_norm": 1.1040759086608887, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 37698 + }, + { + "epoch": 0.37699, + "grad_norm": 0.9835035800933838, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 37699 + }, + { + "epoch": 0.377, + "grad_norm": 0.9440668225288391, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 37700 + }, + { + "epoch": 0.37701, + "grad_norm": 1.0800548791885376, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 37701 + }, + { + "epoch": 0.37702, + "grad_norm": 1.6097608804702759, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 37702 + }, + { + "epoch": 0.37703, + "grad_norm": 0.5953732132911682, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 37703 + }, + { + "epoch": 0.37704, + "grad_norm": 1.1549429893493652, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 37704 + }, + { + "epoch": 0.37705, + "grad_norm": 1.0096031427383423, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 37705 + }, + { + "epoch": 0.37706, + "grad_norm": 0.9868764877319336, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 37706 + }, + { + "epoch": 0.37707, + "grad_norm": 1.101157546043396, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 37707 + }, + { + "epoch": 0.37708, + "grad_norm": 0.9946410655975342, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 37708 + }, + { + "epoch": 0.37709, + "grad_norm": 1.0641334056854248, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 37709 + }, + { + "epoch": 0.3771, + "grad_norm": 0.977350652217865, + "learning_rate": 0.003, + "loss": 3.977, + "step": 37710 + }, + { + "epoch": 0.37711, + "grad_norm": 0.9273620247840881, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 37711 + }, + { + "epoch": 0.37712, + "grad_norm": 0.9628018140792847, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 37712 + }, + { + "epoch": 0.37713, + "grad_norm": 1.0149317979812622, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 37713 + }, + { + "epoch": 0.37714, + "grad_norm": 0.9316555261611938, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 37714 + }, + { + "epoch": 0.37715, + "grad_norm": 1.1689215898513794, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 37715 + }, + { + "epoch": 0.37716, + "grad_norm": 0.8206871747970581, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 37716 + }, + { + "epoch": 0.37717, + "grad_norm": 0.8588489890098572, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 37717 + }, + { + "epoch": 0.37718, + "grad_norm": 0.9787280559539795, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 37718 + }, + { + "epoch": 0.37719, + "grad_norm": 1.0925178527832031, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 37719 + }, + { + "epoch": 0.3772, + "grad_norm": 1.111255168914795, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 37720 + }, + { + "epoch": 0.37721, + "grad_norm": 0.8474717140197754, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 37721 + }, + { + "epoch": 0.37722, + "grad_norm": 0.7243515253067017, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 37722 + }, + { + "epoch": 0.37723, + "grad_norm": 1.0017725229263306, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 37723 + }, + { + "epoch": 0.37724, + "grad_norm": 1.2530094385147095, + "learning_rate": 0.003, + "loss": 3.998, + "step": 37724 + }, + { + "epoch": 0.37725, + "grad_norm": 0.7582256197929382, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 37725 + }, + { + "epoch": 0.37726, + "grad_norm": 0.6675663590431213, + "learning_rate": 0.003, + "loss": 3.945, + "step": 37726 + }, + { + "epoch": 0.37727, + "grad_norm": 0.6812368035316467, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 37727 + }, + { + "epoch": 0.37728, + "grad_norm": 0.5415281057357788, + "learning_rate": 0.003, + "loss": 3.9471, + "step": 37728 + }, + { + "epoch": 0.37729, + "grad_norm": 0.5611748695373535, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 37729 + }, + { + "epoch": 0.3773, + "grad_norm": 0.5425733327865601, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 37730 + }, + { + "epoch": 0.37731, + "grad_norm": 0.5994439125061035, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 37731 + }, + { + "epoch": 0.37732, + "grad_norm": 0.6195065379142761, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 37732 + }, + { + "epoch": 0.37733, + "grad_norm": 0.6765417456626892, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 37733 + }, + { + "epoch": 0.37734, + "grad_norm": 0.8177983164787292, + "learning_rate": 0.003, + "loss": 3.9383, + "step": 37734 + }, + { + "epoch": 0.37735, + "grad_norm": 1.1105151176452637, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 37735 + }, + { + "epoch": 0.37736, + "grad_norm": 1.1332051753997803, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 37736 + }, + { + "epoch": 0.37737, + "grad_norm": 0.6992563009262085, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 37737 + }, + { + "epoch": 0.37738, + "grad_norm": 0.6344037652015686, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 37738 + }, + { + "epoch": 0.37739, + "grad_norm": 0.7558230757713318, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 37739 + }, + { + "epoch": 0.3774, + "grad_norm": 0.8564885258674622, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 37740 + }, + { + "epoch": 0.37741, + "grad_norm": 0.9973117113113403, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 37741 + }, + { + "epoch": 0.37742, + "grad_norm": 0.8416485786437988, + "learning_rate": 0.003, + "loss": 3.987, + "step": 37742 + }, + { + "epoch": 0.37743, + "grad_norm": 0.7659921050071716, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 37743 + }, + { + "epoch": 0.37744, + "grad_norm": 0.8619877099990845, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 37744 + }, + { + "epoch": 0.37745, + "grad_norm": 0.8906664848327637, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 37745 + }, + { + "epoch": 0.37746, + "grad_norm": 0.9324166774749756, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 37746 + }, + { + "epoch": 0.37747, + "grad_norm": 1.0567771196365356, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 37747 + }, + { + "epoch": 0.37748, + "grad_norm": 1.1253832578659058, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 37748 + }, + { + "epoch": 0.37749, + "grad_norm": 0.8254684805870056, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 37749 + }, + { + "epoch": 0.3775, + "grad_norm": 0.7231931090354919, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 37750 + }, + { + "epoch": 0.37751, + "grad_norm": 0.784974992275238, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 37751 + }, + { + "epoch": 0.37752, + "grad_norm": 0.6996989250183105, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 37752 + }, + { + "epoch": 0.37753, + "grad_norm": 0.695166826248169, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 37753 + }, + { + "epoch": 0.37754, + "grad_norm": 0.8200963139533997, + "learning_rate": 0.003, + "loss": 3.955, + "step": 37754 + }, + { + "epoch": 0.37755, + "grad_norm": 1.1728752851486206, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 37755 + }, + { + "epoch": 0.37756, + "grad_norm": 1.2449055910110474, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 37756 + }, + { + "epoch": 0.37757, + "grad_norm": 0.6749657988548279, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 37757 + }, + { + "epoch": 0.37758, + "grad_norm": 0.656760036945343, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 37758 + }, + { + "epoch": 0.37759, + "grad_norm": 0.7068572640419006, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 37759 + }, + { + "epoch": 0.3776, + "grad_norm": 0.8348527550697327, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 37760 + }, + { + "epoch": 0.37761, + "grad_norm": 0.7228893637657166, + "learning_rate": 0.003, + "loss": 3.998, + "step": 37761 + }, + { + "epoch": 0.37762, + "grad_norm": 0.7709391713142395, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 37762 + }, + { + "epoch": 0.37763, + "grad_norm": 0.7661150097846985, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 37763 + }, + { + "epoch": 0.37764, + "grad_norm": 0.8891463279724121, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 37764 + }, + { + "epoch": 0.37765, + "grad_norm": 1.06771981716156, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 37765 + }, + { + "epoch": 0.37766, + "grad_norm": 1.1125109195709229, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 37766 + }, + { + "epoch": 0.37767, + "grad_norm": 0.9836359620094299, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 37767 + }, + { + "epoch": 0.37768, + "grad_norm": 1.1145200729370117, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 37768 + }, + { + "epoch": 0.37769, + "grad_norm": 1.0462067127227783, + "learning_rate": 0.003, + "loss": 3.983, + "step": 37769 + }, + { + "epoch": 0.3777, + "grad_norm": 1.069855809211731, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 37770 + }, + { + "epoch": 0.37771, + "grad_norm": 0.9622517228126526, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 37771 + }, + { + "epoch": 0.37772, + "grad_norm": 0.9783876538276672, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 37772 + }, + { + "epoch": 0.37773, + "grad_norm": 1.1410831212997437, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 37773 + }, + { + "epoch": 0.37774, + "grad_norm": 0.8275396227836609, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 37774 + }, + { + "epoch": 0.37775, + "grad_norm": 0.7980839014053345, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 37775 + }, + { + "epoch": 0.37776, + "grad_norm": 0.9026121497154236, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 37776 + }, + { + "epoch": 0.37777, + "grad_norm": 1.1189794540405273, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 37777 + }, + { + "epoch": 0.37778, + "grad_norm": 1.1391113996505737, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 37778 + }, + { + "epoch": 0.37779, + "grad_norm": 1.0122357606887817, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 37779 + }, + { + "epoch": 0.3778, + "grad_norm": 1.1122875213623047, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 37780 + }, + { + "epoch": 0.37781, + "grad_norm": 0.816095769405365, + "learning_rate": 0.003, + "loss": 3.99, + "step": 37781 + }, + { + "epoch": 0.37782, + "grad_norm": 0.652189314365387, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 37782 + }, + { + "epoch": 0.37783, + "grad_norm": 0.6071068048477173, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 37783 + }, + { + "epoch": 0.37784, + "grad_norm": 0.558743417263031, + "learning_rate": 0.003, + "loss": 3.963, + "step": 37784 + }, + { + "epoch": 0.37785, + "grad_norm": 0.665133535861969, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 37785 + }, + { + "epoch": 0.37786, + "grad_norm": 0.7404493093490601, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 37786 + }, + { + "epoch": 0.37787, + "grad_norm": 1.014377236366272, + "learning_rate": 0.003, + "loss": 3.9338, + "step": 37787 + }, + { + "epoch": 0.37788, + "grad_norm": 1.3855652809143066, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 37788 + }, + { + "epoch": 0.37789, + "grad_norm": 0.6115443706512451, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 37789 + }, + { + "epoch": 0.3779, + "grad_norm": 0.6232308745384216, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 37790 + }, + { + "epoch": 0.37791, + "grad_norm": 0.7556326985359192, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 37791 + }, + { + "epoch": 0.37792, + "grad_norm": 0.8537203669548035, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 37792 + }, + { + "epoch": 0.37793, + "grad_norm": 1.1703099012374878, + "learning_rate": 0.003, + "loss": 3.988, + "step": 37793 + }, + { + "epoch": 0.37794, + "grad_norm": 1.000394582748413, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 37794 + }, + { + "epoch": 0.37795, + "grad_norm": 0.8797861933708191, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 37795 + }, + { + "epoch": 0.37796, + "grad_norm": 0.7498100996017456, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 37796 + }, + { + "epoch": 0.37797, + "grad_norm": 0.7677626609802246, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 37797 + }, + { + "epoch": 0.37798, + "grad_norm": 0.6333097219467163, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 37798 + }, + { + "epoch": 0.37799, + "grad_norm": 0.790778636932373, + "learning_rate": 0.003, + "loss": 3.967, + "step": 37799 + }, + { + "epoch": 0.378, + "grad_norm": 1.0850595235824585, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 37800 + }, + { + "epoch": 0.37801, + "grad_norm": 0.9114550948143005, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 37801 + }, + { + "epoch": 0.37802, + "grad_norm": 1.0265401601791382, + "learning_rate": 0.003, + "loss": 3.965, + "step": 37802 + }, + { + "epoch": 0.37803, + "grad_norm": 1.1373121738433838, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 37803 + }, + { + "epoch": 0.37804, + "grad_norm": 1.134454369544983, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 37804 + }, + { + "epoch": 0.37805, + "grad_norm": 0.9155842065811157, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 37805 + }, + { + "epoch": 0.37806, + "grad_norm": 0.9852800965309143, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 37806 + }, + { + "epoch": 0.37807, + "grad_norm": 1.2567791938781738, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 37807 + }, + { + "epoch": 0.37808, + "grad_norm": 1.1182547807693481, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 37808 + }, + { + "epoch": 0.37809, + "grad_norm": 0.9299301505088806, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 37809 + }, + { + "epoch": 0.3781, + "grad_norm": 0.9433020353317261, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 37810 + }, + { + "epoch": 0.37811, + "grad_norm": 1.2887420654296875, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 37811 + }, + { + "epoch": 0.37812, + "grad_norm": 0.743301272392273, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 37812 + }, + { + "epoch": 0.37813, + "grad_norm": 0.5887417197227478, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 37813 + }, + { + "epoch": 0.37814, + "grad_norm": 0.7386378645896912, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 37814 + }, + { + "epoch": 0.37815, + "grad_norm": 0.8772439956665039, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 37815 + }, + { + "epoch": 0.37816, + "grad_norm": 1.0359402894973755, + "learning_rate": 0.003, + "loss": 3.986, + "step": 37816 + }, + { + "epoch": 0.37817, + "grad_norm": 1.1120389699935913, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 37817 + }, + { + "epoch": 0.37818, + "grad_norm": 0.9956570863723755, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 37818 + }, + { + "epoch": 0.37819, + "grad_norm": 0.9823620915412903, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 37819 + }, + { + "epoch": 0.3782, + "grad_norm": 1.0690981149673462, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 37820 + }, + { + "epoch": 0.37821, + "grad_norm": 0.8875265717506409, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 37821 + }, + { + "epoch": 0.37822, + "grad_norm": 0.720871090888977, + "learning_rate": 0.003, + "loss": 4.0, + "step": 37822 + }, + { + "epoch": 0.37823, + "grad_norm": 0.7054978013038635, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 37823 + }, + { + "epoch": 0.37824, + "grad_norm": 0.6757338643074036, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 37824 + }, + { + "epoch": 0.37825, + "grad_norm": 0.7682034373283386, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 37825 + }, + { + "epoch": 0.37826, + "grad_norm": 0.9007710814476013, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 37826 + }, + { + "epoch": 0.37827, + "grad_norm": 0.9967650175094604, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 37827 + }, + { + "epoch": 0.37828, + "grad_norm": 1.2019203901290894, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 37828 + }, + { + "epoch": 0.37829, + "grad_norm": 0.7060468792915344, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 37829 + }, + { + "epoch": 0.3783, + "grad_norm": 0.6043797135353088, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 37830 + }, + { + "epoch": 0.37831, + "grad_norm": 0.6076478362083435, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 37831 + }, + { + "epoch": 0.37832, + "grad_norm": 0.7258281707763672, + "learning_rate": 0.003, + "loss": 3.978, + "step": 37832 + }, + { + "epoch": 0.37833, + "grad_norm": 0.8251006603240967, + "learning_rate": 0.003, + "loss": 3.99, + "step": 37833 + }, + { + "epoch": 0.37834, + "grad_norm": 0.9191389083862305, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 37834 + }, + { + "epoch": 0.37835, + "grad_norm": 0.8611044883728027, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 37835 + }, + { + "epoch": 0.37836, + "grad_norm": 0.7890713214874268, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 37836 + }, + { + "epoch": 0.37837, + "grad_norm": 0.8478959798812866, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 37837 + }, + { + "epoch": 0.37838, + "grad_norm": 0.9024728536605835, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 37838 + }, + { + "epoch": 0.37839, + "grad_norm": 0.8611441254615784, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 37839 + }, + { + "epoch": 0.3784, + "grad_norm": 0.7273645997047424, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 37840 + }, + { + "epoch": 0.37841, + "grad_norm": 0.712873101234436, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 37841 + }, + { + "epoch": 0.37842, + "grad_norm": 0.7557390928268433, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 37842 + }, + { + "epoch": 0.37843, + "grad_norm": 0.935295045375824, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 37843 + }, + { + "epoch": 0.37844, + "grad_norm": 1.2952524423599243, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 37844 + }, + { + "epoch": 0.37845, + "grad_norm": 0.8081202507019043, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 37845 + }, + { + "epoch": 0.37846, + "grad_norm": 0.8616557717323303, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 37846 + }, + { + "epoch": 0.37847, + "grad_norm": 0.8599599003791809, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 37847 + }, + { + "epoch": 0.37848, + "grad_norm": 0.9448103308677673, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 37848 + }, + { + "epoch": 0.37849, + "grad_norm": 1.0068132877349854, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 37849 + }, + { + "epoch": 0.3785, + "grad_norm": 1.0153762102127075, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 37850 + }, + { + "epoch": 0.37851, + "grad_norm": 1.1616793870925903, + "learning_rate": 0.003, + "loss": 4.032, + "step": 37851 + }, + { + "epoch": 0.37852, + "grad_norm": 1.0763707160949707, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 37852 + }, + { + "epoch": 0.37853, + "grad_norm": 1.4277452230453491, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 37853 + }, + { + "epoch": 0.37854, + "grad_norm": 0.7707440853118896, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 37854 + }, + { + "epoch": 0.37855, + "grad_norm": 0.8035808801651001, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 37855 + }, + { + "epoch": 0.37856, + "grad_norm": 1.012397289276123, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 37856 + }, + { + "epoch": 0.37857, + "grad_norm": 1.1474157571792603, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 37857 + }, + { + "epoch": 0.37858, + "grad_norm": 0.8909511566162109, + "learning_rate": 0.003, + "loss": 3.981, + "step": 37858 + }, + { + "epoch": 0.37859, + "grad_norm": 0.9946690201759338, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 37859 + }, + { + "epoch": 0.3786, + "grad_norm": 1.1623259782791138, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 37860 + }, + { + "epoch": 0.37861, + "grad_norm": 0.8994956016540527, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 37861 + }, + { + "epoch": 0.37862, + "grad_norm": 1.0064650774002075, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 37862 + }, + { + "epoch": 0.37863, + "grad_norm": 1.1358994245529175, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 37863 + }, + { + "epoch": 0.37864, + "grad_norm": 0.7946521043777466, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 37864 + }, + { + "epoch": 0.37865, + "grad_norm": 0.7938322424888611, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 37865 + }, + { + "epoch": 0.37866, + "grad_norm": 0.811851441860199, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 37866 + }, + { + "epoch": 0.37867, + "grad_norm": 0.8259595036506653, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 37867 + }, + { + "epoch": 0.37868, + "grad_norm": 0.7189704179763794, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 37868 + }, + { + "epoch": 0.37869, + "grad_norm": 0.8929141759872437, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 37869 + }, + { + "epoch": 0.3787, + "grad_norm": 0.9622849822044373, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 37870 + }, + { + "epoch": 0.37871, + "grad_norm": 1.0491725206375122, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 37871 + }, + { + "epoch": 0.37872, + "grad_norm": 1.1116244792938232, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 37872 + }, + { + "epoch": 0.37873, + "grad_norm": 0.7869846820831299, + "learning_rate": 0.003, + "loss": 3.973, + "step": 37873 + }, + { + "epoch": 0.37874, + "grad_norm": 0.7616906762123108, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 37874 + }, + { + "epoch": 0.37875, + "grad_norm": 0.7650604248046875, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 37875 + }, + { + "epoch": 0.37876, + "grad_norm": 0.865782618522644, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 37876 + }, + { + "epoch": 0.37877, + "grad_norm": 1.0363314151763916, + "learning_rate": 0.003, + "loss": 3.989, + "step": 37877 + }, + { + "epoch": 0.37878, + "grad_norm": 1.042575716972351, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 37878 + }, + { + "epoch": 0.37879, + "grad_norm": 1.0384796857833862, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 37879 + }, + { + "epoch": 0.3788, + "grad_norm": 0.9919580221176147, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 37880 + }, + { + "epoch": 0.37881, + "grad_norm": 1.151999592781067, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 37881 + }, + { + "epoch": 0.37882, + "grad_norm": 0.9414048194885254, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 37882 + }, + { + "epoch": 0.37883, + "grad_norm": 0.9266946315765381, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 37883 + }, + { + "epoch": 0.37884, + "grad_norm": 0.9956357479095459, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 37884 + }, + { + "epoch": 0.37885, + "grad_norm": 1.11869215965271, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 37885 + }, + { + "epoch": 0.37886, + "grad_norm": 1.0329622030258179, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 37886 + }, + { + "epoch": 0.37887, + "grad_norm": 0.8483553528785706, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 37887 + }, + { + "epoch": 0.37888, + "grad_norm": 0.8886428475379944, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 37888 + }, + { + "epoch": 0.37889, + "grad_norm": 0.7904688119888306, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 37889 + }, + { + "epoch": 0.3789, + "grad_norm": 0.876340389251709, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 37890 + }, + { + "epoch": 0.37891, + "grad_norm": 0.8917452692985535, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 37891 + }, + { + "epoch": 0.37892, + "grad_norm": 0.8983749151229858, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 37892 + }, + { + "epoch": 0.37893, + "grad_norm": 0.8075273036956787, + "learning_rate": 0.003, + "loss": 3.975, + "step": 37893 + }, + { + "epoch": 0.37894, + "grad_norm": 1.0851575136184692, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 37894 + }, + { + "epoch": 0.37895, + "grad_norm": 1.127424716949463, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 37895 + }, + { + "epoch": 0.37896, + "grad_norm": 0.7927245497703552, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 37896 + }, + { + "epoch": 0.37897, + "grad_norm": 0.8035668134689331, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 37897 + }, + { + "epoch": 0.37898, + "grad_norm": 0.8847108483314514, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 37898 + }, + { + "epoch": 0.37899, + "grad_norm": 1.1407302618026733, + "learning_rate": 0.003, + "loss": 3.977, + "step": 37899 + }, + { + "epoch": 0.379, + "grad_norm": 1.0190521478652954, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 37900 + }, + { + "epoch": 0.37901, + "grad_norm": 0.9864888787269592, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 37901 + }, + { + "epoch": 0.37902, + "grad_norm": 1.095169186592102, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 37902 + }, + { + "epoch": 0.37903, + "grad_norm": 0.8676010966300964, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 37903 + }, + { + "epoch": 0.37904, + "grad_norm": 0.9619213342666626, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 37904 + }, + { + "epoch": 0.37905, + "grad_norm": 1.159039855003357, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 37905 + }, + { + "epoch": 0.37906, + "grad_norm": 0.8435835242271423, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 37906 + }, + { + "epoch": 0.37907, + "grad_norm": 0.7062245011329651, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 37907 + }, + { + "epoch": 0.37908, + "grad_norm": 0.5595681667327881, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 37908 + }, + { + "epoch": 0.37909, + "grad_norm": 0.5187668800354004, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 37909 + }, + { + "epoch": 0.3791, + "grad_norm": 0.5482708215713501, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 37910 + }, + { + "epoch": 0.37911, + "grad_norm": 0.5596250891685486, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 37911 + }, + { + "epoch": 0.37912, + "grad_norm": 0.5372951626777649, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 37912 + }, + { + "epoch": 0.37913, + "grad_norm": 0.5568508505821228, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 37913 + }, + { + "epoch": 0.37914, + "grad_norm": 0.7017006278038025, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 37914 + }, + { + "epoch": 0.37915, + "grad_norm": 1.077378511428833, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 37915 + }, + { + "epoch": 0.37916, + "grad_norm": 1.104353427886963, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 37916 + }, + { + "epoch": 0.37917, + "grad_norm": 0.9986782073974609, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 37917 + }, + { + "epoch": 0.37918, + "grad_norm": 1.1893829107284546, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 37918 + }, + { + "epoch": 0.37919, + "grad_norm": 0.8019043207168579, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 37919 + }, + { + "epoch": 0.3792, + "grad_norm": 0.6669702529907227, + "learning_rate": 0.003, + "loss": 3.9382, + "step": 37920 + }, + { + "epoch": 0.37921, + "grad_norm": 0.6721321940422058, + "learning_rate": 0.003, + "loss": 3.9364, + "step": 37921 + }, + { + "epoch": 0.37922, + "grad_norm": 0.7291595339775085, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 37922 + }, + { + "epoch": 0.37923, + "grad_norm": 0.7739689946174622, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 37923 + }, + { + "epoch": 0.37924, + "grad_norm": 1.0433090925216675, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 37924 + }, + { + "epoch": 0.37925, + "grad_norm": 1.1100181341171265, + "learning_rate": 0.003, + "loss": 3.999, + "step": 37925 + }, + { + "epoch": 0.37926, + "grad_norm": 0.9919162392616272, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 37926 + }, + { + "epoch": 0.37927, + "grad_norm": 1.1848030090332031, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 37927 + }, + { + "epoch": 0.37928, + "grad_norm": 0.8832109570503235, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 37928 + }, + { + "epoch": 0.37929, + "grad_norm": 1.0576512813568115, + "learning_rate": 0.003, + "loss": 3.9363, + "step": 37929 + }, + { + "epoch": 0.3793, + "grad_norm": 0.9934818148612976, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 37930 + }, + { + "epoch": 0.37931, + "grad_norm": 0.8410605788230896, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 37931 + }, + { + "epoch": 0.37932, + "grad_norm": 0.7532873153686523, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 37932 + }, + { + "epoch": 0.37933, + "grad_norm": 0.758246123790741, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 37933 + }, + { + "epoch": 0.37934, + "grad_norm": 0.8558342456817627, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 37934 + }, + { + "epoch": 0.37935, + "grad_norm": 1.1373242139816284, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 37935 + }, + { + "epoch": 0.37936, + "grad_norm": 0.888529896736145, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 37936 + }, + { + "epoch": 0.37937, + "grad_norm": 0.8749808669090271, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 37937 + }, + { + "epoch": 0.37938, + "grad_norm": 1.1366137266159058, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 37938 + }, + { + "epoch": 0.37939, + "grad_norm": 1.0371944904327393, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 37939 + }, + { + "epoch": 0.3794, + "grad_norm": 0.8182951211929321, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 37940 + }, + { + "epoch": 0.37941, + "grad_norm": 0.8604822158813477, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 37941 + }, + { + "epoch": 0.37942, + "grad_norm": 1.0292224884033203, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 37942 + }, + { + "epoch": 0.37943, + "grad_norm": 0.9936243891716003, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 37943 + }, + { + "epoch": 0.37944, + "grad_norm": 0.8926070332527161, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 37944 + }, + { + "epoch": 0.37945, + "grad_norm": 0.9095268845558167, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 37945 + }, + { + "epoch": 0.37946, + "grad_norm": 1.0557724237442017, + "learning_rate": 0.003, + "loss": 3.999, + "step": 37946 + }, + { + "epoch": 0.37947, + "grad_norm": 1.0443601608276367, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 37947 + }, + { + "epoch": 0.37948, + "grad_norm": 0.9406717419624329, + "learning_rate": 0.003, + "loss": 3.962, + "step": 37948 + }, + { + "epoch": 0.37949, + "grad_norm": 0.8972463011741638, + "learning_rate": 0.003, + "loss": 3.971, + "step": 37949 + }, + { + "epoch": 0.3795, + "grad_norm": 0.7541260719299316, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 37950 + }, + { + "epoch": 0.37951, + "grad_norm": 0.8338117599487305, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 37951 + }, + { + "epoch": 0.37952, + "grad_norm": 0.8509258031845093, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 37952 + }, + { + "epoch": 0.37953, + "grad_norm": 0.8822827935218811, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 37953 + }, + { + "epoch": 0.37954, + "grad_norm": 1.0357047319412231, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 37954 + }, + { + "epoch": 0.37955, + "grad_norm": 1.113562822341919, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 37955 + }, + { + "epoch": 0.37956, + "grad_norm": 0.943831741809845, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 37956 + }, + { + "epoch": 0.37957, + "grad_norm": 0.9194769859313965, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 37957 + }, + { + "epoch": 0.37958, + "grad_norm": 1.0297421216964722, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 37958 + }, + { + "epoch": 0.37959, + "grad_norm": 1.0895311832427979, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 37959 + }, + { + "epoch": 0.3796, + "grad_norm": 0.8306035399436951, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 37960 + }, + { + "epoch": 0.37961, + "grad_norm": 0.7723745703697205, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 37961 + }, + { + "epoch": 0.37962, + "grad_norm": 0.7590087652206421, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 37962 + }, + { + "epoch": 0.37963, + "grad_norm": 0.7775694727897644, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 37963 + }, + { + "epoch": 0.37964, + "grad_norm": 0.8550145626068115, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 37964 + }, + { + "epoch": 0.37965, + "grad_norm": 1.1117143630981445, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 37965 + }, + { + "epoch": 0.37966, + "grad_norm": 1.5150201320648193, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 37966 + }, + { + "epoch": 0.37967, + "grad_norm": 0.5683659911155701, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 37967 + }, + { + "epoch": 0.37968, + "grad_norm": 0.7562565207481384, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 37968 + }, + { + "epoch": 0.37969, + "grad_norm": 1.1569716930389404, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 37969 + }, + { + "epoch": 0.3797, + "grad_norm": 1.1551189422607422, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 37970 + }, + { + "epoch": 0.37971, + "grad_norm": 0.7959310412406921, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 37971 + }, + { + "epoch": 0.37972, + "grad_norm": 0.7061701416969299, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 37972 + }, + { + "epoch": 0.37973, + "grad_norm": 0.6734079718589783, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 37973 + }, + { + "epoch": 0.37974, + "grad_norm": 0.7398442625999451, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 37974 + }, + { + "epoch": 0.37975, + "grad_norm": 0.9481499195098877, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 37975 + }, + { + "epoch": 0.37976, + "grad_norm": 1.075081467628479, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 37976 + }, + { + "epoch": 0.37977, + "grad_norm": 0.863917350769043, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 37977 + }, + { + "epoch": 0.37978, + "grad_norm": 0.8958925604820251, + "learning_rate": 0.003, + "loss": 3.965, + "step": 37978 + }, + { + "epoch": 0.37979, + "grad_norm": 1.2039726972579956, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 37979 + }, + { + "epoch": 0.3798, + "grad_norm": 1.0813606977462769, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 37980 + }, + { + "epoch": 0.37981, + "grad_norm": 0.9553893208503723, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 37981 + }, + { + "epoch": 0.37982, + "grad_norm": 0.8894033432006836, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 37982 + }, + { + "epoch": 0.37983, + "grad_norm": 0.7245937585830688, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 37983 + }, + { + "epoch": 0.37984, + "grad_norm": 0.7396060824394226, + "learning_rate": 0.003, + "loss": 3.942, + "step": 37984 + }, + { + "epoch": 0.37985, + "grad_norm": 0.798621654510498, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 37985 + }, + { + "epoch": 0.37986, + "grad_norm": 1.2204145193099976, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 37986 + }, + { + "epoch": 0.37987, + "grad_norm": 0.9855758547782898, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 37987 + }, + { + "epoch": 0.37988, + "grad_norm": 0.7935481071472168, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 37988 + }, + { + "epoch": 0.37989, + "grad_norm": 0.6916174292564392, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 37989 + }, + { + "epoch": 0.3799, + "grad_norm": 0.6352143883705139, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 37990 + }, + { + "epoch": 0.37991, + "grad_norm": 0.6785799264907837, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 37991 + }, + { + "epoch": 0.37992, + "grad_norm": 0.7395979166030884, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 37992 + }, + { + "epoch": 0.37993, + "grad_norm": 0.9002166986465454, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 37993 + }, + { + "epoch": 0.37994, + "grad_norm": 1.241751790046692, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 37994 + }, + { + "epoch": 0.37995, + "grad_norm": 0.757758378982544, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 37995 + }, + { + "epoch": 0.37996, + "grad_norm": 0.6617196202278137, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 37996 + }, + { + "epoch": 0.37997, + "grad_norm": 0.6367440223693848, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 37997 + }, + { + "epoch": 0.37998, + "grad_norm": 0.7901639342308044, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 37998 + }, + { + "epoch": 0.37999, + "grad_norm": 0.7835712432861328, + "learning_rate": 0.003, + "loss": 3.9342, + "step": 37999 + }, + { + "epoch": 0.38, + "grad_norm": 0.9771571159362793, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 38000 + }, + { + "epoch": 0.38001, + "grad_norm": 1.4593336582183838, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 38001 + }, + { + "epoch": 0.38002, + "grad_norm": 0.5593804717063904, + "learning_rate": 0.003, + "loss": 3.9489, + "step": 38002 + }, + { + "epoch": 0.38003, + "grad_norm": 0.7649013996124268, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 38003 + }, + { + "epoch": 0.38004, + "grad_norm": 0.8708294034004211, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 38004 + }, + { + "epoch": 0.38005, + "grad_norm": 0.9952723383903503, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 38005 + }, + { + "epoch": 0.38006, + "grad_norm": 1.0905120372772217, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 38006 + }, + { + "epoch": 0.38007, + "grad_norm": 0.917944610118866, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 38007 + }, + { + "epoch": 0.38008, + "grad_norm": 0.7897115349769592, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 38008 + }, + { + "epoch": 0.38009, + "grad_norm": 0.7683362364768982, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 38009 + }, + { + "epoch": 0.3801, + "grad_norm": 0.7294475436210632, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 38010 + }, + { + "epoch": 0.38011, + "grad_norm": 0.8276241421699524, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 38011 + }, + { + "epoch": 0.38012, + "grad_norm": 0.8043729066848755, + "learning_rate": 0.003, + "loss": 3.952, + "step": 38012 + }, + { + "epoch": 0.38013, + "grad_norm": 0.8275504112243652, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 38013 + }, + { + "epoch": 0.38014, + "grad_norm": 1.0374746322631836, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 38014 + }, + { + "epoch": 0.38015, + "grad_norm": 1.3503148555755615, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 38015 + }, + { + "epoch": 0.38016, + "grad_norm": 0.6770234107971191, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 38016 + }, + { + "epoch": 0.38017, + "grad_norm": 0.8829238414764404, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 38017 + }, + { + "epoch": 0.38018, + "grad_norm": 1.1907732486724854, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 38018 + }, + { + "epoch": 0.38019, + "grad_norm": 1.0620548725128174, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 38019 + }, + { + "epoch": 0.3802, + "grad_norm": 1.1636549234390259, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 38020 + }, + { + "epoch": 0.38021, + "grad_norm": 0.8679463863372803, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 38021 + }, + { + "epoch": 0.38022, + "grad_norm": 0.7337862253189087, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 38022 + }, + { + "epoch": 0.38023, + "grad_norm": 0.7434188723564148, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 38023 + }, + { + "epoch": 0.38024, + "grad_norm": 0.8325641751289368, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 38024 + }, + { + "epoch": 0.38025, + "grad_norm": 1.0335570573806763, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 38025 + }, + { + "epoch": 0.38026, + "grad_norm": 1.0276269912719727, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 38026 + }, + { + "epoch": 0.38027, + "grad_norm": 0.8964025378227234, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 38027 + }, + { + "epoch": 0.38028, + "grad_norm": 0.8557948470115662, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 38028 + }, + { + "epoch": 0.38029, + "grad_norm": 1.0760937929153442, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 38029 + }, + { + "epoch": 0.3803, + "grad_norm": 1.450013518333435, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 38030 + }, + { + "epoch": 0.38031, + "grad_norm": 0.5412149429321289, + "learning_rate": 0.003, + "loss": 3.9435, + "step": 38031 + }, + { + "epoch": 0.38032, + "grad_norm": 0.8450503349304199, + "learning_rate": 0.003, + "loss": 3.987, + "step": 38032 + }, + { + "epoch": 0.38033, + "grad_norm": 0.7255046963691711, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 38033 + }, + { + "epoch": 0.38034, + "grad_norm": 0.7551776766777039, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 38034 + }, + { + "epoch": 0.38035, + "grad_norm": 0.8744428157806396, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 38035 + }, + { + "epoch": 0.38036, + "grad_norm": 1.0154309272766113, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 38036 + }, + { + "epoch": 0.38037, + "grad_norm": 1.0985838174819946, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 38037 + }, + { + "epoch": 0.38038, + "grad_norm": 0.8598042726516724, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 38038 + }, + { + "epoch": 0.38039, + "grad_norm": 0.8548157811164856, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 38039 + }, + { + "epoch": 0.3804, + "grad_norm": 0.906045138835907, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 38040 + }, + { + "epoch": 0.38041, + "grad_norm": 0.8254764676094055, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 38041 + }, + { + "epoch": 0.38042, + "grad_norm": 0.8210159540176392, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 38042 + }, + { + "epoch": 0.38043, + "grad_norm": 0.8775815963745117, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 38043 + }, + { + "epoch": 0.38044, + "grad_norm": 0.9703829884529114, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 38044 + }, + { + "epoch": 0.38045, + "grad_norm": 1.1364408731460571, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 38045 + }, + { + "epoch": 0.38046, + "grad_norm": 0.8599440455436707, + "learning_rate": 0.003, + "loss": 3.964, + "step": 38046 + }, + { + "epoch": 0.38047, + "grad_norm": 0.8008179664611816, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 38047 + }, + { + "epoch": 0.38048, + "grad_norm": 0.7038705348968506, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 38048 + }, + { + "epoch": 0.38049, + "grad_norm": 0.7787387371063232, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 38049 + }, + { + "epoch": 0.3805, + "grad_norm": 1.0478307008743286, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 38050 + }, + { + "epoch": 0.38051, + "grad_norm": 1.0409326553344727, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 38051 + }, + { + "epoch": 0.38052, + "grad_norm": 0.915864884853363, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 38052 + }, + { + "epoch": 0.38053, + "grad_norm": 1.0854402780532837, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 38053 + }, + { + "epoch": 0.38054, + "grad_norm": 1.0608572959899902, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 38054 + }, + { + "epoch": 0.38055, + "grad_norm": 0.7881043553352356, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 38055 + }, + { + "epoch": 0.38056, + "grad_norm": 0.9239391684532166, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 38056 + }, + { + "epoch": 0.38057, + "grad_norm": 0.957629919052124, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 38057 + }, + { + "epoch": 0.38058, + "grad_norm": 1.1118443012237549, + "learning_rate": 0.003, + "loss": 4.016, + "step": 38058 + }, + { + "epoch": 0.38059, + "grad_norm": 1.05168879032135, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 38059 + }, + { + "epoch": 0.3806, + "grad_norm": 0.8562440872192383, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 38060 + }, + { + "epoch": 0.38061, + "grad_norm": 0.8460896015167236, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 38061 + }, + { + "epoch": 0.38062, + "grad_norm": 0.9496282935142517, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 38062 + }, + { + "epoch": 0.38063, + "grad_norm": 1.2979583740234375, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 38063 + }, + { + "epoch": 0.38064, + "grad_norm": 0.7271542549133301, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 38064 + }, + { + "epoch": 0.38065, + "grad_norm": 0.6673752665519714, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 38065 + }, + { + "epoch": 0.38066, + "grad_norm": 0.8228026628494263, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 38066 + }, + { + "epoch": 0.38067, + "grad_norm": 1.0758626461029053, + "learning_rate": 0.003, + "loss": 3.987, + "step": 38067 + }, + { + "epoch": 0.38068, + "grad_norm": 1.2057219743728638, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 38068 + }, + { + "epoch": 0.38069, + "grad_norm": 0.9835175275802612, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 38069 + }, + { + "epoch": 0.3807, + "grad_norm": 1.066864252090454, + "learning_rate": 0.003, + "loss": 3.981, + "step": 38070 + }, + { + "epoch": 0.38071, + "grad_norm": 0.9158828854560852, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 38071 + }, + { + "epoch": 0.38072, + "grad_norm": 0.8846721649169922, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 38072 + }, + { + "epoch": 0.38073, + "grad_norm": 0.839199423789978, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 38073 + }, + { + "epoch": 0.38074, + "grad_norm": 0.7881076335906982, + "learning_rate": 0.003, + "loss": 3.965, + "step": 38074 + }, + { + "epoch": 0.38075, + "grad_norm": 0.6955479979515076, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 38075 + }, + { + "epoch": 0.38076, + "grad_norm": 0.6738414764404297, + "learning_rate": 0.003, + "loss": 3.981, + "step": 38076 + }, + { + "epoch": 0.38077, + "grad_norm": 0.7952357530593872, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 38077 + }, + { + "epoch": 0.38078, + "grad_norm": 1.1322767734527588, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 38078 + }, + { + "epoch": 0.38079, + "grad_norm": 1.2054253816604614, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 38079 + }, + { + "epoch": 0.3808, + "grad_norm": 0.7954986691474915, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 38080 + }, + { + "epoch": 0.38081, + "grad_norm": 0.590186357498169, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 38081 + }, + { + "epoch": 0.38082, + "grad_norm": 0.7242714166641235, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 38082 + }, + { + "epoch": 0.38083, + "grad_norm": 1.1410263776779175, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 38083 + }, + { + "epoch": 0.38084, + "grad_norm": 1.2217662334442139, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 38084 + }, + { + "epoch": 0.38085, + "grad_norm": 0.6168562769889832, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 38085 + }, + { + "epoch": 0.38086, + "grad_norm": 0.6435019373893738, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 38086 + }, + { + "epoch": 0.38087, + "grad_norm": 0.9287621974945068, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 38087 + }, + { + "epoch": 0.38088, + "grad_norm": 1.150545358657837, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 38088 + }, + { + "epoch": 0.38089, + "grad_norm": 0.8118289709091187, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 38089 + }, + { + "epoch": 0.3809, + "grad_norm": 0.6606553196907043, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 38090 + }, + { + "epoch": 0.38091, + "grad_norm": 0.8147890567779541, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 38091 + }, + { + "epoch": 0.38092, + "grad_norm": 0.9993979930877686, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 38092 + }, + { + "epoch": 0.38093, + "grad_norm": 1.0112497806549072, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 38093 + }, + { + "epoch": 0.38094, + "grad_norm": 0.9259256720542908, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 38094 + }, + { + "epoch": 0.38095, + "grad_norm": 0.972453236579895, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 38095 + }, + { + "epoch": 0.38096, + "grad_norm": 0.9133829474449158, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 38096 + }, + { + "epoch": 0.38097, + "grad_norm": 1.0185388326644897, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 38097 + }, + { + "epoch": 0.38098, + "grad_norm": 1.2022672891616821, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 38098 + }, + { + "epoch": 0.38099, + "grad_norm": 1.060726523399353, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 38099 + }, + { + "epoch": 0.381, + "grad_norm": 0.9280908107757568, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 38100 + }, + { + "epoch": 0.38101, + "grad_norm": 0.9057480692863464, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 38101 + }, + { + "epoch": 0.38102, + "grad_norm": 1.0151749849319458, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 38102 + }, + { + "epoch": 0.38103, + "grad_norm": 1.0725127458572388, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 38103 + }, + { + "epoch": 0.38104, + "grad_norm": 1.0688763856887817, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 38104 + }, + { + "epoch": 0.38105, + "grad_norm": 1.1857051849365234, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 38105 + }, + { + "epoch": 0.38106, + "grad_norm": 1.0442941188812256, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 38106 + }, + { + "epoch": 0.38107, + "grad_norm": 0.814103901386261, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 38107 + }, + { + "epoch": 0.38108, + "grad_norm": 0.7345314025878906, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 38108 + }, + { + "epoch": 0.38109, + "grad_norm": 0.7341506481170654, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 38109 + }, + { + "epoch": 0.3811, + "grad_norm": 0.7998956441879272, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 38110 + }, + { + "epoch": 0.38111, + "grad_norm": 0.7758166193962097, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 38111 + }, + { + "epoch": 0.38112, + "grad_norm": 0.8256387710571289, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 38112 + }, + { + "epoch": 0.38113, + "grad_norm": 0.8600993752479553, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 38113 + }, + { + "epoch": 0.38114, + "grad_norm": 0.8705414533615112, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 38114 + }, + { + "epoch": 0.38115, + "grad_norm": 0.8991608023643494, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 38115 + }, + { + "epoch": 0.38116, + "grad_norm": 0.8287593126296997, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 38116 + }, + { + "epoch": 0.38117, + "grad_norm": 0.9058568477630615, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 38117 + }, + { + "epoch": 0.38118, + "grad_norm": 0.8821269869804382, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 38118 + }, + { + "epoch": 0.38119, + "grad_norm": 0.8027511239051819, + "learning_rate": 0.003, + "loss": 3.9439, + "step": 38119 + }, + { + "epoch": 0.3812, + "grad_norm": 0.7976489067077637, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 38120 + }, + { + "epoch": 0.38121, + "grad_norm": 0.8143995404243469, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 38121 + }, + { + "epoch": 0.38122, + "grad_norm": 0.8634476065635681, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 38122 + }, + { + "epoch": 0.38123, + "grad_norm": 1.1579554080963135, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 38123 + }, + { + "epoch": 0.38124, + "grad_norm": 1.3370550870895386, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 38124 + }, + { + "epoch": 0.38125, + "grad_norm": 0.6493270397186279, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 38125 + }, + { + "epoch": 0.38126, + "grad_norm": 0.7707735896110535, + "learning_rate": 0.003, + "loss": 3.969, + "step": 38126 + }, + { + "epoch": 0.38127, + "grad_norm": 0.8842806816101074, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 38127 + }, + { + "epoch": 0.38128, + "grad_norm": 1.1297340393066406, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 38128 + }, + { + "epoch": 0.38129, + "grad_norm": 1.198251724243164, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 38129 + }, + { + "epoch": 0.3813, + "grad_norm": 0.654419481754303, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 38130 + }, + { + "epoch": 0.38131, + "grad_norm": 0.6239352822303772, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 38131 + }, + { + "epoch": 0.38132, + "grad_norm": 0.6837000250816345, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 38132 + }, + { + "epoch": 0.38133, + "grad_norm": 0.9704278707504272, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 38133 + }, + { + "epoch": 0.38134, + "grad_norm": 0.9188635945320129, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 38134 + }, + { + "epoch": 0.38135, + "grad_norm": 0.8076731562614441, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 38135 + }, + { + "epoch": 0.38136, + "grad_norm": 0.7313842177391052, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 38136 + }, + { + "epoch": 0.38137, + "grad_norm": 0.718034565448761, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 38137 + }, + { + "epoch": 0.38138, + "grad_norm": 0.9344801306724548, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 38138 + }, + { + "epoch": 0.38139, + "grad_norm": 1.1197185516357422, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 38139 + }, + { + "epoch": 0.3814, + "grad_norm": 1.0107795000076294, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 38140 + }, + { + "epoch": 0.38141, + "grad_norm": 1.1979999542236328, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 38141 + }, + { + "epoch": 0.38142, + "grad_norm": 0.7843431830406189, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 38142 + }, + { + "epoch": 0.38143, + "grad_norm": 0.7210452556610107, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 38143 + }, + { + "epoch": 0.38144, + "grad_norm": 0.8070765733718872, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 38144 + }, + { + "epoch": 0.38145, + "grad_norm": 1.0383696556091309, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 38145 + }, + { + "epoch": 0.38146, + "grad_norm": 1.0818949937820435, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 38146 + }, + { + "epoch": 0.38147, + "grad_norm": 0.9592359662055969, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 38147 + }, + { + "epoch": 0.38148, + "grad_norm": 0.9838439226150513, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 38148 + }, + { + "epoch": 0.38149, + "grad_norm": 1.0839481353759766, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 38149 + }, + { + "epoch": 0.3815, + "grad_norm": 0.8819082975387573, + "learning_rate": 0.003, + "loss": 4.006, + "step": 38150 + }, + { + "epoch": 0.38151, + "grad_norm": 0.9265299439430237, + "learning_rate": 0.003, + "loss": 4.0, + "step": 38151 + }, + { + "epoch": 0.38152, + "grad_norm": 1.1564645767211914, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 38152 + }, + { + "epoch": 0.38153, + "grad_norm": 1.0517091751098633, + "learning_rate": 0.003, + "loss": 3.9423, + "step": 38153 + }, + { + "epoch": 0.38154, + "grad_norm": 1.1207941770553589, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 38154 + }, + { + "epoch": 0.38155, + "grad_norm": 0.9543811678886414, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 38155 + }, + { + "epoch": 0.38156, + "grad_norm": 0.903141438961029, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 38156 + }, + { + "epoch": 0.38157, + "grad_norm": 0.8011553883552551, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 38157 + }, + { + "epoch": 0.38158, + "grad_norm": 0.7282108068466187, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 38158 + }, + { + "epoch": 0.38159, + "grad_norm": 0.8840474486351013, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 38159 + }, + { + "epoch": 0.3816, + "grad_norm": 1.130990982055664, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 38160 + }, + { + "epoch": 0.38161, + "grad_norm": 1.0164761543273926, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 38161 + }, + { + "epoch": 0.38162, + "grad_norm": 1.0485535860061646, + "learning_rate": 0.003, + "loss": 3.975, + "step": 38162 + }, + { + "epoch": 0.38163, + "grad_norm": 1.027864933013916, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 38163 + }, + { + "epoch": 0.38164, + "grad_norm": 0.8561417460441589, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 38164 + }, + { + "epoch": 0.38165, + "grad_norm": 0.6103036403656006, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 38165 + }, + { + "epoch": 0.38166, + "grad_norm": 0.668755292892456, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 38166 + }, + { + "epoch": 0.38167, + "grad_norm": 0.7477791905403137, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 38167 + }, + { + "epoch": 0.38168, + "grad_norm": 0.9029073119163513, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 38168 + }, + { + "epoch": 0.38169, + "grad_norm": 0.8339408040046692, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 38169 + }, + { + "epoch": 0.3817, + "grad_norm": 0.7170405387878418, + "learning_rate": 0.003, + "loss": 3.9441, + "step": 38170 + }, + { + "epoch": 0.38171, + "grad_norm": 0.6972704529762268, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 38171 + }, + { + "epoch": 0.38172, + "grad_norm": 0.652587890625, + "learning_rate": 0.003, + "loss": 3.9469, + "step": 38172 + }, + { + "epoch": 0.38173, + "grad_norm": 0.786607027053833, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 38173 + }, + { + "epoch": 0.38174, + "grad_norm": 0.832467794418335, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 38174 + }, + { + "epoch": 0.38175, + "grad_norm": 0.9863250851631165, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 38175 + }, + { + "epoch": 0.38176, + "grad_norm": 1.2678933143615723, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 38176 + }, + { + "epoch": 0.38177, + "grad_norm": 0.9504258036613464, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 38177 + }, + { + "epoch": 0.38178, + "grad_norm": 0.8909956216812134, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 38178 + }, + { + "epoch": 0.38179, + "grad_norm": 1.0138665437698364, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 38179 + }, + { + "epoch": 0.3818, + "grad_norm": 0.9643624424934387, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 38180 + }, + { + "epoch": 0.38181, + "grad_norm": 1.0648621320724487, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 38181 + }, + { + "epoch": 0.38182, + "grad_norm": 0.8843860030174255, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 38182 + }, + { + "epoch": 0.38183, + "grad_norm": 0.7858692407608032, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 38183 + }, + { + "epoch": 0.38184, + "grad_norm": 0.7227175831794739, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 38184 + }, + { + "epoch": 0.38185, + "grad_norm": 0.7965670228004456, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 38185 + }, + { + "epoch": 0.38186, + "grad_norm": 0.8962279558181763, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 38186 + }, + { + "epoch": 0.38187, + "grad_norm": 0.9036398530006409, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 38187 + }, + { + "epoch": 0.38188, + "grad_norm": 1.133237600326538, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 38188 + }, + { + "epoch": 0.38189, + "grad_norm": 1.0298974514007568, + "learning_rate": 0.003, + "loss": 3.951, + "step": 38189 + }, + { + "epoch": 0.3819, + "grad_norm": 0.9953978657722473, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 38190 + }, + { + "epoch": 0.38191, + "grad_norm": 1.0462439060211182, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 38191 + }, + { + "epoch": 0.38192, + "grad_norm": 1.0741140842437744, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 38192 + }, + { + "epoch": 0.38193, + "grad_norm": 1.0020183324813843, + "learning_rate": 0.003, + "loss": 3.984, + "step": 38193 + }, + { + "epoch": 0.38194, + "grad_norm": 0.8908431529998779, + "learning_rate": 0.003, + "loss": 3.963, + "step": 38194 + }, + { + "epoch": 0.38195, + "grad_norm": 0.995610773563385, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 38195 + }, + { + "epoch": 0.38196, + "grad_norm": 0.9463599920272827, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 38196 + }, + { + "epoch": 0.38197, + "grad_norm": 1.023695707321167, + "learning_rate": 0.003, + "loss": 3.9451, + "step": 38197 + }, + { + "epoch": 0.38198, + "grad_norm": 1.1802637577056885, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 38198 + }, + { + "epoch": 0.38199, + "grad_norm": 0.9696594476699829, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 38199 + }, + { + "epoch": 0.382, + "grad_norm": 1.1216663122177124, + "learning_rate": 0.003, + "loss": 3.999, + "step": 38200 + }, + { + "epoch": 0.38201, + "grad_norm": 0.9980288147926331, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 38201 + }, + { + "epoch": 0.38202, + "grad_norm": 1.0173763036727905, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 38202 + }, + { + "epoch": 0.38203, + "grad_norm": 1.088608980178833, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 38203 + }, + { + "epoch": 0.38204, + "grad_norm": 1.3911231756210327, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 38204 + }, + { + "epoch": 0.38205, + "grad_norm": 0.6603046655654907, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 38205 + }, + { + "epoch": 0.38206, + "grad_norm": 0.940162181854248, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 38206 + }, + { + "epoch": 0.38207, + "grad_norm": 1.1328456401824951, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 38207 + }, + { + "epoch": 0.38208, + "grad_norm": 0.7754746675491333, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 38208 + }, + { + "epoch": 0.38209, + "grad_norm": 0.6977764964103699, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 38209 + }, + { + "epoch": 0.3821, + "grad_norm": 0.6542420387268066, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 38210 + }, + { + "epoch": 0.38211, + "grad_norm": 0.6868578791618347, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 38211 + }, + { + "epoch": 0.38212, + "grad_norm": 0.7190095782279968, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 38212 + }, + { + "epoch": 0.38213, + "grad_norm": 0.771040678024292, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 38213 + }, + { + "epoch": 0.38214, + "grad_norm": 0.9597247242927551, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 38214 + }, + { + "epoch": 0.38215, + "grad_norm": 0.9791871309280396, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 38215 + }, + { + "epoch": 0.38216, + "grad_norm": 0.7252845168113708, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 38216 + }, + { + "epoch": 0.38217, + "grad_norm": 0.6819677352905273, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 38217 + }, + { + "epoch": 0.38218, + "grad_norm": 0.7963882684707642, + "learning_rate": 0.003, + "loss": 3.9393, + "step": 38218 + }, + { + "epoch": 0.38219, + "grad_norm": 0.8075743317604065, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 38219 + }, + { + "epoch": 0.3822, + "grad_norm": 0.8583657145500183, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 38220 + }, + { + "epoch": 0.38221, + "grad_norm": 0.866256058216095, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 38221 + }, + { + "epoch": 0.38222, + "grad_norm": 0.8857022523880005, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 38222 + }, + { + "epoch": 0.38223, + "grad_norm": 1.1557042598724365, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 38223 + }, + { + "epoch": 0.38224, + "grad_norm": 0.9983378648757935, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 38224 + }, + { + "epoch": 0.38225, + "grad_norm": 0.8767260909080505, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 38225 + }, + { + "epoch": 0.38226, + "grad_norm": 0.8421515226364136, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 38226 + }, + { + "epoch": 0.38227, + "grad_norm": 0.8620858192443848, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 38227 + }, + { + "epoch": 0.38228, + "grad_norm": 0.9351505041122437, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 38228 + }, + { + "epoch": 0.38229, + "grad_norm": 1.1339874267578125, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 38229 + }, + { + "epoch": 0.3823, + "grad_norm": 0.7218264937400818, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 38230 + }, + { + "epoch": 0.38231, + "grad_norm": 0.7726501822471619, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 38231 + }, + { + "epoch": 0.38232, + "grad_norm": 0.7935367822647095, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 38232 + }, + { + "epoch": 0.38233, + "grad_norm": 0.9020146131515503, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 38233 + }, + { + "epoch": 0.38234, + "grad_norm": 1.1652573347091675, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 38234 + }, + { + "epoch": 0.38235, + "grad_norm": 1.096869707107544, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 38235 + }, + { + "epoch": 0.38236, + "grad_norm": 0.9771718382835388, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 38236 + }, + { + "epoch": 0.38237, + "grad_norm": 0.9167359471321106, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 38237 + }, + { + "epoch": 0.38238, + "grad_norm": 0.8292696475982666, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 38238 + }, + { + "epoch": 0.38239, + "grad_norm": 0.7535868287086487, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 38239 + }, + { + "epoch": 0.3824, + "grad_norm": 0.6816933751106262, + "learning_rate": 0.003, + "loss": 3.9315, + "step": 38240 + }, + { + "epoch": 0.38241, + "grad_norm": 0.6907483339309692, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 38241 + }, + { + "epoch": 0.38242, + "grad_norm": 0.7200667858123779, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 38242 + }, + { + "epoch": 0.38243, + "grad_norm": 0.8814525604248047, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 38243 + }, + { + "epoch": 0.38244, + "grad_norm": 1.0842466354370117, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 38244 + }, + { + "epoch": 0.38245, + "grad_norm": 1.0902509689331055, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 38245 + }, + { + "epoch": 0.38246, + "grad_norm": 0.9223896861076355, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 38246 + }, + { + "epoch": 0.38247, + "grad_norm": 0.9430326819419861, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 38247 + }, + { + "epoch": 0.38248, + "grad_norm": 1.138801097869873, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 38248 + }, + { + "epoch": 0.38249, + "grad_norm": 0.8567675352096558, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 38249 + }, + { + "epoch": 0.3825, + "grad_norm": 0.8226395845413208, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 38250 + }, + { + "epoch": 0.38251, + "grad_norm": 1.0297198295593262, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 38251 + }, + { + "epoch": 0.38252, + "grad_norm": 1.0605792999267578, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 38252 + }, + { + "epoch": 0.38253, + "grad_norm": 0.9291614294052124, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 38253 + }, + { + "epoch": 0.38254, + "grad_norm": 0.96302729845047, + "learning_rate": 0.003, + "loss": 3.99, + "step": 38254 + }, + { + "epoch": 0.38255, + "grad_norm": 1.0560896396636963, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 38255 + }, + { + "epoch": 0.38256, + "grad_norm": 0.9643657207489014, + "learning_rate": 0.003, + "loss": 3.992, + "step": 38256 + }, + { + "epoch": 0.38257, + "grad_norm": 0.7426228523254395, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 38257 + }, + { + "epoch": 0.38258, + "grad_norm": 0.6231337785720825, + "learning_rate": 0.003, + "loss": 3.974, + "step": 38258 + }, + { + "epoch": 0.38259, + "grad_norm": 0.7651039361953735, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 38259 + }, + { + "epoch": 0.3826, + "grad_norm": 0.7951352596282959, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 38260 + }, + { + "epoch": 0.38261, + "grad_norm": 0.8267971873283386, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 38261 + }, + { + "epoch": 0.38262, + "grad_norm": 0.9772624373435974, + "learning_rate": 0.003, + "loss": 3.993, + "step": 38262 + }, + { + "epoch": 0.38263, + "grad_norm": 1.1919039487838745, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 38263 + }, + { + "epoch": 0.38264, + "grad_norm": 0.8389700055122375, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 38264 + }, + { + "epoch": 0.38265, + "grad_norm": 0.7564305067062378, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 38265 + }, + { + "epoch": 0.38266, + "grad_norm": 0.69765305519104, + "learning_rate": 0.003, + "loss": 3.98, + "step": 38266 + }, + { + "epoch": 0.38267, + "grad_norm": 0.6675229668617249, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 38267 + }, + { + "epoch": 0.38268, + "grad_norm": 0.8251143097877502, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 38268 + }, + { + "epoch": 0.38269, + "grad_norm": 0.9129517078399658, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 38269 + }, + { + "epoch": 0.3827, + "grad_norm": 0.9943205118179321, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 38270 + }, + { + "epoch": 0.38271, + "grad_norm": 1.0641791820526123, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 38271 + }, + { + "epoch": 0.38272, + "grad_norm": 1.013391137123108, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 38272 + }, + { + "epoch": 0.38273, + "grad_norm": 0.9332718253135681, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 38273 + }, + { + "epoch": 0.38274, + "grad_norm": 0.9618353247642517, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 38274 + }, + { + "epoch": 0.38275, + "grad_norm": 1.1594055891036987, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 38275 + }, + { + "epoch": 0.38276, + "grad_norm": 0.8791835308074951, + "learning_rate": 0.003, + "loss": 3.9405, + "step": 38276 + }, + { + "epoch": 0.38277, + "grad_norm": 0.8949288129806519, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 38277 + }, + { + "epoch": 0.38278, + "grad_norm": 0.8963983654975891, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 38278 + }, + { + "epoch": 0.38279, + "grad_norm": 0.7987350821495056, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 38279 + }, + { + "epoch": 0.3828, + "grad_norm": 0.9219951033592224, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 38280 + }, + { + "epoch": 0.38281, + "grad_norm": 0.9730843901634216, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 38281 + }, + { + "epoch": 0.38282, + "grad_norm": 0.851684033870697, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 38282 + }, + { + "epoch": 0.38283, + "grad_norm": 0.8340533971786499, + "learning_rate": 0.003, + "loss": 3.9493, + "step": 38283 + }, + { + "epoch": 0.38284, + "grad_norm": 0.8097660541534424, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 38284 + }, + { + "epoch": 0.38285, + "grad_norm": 0.9675111770629883, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 38285 + }, + { + "epoch": 0.38286, + "grad_norm": 1.4971460103988647, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 38286 + }, + { + "epoch": 0.38287, + "grad_norm": 0.6033962368965149, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 38287 + }, + { + "epoch": 0.38288, + "grad_norm": 0.8082758188247681, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 38288 + }, + { + "epoch": 0.38289, + "grad_norm": 0.8049433827400208, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 38289 + }, + { + "epoch": 0.3829, + "grad_norm": 0.7883219718933105, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 38290 + }, + { + "epoch": 0.38291, + "grad_norm": 0.8350990414619446, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 38291 + }, + { + "epoch": 0.38292, + "grad_norm": 1.1380524635314941, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 38292 + }, + { + "epoch": 0.38293, + "grad_norm": 0.9483348727226257, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 38293 + }, + { + "epoch": 0.38294, + "grad_norm": 0.8099755644798279, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 38294 + }, + { + "epoch": 0.38295, + "grad_norm": 0.7316321730613708, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 38295 + }, + { + "epoch": 0.38296, + "grad_norm": 0.7381266951560974, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 38296 + }, + { + "epoch": 0.38297, + "grad_norm": 0.7258783578872681, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 38297 + }, + { + "epoch": 0.38298, + "grad_norm": 0.720842719078064, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 38298 + }, + { + "epoch": 0.38299, + "grad_norm": 0.719657838344574, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 38299 + }, + { + "epoch": 0.383, + "grad_norm": 0.8441247940063477, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 38300 + }, + { + "epoch": 0.38301, + "grad_norm": 1.142115831375122, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 38301 + }, + { + "epoch": 0.38302, + "grad_norm": 0.8990166187286377, + "learning_rate": 0.003, + "loss": 3.96, + "step": 38302 + }, + { + "epoch": 0.38303, + "grad_norm": 1.1125069856643677, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 38303 + }, + { + "epoch": 0.38304, + "grad_norm": 1.1458330154418945, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 38304 + }, + { + "epoch": 0.38305, + "grad_norm": 1.1539220809936523, + "learning_rate": 0.003, + "loss": 3.9464, + "step": 38305 + }, + { + "epoch": 0.38306, + "grad_norm": 1.030561089515686, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 38306 + }, + { + "epoch": 0.38307, + "grad_norm": 1.01802659034729, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 38307 + }, + { + "epoch": 0.38308, + "grad_norm": 0.9443342089653015, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 38308 + }, + { + "epoch": 0.38309, + "grad_norm": 0.9230753183364868, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 38309 + }, + { + "epoch": 0.3831, + "grad_norm": 0.8927147388458252, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 38310 + }, + { + "epoch": 0.38311, + "grad_norm": 0.9880083203315735, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 38311 + }, + { + "epoch": 0.38312, + "grad_norm": 1.0230660438537598, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 38312 + }, + { + "epoch": 0.38313, + "grad_norm": 0.9426526427268982, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 38313 + }, + { + "epoch": 0.38314, + "grad_norm": 0.690211296081543, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 38314 + }, + { + "epoch": 0.38315, + "grad_norm": 0.5760305523872375, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 38315 + }, + { + "epoch": 0.38316, + "grad_norm": 0.583183765411377, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 38316 + }, + { + "epoch": 0.38317, + "grad_norm": 0.5898782014846802, + "learning_rate": 0.003, + "loss": 3.9462, + "step": 38317 + }, + { + "epoch": 0.38318, + "grad_norm": 0.5695534348487854, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 38318 + }, + { + "epoch": 0.38319, + "grad_norm": 0.6522891521453857, + "learning_rate": 0.003, + "loss": 3.986, + "step": 38319 + }, + { + "epoch": 0.3832, + "grad_norm": 0.7481288909912109, + "learning_rate": 0.003, + "loss": 3.974, + "step": 38320 + }, + { + "epoch": 0.38321, + "grad_norm": 0.8230974674224854, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 38321 + }, + { + "epoch": 0.38322, + "grad_norm": 0.9564329385757446, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 38322 + }, + { + "epoch": 0.38323, + "grad_norm": 1.180891752243042, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 38323 + }, + { + "epoch": 0.38324, + "grad_norm": 1.026199460029602, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 38324 + }, + { + "epoch": 0.38325, + "grad_norm": 1.0447016954421997, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 38325 + }, + { + "epoch": 0.38326, + "grad_norm": 1.0478607416152954, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 38326 + }, + { + "epoch": 0.38327, + "grad_norm": 0.8895474672317505, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 38327 + }, + { + "epoch": 0.38328, + "grad_norm": 0.8803889751434326, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 38328 + }, + { + "epoch": 0.38329, + "grad_norm": 1.0782798528671265, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 38329 + }, + { + "epoch": 0.3833, + "grad_norm": 1.1803159713745117, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 38330 + }, + { + "epoch": 0.38331, + "grad_norm": 0.8478747606277466, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 38331 + }, + { + "epoch": 0.38332, + "grad_norm": 0.8715952038764954, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 38332 + }, + { + "epoch": 0.38333, + "grad_norm": 1.1475248336791992, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 38333 + }, + { + "epoch": 0.38334, + "grad_norm": 1.2263764142990112, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 38334 + }, + { + "epoch": 0.38335, + "grad_norm": 0.7286735773086548, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 38335 + }, + { + "epoch": 0.38336, + "grad_norm": 0.703194260597229, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 38336 + }, + { + "epoch": 0.38337, + "grad_norm": 0.7721149325370789, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 38337 + }, + { + "epoch": 0.38338, + "grad_norm": 0.7756116390228271, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 38338 + }, + { + "epoch": 0.38339, + "grad_norm": 0.7197469472885132, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 38339 + }, + { + "epoch": 0.3834, + "grad_norm": 0.6305176615715027, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 38340 + }, + { + "epoch": 0.38341, + "grad_norm": 0.5886070728302002, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 38341 + }, + { + "epoch": 0.38342, + "grad_norm": 0.5769128203392029, + "learning_rate": 0.003, + "loss": 3.9459, + "step": 38342 + }, + { + "epoch": 0.38343, + "grad_norm": 0.7304250001907349, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 38343 + }, + { + "epoch": 0.38344, + "grad_norm": 1.0237568616867065, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 38344 + }, + { + "epoch": 0.38345, + "grad_norm": 1.2873808145523071, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 38345 + }, + { + "epoch": 0.38346, + "grad_norm": 0.6681447625160217, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 38346 + }, + { + "epoch": 0.38347, + "grad_norm": 0.7324138879776001, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 38347 + }, + { + "epoch": 0.38348, + "grad_norm": 0.8338629603385925, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 38348 + }, + { + "epoch": 0.38349, + "grad_norm": 0.9561827778816223, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 38349 + }, + { + "epoch": 0.3835, + "grad_norm": 1.0907878875732422, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 38350 + }, + { + "epoch": 0.38351, + "grad_norm": 0.8126537203788757, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 38351 + }, + { + "epoch": 0.38352, + "grad_norm": 0.8082467317581177, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 38352 + }, + { + "epoch": 0.38353, + "grad_norm": 0.9627190828323364, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 38353 + }, + { + "epoch": 0.38354, + "grad_norm": 0.9176865220069885, + "learning_rate": 0.003, + "loss": 3.983, + "step": 38354 + }, + { + "epoch": 0.38355, + "grad_norm": 1.142473578453064, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 38355 + }, + { + "epoch": 0.38356, + "grad_norm": 1.0410723686218262, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 38356 + }, + { + "epoch": 0.38357, + "grad_norm": 1.1107512712478638, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 38357 + }, + { + "epoch": 0.38358, + "grad_norm": 0.9712876081466675, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 38358 + }, + { + "epoch": 0.38359, + "grad_norm": 1.051686406135559, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 38359 + }, + { + "epoch": 0.3836, + "grad_norm": 0.9211946725845337, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 38360 + }, + { + "epoch": 0.38361, + "grad_norm": 0.8112687468528748, + "learning_rate": 0.003, + "loss": 3.984, + "step": 38361 + }, + { + "epoch": 0.38362, + "grad_norm": 0.9605982899665833, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 38362 + }, + { + "epoch": 0.38363, + "grad_norm": 1.001365303993225, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 38363 + }, + { + "epoch": 0.38364, + "grad_norm": 0.9353100657463074, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 38364 + }, + { + "epoch": 0.38365, + "grad_norm": 1.154702067375183, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 38365 + }, + { + "epoch": 0.38366, + "grad_norm": 1.1676182746887207, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 38366 + }, + { + "epoch": 0.38367, + "grad_norm": 0.9374552369117737, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 38367 + }, + { + "epoch": 0.38368, + "grad_norm": 0.8828134536743164, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 38368 + }, + { + "epoch": 0.38369, + "grad_norm": 0.8904166221618652, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 38369 + }, + { + "epoch": 0.3837, + "grad_norm": 1.1462745666503906, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 38370 + }, + { + "epoch": 0.38371, + "grad_norm": 0.9670504331588745, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 38371 + }, + { + "epoch": 0.38372, + "grad_norm": 0.9719749689102173, + "learning_rate": 0.003, + "loss": 4.011, + "step": 38372 + }, + { + "epoch": 0.38373, + "grad_norm": 1.0633165836334229, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 38373 + }, + { + "epoch": 0.38374, + "grad_norm": 1.0745080709457397, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 38374 + }, + { + "epoch": 0.38375, + "grad_norm": 0.9290266036987305, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 38375 + }, + { + "epoch": 0.38376, + "grad_norm": 0.8271641135215759, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 38376 + }, + { + "epoch": 0.38377, + "grad_norm": 1.0258231163024902, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 38377 + }, + { + "epoch": 0.38378, + "grad_norm": 0.7859398126602173, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 38378 + }, + { + "epoch": 0.38379, + "grad_norm": 0.7459686398506165, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 38379 + }, + { + "epoch": 0.3838, + "grad_norm": 0.6842153668403625, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 38380 + }, + { + "epoch": 0.38381, + "grad_norm": 0.6891180872917175, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 38381 + }, + { + "epoch": 0.38382, + "grad_norm": 0.6673222780227661, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 38382 + }, + { + "epoch": 0.38383, + "grad_norm": 0.6242796182632446, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 38383 + }, + { + "epoch": 0.38384, + "grad_norm": 0.6753208041191101, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 38384 + }, + { + "epoch": 0.38385, + "grad_norm": 0.7254547476768494, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 38385 + }, + { + "epoch": 0.38386, + "grad_norm": 0.8535836338996887, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 38386 + }, + { + "epoch": 0.38387, + "grad_norm": 1.1593589782714844, + "learning_rate": 0.003, + "loss": 3.993, + "step": 38387 + }, + { + "epoch": 0.38388, + "grad_norm": 0.9543247222900391, + "learning_rate": 0.003, + "loss": 3.975, + "step": 38388 + }, + { + "epoch": 0.38389, + "grad_norm": 0.9655367732048035, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 38389 + }, + { + "epoch": 0.3839, + "grad_norm": 1.1292710304260254, + "learning_rate": 0.003, + "loss": 3.98, + "step": 38390 + }, + { + "epoch": 0.38391, + "grad_norm": 1.234879970550537, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 38391 + }, + { + "epoch": 0.38392, + "grad_norm": 0.807761549949646, + "learning_rate": 0.003, + "loss": 3.964, + "step": 38392 + }, + { + "epoch": 0.38393, + "grad_norm": 0.6061568856239319, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 38393 + }, + { + "epoch": 0.38394, + "grad_norm": 0.6377647519111633, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 38394 + }, + { + "epoch": 0.38395, + "grad_norm": 0.6465964913368225, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 38395 + }, + { + "epoch": 0.38396, + "grad_norm": 0.8228241205215454, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 38396 + }, + { + "epoch": 0.38397, + "grad_norm": 1.0807803869247437, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 38397 + }, + { + "epoch": 0.38398, + "grad_norm": 1.1922880411148071, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 38398 + }, + { + "epoch": 0.38399, + "grad_norm": 0.8458248972892761, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 38399 + }, + { + "epoch": 0.384, + "grad_norm": 0.758871853351593, + "learning_rate": 0.003, + "loss": 3.99, + "step": 38400 + }, + { + "epoch": 0.38401, + "grad_norm": 0.7452692985534668, + "learning_rate": 0.003, + "loss": 3.978, + "step": 38401 + }, + { + "epoch": 0.38402, + "grad_norm": 0.836438000202179, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 38402 + }, + { + "epoch": 0.38403, + "grad_norm": 0.9984953999519348, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 38403 + }, + { + "epoch": 0.38404, + "grad_norm": 0.9426993727684021, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 38404 + }, + { + "epoch": 0.38405, + "grad_norm": 0.8384084105491638, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 38405 + }, + { + "epoch": 0.38406, + "grad_norm": 1.1280039548873901, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 38406 + }, + { + "epoch": 0.38407, + "grad_norm": 1.0940768718719482, + "learning_rate": 0.003, + "loss": 3.9296, + "step": 38407 + }, + { + "epoch": 0.38408, + "grad_norm": 1.075052261352539, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 38408 + }, + { + "epoch": 0.38409, + "grad_norm": 0.9892614483833313, + "learning_rate": 0.003, + "loss": 3.9394, + "step": 38409 + }, + { + "epoch": 0.3841, + "grad_norm": 0.9050381779670715, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 38410 + }, + { + "epoch": 0.38411, + "grad_norm": 0.8876997828483582, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 38411 + }, + { + "epoch": 0.38412, + "grad_norm": 0.7471044659614563, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 38412 + }, + { + "epoch": 0.38413, + "grad_norm": 0.8144605755805969, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 38413 + }, + { + "epoch": 0.38414, + "grad_norm": 1.0515817403793335, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 38414 + }, + { + "epoch": 0.38415, + "grad_norm": 1.0069375038146973, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 38415 + }, + { + "epoch": 0.38416, + "grad_norm": 0.8942800164222717, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 38416 + }, + { + "epoch": 0.38417, + "grad_norm": 0.9391180276870728, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 38417 + }, + { + "epoch": 0.38418, + "grad_norm": 1.117926001548767, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 38418 + }, + { + "epoch": 0.38419, + "grad_norm": 0.8656047582626343, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 38419 + }, + { + "epoch": 0.3842, + "grad_norm": 1.0098416805267334, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 38420 + }, + { + "epoch": 0.38421, + "grad_norm": 1.1534773111343384, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 38421 + }, + { + "epoch": 0.38422, + "grad_norm": 0.799090564250946, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 38422 + }, + { + "epoch": 0.38423, + "grad_norm": 0.762589156627655, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 38423 + }, + { + "epoch": 0.38424, + "grad_norm": 0.7019056081771851, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 38424 + }, + { + "epoch": 0.38425, + "grad_norm": 0.7236169576644897, + "learning_rate": 0.003, + "loss": 3.986, + "step": 38425 + }, + { + "epoch": 0.38426, + "grad_norm": 0.8298882246017456, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 38426 + }, + { + "epoch": 0.38427, + "grad_norm": 0.9490551948547363, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 38427 + }, + { + "epoch": 0.38428, + "grad_norm": 1.033447504043579, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 38428 + }, + { + "epoch": 0.38429, + "grad_norm": 1.2315090894699097, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 38429 + }, + { + "epoch": 0.3843, + "grad_norm": 0.7350283265113831, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 38430 + }, + { + "epoch": 0.38431, + "grad_norm": 0.5931599140167236, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 38431 + }, + { + "epoch": 0.38432, + "grad_norm": 0.6650975346565247, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 38432 + }, + { + "epoch": 0.38433, + "grad_norm": 0.7542621493339539, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 38433 + }, + { + "epoch": 0.38434, + "grad_norm": 0.870905339717865, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 38434 + }, + { + "epoch": 0.38435, + "grad_norm": 1.1517324447631836, + "learning_rate": 0.003, + "loss": 3.989, + "step": 38435 + }, + { + "epoch": 0.38436, + "grad_norm": 1.045875072479248, + "learning_rate": 0.003, + "loss": 3.96, + "step": 38436 + }, + { + "epoch": 0.38437, + "grad_norm": 1.0944727659225464, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 38437 + }, + { + "epoch": 0.38438, + "grad_norm": 1.106195330619812, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 38438 + }, + { + "epoch": 0.38439, + "grad_norm": 0.8590481281280518, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 38439 + }, + { + "epoch": 0.3844, + "grad_norm": 0.7496911287307739, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 38440 + }, + { + "epoch": 0.38441, + "grad_norm": 0.924622654914856, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 38441 + }, + { + "epoch": 0.38442, + "grad_norm": 0.9976384043693542, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 38442 + }, + { + "epoch": 0.38443, + "grad_norm": 0.8726816773414612, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 38443 + }, + { + "epoch": 0.38444, + "grad_norm": 0.8520724177360535, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 38444 + }, + { + "epoch": 0.38445, + "grad_norm": 0.9180564284324646, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 38445 + }, + { + "epoch": 0.38446, + "grad_norm": 1.0336474180221558, + "learning_rate": 0.003, + "loss": 3.961, + "step": 38446 + }, + { + "epoch": 0.38447, + "grad_norm": 1.0148847103118896, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 38447 + }, + { + "epoch": 0.38448, + "grad_norm": 1.0678720474243164, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 38448 + }, + { + "epoch": 0.38449, + "grad_norm": 1.02201509475708, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 38449 + }, + { + "epoch": 0.3845, + "grad_norm": 0.8893022537231445, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 38450 + }, + { + "epoch": 0.38451, + "grad_norm": 0.713106632232666, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 38451 + }, + { + "epoch": 0.38452, + "grad_norm": 0.5821948051452637, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 38452 + }, + { + "epoch": 0.38453, + "grad_norm": 0.567264199256897, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 38453 + }, + { + "epoch": 0.38454, + "grad_norm": 0.5866855978965759, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 38454 + }, + { + "epoch": 0.38455, + "grad_norm": 0.6830258965492249, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 38455 + }, + { + "epoch": 0.38456, + "grad_norm": 0.9152258634567261, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 38456 + }, + { + "epoch": 0.38457, + "grad_norm": 1.1223019361495972, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 38457 + }, + { + "epoch": 0.38458, + "grad_norm": 1.0869678258895874, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 38458 + }, + { + "epoch": 0.38459, + "grad_norm": 1.0536309480667114, + "learning_rate": 0.003, + "loss": 3.972, + "step": 38459 + }, + { + "epoch": 0.3846, + "grad_norm": 1.0492520332336426, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 38460 + }, + { + "epoch": 0.38461, + "grad_norm": 0.7748748064041138, + "learning_rate": 0.003, + "loss": 3.96, + "step": 38461 + }, + { + "epoch": 0.38462, + "grad_norm": 0.7013260126113892, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 38462 + }, + { + "epoch": 0.38463, + "grad_norm": 0.9340742826461792, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 38463 + }, + { + "epoch": 0.38464, + "grad_norm": 1.2623682022094727, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 38464 + }, + { + "epoch": 0.38465, + "grad_norm": 1.0056977272033691, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 38465 + }, + { + "epoch": 0.38466, + "grad_norm": 0.9996942281723022, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 38466 + }, + { + "epoch": 0.38467, + "grad_norm": 1.1063915491104126, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 38467 + }, + { + "epoch": 0.38468, + "grad_norm": 0.9829152226448059, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 38468 + }, + { + "epoch": 0.38469, + "grad_norm": 0.9072749614715576, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 38469 + }, + { + "epoch": 0.3847, + "grad_norm": 0.865738570690155, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 38470 + }, + { + "epoch": 0.38471, + "grad_norm": 0.9651132822036743, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 38471 + }, + { + "epoch": 0.38472, + "grad_norm": 0.9638407230377197, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 38472 + }, + { + "epoch": 0.38473, + "grad_norm": 0.9442453980445862, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 38473 + }, + { + "epoch": 0.38474, + "grad_norm": 0.9965011477470398, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 38474 + }, + { + "epoch": 0.38475, + "grad_norm": 1.1352710723876953, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 38475 + }, + { + "epoch": 0.38476, + "grad_norm": 0.9940089583396912, + "learning_rate": 0.003, + "loss": 3.984, + "step": 38476 + }, + { + "epoch": 0.38477, + "grad_norm": 1.155837059020996, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 38477 + }, + { + "epoch": 0.38478, + "grad_norm": 0.8208484053611755, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 38478 + }, + { + "epoch": 0.38479, + "grad_norm": 0.7580850124359131, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 38479 + }, + { + "epoch": 0.3848, + "grad_norm": 0.7305720448493958, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 38480 + }, + { + "epoch": 0.38481, + "grad_norm": 0.854576051235199, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 38481 + }, + { + "epoch": 0.38482, + "grad_norm": 1.0792075395584106, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 38482 + }, + { + "epoch": 0.38483, + "grad_norm": 0.8910403847694397, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 38483 + }, + { + "epoch": 0.38484, + "grad_norm": 1.023315668106079, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 38484 + }, + { + "epoch": 0.38485, + "grad_norm": 1.1830452680587769, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 38485 + }, + { + "epoch": 0.38486, + "grad_norm": 0.8015562295913696, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 38486 + }, + { + "epoch": 0.38487, + "grad_norm": 0.7222402691841125, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 38487 + }, + { + "epoch": 0.38488, + "grad_norm": 0.6515767574310303, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 38488 + }, + { + "epoch": 0.38489, + "grad_norm": 0.6503364443778992, + "learning_rate": 0.003, + "loss": 3.972, + "step": 38489 + }, + { + "epoch": 0.3849, + "grad_norm": 0.7017368674278259, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 38490 + }, + { + "epoch": 0.38491, + "grad_norm": 0.8128203749656677, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 38491 + }, + { + "epoch": 0.38492, + "grad_norm": 0.8400828242301941, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 38492 + }, + { + "epoch": 0.38493, + "grad_norm": 0.9476539492607117, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 38493 + }, + { + "epoch": 0.38494, + "grad_norm": 1.3184000253677368, + "learning_rate": 0.003, + "loss": 3.984, + "step": 38494 + }, + { + "epoch": 0.38495, + "grad_norm": 0.652495801448822, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 38495 + }, + { + "epoch": 0.38496, + "grad_norm": 0.6434990167617798, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 38496 + }, + { + "epoch": 0.38497, + "grad_norm": 1.042372226715088, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 38497 + }, + { + "epoch": 0.38498, + "grad_norm": 1.089996576309204, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 38498 + }, + { + "epoch": 0.38499, + "grad_norm": 0.8196654319763184, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 38499 + }, + { + "epoch": 0.385, + "grad_norm": 0.7648791670799255, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 38500 + }, + { + "epoch": 0.38501, + "grad_norm": 0.754290759563446, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 38501 + }, + { + "epoch": 0.38502, + "grad_norm": 0.739398717880249, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 38502 + }, + { + "epoch": 0.38503, + "grad_norm": 0.8227025866508484, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 38503 + }, + { + "epoch": 0.38504, + "grad_norm": 0.8199995160102844, + "learning_rate": 0.003, + "loss": 3.9447, + "step": 38504 + }, + { + "epoch": 0.38505, + "grad_norm": 0.9539172649383545, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 38505 + }, + { + "epoch": 0.38506, + "grad_norm": 1.170284628868103, + "learning_rate": 0.003, + "loss": 3.9258, + "step": 38506 + }, + { + "epoch": 0.38507, + "grad_norm": 0.812193751335144, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 38507 + }, + { + "epoch": 0.38508, + "grad_norm": 0.7323534488677979, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 38508 + }, + { + "epoch": 0.38509, + "grad_norm": 0.7423239350318909, + "learning_rate": 0.003, + "loss": 3.975, + "step": 38509 + }, + { + "epoch": 0.3851, + "grad_norm": 0.7764250636100769, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 38510 + }, + { + "epoch": 0.38511, + "grad_norm": 1.0831793546676636, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 38511 + }, + { + "epoch": 0.38512, + "grad_norm": 1.3440861701965332, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 38512 + }, + { + "epoch": 0.38513, + "grad_norm": 0.7473866939544678, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 38513 + }, + { + "epoch": 0.38514, + "grad_norm": 0.6439416408538818, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 38514 + }, + { + "epoch": 0.38515, + "grad_norm": 0.6833723187446594, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 38515 + }, + { + "epoch": 0.38516, + "grad_norm": 0.791093111038208, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 38516 + }, + { + "epoch": 0.38517, + "grad_norm": 0.8358370065689087, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 38517 + }, + { + "epoch": 0.38518, + "grad_norm": 0.8211867213249207, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 38518 + }, + { + "epoch": 0.38519, + "grad_norm": 0.7998371720314026, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 38519 + }, + { + "epoch": 0.3852, + "grad_norm": 1.0818960666656494, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 38520 + }, + { + "epoch": 0.38521, + "grad_norm": 1.2727305889129639, + "learning_rate": 0.003, + "loss": 3.954, + "step": 38521 + }, + { + "epoch": 0.38522, + "grad_norm": 0.6841594576835632, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 38522 + }, + { + "epoch": 0.38523, + "grad_norm": 0.6100098490715027, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 38523 + }, + { + "epoch": 0.38524, + "grad_norm": 0.6801307797431946, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 38524 + }, + { + "epoch": 0.38525, + "grad_norm": 0.6874220371246338, + "learning_rate": 0.003, + "loss": 3.9471, + "step": 38525 + }, + { + "epoch": 0.38526, + "grad_norm": 0.6985601782798767, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 38526 + }, + { + "epoch": 0.38527, + "grad_norm": 0.7927141785621643, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 38527 + }, + { + "epoch": 0.38528, + "grad_norm": 0.876914918422699, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 38528 + }, + { + "epoch": 0.38529, + "grad_norm": 1.0817523002624512, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 38529 + }, + { + "epoch": 0.3853, + "grad_norm": 1.0307248830795288, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 38530 + }, + { + "epoch": 0.38531, + "grad_norm": 1.3144851922988892, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 38531 + }, + { + "epoch": 0.38532, + "grad_norm": 0.9658463001251221, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 38532 + }, + { + "epoch": 0.38533, + "grad_norm": 0.9150742292404175, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 38533 + }, + { + "epoch": 0.38534, + "grad_norm": 0.8327407240867615, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 38534 + }, + { + "epoch": 0.38535, + "grad_norm": 0.9013972878456116, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 38535 + }, + { + "epoch": 0.38536, + "grad_norm": 0.9722267985343933, + "learning_rate": 0.003, + "loss": 3.988, + "step": 38536 + }, + { + "epoch": 0.38537, + "grad_norm": 0.8344037532806396, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 38537 + }, + { + "epoch": 0.38538, + "grad_norm": 0.7434216141700745, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 38538 + }, + { + "epoch": 0.38539, + "grad_norm": 0.8568575382232666, + "learning_rate": 0.003, + "loss": 3.985, + "step": 38539 + }, + { + "epoch": 0.3854, + "grad_norm": 0.8109422326087952, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 38540 + }, + { + "epoch": 0.38541, + "grad_norm": 0.9206395149230957, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 38541 + }, + { + "epoch": 0.38542, + "grad_norm": 1.080257534980774, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 38542 + }, + { + "epoch": 0.38543, + "grad_norm": 1.1319166421890259, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 38543 + }, + { + "epoch": 0.38544, + "grad_norm": 1.0924649238586426, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 38544 + }, + { + "epoch": 0.38545, + "grad_norm": 1.0664780139923096, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 38545 + }, + { + "epoch": 0.38546, + "grad_norm": 0.8787198662757874, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 38546 + }, + { + "epoch": 0.38547, + "grad_norm": 0.8353567123413086, + "learning_rate": 0.003, + "loss": 3.988, + "step": 38547 + }, + { + "epoch": 0.38548, + "grad_norm": 0.9586689472198486, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 38548 + }, + { + "epoch": 0.38549, + "grad_norm": 1.1386511325836182, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 38549 + }, + { + "epoch": 0.3855, + "grad_norm": 1.0978554487228394, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 38550 + }, + { + "epoch": 0.38551, + "grad_norm": 1.1158026456832886, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 38551 + }, + { + "epoch": 0.38552, + "grad_norm": 1.106210470199585, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 38552 + }, + { + "epoch": 0.38553, + "grad_norm": 0.933903694152832, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 38553 + }, + { + "epoch": 0.38554, + "grad_norm": 0.9196327924728394, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 38554 + }, + { + "epoch": 0.38555, + "grad_norm": 0.9610607624053955, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 38555 + }, + { + "epoch": 0.38556, + "grad_norm": 0.8179003596305847, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 38556 + }, + { + "epoch": 0.38557, + "grad_norm": 0.7087590098381042, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 38557 + }, + { + "epoch": 0.38558, + "grad_norm": 0.738856315612793, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 38558 + }, + { + "epoch": 0.38559, + "grad_norm": 0.8735061883926392, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 38559 + }, + { + "epoch": 0.3856, + "grad_norm": 1.0366226434707642, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 38560 + }, + { + "epoch": 0.38561, + "grad_norm": 1.1719951629638672, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 38561 + }, + { + "epoch": 0.38562, + "grad_norm": 0.6686708927154541, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 38562 + }, + { + "epoch": 0.38563, + "grad_norm": 0.5825955271720886, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 38563 + }, + { + "epoch": 0.38564, + "grad_norm": 0.6870729327201843, + "learning_rate": 0.003, + "loss": 3.9488, + "step": 38564 + }, + { + "epoch": 0.38565, + "grad_norm": 0.6571283340454102, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 38565 + }, + { + "epoch": 0.38566, + "grad_norm": 0.7397109270095825, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 38566 + }, + { + "epoch": 0.38567, + "grad_norm": 1.0152555704116821, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 38567 + }, + { + "epoch": 0.38568, + "grad_norm": 1.4559413194656372, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 38568 + }, + { + "epoch": 0.38569, + "grad_norm": 0.5543215870857239, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 38569 + }, + { + "epoch": 0.3857, + "grad_norm": 0.8246952295303345, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 38570 + }, + { + "epoch": 0.38571, + "grad_norm": 0.9759604334831238, + "learning_rate": 0.003, + "loss": 3.992, + "step": 38571 + }, + { + "epoch": 0.38572, + "grad_norm": 0.9922101497650146, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 38572 + }, + { + "epoch": 0.38573, + "grad_norm": 1.0648844242095947, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 38573 + }, + { + "epoch": 0.38574, + "grad_norm": 0.8777400851249695, + "learning_rate": 0.003, + "loss": 3.985, + "step": 38574 + }, + { + "epoch": 0.38575, + "grad_norm": 0.9629040956497192, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 38575 + }, + { + "epoch": 0.38576, + "grad_norm": 1.0928984880447388, + "learning_rate": 0.003, + "loss": 3.9367, + "step": 38576 + }, + { + "epoch": 0.38577, + "grad_norm": 0.9625000953674316, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 38577 + }, + { + "epoch": 0.38578, + "grad_norm": 0.9393990635871887, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 38578 + }, + { + "epoch": 0.38579, + "grad_norm": 1.1628785133361816, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 38579 + }, + { + "epoch": 0.3858, + "grad_norm": 1.0021065473556519, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 38580 + }, + { + "epoch": 0.38581, + "grad_norm": 0.9764000773429871, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 38581 + }, + { + "epoch": 0.38582, + "grad_norm": 0.9858078956604004, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 38582 + }, + { + "epoch": 0.38583, + "grad_norm": 1.0584899187088013, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 38583 + }, + { + "epoch": 0.38584, + "grad_norm": 1.0056687593460083, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 38584 + }, + { + "epoch": 0.38585, + "grad_norm": 1.0030699968338013, + "learning_rate": 0.003, + "loss": 3.974, + "step": 38585 + }, + { + "epoch": 0.38586, + "grad_norm": 0.9141734838485718, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 38586 + }, + { + "epoch": 0.38587, + "grad_norm": 0.8080810308456421, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 38587 + }, + { + "epoch": 0.38588, + "grad_norm": 0.8930972218513489, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 38588 + }, + { + "epoch": 0.38589, + "grad_norm": 0.983237087726593, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 38589 + }, + { + "epoch": 0.3859, + "grad_norm": 1.052630066871643, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 38590 + }, + { + "epoch": 0.38591, + "grad_norm": 1.0034561157226562, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 38591 + }, + { + "epoch": 0.38592, + "grad_norm": 0.9560551047325134, + "learning_rate": 0.003, + "loss": 3.997, + "step": 38592 + }, + { + "epoch": 0.38593, + "grad_norm": 1.1452808380126953, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 38593 + }, + { + "epoch": 0.38594, + "grad_norm": 1.02043879032135, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 38594 + }, + { + "epoch": 0.38595, + "grad_norm": 0.8800731301307678, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 38595 + }, + { + "epoch": 0.38596, + "grad_norm": 0.662095844745636, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 38596 + }, + { + "epoch": 0.38597, + "grad_norm": 0.604276180267334, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 38597 + }, + { + "epoch": 0.38598, + "grad_norm": 0.7030819058418274, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 38598 + }, + { + "epoch": 0.38599, + "grad_norm": 0.8167991042137146, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 38599 + }, + { + "epoch": 0.386, + "grad_norm": 0.9479365944862366, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 38600 + }, + { + "epoch": 0.38601, + "grad_norm": 0.9962188005447388, + "learning_rate": 0.003, + "loss": 4.0, + "step": 38601 + }, + { + "epoch": 0.38602, + "grad_norm": 1.231086254119873, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 38602 + }, + { + "epoch": 0.38603, + "grad_norm": 0.7233859896659851, + "learning_rate": 0.003, + "loss": 3.978, + "step": 38603 + }, + { + "epoch": 0.38604, + "grad_norm": 0.6969234347343445, + "learning_rate": 0.003, + "loss": 3.9483, + "step": 38604 + }, + { + "epoch": 0.38605, + "grad_norm": 0.8414102792739868, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 38605 + }, + { + "epoch": 0.38606, + "grad_norm": 1.027300477027893, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 38606 + }, + { + "epoch": 0.38607, + "grad_norm": 1.103379726409912, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 38607 + }, + { + "epoch": 0.38608, + "grad_norm": 0.8092730045318604, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 38608 + }, + { + "epoch": 0.38609, + "grad_norm": 0.8652795553207397, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 38609 + }, + { + "epoch": 0.3861, + "grad_norm": 0.9257394671440125, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 38610 + }, + { + "epoch": 0.38611, + "grad_norm": 1.0297167301177979, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 38611 + }, + { + "epoch": 0.38612, + "grad_norm": 1.149564266204834, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 38612 + }, + { + "epoch": 0.38613, + "grad_norm": 0.8431012034416199, + "learning_rate": 0.003, + "loss": 3.956, + "step": 38613 + }, + { + "epoch": 0.38614, + "grad_norm": 0.7261300086975098, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 38614 + }, + { + "epoch": 0.38615, + "grad_norm": 0.651387095451355, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 38615 + }, + { + "epoch": 0.38616, + "grad_norm": 0.6981725096702576, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 38616 + }, + { + "epoch": 0.38617, + "grad_norm": 0.7857782244682312, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 38617 + }, + { + "epoch": 0.38618, + "grad_norm": 0.9983789920806885, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 38618 + }, + { + "epoch": 0.38619, + "grad_norm": 1.3327641487121582, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 38619 + }, + { + "epoch": 0.3862, + "grad_norm": 0.5938906073570251, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 38620 + }, + { + "epoch": 0.38621, + "grad_norm": 0.7128418684005737, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 38621 + }, + { + "epoch": 0.38622, + "grad_norm": 0.9934206008911133, + "learning_rate": 0.003, + "loss": 3.97, + "step": 38622 + }, + { + "epoch": 0.38623, + "grad_norm": 1.3068675994873047, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 38623 + }, + { + "epoch": 0.38624, + "grad_norm": 0.8051750659942627, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 38624 + }, + { + "epoch": 0.38625, + "grad_norm": 0.8340904712677002, + "learning_rate": 0.003, + "loss": 3.9378, + "step": 38625 + }, + { + "epoch": 0.38626, + "grad_norm": 0.8563154935836792, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 38626 + }, + { + "epoch": 0.38627, + "grad_norm": 0.7987816333770752, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 38627 + }, + { + "epoch": 0.38628, + "grad_norm": 0.7713271975517273, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 38628 + }, + { + "epoch": 0.38629, + "grad_norm": 0.8240928053855896, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 38629 + }, + { + "epoch": 0.3863, + "grad_norm": 0.7704068422317505, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 38630 + }, + { + "epoch": 0.38631, + "grad_norm": 0.6861779689788818, + "learning_rate": 0.003, + "loss": 3.9489, + "step": 38631 + }, + { + "epoch": 0.38632, + "grad_norm": 0.6697319149971008, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 38632 + }, + { + "epoch": 0.38633, + "grad_norm": 0.8457853198051453, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 38633 + }, + { + "epoch": 0.38634, + "grad_norm": 1.0122599601745605, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 38634 + }, + { + "epoch": 0.38635, + "grad_norm": 0.9440081119537354, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 38635 + }, + { + "epoch": 0.38636, + "grad_norm": 0.901064932346344, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 38636 + }, + { + "epoch": 0.38637, + "grad_norm": 0.9667884707450867, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 38637 + }, + { + "epoch": 0.38638, + "grad_norm": 1.1937124729156494, + "learning_rate": 0.003, + "loss": 3.984, + "step": 38638 + }, + { + "epoch": 0.38639, + "grad_norm": 1.0661896467208862, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 38639 + }, + { + "epoch": 0.3864, + "grad_norm": 1.0616787672042847, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 38640 + }, + { + "epoch": 0.38641, + "grad_norm": 1.090808629989624, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 38641 + }, + { + "epoch": 0.38642, + "grad_norm": 0.8847421407699585, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 38642 + }, + { + "epoch": 0.38643, + "grad_norm": 0.98356032371521, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 38643 + }, + { + "epoch": 0.38644, + "grad_norm": 1.0893596410751343, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 38644 + }, + { + "epoch": 0.38645, + "grad_norm": 1.2311078310012817, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 38645 + }, + { + "epoch": 0.38646, + "grad_norm": 0.9755090475082397, + "learning_rate": 0.003, + "loss": 3.994, + "step": 38646 + }, + { + "epoch": 0.38647, + "grad_norm": 1.0722001791000366, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 38647 + }, + { + "epoch": 0.38648, + "grad_norm": 0.8424494862556458, + "learning_rate": 0.003, + "loss": 3.982, + "step": 38648 + }, + { + "epoch": 0.38649, + "grad_norm": 0.8337389826774597, + "learning_rate": 0.003, + "loss": 3.986, + "step": 38649 + }, + { + "epoch": 0.3865, + "grad_norm": 0.9370608925819397, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 38650 + }, + { + "epoch": 0.38651, + "grad_norm": 1.1036995649337769, + "learning_rate": 0.003, + "loss": 4.0, + "step": 38651 + }, + { + "epoch": 0.38652, + "grad_norm": 1.0246516466140747, + "learning_rate": 0.003, + "loss": 3.998, + "step": 38652 + }, + { + "epoch": 0.38653, + "grad_norm": 1.0471192598342896, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 38653 + }, + { + "epoch": 0.38654, + "grad_norm": 0.9705522060394287, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 38654 + }, + { + "epoch": 0.38655, + "grad_norm": 0.7954696416854858, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 38655 + }, + { + "epoch": 0.38656, + "grad_norm": 0.8204635381698608, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 38656 + }, + { + "epoch": 0.38657, + "grad_norm": 0.9789750576019287, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 38657 + }, + { + "epoch": 0.38658, + "grad_norm": 1.042984962463379, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 38658 + }, + { + "epoch": 0.38659, + "grad_norm": 1.0067189931869507, + "learning_rate": 0.003, + "loss": 3.99, + "step": 38659 + }, + { + "epoch": 0.3866, + "grad_norm": 1.0960693359375, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 38660 + }, + { + "epoch": 0.38661, + "grad_norm": 0.8315709233283997, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 38661 + }, + { + "epoch": 0.38662, + "grad_norm": 0.680634617805481, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 38662 + }, + { + "epoch": 0.38663, + "grad_norm": 0.6197035908699036, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 38663 + }, + { + "epoch": 0.38664, + "grad_norm": 0.6176162958145142, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 38664 + }, + { + "epoch": 0.38665, + "grad_norm": 0.7477765679359436, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 38665 + }, + { + "epoch": 0.38666, + "grad_norm": 0.9129388928413391, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 38666 + }, + { + "epoch": 0.38667, + "grad_norm": 1.2079048156738281, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 38667 + }, + { + "epoch": 0.38668, + "grad_norm": 0.9355792999267578, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 38668 + }, + { + "epoch": 0.38669, + "grad_norm": 0.8369367122650146, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 38669 + }, + { + "epoch": 0.3867, + "grad_norm": 0.8046422600746155, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 38670 + }, + { + "epoch": 0.38671, + "grad_norm": 0.8307945728302002, + "learning_rate": 0.003, + "loss": 3.992, + "step": 38671 + }, + { + "epoch": 0.38672, + "grad_norm": 0.704310953617096, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 38672 + }, + { + "epoch": 0.38673, + "grad_norm": 0.6878207921981812, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 38673 + }, + { + "epoch": 0.38674, + "grad_norm": 0.6115278601646423, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 38674 + }, + { + "epoch": 0.38675, + "grad_norm": 0.6319694519042969, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 38675 + }, + { + "epoch": 0.38676, + "grad_norm": 0.6718556880950928, + "learning_rate": 0.003, + "loss": 3.964, + "step": 38676 + }, + { + "epoch": 0.38677, + "grad_norm": 0.8401044011116028, + "learning_rate": 0.003, + "loss": 3.9474, + "step": 38677 + }, + { + "epoch": 0.38678, + "grad_norm": 1.0141104459762573, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 38678 + }, + { + "epoch": 0.38679, + "grad_norm": 1.1724162101745605, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 38679 + }, + { + "epoch": 0.3868, + "grad_norm": 0.8630922436714172, + "learning_rate": 0.003, + "loss": 3.9337, + "step": 38680 + }, + { + "epoch": 0.38681, + "grad_norm": 0.7584403157234192, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 38681 + }, + { + "epoch": 0.38682, + "grad_norm": 0.8095771074295044, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 38682 + }, + { + "epoch": 0.38683, + "grad_norm": 0.8323969841003418, + "learning_rate": 0.003, + "loss": 3.9373, + "step": 38683 + }, + { + "epoch": 0.38684, + "grad_norm": 0.8326756954193115, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 38684 + }, + { + "epoch": 0.38685, + "grad_norm": 1.0926761627197266, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 38685 + }, + { + "epoch": 0.38686, + "grad_norm": 1.2205461263656616, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 38686 + }, + { + "epoch": 0.38687, + "grad_norm": 0.8584131002426147, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 38687 + }, + { + "epoch": 0.38688, + "grad_norm": 0.889156699180603, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 38688 + }, + { + "epoch": 0.38689, + "grad_norm": 0.8763911724090576, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 38689 + }, + { + "epoch": 0.3869, + "grad_norm": 0.8116421699523926, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 38690 + }, + { + "epoch": 0.38691, + "grad_norm": 0.8249850869178772, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 38691 + }, + { + "epoch": 0.38692, + "grad_norm": 1.0637396574020386, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 38692 + }, + { + "epoch": 0.38693, + "grad_norm": 1.1088933944702148, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 38693 + }, + { + "epoch": 0.38694, + "grad_norm": 0.8537938594818115, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 38694 + }, + { + "epoch": 0.38695, + "grad_norm": 0.7932032942771912, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 38695 + }, + { + "epoch": 0.38696, + "grad_norm": 0.7666738033294678, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 38696 + }, + { + "epoch": 0.38697, + "grad_norm": 0.7524712681770325, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 38697 + }, + { + "epoch": 0.38698, + "grad_norm": 0.8157795071601868, + "learning_rate": 0.003, + "loss": 3.971, + "step": 38698 + }, + { + "epoch": 0.38699, + "grad_norm": 0.9346916675567627, + "learning_rate": 0.003, + "loss": 3.968, + "step": 38699 + }, + { + "epoch": 0.387, + "grad_norm": 0.9176385402679443, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 38700 + }, + { + "epoch": 0.38701, + "grad_norm": 0.8000431060791016, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 38701 + }, + { + "epoch": 0.38702, + "grad_norm": 0.8097980618476868, + "learning_rate": 0.003, + "loss": 3.9319, + "step": 38702 + }, + { + "epoch": 0.38703, + "grad_norm": 1.0555455684661865, + "learning_rate": 0.003, + "loss": 3.9385, + "step": 38703 + }, + { + "epoch": 0.38704, + "grad_norm": 1.2261765003204346, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 38704 + }, + { + "epoch": 0.38705, + "grad_norm": 0.7881658673286438, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 38705 + }, + { + "epoch": 0.38706, + "grad_norm": 0.8303794860839844, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 38706 + }, + { + "epoch": 0.38707, + "grad_norm": 0.899113118648529, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 38707 + }, + { + "epoch": 0.38708, + "grad_norm": 0.9108608365058899, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 38708 + }, + { + "epoch": 0.38709, + "grad_norm": 0.9059717655181885, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 38709 + }, + { + "epoch": 0.3871, + "grad_norm": 0.8282654285430908, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 38710 + }, + { + "epoch": 0.38711, + "grad_norm": 0.8222167491912842, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 38711 + }, + { + "epoch": 0.38712, + "grad_norm": 0.9207999110221863, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 38712 + }, + { + "epoch": 0.38713, + "grad_norm": 0.8657896518707275, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 38713 + }, + { + "epoch": 0.38714, + "grad_norm": 0.8204954266548157, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 38714 + }, + { + "epoch": 0.38715, + "grad_norm": 0.9450134038925171, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 38715 + }, + { + "epoch": 0.38716, + "grad_norm": 1.3547106981277466, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 38716 + }, + { + "epoch": 0.38717, + "grad_norm": 0.9508962631225586, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 38717 + }, + { + "epoch": 0.38718, + "grad_norm": 1.0391947031021118, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 38718 + }, + { + "epoch": 0.38719, + "grad_norm": 1.2121869325637817, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 38719 + }, + { + "epoch": 0.3872, + "grad_norm": 0.7384419441223145, + "learning_rate": 0.003, + "loss": 3.971, + "step": 38720 + }, + { + "epoch": 0.38721, + "grad_norm": 0.6471954584121704, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 38721 + }, + { + "epoch": 0.38722, + "grad_norm": 0.6616935729980469, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 38722 + }, + { + "epoch": 0.38723, + "grad_norm": 0.7164633274078369, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 38723 + }, + { + "epoch": 0.38724, + "grad_norm": 0.7053360342979431, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 38724 + }, + { + "epoch": 0.38725, + "grad_norm": 0.7058894634246826, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 38725 + }, + { + "epoch": 0.38726, + "grad_norm": 0.7139086723327637, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 38726 + }, + { + "epoch": 0.38727, + "grad_norm": 0.8040111064910889, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 38727 + }, + { + "epoch": 0.38728, + "grad_norm": 0.914411723613739, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 38728 + }, + { + "epoch": 0.38729, + "grad_norm": 1.444252610206604, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 38729 + }, + { + "epoch": 0.3873, + "grad_norm": 0.8470168113708496, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 38730 + }, + { + "epoch": 0.38731, + "grad_norm": 0.8429485559463501, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 38731 + }, + { + "epoch": 0.38732, + "grad_norm": 0.9067779779434204, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 38732 + }, + { + "epoch": 0.38733, + "grad_norm": 0.855366587638855, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 38733 + }, + { + "epoch": 0.38734, + "grad_norm": 0.9063640832901001, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 38734 + }, + { + "epoch": 0.38735, + "grad_norm": 0.8696318864822388, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 38735 + }, + { + "epoch": 0.38736, + "grad_norm": 0.9025297164916992, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 38736 + }, + { + "epoch": 0.38737, + "grad_norm": 0.9670706987380981, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 38737 + }, + { + "epoch": 0.38738, + "grad_norm": 0.989773690700531, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 38738 + }, + { + "epoch": 0.38739, + "grad_norm": 0.9456570148468018, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 38739 + }, + { + "epoch": 0.3874, + "grad_norm": 1.0578227043151855, + "learning_rate": 0.003, + "loss": 3.988, + "step": 38740 + }, + { + "epoch": 0.38741, + "grad_norm": 0.9725553393363953, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 38741 + }, + { + "epoch": 0.38742, + "grad_norm": 1.1307238340377808, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 38742 + }, + { + "epoch": 0.38743, + "grad_norm": 0.9671682715415955, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 38743 + }, + { + "epoch": 0.38744, + "grad_norm": 0.7626729011535645, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 38744 + }, + { + "epoch": 0.38745, + "grad_norm": 0.6240810751914978, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 38745 + }, + { + "epoch": 0.38746, + "grad_norm": 0.6764154434204102, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 38746 + }, + { + "epoch": 0.38747, + "grad_norm": 0.7125702500343323, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 38747 + }, + { + "epoch": 0.38748, + "grad_norm": 0.7444522976875305, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 38748 + }, + { + "epoch": 0.38749, + "grad_norm": 0.8885977864265442, + "learning_rate": 0.003, + "loss": 3.9453, + "step": 38749 + }, + { + "epoch": 0.3875, + "grad_norm": 1.3824093341827393, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 38750 + }, + { + "epoch": 0.38751, + "grad_norm": 0.8049081563949585, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 38751 + }, + { + "epoch": 0.38752, + "grad_norm": 0.6922460198402405, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 38752 + }, + { + "epoch": 0.38753, + "grad_norm": 0.7561982870101929, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 38753 + }, + { + "epoch": 0.38754, + "grad_norm": 0.9086295366287231, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 38754 + }, + { + "epoch": 0.38755, + "grad_norm": 0.9362233877182007, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 38755 + }, + { + "epoch": 0.38756, + "grad_norm": 1.0943597555160522, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 38756 + }, + { + "epoch": 0.38757, + "grad_norm": 0.9837122559547424, + "learning_rate": 0.003, + "loss": 4.004, + "step": 38757 + }, + { + "epoch": 0.38758, + "grad_norm": 0.9654169678688049, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 38758 + }, + { + "epoch": 0.38759, + "grad_norm": 0.9569787979125977, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 38759 + }, + { + "epoch": 0.3876, + "grad_norm": 1.054046630859375, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 38760 + }, + { + "epoch": 0.38761, + "grad_norm": 0.9195945262908936, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 38761 + }, + { + "epoch": 0.38762, + "grad_norm": 0.8106651306152344, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 38762 + }, + { + "epoch": 0.38763, + "grad_norm": 0.9038506746292114, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 38763 + }, + { + "epoch": 0.38764, + "grad_norm": 1.0094445943832397, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 38764 + }, + { + "epoch": 0.38765, + "grad_norm": 1.3738707304000854, + "learning_rate": 0.003, + "loss": 3.997, + "step": 38765 + }, + { + "epoch": 0.38766, + "grad_norm": 0.8160242438316345, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 38766 + }, + { + "epoch": 0.38767, + "grad_norm": 0.6940240859985352, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 38767 + }, + { + "epoch": 0.38768, + "grad_norm": 0.7603175044059753, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 38768 + }, + { + "epoch": 0.38769, + "grad_norm": 0.9330071806907654, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 38769 + }, + { + "epoch": 0.3877, + "grad_norm": 1.0022011995315552, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 38770 + }, + { + "epoch": 0.38771, + "grad_norm": 1.2987905740737915, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 38771 + }, + { + "epoch": 0.38772, + "grad_norm": 0.7867817282676697, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 38772 + }, + { + "epoch": 0.38773, + "grad_norm": 0.6630112528800964, + "learning_rate": 0.003, + "loss": 3.971, + "step": 38773 + }, + { + "epoch": 0.38774, + "grad_norm": 0.729417085647583, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 38774 + }, + { + "epoch": 0.38775, + "grad_norm": 0.9015620350837708, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 38775 + }, + { + "epoch": 0.38776, + "grad_norm": 1.0018270015716553, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 38776 + }, + { + "epoch": 0.38777, + "grad_norm": 1.2149490118026733, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 38777 + }, + { + "epoch": 0.38778, + "grad_norm": 0.8259425163269043, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 38778 + }, + { + "epoch": 0.38779, + "grad_norm": 0.6154637932777405, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 38779 + }, + { + "epoch": 0.3878, + "grad_norm": 0.6090474128723145, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 38780 + }, + { + "epoch": 0.38781, + "grad_norm": 0.6422292590141296, + "learning_rate": 0.003, + "loss": 3.9446, + "step": 38781 + }, + { + "epoch": 0.38782, + "grad_norm": 0.6062633395195007, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 38782 + }, + { + "epoch": 0.38783, + "grad_norm": 0.6904019117355347, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 38783 + }, + { + "epoch": 0.38784, + "grad_norm": 0.7940608859062195, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 38784 + }, + { + "epoch": 0.38785, + "grad_norm": 1.0583972930908203, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 38785 + }, + { + "epoch": 0.38786, + "grad_norm": 1.1959130764007568, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 38786 + }, + { + "epoch": 0.38787, + "grad_norm": 0.6676854491233826, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 38787 + }, + { + "epoch": 0.38788, + "grad_norm": 0.5909109711647034, + "learning_rate": 0.003, + "loss": 3.982, + "step": 38788 + }, + { + "epoch": 0.38789, + "grad_norm": 0.5814877152442932, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 38789 + }, + { + "epoch": 0.3879, + "grad_norm": 0.5814769268035889, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 38790 + }, + { + "epoch": 0.38791, + "grad_norm": 0.6568580269813538, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 38791 + }, + { + "epoch": 0.38792, + "grad_norm": 0.7361518144607544, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 38792 + }, + { + "epoch": 0.38793, + "grad_norm": 0.8631564974784851, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 38793 + }, + { + "epoch": 0.38794, + "grad_norm": 0.9255950450897217, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 38794 + }, + { + "epoch": 0.38795, + "grad_norm": 0.9400742053985596, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 38795 + }, + { + "epoch": 0.38796, + "grad_norm": 0.9591965675354004, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 38796 + }, + { + "epoch": 0.38797, + "grad_norm": 1.0276572704315186, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 38797 + }, + { + "epoch": 0.38798, + "grad_norm": 1.076878547668457, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 38798 + }, + { + "epoch": 0.38799, + "grad_norm": 1.1666568517684937, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 38799 + }, + { + "epoch": 0.388, + "grad_norm": 1.1725834608078003, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 38800 + }, + { + "epoch": 0.38801, + "grad_norm": 0.7779600024223328, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 38801 + }, + { + "epoch": 0.38802, + "grad_norm": 0.7215178608894348, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 38802 + }, + { + "epoch": 0.38803, + "grad_norm": 0.9035688042640686, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 38803 + }, + { + "epoch": 0.38804, + "grad_norm": 0.8045915961265564, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 38804 + }, + { + "epoch": 0.38805, + "grad_norm": 0.9189843535423279, + "learning_rate": 0.003, + "loss": 3.9435, + "step": 38805 + }, + { + "epoch": 0.38806, + "grad_norm": 1.1074084043502808, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 38806 + }, + { + "epoch": 0.38807, + "grad_norm": 1.1295907497406006, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 38807 + }, + { + "epoch": 0.38808, + "grad_norm": 1.0376769304275513, + "learning_rate": 0.003, + "loss": 4.004, + "step": 38808 + }, + { + "epoch": 0.38809, + "grad_norm": 0.9611020088195801, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 38809 + }, + { + "epoch": 0.3881, + "grad_norm": 1.0042603015899658, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 38810 + }, + { + "epoch": 0.38811, + "grad_norm": 1.3352019786834717, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 38811 + }, + { + "epoch": 0.38812, + "grad_norm": 0.682758629322052, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 38812 + }, + { + "epoch": 0.38813, + "grad_norm": 0.6473217606544495, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 38813 + }, + { + "epoch": 0.38814, + "grad_norm": 0.9152610898017883, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 38814 + }, + { + "epoch": 0.38815, + "grad_norm": 0.9825630784034729, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 38815 + }, + { + "epoch": 0.38816, + "grad_norm": 1.094529390335083, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 38816 + }, + { + "epoch": 0.38817, + "grad_norm": 1.0144025087356567, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 38817 + }, + { + "epoch": 0.38818, + "grad_norm": 0.8979578018188477, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 38818 + }, + { + "epoch": 0.38819, + "grad_norm": 0.7381201982498169, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 38819 + }, + { + "epoch": 0.3882, + "grad_norm": 0.6752543449401855, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 38820 + }, + { + "epoch": 0.38821, + "grad_norm": 0.6361404061317444, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 38821 + }, + { + "epoch": 0.38822, + "grad_norm": 0.6793997287750244, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 38822 + }, + { + "epoch": 0.38823, + "grad_norm": 0.8335188627243042, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 38823 + }, + { + "epoch": 0.38824, + "grad_norm": 1.1389621496200562, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 38824 + }, + { + "epoch": 0.38825, + "grad_norm": 1.1903460025787354, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 38825 + }, + { + "epoch": 0.38826, + "grad_norm": 0.9221062064170837, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 38826 + }, + { + "epoch": 0.38827, + "grad_norm": 0.9059969186782837, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 38827 + }, + { + "epoch": 0.38828, + "grad_norm": 0.8205886483192444, + "learning_rate": 0.003, + "loss": 3.954, + "step": 38828 + }, + { + "epoch": 0.38829, + "grad_norm": 0.623188853263855, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 38829 + }, + { + "epoch": 0.3883, + "grad_norm": 0.625411868095398, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 38830 + }, + { + "epoch": 0.38831, + "grad_norm": 0.6748117208480835, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 38831 + }, + { + "epoch": 0.38832, + "grad_norm": 0.6861565113067627, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 38832 + }, + { + "epoch": 0.38833, + "grad_norm": 0.6871441602706909, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 38833 + }, + { + "epoch": 0.38834, + "grad_norm": 0.7477957606315613, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 38834 + }, + { + "epoch": 0.38835, + "grad_norm": 1.0044022798538208, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 38835 + }, + { + "epoch": 0.38836, + "grad_norm": 1.4832582473754883, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 38836 + }, + { + "epoch": 0.38837, + "grad_norm": 0.45009875297546387, + "learning_rate": 0.003, + "loss": 3.9473, + "step": 38837 + }, + { + "epoch": 0.38838, + "grad_norm": 0.984885573387146, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 38838 + }, + { + "epoch": 0.38839, + "grad_norm": 1.296791911125183, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 38839 + }, + { + "epoch": 0.3884, + "grad_norm": 0.587407112121582, + "learning_rate": 0.003, + "loss": 3.974, + "step": 38840 + }, + { + "epoch": 0.38841, + "grad_norm": 0.7976229786872864, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 38841 + }, + { + "epoch": 0.38842, + "grad_norm": 0.8063929677009583, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 38842 + }, + { + "epoch": 0.38843, + "grad_norm": 0.8785653114318848, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 38843 + }, + { + "epoch": 0.38844, + "grad_norm": 0.925310492515564, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 38844 + }, + { + "epoch": 0.38845, + "grad_norm": 0.9805254936218262, + "learning_rate": 0.003, + "loss": 3.9455, + "step": 38845 + }, + { + "epoch": 0.38846, + "grad_norm": 1.1166094541549683, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 38846 + }, + { + "epoch": 0.38847, + "grad_norm": 0.9221521019935608, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 38847 + }, + { + "epoch": 0.38848, + "grad_norm": 0.9437494874000549, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 38848 + }, + { + "epoch": 0.38849, + "grad_norm": 1.0022956132888794, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 38849 + }, + { + "epoch": 0.3885, + "grad_norm": 0.9881499409675598, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 38850 + }, + { + "epoch": 0.38851, + "grad_norm": 1.133674144744873, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 38851 + }, + { + "epoch": 0.38852, + "grad_norm": 0.8757269382476807, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 38852 + }, + { + "epoch": 0.38853, + "grad_norm": 0.9263563752174377, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 38853 + }, + { + "epoch": 0.38854, + "grad_norm": 0.9549160003662109, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 38854 + }, + { + "epoch": 0.38855, + "grad_norm": 1.0858711004257202, + "learning_rate": 0.003, + "loss": 3.991, + "step": 38855 + }, + { + "epoch": 0.38856, + "grad_norm": 0.8767210841178894, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 38856 + }, + { + "epoch": 0.38857, + "grad_norm": 0.9422940015792847, + "learning_rate": 0.003, + "loss": 3.986, + "step": 38857 + }, + { + "epoch": 0.38858, + "grad_norm": 1.2365729808807373, + "learning_rate": 0.003, + "loss": 3.998, + "step": 38858 + }, + { + "epoch": 0.38859, + "grad_norm": 0.9732669591903687, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 38859 + }, + { + "epoch": 0.3886, + "grad_norm": 1.0099064111709595, + "learning_rate": 0.003, + "loss": 4.01, + "step": 38860 + }, + { + "epoch": 0.38861, + "grad_norm": 1.3524987697601318, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 38861 + }, + { + "epoch": 0.38862, + "grad_norm": 0.879912257194519, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 38862 + }, + { + "epoch": 0.38863, + "grad_norm": 0.6759223341941833, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 38863 + }, + { + "epoch": 0.38864, + "grad_norm": 0.8656810522079468, + "learning_rate": 0.003, + "loss": 3.997, + "step": 38864 + }, + { + "epoch": 0.38865, + "grad_norm": 0.9679334163665771, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 38865 + }, + { + "epoch": 0.38866, + "grad_norm": 0.9552009105682373, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 38866 + }, + { + "epoch": 0.38867, + "grad_norm": 1.0456242561340332, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 38867 + }, + { + "epoch": 0.38868, + "grad_norm": 1.0211849212646484, + "learning_rate": 0.003, + "loss": 4.011, + "step": 38868 + }, + { + "epoch": 0.38869, + "grad_norm": 0.9963150024414062, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 38869 + }, + { + "epoch": 0.3887, + "grad_norm": 1.2723037004470825, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 38870 + }, + { + "epoch": 0.38871, + "grad_norm": 0.8882833123207092, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 38871 + }, + { + "epoch": 0.38872, + "grad_norm": 0.8180104494094849, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 38872 + }, + { + "epoch": 0.38873, + "grad_norm": 0.8677793145179749, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 38873 + }, + { + "epoch": 0.38874, + "grad_norm": 0.9340067505836487, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 38874 + }, + { + "epoch": 0.38875, + "grad_norm": 1.009913682937622, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 38875 + }, + { + "epoch": 0.38876, + "grad_norm": 0.973212480545044, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 38876 + }, + { + "epoch": 0.38877, + "grad_norm": 0.9952190518379211, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 38877 + }, + { + "epoch": 0.38878, + "grad_norm": 0.8660480976104736, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 38878 + }, + { + "epoch": 0.38879, + "grad_norm": 0.8357144594192505, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 38879 + }, + { + "epoch": 0.3888, + "grad_norm": 0.7979616522789001, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 38880 + }, + { + "epoch": 0.38881, + "grad_norm": 0.7441314458847046, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 38881 + }, + { + "epoch": 0.38882, + "grad_norm": 0.816045880317688, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 38882 + }, + { + "epoch": 0.38883, + "grad_norm": 1.1010335683822632, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 38883 + }, + { + "epoch": 0.38884, + "grad_norm": 1.0355435609817505, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 38884 + }, + { + "epoch": 0.38885, + "grad_norm": 0.9054219722747803, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 38885 + }, + { + "epoch": 0.38886, + "grad_norm": 1.0274218320846558, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 38886 + }, + { + "epoch": 0.38887, + "grad_norm": 1.2017003297805786, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 38887 + }, + { + "epoch": 0.38888, + "grad_norm": 0.8100864291191101, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 38888 + }, + { + "epoch": 0.38889, + "grad_norm": 0.7070009708404541, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 38889 + }, + { + "epoch": 0.3889, + "grad_norm": 0.7825937867164612, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 38890 + }, + { + "epoch": 0.38891, + "grad_norm": 1.1044167280197144, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 38891 + }, + { + "epoch": 0.38892, + "grad_norm": 0.9508754014968872, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 38892 + }, + { + "epoch": 0.38893, + "grad_norm": 1.066591739654541, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 38893 + }, + { + "epoch": 0.38894, + "grad_norm": 1.1308244466781616, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 38894 + }, + { + "epoch": 0.38895, + "grad_norm": 0.8804035186767578, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 38895 + }, + { + "epoch": 0.38896, + "grad_norm": 0.7520185708999634, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 38896 + }, + { + "epoch": 0.38897, + "grad_norm": 0.6450595259666443, + "learning_rate": 0.003, + "loss": 3.9509, + "step": 38897 + }, + { + "epoch": 0.38898, + "grad_norm": 0.5731443762779236, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 38898 + }, + { + "epoch": 0.38899, + "grad_norm": 0.6433382630348206, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 38899 + }, + { + "epoch": 0.389, + "grad_norm": 0.66298907995224, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 38900 + }, + { + "epoch": 0.38901, + "grad_norm": 0.8160093426704407, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 38901 + }, + { + "epoch": 0.38902, + "grad_norm": 0.7721526622772217, + "learning_rate": 0.003, + "loss": 3.9444, + "step": 38902 + }, + { + "epoch": 0.38903, + "grad_norm": 0.700150728225708, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 38903 + }, + { + "epoch": 0.38904, + "grad_norm": 0.7569988965988159, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 38904 + }, + { + "epoch": 0.38905, + "grad_norm": 0.9035890698432922, + "learning_rate": 0.003, + "loss": 3.969, + "step": 38905 + }, + { + "epoch": 0.38906, + "grad_norm": 1.0876492261886597, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 38906 + }, + { + "epoch": 0.38907, + "grad_norm": 0.9978224039077759, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 38907 + }, + { + "epoch": 0.38908, + "grad_norm": 0.9170426726341248, + "learning_rate": 0.003, + "loss": 3.9471, + "step": 38908 + }, + { + "epoch": 0.38909, + "grad_norm": 0.9668936729431152, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 38909 + }, + { + "epoch": 0.3891, + "grad_norm": 0.9257081151008606, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 38910 + }, + { + "epoch": 0.38911, + "grad_norm": 0.9500997066497803, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 38911 + }, + { + "epoch": 0.38912, + "grad_norm": 1.0599216222763062, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 38912 + }, + { + "epoch": 0.38913, + "grad_norm": 1.098394751548767, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 38913 + }, + { + "epoch": 0.38914, + "grad_norm": 0.9483630657196045, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 38914 + }, + { + "epoch": 0.38915, + "grad_norm": 1.0184530019760132, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 38915 + }, + { + "epoch": 0.38916, + "grad_norm": 1.0592931509017944, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 38916 + }, + { + "epoch": 0.38917, + "grad_norm": 1.141789436340332, + "learning_rate": 0.003, + "loss": 3.958, + "step": 38917 + }, + { + "epoch": 0.38918, + "grad_norm": 1.0568453073501587, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 38918 + }, + { + "epoch": 0.38919, + "grad_norm": 1.0081721544265747, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 38919 + }, + { + "epoch": 0.3892, + "grad_norm": 1.142414927482605, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 38920 + }, + { + "epoch": 0.38921, + "grad_norm": 1.169706106185913, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 38921 + }, + { + "epoch": 0.38922, + "grad_norm": 0.8338940143585205, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 38922 + }, + { + "epoch": 0.38923, + "grad_norm": 0.7495743632316589, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 38923 + }, + { + "epoch": 0.38924, + "grad_norm": 0.747981071472168, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 38924 + }, + { + "epoch": 0.38925, + "grad_norm": 0.7191256284713745, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 38925 + }, + { + "epoch": 0.38926, + "grad_norm": 0.7599655985832214, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 38926 + }, + { + "epoch": 0.38927, + "grad_norm": 0.9775967001914978, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 38927 + }, + { + "epoch": 0.38928, + "grad_norm": 1.1093212366104126, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 38928 + }, + { + "epoch": 0.38929, + "grad_norm": 0.8389344215393066, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 38929 + }, + { + "epoch": 0.3893, + "grad_norm": 0.7035770416259766, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 38930 + }, + { + "epoch": 0.38931, + "grad_norm": 0.6678545475006104, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 38931 + }, + { + "epoch": 0.38932, + "grad_norm": 0.7921103239059448, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 38932 + }, + { + "epoch": 0.38933, + "grad_norm": 0.8138514161109924, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 38933 + }, + { + "epoch": 0.38934, + "grad_norm": 0.8587746620178223, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 38934 + }, + { + "epoch": 0.38935, + "grad_norm": 1.0812747478485107, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 38935 + }, + { + "epoch": 0.38936, + "grad_norm": 1.0331244468688965, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 38936 + }, + { + "epoch": 0.38937, + "grad_norm": 0.8649720549583435, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 38937 + }, + { + "epoch": 0.38938, + "grad_norm": 0.8492687940597534, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 38938 + }, + { + "epoch": 0.38939, + "grad_norm": 0.917771577835083, + "learning_rate": 0.003, + "loss": 3.972, + "step": 38939 + }, + { + "epoch": 0.3894, + "grad_norm": 0.9568500518798828, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 38940 + }, + { + "epoch": 0.38941, + "grad_norm": 0.975669801235199, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 38941 + }, + { + "epoch": 0.38942, + "grad_norm": 1.214738368988037, + "learning_rate": 0.003, + "loss": 3.982, + "step": 38942 + }, + { + "epoch": 0.38943, + "grad_norm": 0.7551679015159607, + "learning_rate": 0.003, + "loss": 3.9276, + "step": 38943 + }, + { + "epoch": 0.38944, + "grad_norm": 0.719209611415863, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 38944 + }, + { + "epoch": 0.38945, + "grad_norm": 0.7478774785995483, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 38945 + }, + { + "epoch": 0.38946, + "grad_norm": 0.8361356258392334, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 38946 + }, + { + "epoch": 0.38947, + "grad_norm": 1.067550539970398, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 38947 + }, + { + "epoch": 0.38948, + "grad_norm": 0.9244253039360046, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 38948 + }, + { + "epoch": 0.38949, + "grad_norm": 0.8102492094039917, + "learning_rate": 0.003, + "loss": 4.001, + "step": 38949 + }, + { + "epoch": 0.3895, + "grad_norm": 0.6661200523376465, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 38950 + }, + { + "epoch": 0.38951, + "grad_norm": 0.7102926969528198, + "learning_rate": 0.003, + "loss": 3.9464, + "step": 38951 + }, + { + "epoch": 0.38952, + "grad_norm": 1.054888367652893, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 38952 + }, + { + "epoch": 0.38953, + "grad_norm": 1.1985878944396973, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 38953 + }, + { + "epoch": 0.38954, + "grad_norm": 0.9481074213981628, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 38954 + }, + { + "epoch": 0.38955, + "grad_norm": 0.8962016701698303, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 38955 + }, + { + "epoch": 0.38956, + "grad_norm": 0.8775885701179504, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 38956 + }, + { + "epoch": 0.38957, + "grad_norm": 0.8861585855484009, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 38957 + }, + { + "epoch": 0.38958, + "grad_norm": 0.86421799659729, + "learning_rate": 0.003, + "loss": 3.992, + "step": 38958 + }, + { + "epoch": 0.38959, + "grad_norm": 0.9523035883903503, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 38959 + }, + { + "epoch": 0.3896, + "grad_norm": 1.1445724964141846, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 38960 + }, + { + "epoch": 0.38961, + "grad_norm": 1.0222797393798828, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 38961 + }, + { + "epoch": 0.38962, + "grad_norm": 0.9502788782119751, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 38962 + }, + { + "epoch": 0.38963, + "grad_norm": 0.9525426626205444, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 38963 + }, + { + "epoch": 0.38964, + "grad_norm": 0.9046247601509094, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 38964 + }, + { + "epoch": 0.38965, + "grad_norm": 0.7821720242500305, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 38965 + }, + { + "epoch": 0.38966, + "grad_norm": 0.713989794254303, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 38966 + }, + { + "epoch": 0.38967, + "grad_norm": 0.85662442445755, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 38967 + }, + { + "epoch": 0.38968, + "grad_norm": 1.0566377639770508, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 38968 + }, + { + "epoch": 0.38969, + "grad_norm": 0.9618639349937439, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 38969 + }, + { + "epoch": 0.3897, + "grad_norm": 0.9023075103759766, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 38970 + }, + { + "epoch": 0.38971, + "grad_norm": 0.8474801778793335, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 38971 + }, + { + "epoch": 0.38972, + "grad_norm": 1.0092861652374268, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 38972 + }, + { + "epoch": 0.38973, + "grad_norm": 1.2767752408981323, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 38973 + }, + { + "epoch": 0.38974, + "grad_norm": 0.785193920135498, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 38974 + }, + { + "epoch": 0.38975, + "grad_norm": 0.6895555257797241, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 38975 + }, + { + "epoch": 0.38976, + "grad_norm": 0.7342967391014099, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 38976 + }, + { + "epoch": 0.38977, + "grad_norm": 0.8512099981307983, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 38977 + }, + { + "epoch": 0.38978, + "grad_norm": 0.8398769497871399, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 38978 + }, + { + "epoch": 0.38979, + "grad_norm": 0.8221017122268677, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 38979 + }, + { + "epoch": 0.3898, + "grad_norm": 0.8852145671844482, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 38980 + }, + { + "epoch": 0.38981, + "grad_norm": 1.0465103387832642, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 38981 + }, + { + "epoch": 0.38982, + "grad_norm": 0.9815838932991028, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 38982 + }, + { + "epoch": 0.38983, + "grad_norm": 0.9330195784568787, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 38983 + }, + { + "epoch": 0.38984, + "grad_norm": 0.749336838722229, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 38984 + }, + { + "epoch": 0.38985, + "grad_norm": 0.6849117875099182, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 38985 + }, + { + "epoch": 0.38986, + "grad_norm": 0.6928063631057739, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 38986 + }, + { + "epoch": 0.38987, + "grad_norm": 0.7201590538024902, + "learning_rate": 0.003, + "loss": 3.9424, + "step": 38987 + }, + { + "epoch": 0.38988, + "grad_norm": 0.9178575277328491, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 38988 + }, + { + "epoch": 0.38989, + "grad_norm": 1.229263424873352, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 38989 + }, + { + "epoch": 0.3899, + "grad_norm": 0.7945979833602905, + "learning_rate": 0.003, + "loss": 3.928, + "step": 38990 + }, + { + "epoch": 0.38991, + "grad_norm": 0.6697295308113098, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 38991 + }, + { + "epoch": 0.38992, + "grad_norm": 0.7169021964073181, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 38992 + }, + { + "epoch": 0.38993, + "grad_norm": 0.7995722889900208, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 38993 + }, + { + "epoch": 0.38994, + "grad_norm": 1.0356009006500244, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 38994 + }, + { + "epoch": 0.38995, + "grad_norm": 1.0588363409042358, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 38995 + }, + { + "epoch": 0.38996, + "grad_norm": 0.8608763217926025, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 38996 + }, + { + "epoch": 0.38997, + "grad_norm": 0.8645678162574768, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 38997 + }, + { + "epoch": 0.38998, + "grad_norm": 0.8789684772491455, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 38998 + }, + { + "epoch": 0.38999, + "grad_norm": 1.1445761919021606, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 38999 + }, + { + "epoch": 0.39, + "grad_norm": 0.9432598352432251, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 39000 + }, + { + "epoch": 0.39001, + "grad_norm": 1.0787034034729004, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 39001 + }, + { + "epoch": 0.39002, + "grad_norm": 1.2364238500595093, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 39002 + }, + { + "epoch": 0.39003, + "grad_norm": 0.823710024356842, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 39003 + }, + { + "epoch": 0.39004, + "grad_norm": 0.8022069334983826, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 39004 + }, + { + "epoch": 0.39005, + "grad_norm": 0.9400892853736877, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 39005 + }, + { + "epoch": 0.39006, + "grad_norm": 0.9434930682182312, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 39006 + }, + { + "epoch": 0.39007, + "grad_norm": 0.9437769651412964, + "learning_rate": 0.003, + "loss": 3.984, + "step": 39007 + }, + { + "epoch": 0.39008, + "grad_norm": 1.12667977809906, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 39008 + }, + { + "epoch": 0.39009, + "grad_norm": 0.9674996137619019, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 39009 + }, + { + "epoch": 0.3901, + "grad_norm": 0.8810380101203918, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 39010 + }, + { + "epoch": 0.39011, + "grad_norm": 0.9171406030654907, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 39011 + }, + { + "epoch": 0.39012, + "grad_norm": 1.098192572593689, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 39012 + }, + { + "epoch": 0.39013, + "grad_norm": 1.270060420036316, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 39013 + }, + { + "epoch": 0.39014, + "grad_norm": 0.6726246476173401, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 39014 + }, + { + "epoch": 0.39015, + "grad_norm": 0.6715755462646484, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 39015 + }, + { + "epoch": 0.39016, + "grad_norm": 0.693505585193634, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 39016 + }, + { + "epoch": 0.39017, + "grad_norm": 0.817848801612854, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 39017 + }, + { + "epoch": 0.39018, + "grad_norm": 1.045041799545288, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 39018 + }, + { + "epoch": 0.39019, + "grad_norm": 1.038154125213623, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 39019 + }, + { + "epoch": 0.3902, + "grad_norm": 0.798840343952179, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 39020 + }, + { + "epoch": 0.39021, + "grad_norm": 0.7295784950256348, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 39021 + }, + { + "epoch": 0.39022, + "grad_norm": 0.9636821746826172, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 39022 + }, + { + "epoch": 0.39023, + "grad_norm": 1.0680655241012573, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 39023 + }, + { + "epoch": 0.39024, + "grad_norm": 0.8567080497741699, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 39024 + }, + { + "epoch": 0.39025, + "grad_norm": 0.8557383418083191, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 39025 + }, + { + "epoch": 0.39026, + "grad_norm": 0.853827714920044, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 39026 + }, + { + "epoch": 0.39027, + "grad_norm": 0.7740725874900818, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 39027 + }, + { + "epoch": 0.39028, + "grad_norm": 0.8712995648384094, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 39028 + }, + { + "epoch": 0.39029, + "grad_norm": 1.0271196365356445, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 39029 + }, + { + "epoch": 0.3903, + "grad_norm": 1.1496217250823975, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 39030 + }, + { + "epoch": 0.39031, + "grad_norm": 0.8417750597000122, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 39031 + }, + { + "epoch": 0.39032, + "grad_norm": 0.7551911473274231, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 39032 + }, + { + "epoch": 0.39033, + "grad_norm": 0.9592365622520447, + "learning_rate": 0.003, + "loss": 3.991, + "step": 39033 + }, + { + "epoch": 0.39034, + "grad_norm": 1.2586604356765747, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 39034 + }, + { + "epoch": 0.39035, + "grad_norm": 0.8119566440582275, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 39035 + }, + { + "epoch": 0.39036, + "grad_norm": 0.7924519777297974, + "learning_rate": 0.003, + "loss": 3.978, + "step": 39036 + }, + { + "epoch": 0.39037, + "grad_norm": 0.859775185585022, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 39037 + }, + { + "epoch": 0.39038, + "grad_norm": 1.0960674285888672, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 39038 + }, + { + "epoch": 0.39039, + "grad_norm": 1.0173661708831787, + "learning_rate": 0.003, + "loss": 3.987, + "step": 39039 + }, + { + "epoch": 0.3904, + "grad_norm": 0.9664997458457947, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 39040 + }, + { + "epoch": 0.39041, + "grad_norm": 0.9396675825119019, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 39041 + }, + { + "epoch": 0.39042, + "grad_norm": 0.9508578181266785, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 39042 + }, + { + "epoch": 0.39043, + "grad_norm": 1.0752134323120117, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 39043 + }, + { + "epoch": 0.39044, + "grad_norm": 0.8996550440788269, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 39044 + }, + { + "epoch": 0.39045, + "grad_norm": 0.8792879581451416, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 39045 + }, + { + "epoch": 0.39046, + "grad_norm": 0.8883116841316223, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 39046 + }, + { + "epoch": 0.39047, + "grad_norm": 0.7236500978469849, + "learning_rate": 0.003, + "loss": 3.958, + "step": 39047 + }, + { + "epoch": 0.39048, + "grad_norm": 0.7457318305969238, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 39048 + }, + { + "epoch": 0.39049, + "grad_norm": 0.6322916150093079, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 39049 + }, + { + "epoch": 0.3905, + "grad_norm": 0.8693469166755676, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 39050 + }, + { + "epoch": 0.39051, + "grad_norm": 1.3675215244293213, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 39051 + }, + { + "epoch": 0.39052, + "grad_norm": 0.7819398641586304, + "learning_rate": 0.003, + "loss": 3.991, + "step": 39052 + }, + { + "epoch": 0.39053, + "grad_norm": 0.5948730707168579, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 39053 + }, + { + "epoch": 0.39054, + "grad_norm": 0.5400821566581726, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 39054 + }, + { + "epoch": 0.39055, + "grad_norm": 0.5819209814071655, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 39055 + }, + { + "epoch": 0.39056, + "grad_norm": 0.6774774789810181, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 39056 + }, + { + "epoch": 0.39057, + "grad_norm": 0.939530611038208, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 39057 + }, + { + "epoch": 0.39058, + "grad_norm": 1.124569296836853, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 39058 + }, + { + "epoch": 0.39059, + "grad_norm": 0.8859720230102539, + "learning_rate": 0.003, + "loss": 3.9483, + "step": 39059 + }, + { + "epoch": 0.3906, + "grad_norm": 0.7891148924827576, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 39060 + }, + { + "epoch": 0.39061, + "grad_norm": 0.7646580934524536, + "learning_rate": 0.003, + "loss": 3.986, + "step": 39061 + }, + { + "epoch": 0.39062, + "grad_norm": 0.7736681699752808, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 39062 + }, + { + "epoch": 0.39063, + "grad_norm": 0.7762320041656494, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 39063 + }, + { + "epoch": 0.39064, + "grad_norm": 0.8642066121101379, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 39064 + }, + { + "epoch": 0.39065, + "grad_norm": 0.964796781539917, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 39065 + }, + { + "epoch": 0.39066, + "grad_norm": 1.2479915618896484, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 39066 + }, + { + "epoch": 0.39067, + "grad_norm": 0.9613537192344666, + "learning_rate": 0.003, + "loss": 3.9234, + "step": 39067 + }, + { + "epoch": 0.39068, + "grad_norm": 1.0758825540542603, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 39068 + }, + { + "epoch": 0.39069, + "grad_norm": 0.9748755097389221, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 39069 + }, + { + "epoch": 0.3907, + "grad_norm": 0.9982765316963196, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 39070 + }, + { + "epoch": 0.39071, + "grad_norm": 1.0802075862884521, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 39071 + }, + { + "epoch": 0.39072, + "grad_norm": 0.9854238033294678, + "learning_rate": 0.003, + "loss": 3.9432, + "step": 39072 + }, + { + "epoch": 0.39073, + "grad_norm": 1.0621751546859741, + "learning_rate": 0.003, + "loss": 3.9431, + "step": 39073 + }, + { + "epoch": 0.39074, + "grad_norm": 0.9338923096656799, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 39074 + }, + { + "epoch": 0.39075, + "grad_norm": 1.0616734027862549, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 39075 + }, + { + "epoch": 0.39076, + "grad_norm": 1.2288309335708618, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 39076 + }, + { + "epoch": 0.39077, + "grad_norm": 0.7310962677001953, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 39077 + }, + { + "epoch": 0.39078, + "grad_norm": 0.646567702293396, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 39078 + }, + { + "epoch": 0.39079, + "grad_norm": 0.6731719374656677, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 39079 + }, + { + "epoch": 0.3908, + "grad_norm": 0.6742755174636841, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 39080 + }, + { + "epoch": 0.39081, + "grad_norm": 0.8060929179191589, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 39081 + }, + { + "epoch": 0.39082, + "grad_norm": 0.9833567142486572, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 39082 + }, + { + "epoch": 0.39083, + "grad_norm": 0.9769235849380493, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 39083 + }, + { + "epoch": 0.39084, + "grad_norm": 0.8952370882034302, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 39084 + }, + { + "epoch": 0.39085, + "grad_norm": 0.7919683456420898, + "learning_rate": 0.003, + "loss": 3.9466, + "step": 39085 + }, + { + "epoch": 0.39086, + "grad_norm": 0.7136862277984619, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 39086 + }, + { + "epoch": 0.39087, + "grad_norm": 0.82166588306427, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 39087 + }, + { + "epoch": 0.39088, + "grad_norm": 0.9577913880348206, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 39088 + }, + { + "epoch": 0.39089, + "grad_norm": 0.9260439872741699, + "learning_rate": 0.003, + "loss": 3.948, + "step": 39089 + }, + { + "epoch": 0.3909, + "grad_norm": 0.9133298993110657, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 39090 + }, + { + "epoch": 0.39091, + "grad_norm": 1.0723819732666016, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 39091 + }, + { + "epoch": 0.39092, + "grad_norm": 1.1318162679672241, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 39092 + }, + { + "epoch": 0.39093, + "grad_norm": 1.134751558303833, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 39093 + }, + { + "epoch": 0.39094, + "grad_norm": 1.118722915649414, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 39094 + }, + { + "epoch": 0.39095, + "grad_norm": 0.7993245720863342, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 39095 + }, + { + "epoch": 0.39096, + "grad_norm": 0.7716491222381592, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 39096 + }, + { + "epoch": 0.39097, + "grad_norm": 0.8255627751350403, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 39097 + }, + { + "epoch": 0.39098, + "grad_norm": 0.92741459608078, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 39098 + }, + { + "epoch": 0.39099, + "grad_norm": 0.9884849786758423, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 39099 + }, + { + "epoch": 0.391, + "grad_norm": 0.9382272958755493, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 39100 + }, + { + "epoch": 0.39101, + "grad_norm": 0.9069559574127197, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 39101 + }, + { + "epoch": 0.39102, + "grad_norm": 0.9201146364212036, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 39102 + }, + { + "epoch": 0.39103, + "grad_norm": 0.9447250366210938, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 39103 + }, + { + "epoch": 0.39104, + "grad_norm": 0.9727854132652283, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 39104 + }, + { + "epoch": 0.39105, + "grad_norm": 1.092576265335083, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 39105 + }, + { + "epoch": 0.39106, + "grad_norm": 1.0187749862670898, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 39106 + }, + { + "epoch": 0.39107, + "grad_norm": 1.1501866579055786, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 39107 + }, + { + "epoch": 0.39108, + "grad_norm": 0.9987304210662842, + "learning_rate": 0.003, + "loss": 3.973, + "step": 39108 + }, + { + "epoch": 0.39109, + "grad_norm": 0.9811875820159912, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 39109 + }, + { + "epoch": 0.3911, + "grad_norm": 1.1463655233383179, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 39110 + }, + { + "epoch": 0.39111, + "grad_norm": 0.97467440366745, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 39111 + }, + { + "epoch": 0.39112, + "grad_norm": 0.9335573315620422, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 39112 + }, + { + "epoch": 0.39113, + "grad_norm": 0.8720617890357971, + "learning_rate": 0.003, + "loss": 3.981, + "step": 39113 + }, + { + "epoch": 0.39114, + "grad_norm": 0.8249902129173279, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 39114 + }, + { + "epoch": 0.39115, + "grad_norm": 0.7763603329658508, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 39115 + }, + { + "epoch": 0.39116, + "grad_norm": 0.7893045544624329, + "learning_rate": 0.003, + "loss": 3.968, + "step": 39116 + }, + { + "epoch": 0.39117, + "grad_norm": 0.7042826414108276, + "learning_rate": 0.003, + "loss": 3.966, + "step": 39117 + }, + { + "epoch": 0.39118, + "grad_norm": 0.6667585968971252, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 39118 + }, + { + "epoch": 0.39119, + "grad_norm": 0.6272150874137878, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 39119 + }, + { + "epoch": 0.3912, + "grad_norm": 0.5271295309066772, + "learning_rate": 0.003, + "loss": 3.983, + "step": 39120 + }, + { + "epoch": 0.39121, + "grad_norm": 0.6257014274597168, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 39121 + }, + { + "epoch": 0.39122, + "grad_norm": 0.8789819478988647, + "learning_rate": 0.003, + "loss": 3.964, + "step": 39122 + }, + { + "epoch": 0.39123, + "grad_norm": 1.3614263534545898, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 39123 + }, + { + "epoch": 0.39124, + "grad_norm": 0.8351250886917114, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 39124 + }, + { + "epoch": 0.39125, + "grad_norm": 0.7049562931060791, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 39125 + }, + { + "epoch": 0.39126, + "grad_norm": 0.7389869093894958, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 39126 + }, + { + "epoch": 0.39127, + "grad_norm": 0.6240351796150208, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 39127 + }, + { + "epoch": 0.39128, + "grad_norm": 0.5946695804595947, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 39128 + }, + { + "epoch": 0.39129, + "grad_norm": 0.6062331795692444, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 39129 + }, + { + "epoch": 0.3913, + "grad_norm": 0.6173876523971558, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 39130 + }, + { + "epoch": 0.39131, + "grad_norm": 0.637789249420166, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 39131 + }, + { + "epoch": 0.39132, + "grad_norm": 0.7442302107810974, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 39132 + }, + { + "epoch": 0.39133, + "grad_norm": 0.9327946901321411, + "learning_rate": 0.003, + "loss": 3.9281, + "step": 39133 + }, + { + "epoch": 0.39134, + "grad_norm": 1.1241703033447266, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 39134 + }, + { + "epoch": 0.39135, + "grad_norm": 0.8903172016143799, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 39135 + }, + { + "epoch": 0.39136, + "grad_norm": 0.9431169033050537, + "learning_rate": 0.003, + "loss": 3.9411, + "step": 39136 + }, + { + "epoch": 0.39137, + "grad_norm": 0.9475672245025635, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 39137 + }, + { + "epoch": 0.39138, + "grad_norm": 0.8528732657432556, + "learning_rate": 0.003, + "loss": 3.967, + "step": 39138 + }, + { + "epoch": 0.39139, + "grad_norm": 0.8095344305038452, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 39139 + }, + { + "epoch": 0.3914, + "grad_norm": 0.9062631726264954, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 39140 + }, + { + "epoch": 0.39141, + "grad_norm": 1.0267847776412964, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 39141 + }, + { + "epoch": 0.39142, + "grad_norm": 1.154751181602478, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 39142 + }, + { + "epoch": 0.39143, + "grad_norm": 1.0882891416549683, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 39143 + }, + { + "epoch": 0.39144, + "grad_norm": 0.8344549536705017, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 39144 + }, + { + "epoch": 0.39145, + "grad_norm": 0.8594629168510437, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 39145 + }, + { + "epoch": 0.39146, + "grad_norm": 1.1164331436157227, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 39146 + }, + { + "epoch": 0.39147, + "grad_norm": 1.2649211883544922, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 39147 + }, + { + "epoch": 0.39148, + "grad_norm": 0.8355857133865356, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 39148 + }, + { + "epoch": 0.39149, + "grad_norm": 0.807567834854126, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 39149 + }, + { + "epoch": 0.3915, + "grad_norm": 0.7993524074554443, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 39150 + }, + { + "epoch": 0.39151, + "grad_norm": 0.8792605996131897, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 39151 + }, + { + "epoch": 0.39152, + "grad_norm": 1.0265823602676392, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 39152 + }, + { + "epoch": 0.39153, + "grad_norm": 1.1039146184921265, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 39153 + }, + { + "epoch": 0.39154, + "grad_norm": 0.8934191465377808, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 39154 + }, + { + "epoch": 0.39155, + "grad_norm": 0.9344043135643005, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 39155 + }, + { + "epoch": 0.39156, + "grad_norm": 0.911125659942627, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 39156 + }, + { + "epoch": 0.39157, + "grad_norm": 1.001171588897705, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 39157 + }, + { + "epoch": 0.39158, + "grad_norm": 1.286059856414795, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 39158 + }, + { + "epoch": 0.39159, + "grad_norm": 1.0780457258224487, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 39159 + }, + { + "epoch": 0.3916, + "grad_norm": 1.1017334461212158, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 39160 + }, + { + "epoch": 0.39161, + "grad_norm": 0.8315705060958862, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 39161 + }, + { + "epoch": 0.39162, + "grad_norm": 0.6474281549453735, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 39162 + }, + { + "epoch": 0.39163, + "grad_norm": 0.7348530888557434, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 39163 + }, + { + "epoch": 0.39164, + "grad_norm": 0.7592068910598755, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 39164 + }, + { + "epoch": 0.39165, + "grad_norm": 0.7975964546203613, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 39165 + }, + { + "epoch": 0.39166, + "grad_norm": 0.7025386095046997, + "learning_rate": 0.003, + "loss": 3.962, + "step": 39166 + }, + { + "epoch": 0.39167, + "grad_norm": 0.6404173970222473, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 39167 + }, + { + "epoch": 0.39168, + "grad_norm": 0.6294180750846863, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 39168 + }, + { + "epoch": 0.39169, + "grad_norm": 0.8309274911880493, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 39169 + }, + { + "epoch": 0.3917, + "grad_norm": 1.1096440553665161, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 39170 + }, + { + "epoch": 0.39171, + "grad_norm": 0.9584740400314331, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 39171 + }, + { + "epoch": 0.39172, + "grad_norm": 0.9980972409248352, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 39172 + }, + { + "epoch": 0.39173, + "grad_norm": 1.1789268255233765, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 39173 + }, + { + "epoch": 0.39174, + "grad_norm": 0.8345155119895935, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 39174 + }, + { + "epoch": 0.39175, + "grad_norm": 0.7065041661262512, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 39175 + }, + { + "epoch": 0.39176, + "grad_norm": 0.6782444715499878, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 39176 + }, + { + "epoch": 0.39177, + "grad_norm": 0.72120600938797, + "learning_rate": 0.003, + "loss": 3.966, + "step": 39177 + }, + { + "epoch": 0.39178, + "grad_norm": 0.824923574924469, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 39178 + }, + { + "epoch": 0.39179, + "grad_norm": 0.82206791639328, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 39179 + }, + { + "epoch": 0.3918, + "grad_norm": 1.0108760595321655, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 39180 + }, + { + "epoch": 0.39181, + "grad_norm": 1.108443260192871, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 39181 + }, + { + "epoch": 0.39182, + "grad_norm": 0.68473881483078, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 39182 + }, + { + "epoch": 0.39183, + "grad_norm": 0.8361496329307556, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 39183 + }, + { + "epoch": 0.39184, + "grad_norm": 0.7395591735839844, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 39184 + }, + { + "epoch": 0.39185, + "grad_norm": 0.7354820370674133, + "learning_rate": 0.003, + "loss": 4.003, + "step": 39185 + }, + { + "epoch": 0.39186, + "grad_norm": 0.8562816977500916, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 39186 + }, + { + "epoch": 0.39187, + "grad_norm": 1.1265904903411865, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 39187 + }, + { + "epoch": 0.39188, + "grad_norm": 1.142866849899292, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 39188 + }, + { + "epoch": 0.39189, + "grad_norm": 1.2890610694885254, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 39189 + }, + { + "epoch": 0.3919, + "grad_norm": 0.8303911685943604, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 39190 + }, + { + "epoch": 0.39191, + "grad_norm": 0.6915022134780884, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 39191 + }, + { + "epoch": 0.39192, + "grad_norm": 0.802314817905426, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 39192 + }, + { + "epoch": 0.39193, + "grad_norm": 1.0315626859664917, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 39193 + }, + { + "epoch": 0.39194, + "grad_norm": 1.1874138116836548, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 39194 + }, + { + "epoch": 0.39195, + "grad_norm": 0.7554627656936646, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 39195 + }, + { + "epoch": 0.39196, + "grad_norm": 0.797111451625824, + "learning_rate": 0.003, + "loss": 3.971, + "step": 39196 + }, + { + "epoch": 0.39197, + "grad_norm": 0.9503345489501953, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 39197 + }, + { + "epoch": 0.39198, + "grad_norm": 1.066252589225769, + "learning_rate": 0.003, + "loss": 3.983, + "step": 39198 + }, + { + "epoch": 0.39199, + "grad_norm": 1.132961392402649, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 39199 + }, + { + "epoch": 0.392, + "grad_norm": 0.9599791169166565, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 39200 + }, + { + "epoch": 0.39201, + "grad_norm": 1.0545556545257568, + "learning_rate": 0.003, + "loss": 3.966, + "step": 39201 + }, + { + "epoch": 0.39202, + "grad_norm": 0.8058512806892395, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 39202 + }, + { + "epoch": 0.39203, + "grad_norm": 0.9492311477661133, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 39203 + }, + { + "epoch": 0.39204, + "grad_norm": 1.2249969244003296, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 39204 + }, + { + "epoch": 0.39205, + "grad_norm": 0.9336641430854797, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 39205 + }, + { + "epoch": 0.39206, + "grad_norm": 0.7953923344612122, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 39206 + }, + { + "epoch": 0.39207, + "grad_norm": 0.7835103273391724, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 39207 + }, + { + "epoch": 0.39208, + "grad_norm": 0.8120248317718506, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 39208 + }, + { + "epoch": 0.39209, + "grad_norm": 0.7879859209060669, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 39209 + }, + { + "epoch": 0.3921, + "grad_norm": 0.7894585132598877, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 39210 + }, + { + "epoch": 0.39211, + "grad_norm": 0.7465811967849731, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 39211 + }, + { + "epoch": 0.39212, + "grad_norm": 0.7065538763999939, + "learning_rate": 0.003, + "loss": 3.9315, + "step": 39212 + }, + { + "epoch": 0.39213, + "grad_norm": 0.6982635259628296, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 39213 + }, + { + "epoch": 0.39214, + "grad_norm": 0.8050096035003662, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 39214 + }, + { + "epoch": 0.39215, + "grad_norm": 1.1632565259933472, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 39215 + }, + { + "epoch": 0.39216, + "grad_norm": 1.3281528949737549, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 39216 + }, + { + "epoch": 0.39217, + "grad_norm": 0.6407855749130249, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 39217 + }, + { + "epoch": 0.39218, + "grad_norm": 0.7447514533996582, + "learning_rate": 0.003, + "loss": 4.005, + "step": 39218 + }, + { + "epoch": 0.39219, + "grad_norm": 1.0876920223236084, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 39219 + }, + { + "epoch": 0.3922, + "grad_norm": 1.1319605112075806, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 39220 + }, + { + "epoch": 0.39221, + "grad_norm": 1.059260368347168, + "learning_rate": 0.003, + "loss": 4.003, + "step": 39221 + }, + { + "epoch": 0.39222, + "grad_norm": 1.0043045282363892, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 39222 + }, + { + "epoch": 0.39223, + "grad_norm": 0.9940665364265442, + "learning_rate": 0.003, + "loss": 3.963, + "step": 39223 + }, + { + "epoch": 0.39224, + "grad_norm": 0.9494638442993164, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 39224 + }, + { + "epoch": 0.39225, + "grad_norm": 0.9669984579086304, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 39225 + }, + { + "epoch": 0.39226, + "grad_norm": 1.0988389253616333, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 39226 + }, + { + "epoch": 0.39227, + "grad_norm": 0.9911988973617554, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 39227 + }, + { + "epoch": 0.39228, + "grad_norm": 1.0492756366729736, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 39228 + }, + { + "epoch": 0.39229, + "grad_norm": 1.07772696018219, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 39229 + }, + { + "epoch": 0.3923, + "grad_norm": 0.8956929445266724, + "learning_rate": 0.003, + "loss": 3.939, + "step": 39230 + }, + { + "epoch": 0.39231, + "grad_norm": 0.9770172834396362, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 39231 + }, + { + "epoch": 0.39232, + "grad_norm": 1.0766717195510864, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 39232 + }, + { + "epoch": 0.39233, + "grad_norm": 0.9524819254875183, + "learning_rate": 0.003, + "loss": 3.962, + "step": 39233 + }, + { + "epoch": 0.39234, + "grad_norm": 0.7931509017944336, + "learning_rate": 0.003, + "loss": 3.9375, + "step": 39234 + }, + { + "epoch": 0.39235, + "grad_norm": 0.8549554944038391, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 39235 + }, + { + "epoch": 0.39236, + "grad_norm": 0.8121697306632996, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 39236 + }, + { + "epoch": 0.39237, + "grad_norm": 0.7206552624702454, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 39237 + }, + { + "epoch": 0.39238, + "grad_norm": 0.8088582754135132, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 39238 + }, + { + "epoch": 0.39239, + "grad_norm": 0.9254950284957886, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 39239 + }, + { + "epoch": 0.3924, + "grad_norm": 1.1897507905960083, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 39240 + }, + { + "epoch": 0.39241, + "grad_norm": 1.0464847087860107, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 39241 + }, + { + "epoch": 0.39242, + "grad_norm": 1.060720443725586, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 39242 + }, + { + "epoch": 0.39243, + "grad_norm": 1.0127698183059692, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 39243 + }, + { + "epoch": 0.39244, + "grad_norm": 1.0817837715148926, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 39244 + }, + { + "epoch": 0.39245, + "grad_norm": 0.853442370891571, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 39245 + }, + { + "epoch": 0.39246, + "grad_norm": 0.6797158718109131, + "learning_rate": 0.003, + "loss": 3.942, + "step": 39246 + }, + { + "epoch": 0.39247, + "grad_norm": 0.7023411989212036, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 39247 + }, + { + "epoch": 0.39248, + "grad_norm": 0.7314286828041077, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 39248 + }, + { + "epoch": 0.39249, + "grad_norm": 0.5952747464179993, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 39249 + }, + { + "epoch": 0.3925, + "grad_norm": 0.6251025795936584, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 39250 + }, + { + "epoch": 0.39251, + "grad_norm": 0.6269869208335876, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 39251 + }, + { + "epoch": 0.39252, + "grad_norm": 0.7676891684532166, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 39252 + }, + { + "epoch": 0.39253, + "grad_norm": 1.1658364534378052, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 39253 + }, + { + "epoch": 0.39254, + "grad_norm": 1.1679474115371704, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 39254 + }, + { + "epoch": 0.39255, + "grad_norm": 0.647203803062439, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 39255 + }, + { + "epoch": 0.39256, + "grad_norm": 0.5552818775177002, + "learning_rate": 0.003, + "loss": 3.975, + "step": 39256 + }, + { + "epoch": 0.39257, + "grad_norm": 0.778101921081543, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 39257 + }, + { + "epoch": 0.39258, + "grad_norm": 1.0626472234725952, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 39258 + }, + { + "epoch": 0.39259, + "grad_norm": 1.2140820026397705, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 39259 + }, + { + "epoch": 0.3926, + "grad_norm": 0.582247257232666, + "learning_rate": 0.003, + "loss": 3.9458, + "step": 39260 + }, + { + "epoch": 0.39261, + "grad_norm": 0.6324598789215088, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 39261 + }, + { + "epoch": 0.39262, + "grad_norm": 0.6896337866783142, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 39262 + }, + { + "epoch": 0.39263, + "grad_norm": 0.7167485356330872, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 39263 + }, + { + "epoch": 0.39264, + "grad_norm": 0.801079511642456, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 39264 + }, + { + "epoch": 0.39265, + "grad_norm": 1.079479694366455, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 39265 + }, + { + "epoch": 0.39266, + "grad_norm": 1.3545823097229004, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 39266 + }, + { + "epoch": 0.39267, + "grad_norm": 0.635106086730957, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 39267 + }, + { + "epoch": 0.39268, + "grad_norm": 0.6845577955245972, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 39268 + }, + { + "epoch": 0.39269, + "grad_norm": 0.7082510590553284, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 39269 + }, + { + "epoch": 0.3927, + "grad_norm": 0.7987558841705322, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 39270 + }, + { + "epoch": 0.39271, + "grad_norm": 1.0316284894943237, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 39271 + }, + { + "epoch": 0.39272, + "grad_norm": 1.1126117706298828, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 39272 + }, + { + "epoch": 0.39273, + "grad_norm": 0.9051933288574219, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 39273 + }, + { + "epoch": 0.39274, + "grad_norm": 1.0039539337158203, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 39274 + }, + { + "epoch": 0.39275, + "grad_norm": 1.1453857421875, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 39275 + }, + { + "epoch": 0.39276, + "grad_norm": 0.9567463397979736, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 39276 + }, + { + "epoch": 0.39277, + "grad_norm": 0.9469505548477173, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 39277 + }, + { + "epoch": 0.39278, + "grad_norm": 1.0475108623504639, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 39278 + }, + { + "epoch": 0.39279, + "grad_norm": 1.1789768934249878, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 39279 + }, + { + "epoch": 0.3928, + "grad_norm": 0.8600274324417114, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 39280 + }, + { + "epoch": 0.39281, + "grad_norm": 0.9131070375442505, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 39281 + }, + { + "epoch": 0.39282, + "grad_norm": 1.1965506076812744, + "learning_rate": 0.003, + "loss": 3.992, + "step": 39282 + }, + { + "epoch": 0.39283, + "grad_norm": 0.9762791991233826, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 39283 + }, + { + "epoch": 0.39284, + "grad_norm": 1.03865385055542, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 39284 + }, + { + "epoch": 0.39285, + "grad_norm": 1.0431745052337646, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 39285 + }, + { + "epoch": 0.39286, + "grad_norm": 1.0045287609100342, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 39286 + }, + { + "epoch": 0.39287, + "grad_norm": 0.8489604592323303, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 39287 + }, + { + "epoch": 0.39288, + "grad_norm": 0.7786408066749573, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 39288 + }, + { + "epoch": 0.39289, + "grad_norm": 0.8725773692131042, + "learning_rate": 0.003, + "loss": 3.996, + "step": 39289 + }, + { + "epoch": 0.3929, + "grad_norm": 0.8706948161125183, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 39290 + }, + { + "epoch": 0.39291, + "grad_norm": 0.8296987414360046, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 39291 + }, + { + "epoch": 0.39292, + "grad_norm": 0.8508706092834473, + "learning_rate": 0.003, + "loss": 4.007, + "step": 39292 + }, + { + "epoch": 0.39293, + "grad_norm": 0.927272617816925, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 39293 + }, + { + "epoch": 0.39294, + "grad_norm": 0.9708913564682007, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 39294 + }, + { + "epoch": 0.39295, + "grad_norm": 0.9106818437576294, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 39295 + }, + { + "epoch": 0.39296, + "grad_norm": 0.8319652676582336, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 39296 + }, + { + "epoch": 0.39297, + "grad_norm": 0.7718855142593384, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 39297 + }, + { + "epoch": 0.39298, + "grad_norm": 0.7999560236930847, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 39298 + }, + { + "epoch": 0.39299, + "grad_norm": 0.8627107739448547, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 39299 + }, + { + "epoch": 0.393, + "grad_norm": 0.9119279384613037, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 39300 + }, + { + "epoch": 0.39301, + "grad_norm": 0.8125364184379578, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 39301 + }, + { + "epoch": 0.39302, + "grad_norm": 0.8741961717605591, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 39302 + }, + { + "epoch": 0.39303, + "grad_norm": 0.7844623923301697, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 39303 + }, + { + "epoch": 0.39304, + "grad_norm": 0.7046079635620117, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 39304 + }, + { + "epoch": 0.39305, + "grad_norm": 0.7025046944618225, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 39305 + }, + { + "epoch": 0.39306, + "grad_norm": 0.6438096761703491, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 39306 + }, + { + "epoch": 0.39307, + "grad_norm": 0.647989809513092, + "learning_rate": 0.003, + "loss": 3.96, + "step": 39307 + }, + { + "epoch": 0.39308, + "grad_norm": 0.7274235486984253, + "learning_rate": 0.003, + "loss": 3.9489, + "step": 39308 + }, + { + "epoch": 0.39309, + "grad_norm": 0.907917857170105, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 39309 + }, + { + "epoch": 0.3931, + "grad_norm": 1.1380130052566528, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 39310 + }, + { + "epoch": 0.39311, + "grad_norm": 1.300797462463379, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 39311 + }, + { + "epoch": 0.39312, + "grad_norm": 0.7556284666061401, + "learning_rate": 0.003, + "loss": 4.001, + "step": 39312 + }, + { + "epoch": 0.39313, + "grad_norm": 0.6375650763511658, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 39313 + }, + { + "epoch": 0.39314, + "grad_norm": 0.7233129739761353, + "learning_rate": 0.003, + "loss": 3.979, + "step": 39314 + }, + { + "epoch": 0.39315, + "grad_norm": 0.9215021133422852, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 39315 + }, + { + "epoch": 0.39316, + "grad_norm": 1.2224154472351074, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 39316 + }, + { + "epoch": 0.39317, + "grad_norm": 0.8545042872428894, + "learning_rate": 0.003, + "loss": 3.966, + "step": 39317 + }, + { + "epoch": 0.39318, + "grad_norm": 0.7666043043136597, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 39318 + }, + { + "epoch": 0.39319, + "grad_norm": 0.7172322273254395, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 39319 + }, + { + "epoch": 0.3932, + "grad_norm": 0.7883605360984802, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 39320 + }, + { + "epoch": 0.39321, + "grad_norm": 1.0819430351257324, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 39321 + }, + { + "epoch": 0.39322, + "grad_norm": 1.2101227045059204, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 39322 + }, + { + "epoch": 0.39323, + "grad_norm": 0.811102569103241, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 39323 + }, + { + "epoch": 0.39324, + "grad_norm": 0.8041086196899414, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 39324 + }, + { + "epoch": 0.39325, + "grad_norm": 1.0747673511505127, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 39325 + }, + { + "epoch": 0.39326, + "grad_norm": 1.0708130598068237, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 39326 + }, + { + "epoch": 0.39327, + "grad_norm": 0.8537975549697876, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 39327 + }, + { + "epoch": 0.39328, + "grad_norm": 1.1108163595199585, + "learning_rate": 0.003, + "loss": 3.998, + "step": 39328 + }, + { + "epoch": 0.39329, + "grad_norm": 1.0000139474868774, + "learning_rate": 0.003, + "loss": 4.016, + "step": 39329 + }, + { + "epoch": 0.3933, + "grad_norm": 1.1657319068908691, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 39330 + }, + { + "epoch": 0.39331, + "grad_norm": 1.0733345746994019, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 39331 + }, + { + "epoch": 0.39332, + "grad_norm": 1.1524747610092163, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 39332 + }, + { + "epoch": 0.39333, + "grad_norm": 1.053207278251648, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 39333 + }, + { + "epoch": 0.39334, + "grad_norm": 1.1601238250732422, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 39334 + }, + { + "epoch": 0.39335, + "grad_norm": 1.218551516532898, + "learning_rate": 0.003, + "loss": 4.008, + "step": 39335 + }, + { + "epoch": 0.39336, + "grad_norm": 0.6182349324226379, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 39336 + }, + { + "epoch": 0.39337, + "grad_norm": 0.6540300846099854, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 39337 + }, + { + "epoch": 0.39338, + "grad_norm": 0.8136763572692871, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 39338 + }, + { + "epoch": 0.39339, + "grad_norm": 0.9177677631378174, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 39339 + }, + { + "epoch": 0.3934, + "grad_norm": 1.048399806022644, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 39340 + }, + { + "epoch": 0.39341, + "grad_norm": 0.8671581745147705, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 39341 + }, + { + "epoch": 0.39342, + "grad_norm": 0.7644180655479431, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 39342 + }, + { + "epoch": 0.39343, + "grad_norm": 0.9020448923110962, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 39343 + }, + { + "epoch": 0.39344, + "grad_norm": 1.0508159399032593, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 39344 + }, + { + "epoch": 0.39345, + "grad_norm": 1.1386260986328125, + "learning_rate": 0.003, + "loss": 3.995, + "step": 39345 + }, + { + "epoch": 0.39346, + "grad_norm": 0.9430039525032043, + "learning_rate": 0.003, + "loss": 3.994, + "step": 39346 + }, + { + "epoch": 0.39347, + "grad_norm": 0.9350953698158264, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 39347 + }, + { + "epoch": 0.39348, + "grad_norm": 1.027167558670044, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 39348 + }, + { + "epoch": 0.39349, + "grad_norm": 0.9584885239601135, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 39349 + }, + { + "epoch": 0.3935, + "grad_norm": 0.9180642366409302, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 39350 + }, + { + "epoch": 0.39351, + "grad_norm": 0.990328311920166, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 39351 + }, + { + "epoch": 0.39352, + "grad_norm": 1.0396742820739746, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 39352 + }, + { + "epoch": 0.39353, + "grad_norm": 0.976937472820282, + "learning_rate": 0.003, + "loss": 3.9462, + "step": 39353 + }, + { + "epoch": 0.39354, + "grad_norm": 1.2055895328521729, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 39354 + }, + { + "epoch": 0.39355, + "grad_norm": 0.6632953882217407, + "learning_rate": 0.003, + "loss": 3.955, + "step": 39355 + }, + { + "epoch": 0.39356, + "grad_norm": 0.5480393767356873, + "learning_rate": 0.003, + "loss": 3.992, + "step": 39356 + }, + { + "epoch": 0.39357, + "grad_norm": 0.7236352562904358, + "learning_rate": 0.003, + "loss": 3.974, + "step": 39357 + }, + { + "epoch": 0.39358, + "grad_norm": 0.8928993940353394, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 39358 + }, + { + "epoch": 0.39359, + "grad_norm": 1.1083531379699707, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 39359 + }, + { + "epoch": 0.3936, + "grad_norm": 0.9013527035713196, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 39360 + }, + { + "epoch": 0.39361, + "grad_norm": 0.775601863861084, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 39361 + }, + { + "epoch": 0.39362, + "grad_norm": 0.8294193744659424, + "learning_rate": 0.003, + "loss": 3.978, + "step": 39362 + }, + { + "epoch": 0.39363, + "grad_norm": 1.030385136604309, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 39363 + }, + { + "epoch": 0.39364, + "grad_norm": 1.243065357208252, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 39364 + }, + { + "epoch": 0.39365, + "grad_norm": 0.8885101675987244, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 39365 + }, + { + "epoch": 0.39366, + "grad_norm": 0.8989669680595398, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 39366 + }, + { + "epoch": 0.39367, + "grad_norm": 1.0644614696502686, + "learning_rate": 0.003, + "loss": 3.977, + "step": 39367 + }, + { + "epoch": 0.39368, + "grad_norm": 1.0360774993896484, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 39368 + }, + { + "epoch": 0.39369, + "grad_norm": 0.8375632762908936, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 39369 + }, + { + "epoch": 0.3937, + "grad_norm": 0.7882668375968933, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 39370 + }, + { + "epoch": 0.39371, + "grad_norm": 0.7511681318283081, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 39371 + }, + { + "epoch": 0.39372, + "grad_norm": 0.8342697620391846, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 39372 + }, + { + "epoch": 0.39373, + "grad_norm": 0.9516710638999939, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 39373 + }, + { + "epoch": 0.39374, + "grad_norm": 0.928092360496521, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 39374 + }, + { + "epoch": 0.39375, + "grad_norm": 0.8209319114685059, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 39375 + }, + { + "epoch": 0.39376, + "grad_norm": 0.9645319581031799, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 39376 + }, + { + "epoch": 0.39377, + "grad_norm": 1.4168814420700073, + "learning_rate": 0.003, + "loss": 3.999, + "step": 39377 + }, + { + "epoch": 0.39378, + "grad_norm": 0.6777527928352356, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 39378 + }, + { + "epoch": 0.39379, + "grad_norm": 0.6972100138664246, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 39379 + }, + { + "epoch": 0.3938, + "grad_norm": 0.9467665553092957, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 39380 + }, + { + "epoch": 0.39381, + "grad_norm": 1.0077967643737793, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 39381 + }, + { + "epoch": 0.39382, + "grad_norm": 0.9943060278892517, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 39382 + }, + { + "epoch": 0.39383, + "grad_norm": 1.025320053100586, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 39383 + }, + { + "epoch": 0.39384, + "grad_norm": 0.9689584970474243, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 39384 + }, + { + "epoch": 0.39385, + "grad_norm": 0.8578867316246033, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 39385 + }, + { + "epoch": 0.39386, + "grad_norm": 0.8661893606185913, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 39386 + }, + { + "epoch": 0.39387, + "grad_norm": 0.8456808924674988, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 39387 + }, + { + "epoch": 0.39388, + "grad_norm": 1.0050028562545776, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 39388 + }, + { + "epoch": 0.39389, + "grad_norm": 1.1173486709594727, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 39389 + }, + { + "epoch": 0.3939, + "grad_norm": 0.8918794393539429, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 39390 + }, + { + "epoch": 0.39391, + "grad_norm": 0.8513628244400024, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 39391 + }, + { + "epoch": 0.39392, + "grad_norm": 0.9021734595298767, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 39392 + }, + { + "epoch": 0.39393, + "grad_norm": 0.998335063457489, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 39393 + }, + { + "epoch": 0.39394, + "grad_norm": 1.0527180433273315, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 39394 + }, + { + "epoch": 0.39395, + "grad_norm": 0.9227848052978516, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 39395 + }, + { + "epoch": 0.39396, + "grad_norm": 1.0183171033859253, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 39396 + }, + { + "epoch": 0.39397, + "grad_norm": 1.064790964126587, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 39397 + }, + { + "epoch": 0.39398, + "grad_norm": 0.9703193306922913, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 39398 + }, + { + "epoch": 0.39399, + "grad_norm": 1.0073703527450562, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 39399 + }, + { + "epoch": 0.394, + "grad_norm": 1.0049622058868408, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 39400 + }, + { + "epoch": 0.39401, + "grad_norm": 1.030996322631836, + "learning_rate": 0.003, + "loss": 3.983, + "step": 39401 + }, + { + "epoch": 0.39402, + "grad_norm": 0.8245805501937866, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 39402 + }, + { + "epoch": 0.39403, + "grad_norm": 0.7286421656608582, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 39403 + }, + { + "epoch": 0.39404, + "grad_norm": 0.7650429010391235, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 39404 + }, + { + "epoch": 0.39405, + "grad_norm": 0.7022450566291809, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 39405 + }, + { + "epoch": 0.39406, + "grad_norm": 0.5807328224182129, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 39406 + }, + { + "epoch": 0.39407, + "grad_norm": 0.6531258225440979, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 39407 + }, + { + "epoch": 0.39408, + "grad_norm": 0.7178402543067932, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 39408 + }, + { + "epoch": 0.39409, + "grad_norm": 0.7331390976905823, + "learning_rate": 0.003, + "loss": 3.963, + "step": 39409 + }, + { + "epoch": 0.3941, + "grad_norm": 1.037203311920166, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 39410 + }, + { + "epoch": 0.39411, + "grad_norm": 1.4137500524520874, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 39411 + }, + { + "epoch": 0.39412, + "grad_norm": 0.6833665370941162, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 39412 + }, + { + "epoch": 0.39413, + "grad_norm": 0.7765567302703857, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 39413 + }, + { + "epoch": 0.39414, + "grad_norm": 0.8307558298110962, + "learning_rate": 0.003, + "loss": 3.982, + "step": 39414 + }, + { + "epoch": 0.39415, + "grad_norm": 0.9397116899490356, + "learning_rate": 0.003, + "loss": 3.996, + "step": 39415 + }, + { + "epoch": 0.39416, + "grad_norm": 0.8748547434806824, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 39416 + }, + { + "epoch": 0.39417, + "grad_norm": 0.7318115830421448, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 39417 + }, + { + "epoch": 0.39418, + "grad_norm": 0.6280592679977417, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 39418 + }, + { + "epoch": 0.39419, + "grad_norm": 0.7706503868103027, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 39419 + }, + { + "epoch": 0.3942, + "grad_norm": 0.8746979832649231, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 39420 + }, + { + "epoch": 0.39421, + "grad_norm": 1.1740241050720215, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 39421 + }, + { + "epoch": 0.39422, + "grad_norm": 0.9855701923370361, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 39422 + }, + { + "epoch": 0.39423, + "grad_norm": 0.935369610786438, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 39423 + }, + { + "epoch": 0.39424, + "grad_norm": 0.9625263214111328, + "learning_rate": 0.003, + "loss": 3.979, + "step": 39424 + }, + { + "epoch": 0.39425, + "grad_norm": 0.8913097381591797, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 39425 + }, + { + "epoch": 0.39426, + "grad_norm": 0.9076449871063232, + "learning_rate": 0.003, + "loss": 3.947, + "step": 39426 + }, + { + "epoch": 0.39427, + "grad_norm": 1.055273175239563, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 39427 + }, + { + "epoch": 0.39428, + "grad_norm": 0.9126793742179871, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 39428 + }, + { + "epoch": 0.39429, + "grad_norm": 0.7121003270149231, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 39429 + }, + { + "epoch": 0.3943, + "grad_norm": 0.722160279750824, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 39430 + }, + { + "epoch": 0.39431, + "grad_norm": 0.7218068838119507, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 39431 + }, + { + "epoch": 0.39432, + "grad_norm": 0.6209065318107605, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 39432 + }, + { + "epoch": 0.39433, + "grad_norm": 0.7514164447784424, + "learning_rate": 0.003, + "loss": 3.968, + "step": 39433 + }, + { + "epoch": 0.39434, + "grad_norm": 0.9989203214645386, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 39434 + }, + { + "epoch": 0.39435, + "grad_norm": 1.080823302268982, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 39435 + }, + { + "epoch": 0.39436, + "grad_norm": 0.8974745273590088, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 39436 + }, + { + "epoch": 0.39437, + "grad_norm": 0.8451781868934631, + "learning_rate": 0.003, + "loss": 3.9424, + "step": 39437 + }, + { + "epoch": 0.39438, + "grad_norm": 0.9818118810653687, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 39438 + }, + { + "epoch": 0.39439, + "grad_norm": 1.1832873821258545, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 39439 + }, + { + "epoch": 0.3944, + "grad_norm": 0.8620732426643372, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 39440 + }, + { + "epoch": 0.39441, + "grad_norm": 0.7298000454902649, + "learning_rate": 0.003, + "loss": 3.994, + "step": 39441 + }, + { + "epoch": 0.39442, + "grad_norm": 0.7486200928688049, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 39442 + }, + { + "epoch": 0.39443, + "grad_norm": 0.9341102838516235, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 39443 + }, + { + "epoch": 0.39444, + "grad_norm": 1.0767278671264648, + "learning_rate": 0.003, + "loss": 4.007, + "step": 39444 + }, + { + "epoch": 0.39445, + "grad_norm": 0.9739298820495605, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 39445 + }, + { + "epoch": 0.39446, + "grad_norm": 0.9797956347465515, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 39446 + }, + { + "epoch": 0.39447, + "grad_norm": 0.9776952862739563, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 39447 + }, + { + "epoch": 0.39448, + "grad_norm": 0.9918183088302612, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 39448 + }, + { + "epoch": 0.39449, + "grad_norm": 1.1700165271759033, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 39449 + }, + { + "epoch": 0.3945, + "grad_norm": 0.8985574245452881, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 39450 + }, + { + "epoch": 0.39451, + "grad_norm": 0.9102829694747925, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 39451 + }, + { + "epoch": 0.39452, + "grad_norm": 1.0023086071014404, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 39452 + }, + { + "epoch": 0.39453, + "grad_norm": 1.125227689743042, + "learning_rate": 0.003, + "loss": 3.982, + "step": 39453 + }, + { + "epoch": 0.39454, + "grad_norm": 0.7337061762809753, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 39454 + }, + { + "epoch": 0.39455, + "grad_norm": 0.7245105504989624, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 39455 + }, + { + "epoch": 0.39456, + "grad_norm": 0.8318695425987244, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 39456 + }, + { + "epoch": 0.39457, + "grad_norm": 1.0898563861846924, + "learning_rate": 0.003, + "loss": 3.997, + "step": 39457 + }, + { + "epoch": 0.39458, + "grad_norm": 0.997572124004364, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 39458 + }, + { + "epoch": 0.39459, + "grad_norm": 0.9065873026847839, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 39459 + }, + { + "epoch": 0.3946, + "grad_norm": 0.8426827192306519, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 39460 + }, + { + "epoch": 0.39461, + "grad_norm": 0.8361157178878784, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 39461 + }, + { + "epoch": 0.39462, + "grad_norm": 0.853642463684082, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 39462 + }, + { + "epoch": 0.39463, + "grad_norm": 0.9343138337135315, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 39463 + }, + { + "epoch": 0.39464, + "grad_norm": 1.0332677364349365, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 39464 + }, + { + "epoch": 0.39465, + "grad_norm": 1.049599051475525, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 39465 + }, + { + "epoch": 0.39466, + "grad_norm": 0.9307602643966675, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 39466 + }, + { + "epoch": 0.39467, + "grad_norm": 0.7361392974853516, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 39467 + }, + { + "epoch": 0.39468, + "grad_norm": 0.623975932598114, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 39468 + }, + { + "epoch": 0.39469, + "grad_norm": 0.721177875995636, + "learning_rate": 0.003, + "loss": 3.9458, + "step": 39469 + }, + { + "epoch": 0.3947, + "grad_norm": 0.7714614272117615, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 39470 + }, + { + "epoch": 0.39471, + "grad_norm": 0.9712254405021667, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 39471 + }, + { + "epoch": 0.39472, + "grad_norm": 1.212600588798523, + "learning_rate": 0.003, + "loss": 4.004, + "step": 39472 + }, + { + "epoch": 0.39473, + "grad_norm": 1.0219109058380127, + "learning_rate": 0.003, + "loss": 4.007, + "step": 39473 + }, + { + "epoch": 0.39474, + "grad_norm": 0.9929548501968384, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 39474 + }, + { + "epoch": 0.39475, + "grad_norm": 1.107399582862854, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 39475 + }, + { + "epoch": 0.39476, + "grad_norm": 0.9993430972099304, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 39476 + }, + { + "epoch": 0.39477, + "grad_norm": 0.8320327401161194, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 39477 + }, + { + "epoch": 0.39478, + "grad_norm": 0.8553779125213623, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 39478 + }, + { + "epoch": 0.39479, + "grad_norm": 1.0917407274246216, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 39479 + }, + { + "epoch": 0.3948, + "grad_norm": 0.9448891282081604, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 39480 + }, + { + "epoch": 0.39481, + "grad_norm": 0.8555961847305298, + "learning_rate": 0.003, + "loss": 3.944, + "step": 39481 + }, + { + "epoch": 0.39482, + "grad_norm": 0.7689669132232666, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 39482 + }, + { + "epoch": 0.39483, + "grad_norm": 0.7164962887763977, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 39483 + }, + { + "epoch": 0.39484, + "grad_norm": 0.6586937308311462, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 39484 + }, + { + "epoch": 0.39485, + "grad_norm": 0.858056366443634, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 39485 + }, + { + "epoch": 0.39486, + "grad_norm": 1.1136893033981323, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 39486 + }, + { + "epoch": 0.39487, + "grad_norm": 0.9356541633605957, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 39487 + }, + { + "epoch": 0.39488, + "grad_norm": 1.0323816537857056, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 39488 + }, + { + "epoch": 0.39489, + "grad_norm": 0.9550389647483826, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 39489 + }, + { + "epoch": 0.3949, + "grad_norm": 0.9424618482589722, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 39490 + }, + { + "epoch": 0.39491, + "grad_norm": 1.1587949991226196, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 39491 + }, + { + "epoch": 0.39492, + "grad_norm": 1.2876251935958862, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 39492 + }, + { + "epoch": 0.39493, + "grad_norm": 0.6790644526481628, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 39493 + }, + { + "epoch": 0.39494, + "grad_norm": 0.688975989818573, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 39494 + }, + { + "epoch": 0.39495, + "grad_norm": 0.8053507804870605, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 39495 + }, + { + "epoch": 0.39496, + "grad_norm": 0.8592070937156677, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 39496 + }, + { + "epoch": 0.39497, + "grad_norm": 1.010407567024231, + "learning_rate": 0.003, + "loss": 4.002, + "step": 39497 + }, + { + "epoch": 0.39498, + "grad_norm": 0.883693516254425, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 39498 + }, + { + "epoch": 0.39499, + "grad_norm": 0.8109884262084961, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 39499 + }, + { + "epoch": 0.395, + "grad_norm": 0.8186984658241272, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 39500 + }, + { + "epoch": 0.39501, + "grad_norm": 0.8826310038566589, + "learning_rate": 0.003, + "loss": 3.968, + "step": 39501 + }, + { + "epoch": 0.39502, + "grad_norm": 1.0486481189727783, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 39502 + }, + { + "epoch": 0.39503, + "grad_norm": 0.9615862369537354, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 39503 + }, + { + "epoch": 0.39504, + "grad_norm": 0.8275964260101318, + "learning_rate": 0.003, + "loss": 3.972, + "step": 39504 + }, + { + "epoch": 0.39505, + "grad_norm": 0.6939815282821655, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 39505 + }, + { + "epoch": 0.39506, + "grad_norm": 0.7050009965896606, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 39506 + }, + { + "epoch": 0.39507, + "grad_norm": 0.7544739842414856, + "learning_rate": 0.003, + "loss": 3.97, + "step": 39507 + }, + { + "epoch": 0.39508, + "grad_norm": 0.7793416380882263, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 39508 + }, + { + "epoch": 0.39509, + "grad_norm": 0.6496865749359131, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 39509 + }, + { + "epoch": 0.3951, + "grad_norm": 0.666163444519043, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 39510 + }, + { + "epoch": 0.39511, + "grad_norm": 0.9191520810127258, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 39511 + }, + { + "epoch": 0.39512, + "grad_norm": 1.2494332790374756, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 39512 + }, + { + "epoch": 0.39513, + "grad_norm": 0.8363098502159119, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 39513 + }, + { + "epoch": 0.39514, + "grad_norm": 0.6931872367858887, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 39514 + }, + { + "epoch": 0.39515, + "grad_norm": 0.6765375733375549, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 39515 + }, + { + "epoch": 0.39516, + "grad_norm": 0.7022161483764648, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 39516 + }, + { + "epoch": 0.39517, + "grad_norm": 0.840650737285614, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 39517 + }, + { + "epoch": 0.39518, + "grad_norm": 1.1124837398529053, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 39518 + }, + { + "epoch": 0.39519, + "grad_norm": 1.0633612871170044, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 39519 + }, + { + "epoch": 0.3952, + "grad_norm": 0.8386557698249817, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 39520 + }, + { + "epoch": 0.39521, + "grad_norm": 0.862890899181366, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 39521 + }, + { + "epoch": 0.39522, + "grad_norm": 0.8874621987342834, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 39522 + }, + { + "epoch": 0.39523, + "grad_norm": 0.854511559009552, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 39523 + }, + { + "epoch": 0.39524, + "grad_norm": 0.8622159361839294, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 39524 + }, + { + "epoch": 0.39525, + "grad_norm": 1.059640645980835, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 39525 + }, + { + "epoch": 0.39526, + "grad_norm": 0.9420422911643982, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 39526 + }, + { + "epoch": 0.39527, + "grad_norm": 1.042703628540039, + "learning_rate": 0.003, + "loss": 4.013, + "step": 39527 + }, + { + "epoch": 0.39528, + "grad_norm": 0.792231559753418, + "learning_rate": 0.003, + "loss": 3.965, + "step": 39528 + }, + { + "epoch": 0.39529, + "grad_norm": 0.6881149411201477, + "learning_rate": 0.003, + "loss": 3.999, + "step": 39529 + }, + { + "epoch": 0.3953, + "grad_norm": 0.6487579345703125, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 39530 + }, + { + "epoch": 0.39531, + "grad_norm": 0.8916478157043457, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 39531 + }, + { + "epoch": 0.39532, + "grad_norm": 1.1603126525878906, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 39532 + }, + { + "epoch": 0.39533, + "grad_norm": 1.2496081590652466, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 39533 + }, + { + "epoch": 0.39534, + "grad_norm": 0.7631696462631226, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 39534 + }, + { + "epoch": 0.39535, + "grad_norm": 0.7852368354797363, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 39535 + }, + { + "epoch": 0.39536, + "grad_norm": 0.9262005686759949, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 39536 + }, + { + "epoch": 0.39537, + "grad_norm": 1.0157461166381836, + "learning_rate": 0.003, + "loss": 3.984, + "step": 39537 + }, + { + "epoch": 0.39538, + "grad_norm": 0.9331406354904175, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 39538 + }, + { + "epoch": 0.39539, + "grad_norm": 0.88294517993927, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 39539 + }, + { + "epoch": 0.3954, + "grad_norm": 0.7647422552108765, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 39540 + }, + { + "epoch": 0.39541, + "grad_norm": 0.8591750860214233, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 39541 + }, + { + "epoch": 0.39542, + "grad_norm": 0.8520599007606506, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 39542 + }, + { + "epoch": 0.39543, + "grad_norm": 1.0307492017745972, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 39543 + }, + { + "epoch": 0.39544, + "grad_norm": 1.174776554107666, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 39544 + }, + { + "epoch": 0.39545, + "grad_norm": 0.9539943933486938, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 39545 + }, + { + "epoch": 0.39546, + "grad_norm": 0.8113473057746887, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 39546 + }, + { + "epoch": 0.39547, + "grad_norm": 0.6645479798316956, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 39547 + }, + { + "epoch": 0.39548, + "grad_norm": 0.6258794069290161, + "learning_rate": 0.003, + "loss": 3.9409, + "step": 39548 + }, + { + "epoch": 0.39549, + "grad_norm": 0.7668212056159973, + "learning_rate": 0.003, + "loss": 3.9446, + "step": 39549 + }, + { + "epoch": 0.3955, + "grad_norm": 1.048069953918457, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 39550 + }, + { + "epoch": 0.39551, + "grad_norm": 1.4918259382247925, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 39551 + }, + { + "epoch": 0.39552, + "grad_norm": 0.6936498284339905, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 39552 + }, + { + "epoch": 0.39553, + "grad_norm": 0.8255947232246399, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 39553 + }, + { + "epoch": 0.39554, + "grad_norm": 0.8296370506286621, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 39554 + }, + { + "epoch": 0.39555, + "grad_norm": 0.9308274984359741, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 39555 + }, + { + "epoch": 0.39556, + "grad_norm": 1.1411813497543335, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 39556 + }, + { + "epoch": 0.39557, + "grad_norm": 0.9100428819656372, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 39557 + }, + { + "epoch": 0.39558, + "grad_norm": 1.0502036809921265, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 39558 + }, + { + "epoch": 0.39559, + "grad_norm": 1.229467511177063, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 39559 + }, + { + "epoch": 0.3956, + "grad_norm": 0.6453992128372192, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 39560 + }, + { + "epoch": 0.39561, + "grad_norm": 0.7393035292625427, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 39561 + }, + { + "epoch": 0.39562, + "grad_norm": 0.9768728017807007, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 39562 + }, + { + "epoch": 0.39563, + "grad_norm": 1.1046113967895508, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 39563 + }, + { + "epoch": 0.39564, + "grad_norm": 0.9618154764175415, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 39564 + }, + { + "epoch": 0.39565, + "grad_norm": 0.9429236054420471, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 39565 + }, + { + "epoch": 0.39566, + "grad_norm": 0.9071834087371826, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 39566 + }, + { + "epoch": 0.39567, + "grad_norm": 0.9804063439369202, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 39567 + }, + { + "epoch": 0.39568, + "grad_norm": 0.9222700595855713, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 39568 + }, + { + "epoch": 0.39569, + "grad_norm": 0.9218378067016602, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 39569 + }, + { + "epoch": 0.3957, + "grad_norm": 0.8767896294593811, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 39570 + }, + { + "epoch": 0.39571, + "grad_norm": 0.7576056718826294, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 39571 + }, + { + "epoch": 0.39572, + "grad_norm": 0.7836761474609375, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 39572 + }, + { + "epoch": 0.39573, + "grad_norm": 0.8476893901824951, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 39573 + }, + { + "epoch": 0.39574, + "grad_norm": 1.0869081020355225, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 39574 + }, + { + "epoch": 0.39575, + "grad_norm": 1.0228351354599, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 39575 + }, + { + "epoch": 0.39576, + "grad_norm": 0.8889255523681641, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 39576 + }, + { + "epoch": 0.39577, + "grad_norm": 0.9857422113418579, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 39577 + }, + { + "epoch": 0.39578, + "grad_norm": 1.0330357551574707, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 39578 + }, + { + "epoch": 0.39579, + "grad_norm": 1.1704363822937012, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 39579 + }, + { + "epoch": 0.3958, + "grad_norm": 1.1002787351608276, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 39580 + }, + { + "epoch": 0.39581, + "grad_norm": 1.0843886137008667, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 39581 + }, + { + "epoch": 0.39582, + "grad_norm": 0.9508997797966003, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 39582 + }, + { + "epoch": 0.39583, + "grad_norm": 0.8477262258529663, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 39583 + }, + { + "epoch": 0.39584, + "grad_norm": 0.8788061738014221, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 39584 + }, + { + "epoch": 0.39585, + "grad_norm": 0.9082399606704712, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 39585 + }, + { + "epoch": 0.39586, + "grad_norm": 0.962428867816925, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 39586 + }, + { + "epoch": 0.39587, + "grad_norm": 0.8474539518356323, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 39587 + }, + { + "epoch": 0.39588, + "grad_norm": 0.7586913704872131, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 39588 + }, + { + "epoch": 0.39589, + "grad_norm": 0.9515244960784912, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 39589 + }, + { + "epoch": 0.3959, + "grad_norm": 1.5722540616989136, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 39590 + }, + { + "epoch": 0.39591, + "grad_norm": 0.4680001735687256, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 39591 + }, + { + "epoch": 0.39592, + "grad_norm": 1.06490957736969, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 39592 + }, + { + "epoch": 0.39593, + "grad_norm": 1.2269724607467651, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 39593 + }, + { + "epoch": 0.39594, + "grad_norm": 0.6161378026008606, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 39594 + }, + { + "epoch": 0.39595, + "grad_norm": 0.5548655986785889, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 39595 + }, + { + "epoch": 0.39596, + "grad_norm": 0.6426730155944824, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 39596 + }, + { + "epoch": 0.39597, + "grad_norm": 0.6286534667015076, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 39597 + }, + { + "epoch": 0.39598, + "grad_norm": 0.6915296316146851, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 39598 + }, + { + "epoch": 0.39599, + "grad_norm": 0.758804440498352, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 39599 + }, + { + "epoch": 0.396, + "grad_norm": 0.7696107029914856, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 39600 + }, + { + "epoch": 0.39601, + "grad_norm": 0.7963215112686157, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 39601 + }, + { + "epoch": 0.39602, + "grad_norm": 0.8771117329597473, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 39602 + }, + { + "epoch": 0.39603, + "grad_norm": 1.05582857131958, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 39603 + }, + { + "epoch": 0.39604, + "grad_norm": 1.1230361461639404, + "learning_rate": 0.003, + "loss": 3.977, + "step": 39604 + }, + { + "epoch": 0.39605, + "grad_norm": 0.8642532825469971, + "learning_rate": 0.003, + "loss": 3.981, + "step": 39605 + }, + { + "epoch": 0.39606, + "grad_norm": 0.8860926032066345, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 39606 + }, + { + "epoch": 0.39607, + "grad_norm": 0.9677561521530151, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 39607 + }, + { + "epoch": 0.39608, + "grad_norm": 1.140480637550354, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 39608 + }, + { + "epoch": 0.39609, + "grad_norm": 0.7512319087982178, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 39609 + }, + { + "epoch": 0.3961, + "grad_norm": 0.8493725657463074, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 39610 + }, + { + "epoch": 0.39611, + "grad_norm": 0.8733487725257874, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 39611 + }, + { + "epoch": 0.39612, + "grad_norm": 0.8025281429290771, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 39612 + }, + { + "epoch": 0.39613, + "grad_norm": 0.7720041871070862, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 39613 + }, + { + "epoch": 0.39614, + "grad_norm": 0.8223550319671631, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 39614 + }, + { + "epoch": 0.39615, + "grad_norm": 0.883054256439209, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 39615 + }, + { + "epoch": 0.39616, + "grad_norm": 0.9377625584602356, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 39616 + }, + { + "epoch": 0.39617, + "grad_norm": 1.2634916305541992, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 39617 + }, + { + "epoch": 0.39618, + "grad_norm": 0.9735322594642639, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 39618 + }, + { + "epoch": 0.39619, + "grad_norm": 0.9703618288040161, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 39619 + }, + { + "epoch": 0.3962, + "grad_norm": 1.0584808588027954, + "learning_rate": 0.003, + "loss": 3.992, + "step": 39620 + }, + { + "epoch": 0.39621, + "grad_norm": 0.9568134546279907, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 39621 + }, + { + "epoch": 0.39622, + "grad_norm": 0.9786168336868286, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 39622 + }, + { + "epoch": 0.39623, + "grad_norm": 0.865325927734375, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 39623 + }, + { + "epoch": 0.39624, + "grad_norm": 0.9499419927597046, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 39624 + }, + { + "epoch": 0.39625, + "grad_norm": 1.1112092733383179, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 39625 + }, + { + "epoch": 0.39626, + "grad_norm": 0.9499293565750122, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 39626 + }, + { + "epoch": 0.39627, + "grad_norm": 0.7832777500152588, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 39627 + }, + { + "epoch": 0.39628, + "grad_norm": 0.7499008774757385, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 39628 + }, + { + "epoch": 0.39629, + "grad_norm": 0.7417532801628113, + "learning_rate": 0.003, + "loss": 3.9431, + "step": 39629 + }, + { + "epoch": 0.3963, + "grad_norm": 0.8195115327835083, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 39630 + }, + { + "epoch": 0.39631, + "grad_norm": 0.9213013052940369, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 39631 + }, + { + "epoch": 0.39632, + "grad_norm": 0.9378427267074585, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 39632 + }, + { + "epoch": 0.39633, + "grad_norm": 1.0062326192855835, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 39633 + }, + { + "epoch": 0.39634, + "grad_norm": 1.3074791431427002, + "learning_rate": 0.003, + "loss": 3.996, + "step": 39634 + }, + { + "epoch": 0.39635, + "grad_norm": 0.5489073991775513, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 39635 + }, + { + "epoch": 0.39636, + "grad_norm": 0.8519502878189087, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 39636 + }, + { + "epoch": 0.39637, + "grad_norm": 1.2486897706985474, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 39637 + }, + { + "epoch": 0.39638, + "grad_norm": 0.5285201072692871, + "learning_rate": 0.003, + "loss": 3.9361, + "step": 39638 + }, + { + "epoch": 0.39639, + "grad_norm": 0.7496002316474915, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 39639 + }, + { + "epoch": 0.3964, + "grad_norm": 1.0003018379211426, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 39640 + }, + { + "epoch": 0.39641, + "grad_norm": 0.8916643857955933, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 39641 + }, + { + "epoch": 0.39642, + "grad_norm": 0.7778835892677307, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 39642 + }, + { + "epoch": 0.39643, + "grad_norm": 0.8122524619102478, + "learning_rate": 0.003, + "loss": 3.9458, + "step": 39643 + }, + { + "epoch": 0.39644, + "grad_norm": 0.8961004018783569, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 39644 + }, + { + "epoch": 0.39645, + "grad_norm": 0.9349992275238037, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 39645 + }, + { + "epoch": 0.39646, + "grad_norm": 1.007212519645691, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 39646 + }, + { + "epoch": 0.39647, + "grad_norm": 1.0167316198349, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 39647 + }, + { + "epoch": 0.39648, + "grad_norm": 1.0308841466903687, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 39648 + }, + { + "epoch": 0.39649, + "grad_norm": 0.9617045521736145, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 39649 + }, + { + "epoch": 0.3965, + "grad_norm": 1.0186128616333008, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 39650 + }, + { + "epoch": 0.39651, + "grad_norm": 0.9687207937240601, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 39651 + }, + { + "epoch": 0.39652, + "grad_norm": 0.8948940634727478, + "learning_rate": 0.003, + "loss": 3.959, + "step": 39652 + }, + { + "epoch": 0.39653, + "grad_norm": 0.745856523513794, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 39653 + }, + { + "epoch": 0.39654, + "grad_norm": 0.7870029211044312, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 39654 + }, + { + "epoch": 0.39655, + "grad_norm": 0.7619654536247253, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 39655 + }, + { + "epoch": 0.39656, + "grad_norm": 0.6862085461616516, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 39656 + }, + { + "epoch": 0.39657, + "grad_norm": 0.8390547037124634, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 39657 + }, + { + "epoch": 0.39658, + "grad_norm": 0.9751468896865845, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 39658 + }, + { + "epoch": 0.39659, + "grad_norm": 1.516250729560852, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 39659 + }, + { + "epoch": 0.3966, + "grad_norm": 0.6675020456314087, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 39660 + }, + { + "epoch": 0.39661, + "grad_norm": 0.7456993460655212, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 39661 + }, + { + "epoch": 0.39662, + "grad_norm": 0.85455322265625, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 39662 + }, + { + "epoch": 0.39663, + "grad_norm": 1.0907742977142334, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 39663 + }, + { + "epoch": 0.39664, + "grad_norm": 1.1919914484024048, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 39664 + }, + { + "epoch": 0.39665, + "grad_norm": 0.8207470178604126, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 39665 + }, + { + "epoch": 0.39666, + "grad_norm": 0.8162991404533386, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 39666 + }, + { + "epoch": 0.39667, + "grad_norm": 0.8042922019958496, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 39667 + }, + { + "epoch": 0.39668, + "grad_norm": 0.9630677700042725, + "learning_rate": 0.003, + "loss": 3.983, + "step": 39668 + }, + { + "epoch": 0.39669, + "grad_norm": 0.8971039652824402, + "learning_rate": 0.003, + "loss": 3.967, + "step": 39669 + }, + { + "epoch": 0.3967, + "grad_norm": 0.873341977596283, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 39670 + }, + { + "epoch": 0.39671, + "grad_norm": 0.907477617263794, + "learning_rate": 0.003, + "loss": 3.984, + "step": 39671 + }, + { + "epoch": 0.39672, + "grad_norm": 1.048374056816101, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 39672 + }, + { + "epoch": 0.39673, + "grad_norm": 1.1414794921875, + "learning_rate": 0.003, + "loss": 3.993, + "step": 39673 + }, + { + "epoch": 0.39674, + "grad_norm": 0.9157114028930664, + "learning_rate": 0.003, + "loss": 3.99, + "step": 39674 + }, + { + "epoch": 0.39675, + "grad_norm": 1.3411933183670044, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 39675 + }, + { + "epoch": 0.39676, + "grad_norm": 0.9141484498977661, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 39676 + }, + { + "epoch": 0.39677, + "grad_norm": 1.233206868171692, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 39677 + }, + { + "epoch": 0.39678, + "grad_norm": 1.0068484544754028, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 39678 + }, + { + "epoch": 0.39679, + "grad_norm": 0.9143928289413452, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 39679 + }, + { + "epoch": 0.3968, + "grad_norm": 0.8943551182746887, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 39680 + }, + { + "epoch": 0.39681, + "grad_norm": 0.8850294947624207, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 39681 + }, + { + "epoch": 0.39682, + "grad_norm": 0.8464421033859253, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 39682 + }, + { + "epoch": 0.39683, + "grad_norm": 0.866225004196167, + "learning_rate": 0.003, + "loss": 3.951, + "step": 39683 + }, + { + "epoch": 0.39684, + "grad_norm": 0.7694737911224365, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 39684 + }, + { + "epoch": 0.39685, + "grad_norm": 0.6975412964820862, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 39685 + }, + { + "epoch": 0.39686, + "grad_norm": 0.652955174446106, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 39686 + }, + { + "epoch": 0.39687, + "grad_norm": 0.7090306282043457, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 39687 + }, + { + "epoch": 0.39688, + "grad_norm": 0.7160660624504089, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 39688 + }, + { + "epoch": 0.39689, + "grad_norm": 0.6628866791725159, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 39689 + }, + { + "epoch": 0.3969, + "grad_norm": 0.6821931600570679, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 39690 + }, + { + "epoch": 0.39691, + "grad_norm": 0.8955259919166565, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 39691 + }, + { + "epoch": 0.39692, + "grad_norm": 1.1553659439086914, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 39692 + }, + { + "epoch": 0.39693, + "grad_norm": 0.9776885509490967, + "learning_rate": 0.003, + "loss": 3.9165, + "step": 39693 + }, + { + "epoch": 0.39694, + "grad_norm": 1.032961130142212, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 39694 + }, + { + "epoch": 0.39695, + "grad_norm": 1.1399370431900024, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 39695 + }, + { + "epoch": 0.39696, + "grad_norm": 0.7927933931350708, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 39696 + }, + { + "epoch": 0.39697, + "grad_norm": 0.6046369671821594, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 39697 + }, + { + "epoch": 0.39698, + "grad_norm": 0.7371370792388916, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 39698 + }, + { + "epoch": 0.39699, + "grad_norm": 0.8566108345985413, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 39699 + }, + { + "epoch": 0.397, + "grad_norm": 0.961521565914154, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 39700 + }, + { + "epoch": 0.39701, + "grad_norm": 0.9591478109359741, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 39701 + }, + { + "epoch": 0.39702, + "grad_norm": 0.9157148599624634, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 39702 + }, + { + "epoch": 0.39703, + "grad_norm": 0.8524657487869263, + "learning_rate": 0.003, + "loss": 3.97, + "step": 39703 + }, + { + "epoch": 0.39704, + "grad_norm": 0.8336542844772339, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 39704 + }, + { + "epoch": 0.39705, + "grad_norm": 0.9032055735588074, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 39705 + }, + { + "epoch": 0.39706, + "grad_norm": 1.0058033466339111, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 39706 + }, + { + "epoch": 0.39707, + "grad_norm": 1.1773144006729126, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 39707 + }, + { + "epoch": 0.39708, + "grad_norm": 0.8730186820030212, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 39708 + }, + { + "epoch": 0.39709, + "grad_norm": 0.8041767477989197, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 39709 + }, + { + "epoch": 0.3971, + "grad_norm": 0.8619558215141296, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 39710 + }, + { + "epoch": 0.39711, + "grad_norm": 1.0245304107666016, + "learning_rate": 0.003, + "loss": 3.979, + "step": 39711 + }, + { + "epoch": 0.39712, + "grad_norm": 1.2109333276748657, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 39712 + }, + { + "epoch": 0.39713, + "grad_norm": 0.9032860994338989, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 39713 + }, + { + "epoch": 0.39714, + "grad_norm": 0.989656388759613, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 39714 + }, + { + "epoch": 0.39715, + "grad_norm": 1.0307952165603638, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 39715 + }, + { + "epoch": 0.39716, + "grad_norm": 1.1846321821212769, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 39716 + }, + { + "epoch": 0.39717, + "grad_norm": 0.8752274513244629, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 39717 + }, + { + "epoch": 0.39718, + "grad_norm": 0.7708219885826111, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 39718 + }, + { + "epoch": 0.39719, + "grad_norm": 0.9001410007476807, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 39719 + }, + { + "epoch": 0.3972, + "grad_norm": 1.1987162828445435, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 39720 + }, + { + "epoch": 0.39721, + "grad_norm": 0.8038651943206787, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 39721 + }, + { + "epoch": 0.39722, + "grad_norm": 0.7949161529541016, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 39722 + }, + { + "epoch": 0.39723, + "grad_norm": 0.90280681848526, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 39723 + }, + { + "epoch": 0.39724, + "grad_norm": 1.201714277267456, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 39724 + }, + { + "epoch": 0.39725, + "grad_norm": 0.9792006611824036, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 39725 + }, + { + "epoch": 0.39726, + "grad_norm": 0.8824188709259033, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 39726 + }, + { + "epoch": 0.39727, + "grad_norm": 0.822753369808197, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 39727 + }, + { + "epoch": 0.39728, + "grad_norm": 0.8775109052658081, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 39728 + }, + { + "epoch": 0.39729, + "grad_norm": 0.8654429912567139, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 39729 + }, + { + "epoch": 0.3973, + "grad_norm": 0.7374911904335022, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 39730 + }, + { + "epoch": 0.39731, + "grad_norm": 0.8206508159637451, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 39731 + }, + { + "epoch": 0.39732, + "grad_norm": 0.8381550312042236, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 39732 + }, + { + "epoch": 0.39733, + "grad_norm": 0.7256451845169067, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 39733 + }, + { + "epoch": 0.39734, + "grad_norm": 0.722194492816925, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 39734 + }, + { + "epoch": 0.39735, + "grad_norm": 0.716968834400177, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 39735 + }, + { + "epoch": 0.39736, + "grad_norm": 0.7211662530899048, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 39736 + }, + { + "epoch": 0.39737, + "grad_norm": 0.8822270631790161, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 39737 + }, + { + "epoch": 0.39738, + "grad_norm": 1.08498215675354, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 39738 + }, + { + "epoch": 0.39739, + "grad_norm": 0.9944210648536682, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 39739 + }, + { + "epoch": 0.3974, + "grad_norm": 0.9878964424133301, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 39740 + }, + { + "epoch": 0.39741, + "grad_norm": 1.1690438985824585, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 39741 + }, + { + "epoch": 0.39742, + "grad_norm": 0.9155958890914917, + "learning_rate": 0.003, + "loss": 3.948, + "step": 39742 + }, + { + "epoch": 0.39743, + "grad_norm": 0.9037808179855347, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 39743 + }, + { + "epoch": 0.39744, + "grad_norm": 1.0211377143859863, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 39744 + }, + { + "epoch": 0.39745, + "grad_norm": 1.0788112878799438, + "learning_rate": 0.003, + "loss": 3.9404, + "step": 39745 + }, + { + "epoch": 0.39746, + "grad_norm": 0.8350699543952942, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 39746 + }, + { + "epoch": 0.39747, + "grad_norm": 0.6965958476066589, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 39747 + }, + { + "epoch": 0.39748, + "grad_norm": 0.7241303324699402, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 39748 + }, + { + "epoch": 0.39749, + "grad_norm": 0.8872988820075989, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 39749 + }, + { + "epoch": 0.3975, + "grad_norm": 0.9421464800834656, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 39750 + }, + { + "epoch": 0.39751, + "grad_norm": 0.9969943761825562, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 39751 + }, + { + "epoch": 0.39752, + "grad_norm": 1.013429880142212, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 39752 + }, + { + "epoch": 0.39753, + "grad_norm": 0.982036828994751, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 39753 + }, + { + "epoch": 0.39754, + "grad_norm": 0.8421829342842102, + "learning_rate": 0.003, + "loss": 3.976, + "step": 39754 + }, + { + "epoch": 0.39755, + "grad_norm": 0.7980527877807617, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 39755 + }, + { + "epoch": 0.39756, + "grad_norm": 0.8371139764785767, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 39756 + }, + { + "epoch": 0.39757, + "grad_norm": 0.9615294933319092, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 39757 + }, + { + "epoch": 0.39758, + "grad_norm": 1.3084733486175537, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 39758 + }, + { + "epoch": 0.39759, + "grad_norm": 0.8092347979545593, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 39759 + }, + { + "epoch": 0.3976, + "grad_norm": 0.6902133226394653, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 39760 + }, + { + "epoch": 0.39761, + "grad_norm": 0.8917631506919861, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 39761 + }, + { + "epoch": 0.39762, + "grad_norm": 0.9618569612503052, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 39762 + }, + { + "epoch": 0.39763, + "grad_norm": 1.0180050134658813, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 39763 + }, + { + "epoch": 0.39764, + "grad_norm": 1.0213662385940552, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 39764 + }, + { + "epoch": 0.39765, + "grad_norm": 1.1314170360565186, + "learning_rate": 0.003, + "loss": 3.987, + "step": 39765 + }, + { + "epoch": 0.39766, + "grad_norm": 0.955659806728363, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 39766 + }, + { + "epoch": 0.39767, + "grad_norm": 1.1795176267623901, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 39767 + }, + { + "epoch": 0.39768, + "grad_norm": 0.9179406762123108, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 39768 + }, + { + "epoch": 0.39769, + "grad_norm": 0.932865560054779, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 39769 + }, + { + "epoch": 0.3977, + "grad_norm": 1.0162296295166016, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 39770 + }, + { + "epoch": 0.39771, + "grad_norm": 1.027442216873169, + "learning_rate": 0.003, + "loss": 4.012, + "step": 39771 + }, + { + "epoch": 0.39772, + "grad_norm": 0.8971073031425476, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 39772 + }, + { + "epoch": 0.39773, + "grad_norm": 0.9950178265571594, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 39773 + }, + { + "epoch": 0.39774, + "grad_norm": 1.2918205261230469, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 39774 + }, + { + "epoch": 0.39775, + "grad_norm": 0.8385719656944275, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 39775 + }, + { + "epoch": 0.39776, + "grad_norm": 0.6898853182792664, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 39776 + }, + { + "epoch": 0.39777, + "grad_norm": 0.7234213948249817, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 39777 + }, + { + "epoch": 0.39778, + "grad_norm": 0.9145990610122681, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 39778 + }, + { + "epoch": 0.39779, + "grad_norm": 0.9339357018470764, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 39779 + }, + { + "epoch": 0.3978, + "grad_norm": 0.9327004551887512, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 39780 + }, + { + "epoch": 0.39781, + "grad_norm": 1.0537234544754028, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 39781 + }, + { + "epoch": 0.39782, + "grad_norm": 0.9587725400924683, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 39782 + }, + { + "epoch": 0.39783, + "grad_norm": 0.8352639675140381, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 39783 + }, + { + "epoch": 0.39784, + "grad_norm": 0.9013245701789856, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 39784 + }, + { + "epoch": 0.39785, + "grad_norm": 1.0438411235809326, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 39785 + }, + { + "epoch": 0.39786, + "grad_norm": 1.0420005321502686, + "learning_rate": 0.003, + "loss": 3.97, + "step": 39786 + }, + { + "epoch": 0.39787, + "grad_norm": 0.9220516085624695, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 39787 + }, + { + "epoch": 0.39788, + "grad_norm": 1.0491702556610107, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 39788 + }, + { + "epoch": 0.39789, + "grad_norm": 1.2251355648040771, + "learning_rate": 0.003, + "loss": 3.993, + "step": 39789 + }, + { + "epoch": 0.3979, + "grad_norm": 0.7914882302284241, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 39790 + }, + { + "epoch": 0.39791, + "grad_norm": 0.6983063817024231, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 39791 + }, + { + "epoch": 0.39792, + "grad_norm": 0.7894557118415833, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 39792 + }, + { + "epoch": 0.39793, + "grad_norm": 0.8279093503952026, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 39793 + }, + { + "epoch": 0.39794, + "grad_norm": 1.0181519985198975, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 39794 + }, + { + "epoch": 0.39795, + "grad_norm": 1.0474472045898438, + "learning_rate": 0.003, + "loss": 3.972, + "step": 39795 + }, + { + "epoch": 0.39796, + "grad_norm": 0.8597697019577026, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 39796 + }, + { + "epoch": 0.39797, + "grad_norm": 0.7206594347953796, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 39797 + }, + { + "epoch": 0.39798, + "grad_norm": 0.6965987682342529, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 39798 + }, + { + "epoch": 0.39799, + "grad_norm": 0.6443347930908203, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 39799 + }, + { + "epoch": 0.398, + "grad_norm": 0.728515088558197, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 39800 + }, + { + "epoch": 0.39801, + "grad_norm": 0.9850038290023804, + "learning_rate": 0.003, + "loss": 3.956, + "step": 39801 + }, + { + "epoch": 0.39802, + "grad_norm": 1.246987223625183, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 39802 + }, + { + "epoch": 0.39803, + "grad_norm": 0.837175726890564, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 39803 + }, + { + "epoch": 0.39804, + "grad_norm": 0.7643887996673584, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 39804 + }, + { + "epoch": 0.39805, + "grad_norm": 0.7096662521362305, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 39805 + }, + { + "epoch": 0.39806, + "grad_norm": 0.6198856234550476, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 39806 + }, + { + "epoch": 0.39807, + "grad_norm": 0.6440097689628601, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 39807 + }, + { + "epoch": 0.39808, + "grad_norm": 0.7409360408782959, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 39808 + }, + { + "epoch": 0.39809, + "grad_norm": 0.8640613555908203, + "learning_rate": 0.003, + "loss": 3.9415, + "step": 39809 + }, + { + "epoch": 0.3981, + "grad_norm": 0.8291612863540649, + "learning_rate": 0.003, + "loss": 3.959, + "step": 39810 + }, + { + "epoch": 0.39811, + "grad_norm": 0.9103133082389832, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 39811 + }, + { + "epoch": 0.39812, + "grad_norm": 0.9873570203781128, + "learning_rate": 0.003, + "loss": 3.9283, + "step": 39812 + }, + { + "epoch": 0.39813, + "grad_norm": 0.8890622854232788, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 39813 + }, + { + "epoch": 0.39814, + "grad_norm": 0.8059793710708618, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 39814 + }, + { + "epoch": 0.39815, + "grad_norm": 0.8065513372421265, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 39815 + }, + { + "epoch": 0.39816, + "grad_norm": 0.8681493997573853, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 39816 + }, + { + "epoch": 0.39817, + "grad_norm": 1.0211621522903442, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 39817 + }, + { + "epoch": 0.39818, + "grad_norm": 0.9241937398910522, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 39818 + }, + { + "epoch": 0.39819, + "grad_norm": 0.9533343315124512, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 39819 + }, + { + "epoch": 0.3982, + "grad_norm": 0.9894739985466003, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 39820 + }, + { + "epoch": 0.39821, + "grad_norm": 1.0425362586975098, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 39821 + }, + { + "epoch": 0.39822, + "grad_norm": 1.0951696634292603, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 39822 + }, + { + "epoch": 0.39823, + "grad_norm": 1.233198642730713, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 39823 + }, + { + "epoch": 0.39824, + "grad_norm": 0.8418209552764893, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 39824 + }, + { + "epoch": 0.39825, + "grad_norm": 0.7336748838424683, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 39825 + }, + { + "epoch": 0.39826, + "grad_norm": 0.8809176683425903, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 39826 + }, + { + "epoch": 0.39827, + "grad_norm": 0.9786887168884277, + "learning_rate": 0.003, + "loss": 3.989, + "step": 39827 + }, + { + "epoch": 0.39828, + "grad_norm": 1.0458312034606934, + "learning_rate": 0.003, + "loss": 3.9397, + "step": 39828 + }, + { + "epoch": 0.39829, + "grad_norm": 0.9626274704933167, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 39829 + }, + { + "epoch": 0.3983, + "grad_norm": 0.8264726400375366, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 39830 + }, + { + "epoch": 0.39831, + "grad_norm": 0.8367471098899841, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 39831 + }, + { + "epoch": 0.39832, + "grad_norm": 1.0884376764297485, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 39832 + }, + { + "epoch": 0.39833, + "grad_norm": 1.2696547508239746, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 39833 + }, + { + "epoch": 0.39834, + "grad_norm": 0.9122260212898254, + "learning_rate": 0.003, + "loss": 3.987, + "step": 39834 + }, + { + "epoch": 0.39835, + "grad_norm": 0.8748539686203003, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 39835 + }, + { + "epoch": 0.39836, + "grad_norm": 0.779360830783844, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 39836 + }, + { + "epoch": 0.39837, + "grad_norm": 0.733674168586731, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 39837 + }, + { + "epoch": 0.39838, + "grad_norm": 0.8426083922386169, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 39838 + }, + { + "epoch": 0.39839, + "grad_norm": 0.9235841035842896, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 39839 + }, + { + "epoch": 0.3984, + "grad_norm": 0.8091062903404236, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 39840 + }, + { + "epoch": 0.39841, + "grad_norm": 0.8031152486801147, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 39841 + }, + { + "epoch": 0.39842, + "grad_norm": 0.7231667637825012, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 39842 + }, + { + "epoch": 0.39843, + "grad_norm": 0.7294413447380066, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 39843 + }, + { + "epoch": 0.39844, + "grad_norm": 0.7884652018547058, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 39844 + }, + { + "epoch": 0.39845, + "grad_norm": 1.011299729347229, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 39845 + }, + { + "epoch": 0.39846, + "grad_norm": 1.0844249725341797, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 39846 + }, + { + "epoch": 0.39847, + "grad_norm": 0.8813523054122925, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 39847 + }, + { + "epoch": 0.39848, + "grad_norm": 0.936109185218811, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 39848 + }, + { + "epoch": 0.39849, + "grad_norm": 1.0504727363586426, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 39849 + }, + { + "epoch": 0.3985, + "grad_norm": 0.9289093017578125, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 39850 + }, + { + "epoch": 0.39851, + "grad_norm": 1.0239193439483643, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 39851 + }, + { + "epoch": 0.39852, + "grad_norm": 1.0875266790390015, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 39852 + }, + { + "epoch": 0.39853, + "grad_norm": 0.9194095730781555, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 39853 + }, + { + "epoch": 0.39854, + "grad_norm": 0.7665318250656128, + "learning_rate": 0.003, + "loss": 3.996, + "step": 39854 + }, + { + "epoch": 0.39855, + "grad_norm": 0.7598898410797119, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 39855 + }, + { + "epoch": 0.39856, + "grad_norm": 0.6762543320655823, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 39856 + }, + { + "epoch": 0.39857, + "grad_norm": 0.730301558971405, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 39857 + }, + { + "epoch": 0.39858, + "grad_norm": 0.8243348002433777, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 39858 + }, + { + "epoch": 0.39859, + "grad_norm": 0.9482315182685852, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 39859 + }, + { + "epoch": 0.3986, + "grad_norm": 1.0360287427902222, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 39860 + }, + { + "epoch": 0.39861, + "grad_norm": 1.0484423637390137, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 39861 + }, + { + "epoch": 0.39862, + "grad_norm": 0.9859178066253662, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 39862 + }, + { + "epoch": 0.39863, + "grad_norm": 0.9314345717430115, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 39863 + }, + { + "epoch": 0.39864, + "grad_norm": 1.0445045232772827, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 39864 + }, + { + "epoch": 0.39865, + "grad_norm": 1.1640139818191528, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 39865 + }, + { + "epoch": 0.39866, + "grad_norm": 0.9948902130126953, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 39866 + }, + { + "epoch": 0.39867, + "grad_norm": 1.2148830890655518, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 39867 + }, + { + "epoch": 0.39868, + "grad_norm": 0.8964214324951172, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 39868 + }, + { + "epoch": 0.39869, + "grad_norm": 0.786188542842865, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 39869 + }, + { + "epoch": 0.3987, + "grad_norm": 0.8052703142166138, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 39870 + }, + { + "epoch": 0.39871, + "grad_norm": 0.9457448124885559, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 39871 + }, + { + "epoch": 0.39872, + "grad_norm": 1.1437238454818726, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 39872 + }, + { + "epoch": 0.39873, + "grad_norm": 1.0373599529266357, + "learning_rate": 0.003, + "loss": 3.9359, + "step": 39873 + }, + { + "epoch": 0.39874, + "grad_norm": 0.9049200415611267, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 39874 + }, + { + "epoch": 0.39875, + "grad_norm": 0.8166525959968567, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 39875 + }, + { + "epoch": 0.39876, + "grad_norm": 0.8457248210906982, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 39876 + }, + { + "epoch": 0.39877, + "grad_norm": 0.8150965571403503, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 39877 + }, + { + "epoch": 0.39878, + "grad_norm": 0.7379161715507507, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 39878 + }, + { + "epoch": 0.39879, + "grad_norm": 0.6905511021614075, + "learning_rate": 0.003, + "loss": 3.996, + "step": 39879 + }, + { + "epoch": 0.3988, + "grad_norm": 0.6602028012275696, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 39880 + }, + { + "epoch": 0.39881, + "grad_norm": 0.6963975429534912, + "learning_rate": 0.003, + "loss": 3.969, + "step": 39881 + }, + { + "epoch": 0.39882, + "grad_norm": 0.7280617356300354, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 39882 + }, + { + "epoch": 0.39883, + "grad_norm": 0.7553749084472656, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 39883 + }, + { + "epoch": 0.39884, + "grad_norm": 0.9680041074752808, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 39884 + }, + { + "epoch": 0.39885, + "grad_norm": 1.3016982078552246, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 39885 + }, + { + "epoch": 0.39886, + "grad_norm": 0.8009873628616333, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 39886 + }, + { + "epoch": 0.39887, + "grad_norm": 0.8149824142456055, + "learning_rate": 0.003, + "loss": 3.97, + "step": 39887 + }, + { + "epoch": 0.39888, + "grad_norm": 0.834967851638794, + "learning_rate": 0.003, + "loss": 3.976, + "step": 39888 + }, + { + "epoch": 0.39889, + "grad_norm": 0.9221883416175842, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 39889 + }, + { + "epoch": 0.3989, + "grad_norm": 1.0189982652664185, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 39890 + }, + { + "epoch": 0.39891, + "grad_norm": 1.1375691890716553, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 39891 + }, + { + "epoch": 0.39892, + "grad_norm": 0.7891121506690979, + "learning_rate": 0.003, + "loss": 4.006, + "step": 39892 + }, + { + "epoch": 0.39893, + "grad_norm": 0.7641212344169617, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 39893 + }, + { + "epoch": 0.39894, + "grad_norm": 1.1933813095092773, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 39894 + }, + { + "epoch": 0.39895, + "grad_norm": 0.7248045206069946, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 39895 + }, + { + "epoch": 0.39896, + "grad_norm": 0.5344306230545044, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 39896 + }, + { + "epoch": 0.39897, + "grad_norm": 0.5913020372390747, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 39897 + }, + { + "epoch": 0.39898, + "grad_norm": 0.7536994218826294, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 39898 + }, + { + "epoch": 0.39899, + "grad_norm": 1.0077022314071655, + "learning_rate": 0.003, + "loss": 3.995, + "step": 39899 + }, + { + "epoch": 0.399, + "grad_norm": 1.081622838973999, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 39900 + }, + { + "epoch": 0.39901, + "grad_norm": 0.8391085863113403, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 39901 + }, + { + "epoch": 0.39902, + "grad_norm": 0.8232147693634033, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 39902 + }, + { + "epoch": 0.39903, + "grad_norm": 0.8506627678871155, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 39903 + }, + { + "epoch": 0.39904, + "grad_norm": 0.8686240911483765, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 39904 + }, + { + "epoch": 0.39905, + "grad_norm": 0.9731010794639587, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 39905 + }, + { + "epoch": 0.39906, + "grad_norm": 1.3022472858428955, + "learning_rate": 0.003, + "loss": 4.018, + "step": 39906 + }, + { + "epoch": 0.39907, + "grad_norm": 0.8332622647285461, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 39907 + }, + { + "epoch": 0.39908, + "grad_norm": 0.8626826405525208, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 39908 + }, + { + "epoch": 0.39909, + "grad_norm": 0.9217769503593445, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 39909 + }, + { + "epoch": 0.3991, + "grad_norm": 1.0832535028457642, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 39910 + }, + { + "epoch": 0.39911, + "grad_norm": 1.011770486831665, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 39911 + }, + { + "epoch": 0.39912, + "grad_norm": 0.9801819920539856, + "learning_rate": 0.003, + "loss": 3.948, + "step": 39912 + }, + { + "epoch": 0.39913, + "grad_norm": 0.9713297486305237, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 39913 + }, + { + "epoch": 0.39914, + "grad_norm": 1.2103403806686401, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 39914 + }, + { + "epoch": 0.39915, + "grad_norm": 0.8930085897445679, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 39915 + }, + { + "epoch": 0.39916, + "grad_norm": 0.7189071178436279, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 39916 + }, + { + "epoch": 0.39917, + "grad_norm": 0.7896829843521118, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 39917 + }, + { + "epoch": 0.39918, + "grad_norm": 0.8641054630279541, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 39918 + }, + { + "epoch": 0.39919, + "grad_norm": 0.956608772277832, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 39919 + }, + { + "epoch": 0.3992, + "grad_norm": 1.1110707521438599, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 39920 + }, + { + "epoch": 0.39921, + "grad_norm": 0.9947924017906189, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 39921 + }, + { + "epoch": 0.39922, + "grad_norm": 1.1938306093215942, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 39922 + }, + { + "epoch": 0.39923, + "grad_norm": 0.7399154901504517, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 39923 + }, + { + "epoch": 0.39924, + "grad_norm": 0.6738746762275696, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 39924 + }, + { + "epoch": 0.39925, + "grad_norm": 0.780916690826416, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 39925 + }, + { + "epoch": 0.39926, + "grad_norm": 1.0954943895339966, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 39926 + }, + { + "epoch": 0.39927, + "grad_norm": 1.4050183296203613, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 39927 + }, + { + "epoch": 0.39928, + "grad_norm": 0.6049776673316956, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 39928 + }, + { + "epoch": 0.39929, + "grad_norm": 0.780720591545105, + "learning_rate": 0.003, + "loss": 3.9268, + "step": 39929 + }, + { + "epoch": 0.3993, + "grad_norm": 0.7294303774833679, + "learning_rate": 0.003, + "loss": 3.973, + "step": 39930 + }, + { + "epoch": 0.39931, + "grad_norm": 0.5469682812690735, + "learning_rate": 0.003, + "loss": 3.973, + "step": 39931 + }, + { + "epoch": 0.39932, + "grad_norm": 0.6826446652412415, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 39932 + }, + { + "epoch": 0.39933, + "grad_norm": 0.9141577482223511, + "learning_rate": 0.003, + "loss": 3.973, + "step": 39933 + }, + { + "epoch": 0.39934, + "grad_norm": 1.0786261558532715, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 39934 + }, + { + "epoch": 0.39935, + "grad_norm": 1.0396144390106201, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 39935 + }, + { + "epoch": 0.39936, + "grad_norm": 1.008193850517273, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 39936 + }, + { + "epoch": 0.39937, + "grad_norm": 0.9368191361427307, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 39937 + }, + { + "epoch": 0.39938, + "grad_norm": 0.8206140398979187, + "learning_rate": 0.003, + "loss": 3.978, + "step": 39938 + }, + { + "epoch": 0.39939, + "grad_norm": 0.8073264360427856, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 39939 + }, + { + "epoch": 0.3994, + "grad_norm": 1.0340800285339355, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 39940 + }, + { + "epoch": 0.39941, + "grad_norm": 0.9792513847351074, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 39941 + }, + { + "epoch": 0.39942, + "grad_norm": 0.7890295386314392, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 39942 + }, + { + "epoch": 0.39943, + "grad_norm": 0.742499589920044, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 39943 + }, + { + "epoch": 0.39944, + "grad_norm": 0.802399754524231, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 39944 + }, + { + "epoch": 0.39945, + "grad_norm": 0.8720385432243347, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 39945 + }, + { + "epoch": 0.39946, + "grad_norm": 0.9842882752418518, + "learning_rate": 0.003, + "loss": 3.985, + "step": 39946 + }, + { + "epoch": 0.39947, + "grad_norm": 1.307361125946045, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 39947 + }, + { + "epoch": 0.39948, + "grad_norm": 0.8089094758033752, + "learning_rate": 0.003, + "loss": 4.003, + "step": 39948 + }, + { + "epoch": 0.39949, + "grad_norm": 0.8702298998832703, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 39949 + }, + { + "epoch": 0.3995, + "grad_norm": 0.9962202310562134, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 39950 + }, + { + "epoch": 0.39951, + "grad_norm": 1.047991156578064, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 39951 + }, + { + "epoch": 0.39952, + "grad_norm": 0.9929895997047424, + "learning_rate": 0.003, + "loss": 3.9359, + "step": 39952 + }, + { + "epoch": 0.39953, + "grad_norm": 1.4129058122634888, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 39953 + }, + { + "epoch": 0.39954, + "grad_norm": 0.9776272773742676, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 39954 + }, + { + "epoch": 0.39955, + "grad_norm": 0.856248140335083, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 39955 + }, + { + "epoch": 0.39956, + "grad_norm": 0.7928394675254822, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 39956 + }, + { + "epoch": 0.39957, + "grad_norm": 0.7104604840278625, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 39957 + }, + { + "epoch": 0.39958, + "grad_norm": 0.6750180125236511, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 39958 + }, + { + "epoch": 0.39959, + "grad_norm": 0.7426301836967468, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 39959 + }, + { + "epoch": 0.3996, + "grad_norm": 0.7395474314689636, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 39960 + }, + { + "epoch": 0.39961, + "grad_norm": 0.8029334545135498, + "learning_rate": 0.003, + "loss": 3.9457, + "step": 39961 + }, + { + "epoch": 0.39962, + "grad_norm": 1.0095174312591553, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 39962 + }, + { + "epoch": 0.39963, + "grad_norm": 1.2229318618774414, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 39963 + }, + { + "epoch": 0.39964, + "grad_norm": 0.9211933612823486, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 39964 + }, + { + "epoch": 0.39965, + "grad_norm": 0.8700349926948547, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 39965 + }, + { + "epoch": 0.39966, + "grad_norm": 1.0036672353744507, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 39966 + }, + { + "epoch": 0.39967, + "grad_norm": 1.1236369609832764, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 39967 + }, + { + "epoch": 0.39968, + "grad_norm": 0.9381220936775208, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 39968 + }, + { + "epoch": 0.39969, + "grad_norm": 0.9562534689903259, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 39969 + }, + { + "epoch": 0.3997, + "grad_norm": 1.051042914390564, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 39970 + }, + { + "epoch": 0.39971, + "grad_norm": 0.9154697060585022, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 39971 + }, + { + "epoch": 0.39972, + "grad_norm": 0.7985889315605164, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 39972 + }, + { + "epoch": 0.39973, + "grad_norm": 0.7026824951171875, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 39973 + }, + { + "epoch": 0.39974, + "grad_norm": 0.6679714918136597, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 39974 + }, + { + "epoch": 0.39975, + "grad_norm": 0.6883552670478821, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 39975 + }, + { + "epoch": 0.39976, + "grad_norm": 0.7167072296142578, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 39976 + }, + { + "epoch": 0.39977, + "grad_norm": 0.6382402777671814, + "learning_rate": 0.003, + "loss": 3.97, + "step": 39977 + }, + { + "epoch": 0.39978, + "grad_norm": 0.7536785006523132, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 39978 + }, + { + "epoch": 0.39979, + "grad_norm": 0.9399287104606628, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 39979 + }, + { + "epoch": 0.3998, + "grad_norm": 1.086507797241211, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 39980 + }, + { + "epoch": 0.39981, + "grad_norm": 1.1604596376419067, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 39981 + }, + { + "epoch": 0.39982, + "grad_norm": 0.8544206619262695, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 39982 + }, + { + "epoch": 0.39983, + "grad_norm": 0.7604381442070007, + "learning_rate": 0.003, + "loss": 3.977, + "step": 39983 + }, + { + "epoch": 0.39984, + "grad_norm": 0.6734378933906555, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 39984 + }, + { + "epoch": 0.39985, + "grad_norm": 0.7307486534118652, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 39985 + }, + { + "epoch": 0.39986, + "grad_norm": 0.8809465765953064, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 39986 + }, + { + "epoch": 0.39987, + "grad_norm": 1.0474919080734253, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 39987 + }, + { + "epoch": 0.39988, + "grad_norm": 1.0166571140289307, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 39988 + }, + { + "epoch": 0.39989, + "grad_norm": 0.9899111390113831, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 39989 + }, + { + "epoch": 0.3999, + "grad_norm": 0.8816802501678467, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 39990 + }, + { + "epoch": 0.39991, + "grad_norm": 0.9296420216560364, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 39991 + }, + { + "epoch": 0.39992, + "grad_norm": 1.1288819313049316, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 39992 + }, + { + "epoch": 0.39993, + "grad_norm": 1.0015040636062622, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 39993 + }, + { + "epoch": 0.39994, + "grad_norm": 0.9085958003997803, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 39994 + }, + { + "epoch": 0.39995, + "grad_norm": 0.9194366335868835, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 39995 + }, + { + "epoch": 0.39996, + "grad_norm": 0.9264000058174133, + "learning_rate": 0.003, + "loss": 3.99, + "step": 39996 + }, + { + "epoch": 0.39997, + "grad_norm": 1.0023447275161743, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 39997 + }, + { + "epoch": 0.39998, + "grad_norm": 1.4293749332427979, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 39998 + }, + { + "epoch": 0.39999, + "grad_norm": 0.6779047846794128, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 39999 + }, + { + "epoch": 0.4, + "grad_norm": 0.7162042856216431, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 40000 + }, + { + "epoch": 0.40001, + "grad_norm": 0.8121765851974487, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 40001 + }, + { + "epoch": 0.40002, + "grad_norm": 0.9027695655822754, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 40002 + }, + { + "epoch": 0.40003, + "grad_norm": 0.9661027193069458, + "learning_rate": 0.003, + "loss": 4.019, + "step": 40003 + }, + { + "epoch": 0.40004, + "grad_norm": 1.2159117460250854, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 40004 + }, + { + "epoch": 0.40005, + "grad_norm": 1.0859593152999878, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 40005 + }, + { + "epoch": 0.40006, + "grad_norm": 0.9011837840080261, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 40006 + }, + { + "epoch": 0.40007, + "grad_norm": 0.7898913621902466, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 40007 + }, + { + "epoch": 0.40008, + "grad_norm": 0.8522996306419373, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 40008 + }, + { + "epoch": 0.40009, + "grad_norm": 1.0376102924346924, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 40009 + }, + { + "epoch": 0.4001, + "grad_norm": 1.098236322402954, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 40010 + }, + { + "epoch": 0.40011, + "grad_norm": 0.8408608436584473, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 40011 + }, + { + "epoch": 0.40012, + "grad_norm": 0.7358251810073853, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 40012 + }, + { + "epoch": 0.40013, + "grad_norm": 0.8375370502471924, + "learning_rate": 0.003, + "loss": 3.9387, + "step": 40013 + }, + { + "epoch": 0.40014, + "grad_norm": 0.9703590273857117, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 40014 + }, + { + "epoch": 0.40015, + "grad_norm": 0.9778888821601868, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 40015 + }, + { + "epoch": 0.40016, + "grad_norm": 0.8980645537376404, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 40016 + }, + { + "epoch": 0.40017, + "grad_norm": 0.9393784403800964, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 40017 + }, + { + "epoch": 0.40018, + "grad_norm": 0.9120060801506042, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 40018 + }, + { + "epoch": 0.40019, + "grad_norm": 0.9196029305458069, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 40019 + }, + { + "epoch": 0.4002, + "grad_norm": 0.9602447152137756, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 40020 + }, + { + "epoch": 0.40021, + "grad_norm": 1.1539093255996704, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 40021 + }, + { + "epoch": 0.40022, + "grad_norm": 0.9202343821525574, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 40022 + }, + { + "epoch": 0.40023, + "grad_norm": 0.747663140296936, + "learning_rate": 0.003, + "loss": 4.011, + "step": 40023 + }, + { + "epoch": 0.40024, + "grad_norm": 0.6472582221031189, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 40024 + }, + { + "epoch": 0.40025, + "grad_norm": 0.6293637156486511, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 40025 + }, + { + "epoch": 0.40026, + "grad_norm": 0.7925937175750732, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 40026 + }, + { + "epoch": 0.40027, + "grad_norm": 1.039599061012268, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 40027 + }, + { + "epoch": 0.40028, + "grad_norm": 1.0562916994094849, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 40028 + }, + { + "epoch": 0.40029, + "grad_norm": 0.8909109234809875, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 40029 + }, + { + "epoch": 0.4003, + "grad_norm": 0.8538114428520203, + "learning_rate": 0.003, + "loss": 3.964, + "step": 40030 + }, + { + "epoch": 0.40031, + "grad_norm": 0.7831065654754639, + "learning_rate": 0.003, + "loss": 3.9321, + "step": 40031 + }, + { + "epoch": 0.40032, + "grad_norm": 0.7991322875022888, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 40032 + }, + { + "epoch": 0.40033, + "grad_norm": 0.8323495388031006, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 40033 + }, + { + "epoch": 0.40034, + "grad_norm": 0.9604739546775818, + "learning_rate": 0.003, + "loss": 3.992, + "step": 40034 + }, + { + "epoch": 0.40035, + "grad_norm": 1.2324492931365967, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 40035 + }, + { + "epoch": 0.40036, + "grad_norm": 0.7693159580230713, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 40036 + }, + { + "epoch": 0.40037, + "grad_norm": 0.696845293045044, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 40037 + }, + { + "epoch": 0.40038, + "grad_norm": 0.6637425422668457, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 40038 + }, + { + "epoch": 0.40039, + "grad_norm": 0.6294717192649841, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 40039 + }, + { + "epoch": 0.4004, + "grad_norm": 0.6475238800048828, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 40040 + }, + { + "epoch": 0.40041, + "grad_norm": 0.5723147392272949, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 40041 + }, + { + "epoch": 0.40042, + "grad_norm": 0.6375605463981628, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 40042 + }, + { + "epoch": 0.40043, + "grad_norm": 0.9453613758087158, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 40043 + }, + { + "epoch": 0.40044, + "grad_norm": 1.2483776807785034, + "learning_rate": 0.003, + "loss": 3.968, + "step": 40044 + }, + { + "epoch": 0.40045, + "grad_norm": 0.7507554292678833, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 40045 + }, + { + "epoch": 0.40046, + "grad_norm": 0.8234479427337646, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 40046 + }, + { + "epoch": 0.40047, + "grad_norm": 0.9317229986190796, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 40047 + }, + { + "epoch": 0.40048, + "grad_norm": 1.193703293800354, + "learning_rate": 0.003, + "loss": 3.962, + "step": 40048 + }, + { + "epoch": 0.40049, + "grad_norm": 1.0383083820343018, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 40049 + }, + { + "epoch": 0.4005, + "grad_norm": 0.906015157699585, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 40050 + }, + { + "epoch": 0.40051, + "grad_norm": 0.74274742603302, + "learning_rate": 0.003, + "loss": 3.991, + "step": 40051 + }, + { + "epoch": 0.40052, + "grad_norm": 0.7445491552352905, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 40052 + }, + { + "epoch": 0.40053, + "grad_norm": 0.8090980648994446, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 40053 + }, + { + "epoch": 0.40054, + "grad_norm": 0.7613006830215454, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 40054 + }, + { + "epoch": 0.40055, + "grad_norm": 0.859754741191864, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 40055 + }, + { + "epoch": 0.40056, + "grad_norm": 0.8811156153678894, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 40056 + }, + { + "epoch": 0.40057, + "grad_norm": 0.8611463904380798, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 40057 + }, + { + "epoch": 0.40058, + "grad_norm": 0.8933367133140564, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 40058 + }, + { + "epoch": 0.40059, + "grad_norm": 0.8614692091941833, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 40059 + }, + { + "epoch": 0.4006, + "grad_norm": 0.8439565896987915, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 40060 + }, + { + "epoch": 0.40061, + "grad_norm": 0.8995826244354248, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 40061 + }, + { + "epoch": 0.40062, + "grad_norm": 0.8963746428489685, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 40062 + }, + { + "epoch": 0.40063, + "grad_norm": 0.8795823454856873, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 40063 + }, + { + "epoch": 0.40064, + "grad_norm": 1.1042197942733765, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 40064 + }, + { + "epoch": 0.40065, + "grad_norm": 1.1579493284225464, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 40065 + }, + { + "epoch": 0.40066, + "grad_norm": 1.032400369644165, + "learning_rate": 0.003, + "loss": 3.957, + "step": 40066 + }, + { + "epoch": 0.40067, + "grad_norm": 1.4191397428512573, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 40067 + }, + { + "epoch": 0.40068, + "grad_norm": 0.917443037033081, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 40068 + }, + { + "epoch": 0.40069, + "grad_norm": 0.8339070677757263, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 40069 + }, + { + "epoch": 0.4007, + "grad_norm": 0.9139005541801453, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 40070 + }, + { + "epoch": 0.40071, + "grad_norm": 1.1194454431533813, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 40071 + }, + { + "epoch": 0.40072, + "grad_norm": 1.149724006652832, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 40072 + }, + { + "epoch": 0.40073, + "grad_norm": 0.7983490228652954, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 40073 + }, + { + "epoch": 0.40074, + "grad_norm": 0.7053617835044861, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 40074 + }, + { + "epoch": 0.40075, + "grad_norm": 0.7970131039619446, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 40075 + }, + { + "epoch": 0.40076, + "grad_norm": 0.8820486068725586, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 40076 + }, + { + "epoch": 0.40077, + "grad_norm": 1.0109707117080688, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 40077 + }, + { + "epoch": 0.40078, + "grad_norm": 1.101170301437378, + "learning_rate": 0.003, + "loss": 3.961, + "step": 40078 + }, + { + "epoch": 0.40079, + "grad_norm": 0.9935923218727112, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 40079 + }, + { + "epoch": 0.4008, + "grad_norm": 1.066463828086853, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 40080 + }, + { + "epoch": 0.40081, + "grad_norm": 1.0438342094421387, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 40081 + }, + { + "epoch": 0.40082, + "grad_norm": 0.9832364916801453, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 40082 + }, + { + "epoch": 0.40083, + "grad_norm": 0.9974456429481506, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 40083 + }, + { + "epoch": 0.40084, + "grad_norm": 0.9359449744224548, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 40084 + }, + { + "epoch": 0.40085, + "grad_norm": 1.0582484006881714, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 40085 + }, + { + "epoch": 0.40086, + "grad_norm": 0.9656417965888977, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 40086 + }, + { + "epoch": 0.40087, + "grad_norm": 0.9169012904167175, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 40087 + }, + { + "epoch": 0.40088, + "grad_norm": 0.950782835483551, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 40088 + }, + { + "epoch": 0.40089, + "grad_norm": 1.0430448055267334, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 40089 + }, + { + "epoch": 0.4009, + "grad_norm": 0.8388198018074036, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 40090 + }, + { + "epoch": 0.40091, + "grad_norm": 0.8297619223594666, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 40091 + }, + { + "epoch": 0.40092, + "grad_norm": 0.9238507747650146, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 40092 + }, + { + "epoch": 0.40093, + "grad_norm": 0.9436755776405334, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 40093 + }, + { + "epoch": 0.40094, + "grad_norm": 1.1152275800704956, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 40094 + }, + { + "epoch": 0.40095, + "grad_norm": 1.0533523559570312, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 40095 + }, + { + "epoch": 0.40096, + "grad_norm": 0.9641923308372498, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 40096 + }, + { + "epoch": 0.40097, + "grad_norm": 0.888610303401947, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 40097 + }, + { + "epoch": 0.40098, + "grad_norm": 0.7873759269714355, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 40098 + }, + { + "epoch": 0.40099, + "grad_norm": 0.7906292676925659, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 40099 + }, + { + "epoch": 0.401, + "grad_norm": 0.8628591895103455, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 40100 + }, + { + "epoch": 0.40101, + "grad_norm": 0.7970002889633179, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 40101 + }, + { + "epoch": 0.40102, + "grad_norm": 0.9335940480232239, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 40102 + }, + { + "epoch": 0.40103, + "grad_norm": 0.944437563419342, + "learning_rate": 0.003, + "loss": 3.984, + "step": 40103 + }, + { + "epoch": 0.40104, + "grad_norm": 0.9406269788742065, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 40104 + }, + { + "epoch": 0.40105, + "grad_norm": 1.0006083250045776, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 40105 + }, + { + "epoch": 0.40106, + "grad_norm": 0.9608407020568848, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 40106 + }, + { + "epoch": 0.40107, + "grad_norm": 1.0856961011886597, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 40107 + }, + { + "epoch": 0.40108, + "grad_norm": 1.2321244478225708, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 40108 + }, + { + "epoch": 0.40109, + "grad_norm": 0.7572892308235168, + "learning_rate": 0.003, + "loss": 3.992, + "step": 40109 + }, + { + "epoch": 0.4011, + "grad_norm": 0.5868052840232849, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 40110 + }, + { + "epoch": 0.40111, + "grad_norm": 0.5714976787567139, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 40111 + }, + { + "epoch": 0.40112, + "grad_norm": 0.49084964394569397, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 40112 + }, + { + "epoch": 0.40113, + "grad_norm": 0.5001804232597351, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 40113 + }, + { + "epoch": 0.40114, + "grad_norm": 0.5767372250556946, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 40114 + }, + { + "epoch": 0.40115, + "grad_norm": 0.6649559736251831, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 40115 + }, + { + "epoch": 0.40116, + "grad_norm": 0.9025324583053589, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 40116 + }, + { + "epoch": 0.40117, + "grad_norm": 1.3555569648742676, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 40117 + }, + { + "epoch": 0.40118, + "grad_norm": 0.6694959402084351, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 40118 + }, + { + "epoch": 0.40119, + "grad_norm": 0.6417465209960938, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 40119 + }, + { + "epoch": 0.4012, + "grad_norm": 0.7747312188148499, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 40120 + }, + { + "epoch": 0.40121, + "grad_norm": 0.8067141175270081, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 40121 + }, + { + "epoch": 0.40122, + "grad_norm": 0.7540450692176819, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 40122 + }, + { + "epoch": 0.40123, + "grad_norm": 0.6759268045425415, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 40123 + }, + { + "epoch": 0.40124, + "grad_norm": 0.7137091159820557, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 40124 + }, + { + "epoch": 0.40125, + "grad_norm": 0.9618066549301147, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 40125 + }, + { + "epoch": 0.40126, + "grad_norm": 1.2438068389892578, + "learning_rate": 0.003, + "loss": 3.993, + "step": 40126 + }, + { + "epoch": 0.40127, + "grad_norm": 0.845868706703186, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 40127 + }, + { + "epoch": 0.40128, + "grad_norm": 0.8165836334228516, + "learning_rate": 0.003, + "loss": 3.978, + "step": 40128 + }, + { + "epoch": 0.40129, + "grad_norm": 0.9554463028907776, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 40129 + }, + { + "epoch": 0.4013, + "grad_norm": 0.9888679385185242, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 40130 + }, + { + "epoch": 0.40131, + "grad_norm": 0.8868873119354248, + "learning_rate": 0.003, + "loss": 3.993, + "step": 40131 + }, + { + "epoch": 0.40132, + "grad_norm": 0.7912784218788147, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 40132 + }, + { + "epoch": 0.40133, + "grad_norm": 0.8493282794952393, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 40133 + }, + { + "epoch": 0.40134, + "grad_norm": 0.8013893961906433, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 40134 + }, + { + "epoch": 0.40135, + "grad_norm": 0.8237121105194092, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 40135 + }, + { + "epoch": 0.40136, + "grad_norm": 0.8959832787513733, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 40136 + }, + { + "epoch": 0.40137, + "grad_norm": 1.082133412361145, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 40137 + }, + { + "epoch": 0.40138, + "grad_norm": 1.2412053346633911, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 40138 + }, + { + "epoch": 0.40139, + "grad_norm": 1.1949775218963623, + "learning_rate": 0.003, + "loss": 3.961, + "step": 40139 + }, + { + "epoch": 0.4014, + "grad_norm": 1.0192748308181763, + "learning_rate": 0.003, + "loss": 4.026, + "step": 40140 + }, + { + "epoch": 0.40141, + "grad_norm": 1.0578135251998901, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 40141 + }, + { + "epoch": 0.40142, + "grad_norm": 0.8499462008476257, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 40142 + }, + { + "epoch": 0.40143, + "grad_norm": 0.8465281128883362, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 40143 + }, + { + "epoch": 0.40144, + "grad_norm": 0.8621542453765869, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 40144 + }, + { + "epoch": 0.40145, + "grad_norm": 1.0328567028045654, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 40145 + }, + { + "epoch": 0.40146, + "grad_norm": 1.2066646814346313, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 40146 + }, + { + "epoch": 0.40147, + "grad_norm": 0.8926025629043579, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 40147 + }, + { + "epoch": 0.40148, + "grad_norm": 0.8776819109916687, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 40148 + }, + { + "epoch": 0.40149, + "grad_norm": 0.8178439140319824, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 40149 + }, + { + "epoch": 0.4015, + "grad_norm": 0.8466310501098633, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 40150 + }, + { + "epoch": 0.40151, + "grad_norm": 0.8211356401443481, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 40151 + }, + { + "epoch": 0.40152, + "grad_norm": 0.7526400089263916, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 40152 + }, + { + "epoch": 0.40153, + "grad_norm": 0.7697922587394714, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 40153 + }, + { + "epoch": 0.40154, + "grad_norm": 0.6889259815216064, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 40154 + }, + { + "epoch": 0.40155, + "grad_norm": 0.7806645035743713, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 40155 + }, + { + "epoch": 0.40156, + "grad_norm": 1.1391676664352417, + "learning_rate": 0.003, + "loss": 3.9362, + "step": 40156 + }, + { + "epoch": 0.40157, + "grad_norm": 1.1494812965393066, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 40157 + }, + { + "epoch": 0.40158, + "grad_norm": 0.858216404914856, + "learning_rate": 0.003, + "loss": 3.9417, + "step": 40158 + }, + { + "epoch": 0.40159, + "grad_norm": 0.7509071230888367, + "learning_rate": 0.003, + "loss": 3.998, + "step": 40159 + }, + { + "epoch": 0.4016, + "grad_norm": 0.5859501361846924, + "learning_rate": 0.003, + "loss": 3.944, + "step": 40160 + }, + { + "epoch": 0.40161, + "grad_norm": 0.8585663437843323, + "learning_rate": 0.003, + "loss": 3.9237, + "step": 40161 + }, + { + "epoch": 0.40162, + "grad_norm": 1.295516848564148, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 40162 + }, + { + "epoch": 0.40163, + "grad_norm": 0.7308545112609863, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 40163 + }, + { + "epoch": 0.40164, + "grad_norm": 0.7267751097679138, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 40164 + }, + { + "epoch": 0.40165, + "grad_norm": 0.6843788623809814, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 40165 + }, + { + "epoch": 0.40166, + "grad_norm": 0.7182525992393494, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 40166 + }, + { + "epoch": 0.40167, + "grad_norm": 0.8266682624816895, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 40167 + }, + { + "epoch": 0.40168, + "grad_norm": 0.8032686710357666, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 40168 + }, + { + "epoch": 0.40169, + "grad_norm": 0.9306706190109253, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 40169 + }, + { + "epoch": 0.4017, + "grad_norm": 1.1902459859848022, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 40170 + }, + { + "epoch": 0.40171, + "grad_norm": 1.1178994178771973, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 40171 + }, + { + "epoch": 0.40172, + "grad_norm": 0.8561018705368042, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 40172 + }, + { + "epoch": 0.40173, + "grad_norm": 0.8656502366065979, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 40173 + }, + { + "epoch": 0.40174, + "grad_norm": 1.0058879852294922, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 40174 + }, + { + "epoch": 0.40175, + "grad_norm": 0.978961706161499, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 40175 + }, + { + "epoch": 0.40176, + "grad_norm": 1.2268307209014893, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 40176 + }, + { + "epoch": 0.40177, + "grad_norm": 1.0589441061019897, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 40177 + }, + { + "epoch": 0.40178, + "grad_norm": 0.7658606171607971, + "learning_rate": 0.003, + "loss": 3.9392, + "step": 40178 + }, + { + "epoch": 0.40179, + "grad_norm": 0.9057439565658569, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 40179 + }, + { + "epoch": 0.4018, + "grad_norm": 0.8544208407402039, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 40180 + }, + { + "epoch": 0.40181, + "grad_norm": 0.920807957649231, + "learning_rate": 0.003, + "loss": 3.9399, + "step": 40181 + }, + { + "epoch": 0.40182, + "grad_norm": 1.062771201133728, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 40182 + }, + { + "epoch": 0.40183, + "grad_norm": 1.0940799713134766, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 40183 + }, + { + "epoch": 0.40184, + "grad_norm": 1.0276442766189575, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 40184 + }, + { + "epoch": 0.40185, + "grad_norm": 1.140401840209961, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 40185 + }, + { + "epoch": 0.40186, + "grad_norm": 0.9589017629623413, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 40186 + }, + { + "epoch": 0.40187, + "grad_norm": 1.3215276002883911, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 40187 + }, + { + "epoch": 0.40188, + "grad_norm": 0.9345223307609558, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 40188 + }, + { + "epoch": 0.40189, + "grad_norm": 0.9492669105529785, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 40189 + }, + { + "epoch": 0.4019, + "grad_norm": 0.9240803718566895, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 40190 + }, + { + "epoch": 0.40191, + "grad_norm": 0.8339318633079529, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 40191 + }, + { + "epoch": 0.40192, + "grad_norm": 0.7096352577209473, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 40192 + }, + { + "epoch": 0.40193, + "grad_norm": 0.7683164477348328, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 40193 + }, + { + "epoch": 0.40194, + "grad_norm": 0.9455286264419556, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 40194 + }, + { + "epoch": 0.40195, + "grad_norm": 1.0773111581802368, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 40195 + }, + { + "epoch": 0.40196, + "grad_norm": 1.0317739248275757, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 40196 + }, + { + "epoch": 0.40197, + "grad_norm": 1.1529135704040527, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 40197 + }, + { + "epoch": 0.40198, + "grad_norm": 0.8766962289810181, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 40198 + }, + { + "epoch": 0.40199, + "grad_norm": 0.874167263507843, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 40199 + }, + { + "epoch": 0.402, + "grad_norm": 0.9888526797294617, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 40200 + }, + { + "epoch": 0.40201, + "grad_norm": 1.0757722854614258, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 40201 + }, + { + "epoch": 0.40202, + "grad_norm": 0.9248135089874268, + "learning_rate": 0.003, + "loss": 3.9428, + "step": 40202 + }, + { + "epoch": 0.40203, + "grad_norm": 1.0163815021514893, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 40203 + }, + { + "epoch": 0.40204, + "grad_norm": 1.101116418838501, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 40204 + }, + { + "epoch": 0.40205, + "grad_norm": 0.9505676627159119, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 40205 + }, + { + "epoch": 0.40206, + "grad_norm": 0.8511347770690918, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 40206 + }, + { + "epoch": 0.40207, + "grad_norm": 0.7540643811225891, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 40207 + }, + { + "epoch": 0.40208, + "grad_norm": 0.8769600987434387, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 40208 + }, + { + "epoch": 0.40209, + "grad_norm": 1.013735055923462, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 40209 + }, + { + "epoch": 0.4021, + "grad_norm": 1.178158164024353, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 40210 + }, + { + "epoch": 0.40211, + "grad_norm": 0.7655798196792603, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 40211 + }, + { + "epoch": 0.40212, + "grad_norm": 0.7198166847229004, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 40212 + }, + { + "epoch": 0.40213, + "grad_norm": 0.7597734928131104, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 40213 + }, + { + "epoch": 0.40214, + "grad_norm": 0.7511619329452515, + "learning_rate": 0.003, + "loss": 3.9418, + "step": 40214 + }, + { + "epoch": 0.40215, + "grad_norm": 0.8857923150062561, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 40215 + }, + { + "epoch": 0.40216, + "grad_norm": 1.077701449394226, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 40216 + }, + { + "epoch": 0.40217, + "grad_norm": 1.1002817153930664, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 40217 + }, + { + "epoch": 0.40218, + "grad_norm": 0.9663345217704773, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 40218 + }, + { + "epoch": 0.40219, + "grad_norm": 0.9045743942260742, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 40219 + }, + { + "epoch": 0.4022, + "grad_norm": 0.9400999546051025, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 40220 + }, + { + "epoch": 0.40221, + "grad_norm": 0.9362815022468567, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 40221 + }, + { + "epoch": 0.40222, + "grad_norm": 0.8550698161125183, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 40222 + }, + { + "epoch": 0.40223, + "grad_norm": 0.8553100228309631, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 40223 + }, + { + "epoch": 0.40224, + "grad_norm": 0.8626173138618469, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 40224 + }, + { + "epoch": 0.40225, + "grad_norm": 0.9976007342338562, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 40225 + }, + { + "epoch": 0.40226, + "grad_norm": 1.0935686826705933, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 40226 + }, + { + "epoch": 0.40227, + "grad_norm": 1.0035165548324585, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 40227 + }, + { + "epoch": 0.40228, + "grad_norm": 0.949035108089447, + "learning_rate": 0.003, + "loss": 4.005, + "step": 40228 + }, + { + "epoch": 0.40229, + "grad_norm": 0.9098554849624634, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 40229 + }, + { + "epoch": 0.4023, + "grad_norm": 0.9387121796607971, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 40230 + }, + { + "epoch": 0.40231, + "grad_norm": 0.9783697724342346, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 40231 + }, + { + "epoch": 0.40232, + "grad_norm": 1.1798303127288818, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 40232 + }, + { + "epoch": 0.40233, + "grad_norm": 1.0299533605575562, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 40233 + }, + { + "epoch": 0.40234, + "grad_norm": 0.969823956489563, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 40234 + }, + { + "epoch": 0.40235, + "grad_norm": 0.9929221868515015, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 40235 + }, + { + "epoch": 0.40236, + "grad_norm": 1.117824673652649, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 40236 + }, + { + "epoch": 0.40237, + "grad_norm": 0.8420213460922241, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 40237 + }, + { + "epoch": 0.40238, + "grad_norm": 0.8155110478401184, + "learning_rate": 0.003, + "loss": 3.975, + "step": 40238 + }, + { + "epoch": 0.40239, + "grad_norm": 0.7693844437599182, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 40239 + }, + { + "epoch": 0.4024, + "grad_norm": 0.6445848345756531, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 40240 + }, + { + "epoch": 0.40241, + "grad_norm": 0.6937522292137146, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 40241 + }, + { + "epoch": 0.40242, + "grad_norm": 0.7287824153900146, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 40242 + }, + { + "epoch": 0.40243, + "grad_norm": 0.7347718477249146, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 40243 + }, + { + "epoch": 0.40244, + "grad_norm": 0.7537151575088501, + "learning_rate": 0.003, + "loss": 3.999, + "step": 40244 + }, + { + "epoch": 0.40245, + "grad_norm": 0.8656308054924011, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 40245 + }, + { + "epoch": 0.40246, + "grad_norm": 1.0196274518966675, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 40246 + }, + { + "epoch": 0.40247, + "grad_norm": 0.9993747472763062, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 40247 + }, + { + "epoch": 0.40248, + "grad_norm": 1.057694911956787, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 40248 + }, + { + "epoch": 0.40249, + "grad_norm": 1.1469781398773193, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 40249 + }, + { + "epoch": 0.4025, + "grad_norm": 0.835368812084198, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 40250 + }, + { + "epoch": 0.40251, + "grad_norm": 0.8277757167816162, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 40251 + }, + { + "epoch": 0.40252, + "grad_norm": 0.9732588529586792, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 40252 + }, + { + "epoch": 0.40253, + "grad_norm": 0.8702344298362732, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 40253 + }, + { + "epoch": 0.40254, + "grad_norm": 0.8868075609207153, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 40254 + }, + { + "epoch": 0.40255, + "grad_norm": 0.7259416580200195, + "learning_rate": 0.003, + "loss": 3.965, + "step": 40255 + }, + { + "epoch": 0.40256, + "grad_norm": 0.7216874361038208, + "learning_rate": 0.003, + "loss": 3.963, + "step": 40256 + }, + { + "epoch": 0.40257, + "grad_norm": 0.9324195384979248, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 40257 + }, + { + "epoch": 0.40258, + "grad_norm": 1.4867002964019775, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 40258 + }, + { + "epoch": 0.40259, + "grad_norm": 0.7237091064453125, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 40259 + }, + { + "epoch": 0.4026, + "grad_norm": 0.6902821063995361, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 40260 + }, + { + "epoch": 0.40261, + "grad_norm": 0.5838561058044434, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 40261 + }, + { + "epoch": 0.40262, + "grad_norm": 0.695468544960022, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 40262 + }, + { + "epoch": 0.40263, + "grad_norm": 0.7991549372673035, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 40263 + }, + { + "epoch": 0.40264, + "grad_norm": 0.8259188532829285, + "learning_rate": 0.003, + "loss": 3.9396, + "step": 40264 + }, + { + "epoch": 0.40265, + "grad_norm": 0.8989471197128296, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 40265 + }, + { + "epoch": 0.40266, + "grad_norm": 0.8915395140647888, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 40266 + }, + { + "epoch": 0.40267, + "grad_norm": 1.1771539449691772, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 40267 + }, + { + "epoch": 0.40268, + "grad_norm": 1.129913568496704, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 40268 + }, + { + "epoch": 0.40269, + "grad_norm": 0.8791612982749939, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 40269 + }, + { + "epoch": 0.4027, + "grad_norm": 0.8092635273933411, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 40270 + }, + { + "epoch": 0.40271, + "grad_norm": 0.8707496523857117, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 40271 + }, + { + "epoch": 0.40272, + "grad_norm": 0.8641815185546875, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 40272 + }, + { + "epoch": 0.40273, + "grad_norm": 0.7100216746330261, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 40273 + }, + { + "epoch": 0.40274, + "grad_norm": 0.6460208296775818, + "learning_rate": 0.003, + "loss": 3.954, + "step": 40274 + }, + { + "epoch": 0.40275, + "grad_norm": 0.8270339369773865, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 40275 + }, + { + "epoch": 0.40276, + "grad_norm": 0.9651469588279724, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 40276 + }, + { + "epoch": 0.40277, + "grad_norm": 1.0041635036468506, + "learning_rate": 0.003, + "loss": 3.9372, + "step": 40277 + }, + { + "epoch": 0.40278, + "grad_norm": 1.1863751411437988, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 40278 + }, + { + "epoch": 0.40279, + "grad_norm": 0.7497547268867493, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 40279 + }, + { + "epoch": 0.4028, + "grad_norm": 0.5677827000617981, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 40280 + }, + { + "epoch": 0.40281, + "grad_norm": 0.5943251848220825, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 40281 + }, + { + "epoch": 0.40282, + "grad_norm": 0.6796475052833557, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 40282 + }, + { + "epoch": 0.40283, + "grad_norm": 0.795792281627655, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 40283 + }, + { + "epoch": 0.40284, + "grad_norm": 0.9640011787414551, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 40284 + }, + { + "epoch": 0.40285, + "grad_norm": 1.137912392616272, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 40285 + }, + { + "epoch": 0.40286, + "grad_norm": 0.9373136758804321, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 40286 + }, + { + "epoch": 0.40287, + "grad_norm": 0.9392118453979492, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 40287 + }, + { + "epoch": 0.40288, + "grad_norm": 0.8722360134124756, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 40288 + }, + { + "epoch": 0.40289, + "grad_norm": 0.8801825046539307, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 40289 + }, + { + "epoch": 0.4029, + "grad_norm": 1.1238842010498047, + "learning_rate": 0.003, + "loss": 3.962, + "step": 40290 + }, + { + "epoch": 0.40291, + "grad_norm": 1.057011604309082, + "learning_rate": 0.003, + "loss": 4.002, + "step": 40291 + }, + { + "epoch": 0.40292, + "grad_norm": 0.9798825979232788, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 40292 + }, + { + "epoch": 0.40293, + "grad_norm": 1.014480710029602, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 40293 + }, + { + "epoch": 0.40294, + "grad_norm": 1.0850815773010254, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 40294 + }, + { + "epoch": 0.40295, + "grad_norm": 0.7673048377037048, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 40295 + }, + { + "epoch": 0.40296, + "grad_norm": 0.6586326956748962, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 40296 + }, + { + "epoch": 0.40297, + "grad_norm": 0.7665443420410156, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 40297 + }, + { + "epoch": 0.40298, + "grad_norm": 0.9027659296989441, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 40298 + }, + { + "epoch": 0.40299, + "grad_norm": 1.19182550907135, + "learning_rate": 0.003, + "loss": 4.002, + "step": 40299 + }, + { + "epoch": 0.403, + "grad_norm": 1.0148853063583374, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 40300 + }, + { + "epoch": 0.40301, + "grad_norm": 1.0779414176940918, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 40301 + }, + { + "epoch": 0.40302, + "grad_norm": 0.9300176501274109, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 40302 + }, + { + "epoch": 0.40303, + "grad_norm": 0.8257849216461182, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 40303 + }, + { + "epoch": 0.40304, + "grad_norm": 0.9062145352363586, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 40304 + }, + { + "epoch": 0.40305, + "grad_norm": 1.0306529998779297, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 40305 + }, + { + "epoch": 0.40306, + "grad_norm": 1.0143671035766602, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 40306 + }, + { + "epoch": 0.40307, + "grad_norm": 1.1468392610549927, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 40307 + }, + { + "epoch": 0.40308, + "grad_norm": 0.9792750477790833, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 40308 + }, + { + "epoch": 0.40309, + "grad_norm": 0.9917269945144653, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 40309 + }, + { + "epoch": 0.4031, + "grad_norm": 1.0802127122879028, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 40310 + }, + { + "epoch": 0.40311, + "grad_norm": 0.8839142322540283, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 40311 + }, + { + "epoch": 0.40312, + "grad_norm": 0.7492055892944336, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 40312 + }, + { + "epoch": 0.40313, + "grad_norm": 0.652870237827301, + "learning_rate": 0.003, + "loss": 3.971, + "step": 40313 + }, + { + "epoch": 0.40314, + "grad_norm": 0.6232883930206299, + "learning_rate": 0.003, + "loss": 3.9319, + "step": 40314 + }, + { + "epoch": 0.40315, + "grad_norm": 0.6608443856239319, + "learning_rate": 0.003, + "loss": 3.9455, + "step": 40315 + }, + { + "epoch": 0.40316, + "grad_norm": 0.8088348507881165, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 40316 + }, + { + "epoch": 0.40317, + "grad_norm": 1.146260142326355, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 40317 + }, + { + "epoch": 0.40318, + "grad_norm": 1.1189135313034058, + "learning_rate": 0.003, + "loss": 3.991, + "step": 40318 + }, + { + "epoch": 0.40319, + "grad_norm": 0.9485734105110168, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 40319 + }, + { + "epoch": 0.4032, + "grad_norm": 0.829016923904419, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 40320 + }, + { + "epoch": 0.40321, + "grad_norm": 0.7615781426429749, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 40321 + }, + { + "epoch": 0.40322, + "grad_norm": 0.8743231296539307, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 40322 + }, + { + "epoch": 0.40323, + "grad_norm": 0.9512473940849304, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 40323 + }, + { + "epoch": 0.40324, + "grad_norm": 0.8578618168830872, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 40324 + }, + { + "epoch": 0.40325, + "grad_norm": 0.8163694739341736, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 40325 + }, + { + "epoch": 0.40326, + "grad_norm": 0.7240392565727234, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 40326 + }, + { + "epoch": 0.40327, + "grad_norm": 0.7642767429351807, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 40327 + }, + { + "epoch": 0.40328, + "grad_norm": 0.9519209265708923, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 40328 + }, + { + "epoch": 0.40329, + "grad_norm": 1.1641110181808472, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 40329 + }, + { + "epoch": 0.4033, + "grad_norm": 0.8919271230697632, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 40330 + }, + { + "epoch": 0.40331, + "grad_norm": 0.8787993788719177, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 40331 + }, + { + "epoch": 0.40332, + "grad_norm": 1.0508915185928345, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 40332 + }, + { + "epoch": 0.40333, + "grad_norm": 1.0038068294525146, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 40333 + }, + { + "epoch": 0.40334, + "grad_norm": 0.9086034893989563, + "learning_rate": 0.003, + "loss": 3.958, + "step": 40334 + }, + { + "epoch": 0.40335, + "grad_norm": 0.7699264883995056, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 40335 + }, + { + "epoch": 0.40336, + "grad_norm": 0.6756032109260559, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 40336 + }, + { + "epoch": 0.40337, + "grad_norm": 0.6030412316322327, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 40337 + }, + { + "epoch": 0.40338, + "grad_norm": 0.5835550427436829, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 40338 + }, + { + "epoch": 0.40339, + "grad_norm": 0.6117238998413086, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 40339 + }, + { + "epoch": 0.4034, + "grad_norm": 0.6789363622665405, + "learning_rate": 0.003, + "loss": 3.952, + "step": 40340 + }, + { + "epoch": 0.40341, + "grad_norm": 0.8966889381408691, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 40341 + }, + { + "epoch": 0.40342, + "grad_norm": 1.263900637626648, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 40342 + }, + { + "epoch": 0.40343, + "grad_norm": 0.8787902593612671, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 40343 + }, + { + "epoch": 0.40344, + "grad_norm": 0.7626080513000488, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 40344 + }, + { + "epoch": 0.40345, + "grad_norm": 0.7751805782318115, + "learning_rate": 0.003, + "loss": 3.962, + "step": 40345 + }, + { + "epoch": 0.40346, + "grad_norm": 0.8635360598564148, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 40346 + }, + { + "epoch": 0.40347, + "grad_norm": 0.9679340720176697, + "learning_rate": 0.003, + "loss": 3.969, + "step": 40347 + }, + { + "epoch": 0.40348, + "grad_norm": 0.9519258737564087, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 40348 + }, + { + "epoch": 0.40349, + "grad_norm": 1.1458730697631836, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 40349 + }, + { + "epoch": 0.4035, + "grad_norm": 1.025905728340149, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 40350 + }, + { + "epoch": 0.40351, + "grad_norm": 0.9195383191108704, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 40351 + }, + { + "epoch": 0.40352, + "grad_norm": 0.9221135377883911, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 40352 + }, + { + "epoch": 0.40353, + "grad_norm": 1.103484869003296, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 40353 + }, + { + "epoch": 0.40354, + "grad_norm": 1.1781758069992065, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 40354 + }, + { + "epoch": 0.40355, + "grad_norm": 1.0121190547943115, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 40355 + }, + { + "epoch": 0.40356, + "grad_norm": 1.0462439060211182, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 40356 + }, + { + "epoch": 0.40357, + "grad_norm": 0.9319489002227783, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 40357 + }, + { + "epoch": 0.40358, + "grad_norm": 0.9874815344810486, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 40358 + }, + { + "epoch": 0.40359, + "grad_norm": 1.1686402559280396, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 40359 + }, + { + "epoch": 0.4036, + "grad_norm": 1.0985956192016602, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 40360 + }, + { + "epoch": 0.40361, + "grad_norm": 0.8643165826797485, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 40361 + }, + { + "epoch": 0.40362, + "grad_norm": 0.7862058281898499, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 40362 + }, + { + "epoch": 0.40363, + "grad_norm": 0.8491990566253662, + "learning_rate": 0.003, + "loss": 3.975, + "step": 40363 + }, + { + "epoch": 0.40364, + "grad_norm": 0.9194644093513489, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 40364 + }, + { + "epoch": 0.40365, + "grad_norm": 0.7696728110313416, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 40365 + }, + { + "epoch": 0.40366, + "grad_norm": 0.8150215744972229, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 40366 + }, + { + "epoch": 0.40367, + "grad_norm": 0.8712667226791382, + "learning_rate": 0.003, + "loss": 3.995, + "step": 40367 + }, + { + "epoch": 0.40368, + "grad_norm": 0.8510116338729858, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 40368 + }, + { + "epoch": 0.40369, + "grad_norm": 0.8314531445503235, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 40369 + }, + { + "epoch": 0.4037, + "grad_norm": 0.8495156764984131, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 40370 + }, + { + "epoch": 0.40371, + "grad_norm": 0.8338257074356079, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 40371 + }, + { + "epoch": 0.40372, + "grad_norm": 1.0409116744995117, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 40372 + }, + { + "epoch": 0.40373, + "grad_norm": 1.3372243642807007, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 40373 + }, + { + "epoch": 0.40374, + "grad_norm": 0.8388099074363708, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 40374 + }, + { + "epoch": 0.40375, + "grad_norm": 0.8068958520889282, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 40375 + }, + { + "epoch": 0.40376, + "grad_norm": 1.0544556379318237, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 40376 + }, + { + "epoch": 0.40377, + "grad_norm": 1.136683702468872, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 40377 + }, + { + "epoch": 0.40378, + "grad_norm": 1.0083273649215698, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 40378 + }, + { + "epoch": 0.40379, + "grad_norm": 1.0975998640060425, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 40379 + }, + { + "epoch": 0.4038, + "grad_norm": 0.9828320741653442, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 40380 + }, + { + "epoch": 0.40381, + "grad_norm": 1.0810178518295288, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 40381 + }, + { + "epoch": 0.40382, + "grad_norm": 1.0770021677017212, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 40382 + }, + { + "epoch": 0.40383, + "grad_norm": 0.7854233384132385, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 40383 + }, + { + "epoch": 0.40384, + "grad_norm": 0.6706753373146057, + "learning_rate": 0.003, + "loss": 3.967, + "step": 40384 + }, + { + "epoch": 0.40385, + "grad_norm": 0.7205941677093506, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 40385 + }, + { + "epoch": 0.40386, + "grad_norm": 0.8858914375305176, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 40386 + }, + { + "epoch": 0.40387, + "grad_norm": 1.1307395696640015, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 40387 + }, + { + "epoch": 0.40388, + "grad_norm": 1.0018211603164673, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 40388 + }, + { + "epoch": 0.40389, + "grad_norm": 1.0618492364883423, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 40389 + }, + { + "epoch": 0.4039, + "grad_norm": 0.9315917491912842, + "learning_rate": 0.003, + "loss": 3.9456, + "step": 40390 + }, + { + "epoch": 0.40391, + "grad_norm": 0.9809015393257141, + "learning_rate": 0.003, + "loss": 3.998, + "step": 40391 + }, + { + "epoch": 0.40392, + "grad_norm": 0.980414628982544, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 40392 + }, + { + "epoch": 0.40393, + "grad_norm": 0.8636367917060852, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 40393 + }, + { + "epoch": 0.40394, + "grad_norm": 0.7544284462928772, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 40394 + }, + { + "epoch": 0.40395, + "grad_norm": 0.6479484438896179, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 40395 + }, + { + "epoch": 0.40396, + "grad_norm": 0.6102585196495056, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 40396 + }, + { + "epoch": 0.40397, + "grad_norm": 0.6599613428115845, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 40397 + }, + { + "epoch": 0.40398, + "grad_norm": 0.6848459243774414, + "learning_rate": 0.003, + "loss": 3.965, + "step": 40398 + }, + { + "epoch": 0.40399, + "grad_norm": 0.6897225975990295, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 40399 + }, + { + "epoch": 0.404, + "grad_norm": 0.7521931529045105, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 40400 + }, + { + "epoch": 0.40401, + "grad_norm": 0.7327809929847717, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 40401 + }, + { + "epoch": 0.40402, + "grad_norm": 0.8421659469604492, + "learning_rate": 0.003, + "loss": 3.973, + "step": 40402 + }, + { + "epoch": 0.40403, + "grad_norm": 1.2022699117660522, + "learning_rate": 0.003, + "loss": 3.9462, + "step": 40403 + }, + { + "epoch": 0.40404, + "grad_norm": 1.1621185541152954, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 40404 + }, + { + "epoch": 0.40405, + "grad_norm": 0.6431906223297119, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 40405 + }, + { + "epoch": 0.40406, + "grad_norm": 0.6540682315826416, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 40406 + }, + { + "epoch": 0.40407, + "grad_norm": 0.8671157956123352, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 40407 + }, + { + "epoch": 0.40408, + "grad_norm": 0.8955549597740173, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 40408 + }, + { + "epoch": 0.40409, + "grad_norm": 0.9553894400596619, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 40409 + }, + { + "epoch": 0.4041, + "grad_norm": 1.2745351791381836, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 40410 + }, + { + "epoch": 0.40411, + "grad_norm": 0.7282829284667969, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 40411 + }, + { + "epoch": 0.40412, + "grad_norm": 0.7702677845954895, + "learning_rate": 0.003, + "loss": 3.981, + "step": 40412 + }, + { + "epoch": 0.40413, + "grad_norm": 0.7328338623046875, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 40413 + }, + { + "epoch": 0.40414, + "grad_norm": 0.9021430015563965, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 40414 + }, + { + "epoch": 0.40415, + "grad_norm": 1.0898709297180176, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 40415 + }, + { + "epoch": 0.40416, + "grad_norm": 0.8916219472885132, + "learning_rate": 0.003, + "loss": 3.98, + "step": 40416 + }, + { + "epoch": 0.40417, + "grad_norm": 0.878346860408783, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 40417 + }, + { + "epoch": 0.40418, + "grad_norm": 0.8316139578819275, + "learning_rate": 0.003, + "loss": 3.9231, + "step": 40418 + }, + { + "epoch": 0.40419, + "grad_norm": 0.8174173831939697, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 40419 + }, + { + "epoch": 0.4042, + "grad_norm": 0.8759422898292542, + "learning_rate": 0.003, + "loss": 3.987, + "step": 40420 + }, + { + "epoch": 0.40421, + "grad_norm": 1.0290495157241821, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 40421 + }, + { + "epoch": 0.40422, + "grad_norm": 1.2756950855255127, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 40422 + }, + { + "epoch": 0.40423, + "grad_norm": 0.730744481086731, + "learning_rate": 0.003, + "loss": 3.953, + "step": 40423 + }, + { + "epoch": 0.40424, + "grad_norm": 0.6405521035194397, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 40424 + }, + { + "epoch": 0.40425, + "grad_norm": 0.641980767250061, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 40425 + }, + { + "epoch": 0.40426, + "grad_norm": 0.7408418655395508, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 40426 + }, + { + "epoch": 0.40427, + "grad_norm": 1.0086522102355957, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 40427 + }, + { + "epoch": 0.40428, + "grad_norm": 1.4597880840301514, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 40428 + }, + { + "epoch": 0.40429, + "grad_norm": 0.5724375247955322, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 40429 + }, + { + "epoch": 0.4043, + "grad_norm": 0.7701717615127563, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 40430 + }, + { + "epoch": 0.40431, + "grad_norm": 0.7940741181373596, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 40431 + }, + { + "epoch": 0.40432, + "grad_norm": 0.7264264225959778, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 40432 + }, + { + "epoch": 0.40433, + "grad_norm": 0.6535962224006653, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 40433 + }, + { + "epoch": 0.40434, + "grad_norm": 0.7715522646903992, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 40434 + }, + { + "epoch": 0.40435, + "grad_norm": 0.9041137099266052, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 40435 + }, + { + "epoch": 0.40436, + "grad_norm": 1.1741371154785156, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 40436 + }, + { + "epoch": 0.40437, + "grad_norm": 1.0730903148651123, + "learning_rate": 0.003, + "loss": 3.9247, + "step": 40437 + }, + { + "epoch": 0.40438, + "grad_norm": 1.2413928508758545, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 40438 + }, + { + "epoch": 0.40439, + "grad_norm": 0.761562705039978, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 40439 + }, + { + "epoch": 0.4044, + "grad_norm": 0.7025253176689148, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 40440 + }, + { + "epoch": 0.40441, + "grad_norm": 0.8406286835670471, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 40441 + }, + { + "epoch": 0.40442, + "grad_norm": 0.9385424852371216, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 40442 + }, + { + "epoch": 0.40443, + "grad_norm": 1.1059705018997192, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 40443 + }, + { + "epoch": 0.40444, + "grad_norm": 1.1041431427001953, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 40444 + }, + { + "epoch": 0.40445, + "grad_norm": 1.1347129344940186, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 40445 + }, + { + "epoch": 0.40446, + "grad_norm": 1.1282604932785034, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 40446 + }, + { + "epoch": 0.40447, + "grad_norm": 0.9089956283569336, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 40447 + }, + { + "epoch": 0.40448, + "grad_norm": 0.808849573135376, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 40448 + }, + { + "epoch": 0.40449, + "grad_norm": 0.7809805870056152, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 40449 + }, + { + "epoch": 0.4045, + "grad_norm": 0.7889823913574219, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 40450 + }, + { + "epoch": 0.40451, + "grad_norm": 0.9423097968101501, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 40451 + }, + { + "epoch": 0.40452, + "grad_norm": 1.0494776964187622, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 40452 + }, + { + "epoch": 0.40453, + "grad_norm": 0.7402172684669495, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 40453 + }, + { + "epoch": 0.40454, + "grad_norm": 0.6960288286209106, + "learning_rate": 0.003, + "loss": 3.973, + "step": 40454 + }, + { + "epoch": 0.40455, + "grad_norm": 0.7784504294395447, + "learning_rate": 0.003, + "loss": 3.973, + "step": 40455 + }, + { + "epoch": 0.40456, + "grad_norm": 1.2806684970855713, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 40456 + }, + { + "epoch": 0.40457, + "grad_norm": 0.883298933506012, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 40457 + }, + { + "epoch": 0.40458, + "grad_norm": 0.791825532913208, + "learning_rate": 0.003, + "loss": 3.966, + "step": 40458 + }, + { + "epoch": 0.40459, + "grad_norm": 0.8735390305519104, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 40459 + }, + { + "epoch": 0.4046, + "grad_norm": 1.1302465200424194, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 40460 + }, + { + "epoch": 0.40461, + "grad_norm": 1.1069966554641724, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 40461 + }, + { + "epoch": 0.40462, + "grad_norm": 0.9074348211288452, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 40462 + }, + { + "epoch": 0.40463, + "grad_norm": 0.7568057179450989, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 40463 + }, + { + "epoch": 0.40464, + "grad_norm": 0.6488666534423828, + "learning_rate": 0.003, + "loss": 3.974, + "step": 40464 + }, + { + "epoch": 0.40465, + "grad_norm": 0.6146978735923767, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 40465 + }, + { + "epoch": 0.40466, + "grad_norm": 0.6326691508293152, + "learning_rate": 0.003, + "loss": 3.9434, + "step": 40466 + }, + { + "epoch": 0.40467, + "grad_norm": 0.7034300565719604, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 40467 + }, + { + "epoch": 0.40468, + "grad_norm": 0.7990104556083679, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 40468 + }, + { + "epoch": 0.40469, + "grad_norm": 0.8628001809120178, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 40469 + }, + { + "epoch": 0.4047, + "grad_norm": 0.9514825344085693, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 40470 + }, + { + "epoch": 0.40471, + "grad_norm": 1.0781004428863525, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 40471 + }, + { + "epoch": 0.40472, + "grad_norm": 0.9775097370147705, + "learning_rate": 0.003, + "loss": 3.9488, + "step": 40472 + }, + { + "epoch": 0.40473, + "grad_norm": 0.9947212934494019, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 40473 + }, + { + "epoch": 0.40474, + "grad_norm": 0.8975867629051208, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 40474 + }, + { + "epoch": 0.40475, + "grad_norm": 0.8522964715957642, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 40475 + }, + { + "epoch": 0.40476, + "grad_norm": 0.9282830953598022, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 40476 + }, + { + "epoch": 0.40477, + "grad_norm": 0.9955650568008423, + "learning_rate": 0.003, + "loss": 3.994, + "step": 40477 + }, + { + "epoch": 0.40478, + "grad_norm": 1.0194497108459473, + "learning_rate": 0.003, + "loss": 4.008, + "step": 40478 + }, + { + "epoch": 0.40479, + "grad_norm": 0.8784494400024414, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 40479 + }, + { + "epoch": 0.4048, + "grad_norm": 0.8284628987312317, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 40480 + }, + { + "epoch": 0.40481, + "grad_norm": 0.8790768384933472, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 40481 + }, + { + "epoch": 0.40482, + "grad_norm": 1.0467976331710815, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 40482 + }, + { + "epoch": 0.40483, + "grad_norm": 1.0203853845596313, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 40483 + }, + { + "epoch": 0.40484, + "grad_norm": 1.1536705493927002, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 40484 + }, + { + "epoch": 0.40485, + "grad_norm": 1.0480924844741821, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 40485 + }, + { + "epoch": 0.40486, + "grad_norm": 1.0561819076538086, + "learning_rate": 0.003, + "loss": 3.988, + "step": 40486 + }, + { + "epoch": 0.40487, + "grad_norm": 1.0113022327423096, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 40487 + }, + { + "epoch": 0.40488, + "grad_norm": 1.0530145168304443, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 40488 + }, + { + "epoch": 0.40489, + "grad_norm": 0.8723465204238892, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 40489 + }, + { + "epoch": 0.4049, + "grad_norm": 0.7527973055839539, + "learning_rate": 0.003, + "loss": 3.982, + "step": 40490 + }, + { + "epoch": 0.40491, + "grad_norm": 0.8272317051887512, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 40491 + }, + { + "epoch": 0.40492, + "grad_norm": 0.9380320906639099, + "learning_rate": 0.003, + "loss": 4.003, + "step": 40492 + }, + { + "epoch": 0.40493, + "grad_norm": 1.1025339365005493, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 40493 + }, + { + "epoch": 0.40494, + "grad_norm": 1.1238977909088135, + "learning_rate": 0.003, + "loss": 3.991, + "step": 40494 + }, + { + "epoch": 0.40495, + "grad_norm": 0.8920940160751343, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 40495 + }, + { + "epoch": 0.40496, + "grad_norm": 0.9063047170639038, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 40496 + }, + { + "epoch": 0.40497, + "grad_norm": 0.9802424907684326, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 40497 + }, + { + "epoch": 0.40498, + "grad_norm": 1.2207056283950806, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 40498 + }, + { + "epoch": 0.40499, + "grad_norm": 1.0252774953842163, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 40499 + }, + { + "epoch": 0.405, + "grad_norm": 1.0077083110809326, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 40500 + }, + { + "epoch": 0.40501, + "grad_norm": 0.9892186522483826, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 40501 + }, + { + "epoch": 0.40502, + "grad_norm": 0.9820939898490906, + "learning_rate": 0.003, + "loss": 3.9411, + "step": 40502 + }, + { + "epoch": 0.40503, + "grad_norm": 0.9246253967285156, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 40503 + }, + { + "epoch": 0.40504, + "grad_norm": 0.9120099544525146, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 40504 + }, + { + "epoch": 0.40505, + "grad_norm": 0.873741865158081, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 40505 + }, + { + "epoch": 0.40506, + "grad_norm": 0.7136989831924438, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 40506 + }, + { + "epoch": 0.40507, + "grad_norm": 0.7072519063949585, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 40507 + }, + { + "epoch": 0.40508, + "grad_norm": 0.8108944892883301, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 40508 + }, + { + "epoch": 0.40509, + "grad_norm": 0.7215296030044556, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 40509 + }, + { + "epoch": 0.4051, + "grad_norm": 0.7310380935668945, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 40510 + }, + { + "epoch": 0.40511, + "grad_norm": 0.7019603848457336, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 40511 + }, + { + "epoch": 0.40512, + "grad_norm": 0.8294612169265747, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 40512 + }, + { + "epoch": 0.40513, + "grad_norm": 0.9532233476638794, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 40513 + }, + { + "epoch": 0.40514, + "grad_norm": 1.2209889888763428, + "learning_rate": 0.003, + "loss": 3.988, + "step": 40514 + }, + { + "epoch": 0.40515, + "grad_norm": 0.9943005442619324, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 40515 + }, + { + "epoch": 0.40516, + "grad_norm": 0.9943892955780029, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 40516 + }, + { + "epoch": 0.40517, + "grad_norm": 1.292333722114563, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 40517 + }, + { + "epoch": 0.40518, + "grad_norm": 0.8415559530258179, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 40518 + }, + { + "epoch": 0.40519, + "grad_norm": 0.8172149062156677, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 40519 + }, + { + "epoch": 0.4052, + "grad_norm": 0.8818583488464355, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 40520 + }, + { + "epoch": 0.40521, + "grad_norm": 0.8251470327377319, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 40521 + }, + { + "epoch": 0.40522, + "grad_norm": 0.9160980582237244, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 40522 + }, + { + "epoch": 0.40523, + "grad_norm": 1.1041382551193237, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 40523 + }, + { + "epoch": 0.40524, + "grad_norm": 0.7836965918540955, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 40524 + }, + { + "epoch": 0.40525, + "grad_norm": 0.6986514925956726, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 40525 + }, + { + "epoch": 0.40526, + "grad_norm": 0.6335393786430359, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 40526 + }, + { + "epoch": 0.40527, + "grad_norm": 0.5791330337524414, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 40527 + }, + { + "epoch": 0.40528, + "grad_norm": 0.6505630016326904, + "learning_rate": 0.003, + "loss": 3.968, + "step": 40528 + }, + { + "epoch": 0.40529, + "grad_norm": 0.8054616451263428, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 40529 + }, + { + "epoch": 0.4053, + "grad_norm": 1.0247430801391602, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 40530 + }, + { + "epoch": 0.40531, + "grad_norm": 1.0672688484191895, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 40531 + }, + { + "epoch": 0.40532, + "grad_norm": 0.9296207427978516, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 40532 + }, + { + "epoch": 0.40533, + "grad_norm": 0.9914017915725708, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 40533 + }, + { + "epoch": 0.40534, + "grad_norm": 1.1617145538330078, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 40534 + }, + { + "epoch": 0.40535, + "grad_norm": 1.0755709409713745, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 40535 + }, + { + "epoch": 0.40536, + "grad_norm": 0.9898533225059509, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 40536 + }, + { + "epoch": 0.40537, + "grad_norm": 1.1676592826843262, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 40537 + }, + { + "epoch": 0.40538, + "grad_norm": 1.1953117847442627, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 40538 + }, + { + "epoch": 0.40539, + "grad_norm": 0.7474803924560547, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 40539 + }, + { + "epoch": 0.4054, + "grad_norm": 0.58269864320755, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 40540 + }, + { + "epoch": 0.40541, + "grad_norm": 0.7923963665962219, + "learning_rate": 0.003, + "loss": 3.974, + "step": 40541 + }, + { + "epoch": 0.40542, + "grad_norm": 1.1918554306030273, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 40542 + }, + { + "epoch": 0.40543, + "grad_norm": 1.1315028667449951, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 40543 + }, + { + "epoch": 0.40544, + "grad_norm": 0.7236154079437256, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 40544 + }, + { + "epoch": 0.40545, + "grad_norm": 0.47585734724998474, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 40545 + }, + { + "epoch": 0.40546, + "grad_norm": 0.5306246280670166, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 40546 + }, + { + "epoch": 0.40547, + "grad_norm": 0.5483518838882446, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 40547 + }, + { + "epoch": 0.40548, + "grad_norm": 0.4937819242477417, + "learning_rate": 0.003, + "loss": 3.9371, + "step": 40548 + }, + { + "epoch": 0.40549, + "grad_norm": 0.5287407636642456, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 40549 + }, + { + "epoch": 0.4055, + "grad_norm": 0.6949765682220459, + "learning_rate": 0.003, + "loss": 3.993, + "step": 40550 + }, + { + "epoch": 0.40551, + "grad_norm": 0.9387413859367371, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 40551 + }, + { + "epoch": 0.40552, + "grad_norm": 1.1801241636276245, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 40552 + }, + { + "epoch": 0.40553, + "grad_norm": 0.7936729192733765, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 40553 + }, + { + "epoch": 0.40554, + "grad_norm": 0.9365137815475464, + "learning_rate": 0.003, + "loss": 3.966, + "step": 40554 + }, + { + "epoch": 0.40555, + "grad_norm": 1.1647353172302246, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 40555 + }, + { + "epoch": 0.40556, + "grad_norm": 1.1876846551895142, + "learning_rate": 0.003, + "loss": 3.975, + "step": 40556 + }, + { + "epoch": 0.40557, + "grad_norm": 0.7748845815658569, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 40557 + }, + { + "epoch": 0.40558, + "grad_norm": 0.6815659403800964, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 40558 + }, + { + "epoch": 0.40559, + "grad_norm": 0.6270972490310669, + "learning_rate": 0.003, + "loss": 3.9467, + "step": 40559 + }, + { + "epoch": 0.4056, + "grad_norm": 0.593883216381073, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 40560 + }, + { + "epoch": 0.40561, + "grad_norm": 0.6758987903594971, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 40561 + }, + { + "epoch": 0.40562, + "grad_norm": 0.7514887452125549, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 40562 + }, + { + "epoch": 0.40563, + "grad_norm": 0.7021745443344116, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 40563 + }, + { + "epoch": 0.40564, + "grad_norm": 0.7278351783752441, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 40564 + }, + { + "epoch": 0.40565, + "grad_norm": 0.7159218192100525, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 40565 + }, + { + "epoch": 0.40566, + "grad_norm": 0.658631443977356, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 40566 + }, + { + "epoch": 0.40567, + "grad_norm": 0.7740139365196228, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 40567 + }, + { + "epoch": 0.40568, + "grad_norm": 0.934179425239563, + "learning_rate": 0.003, + "loss": 3.962, + "step": 40568 + }, + { + "epoch": 0.40569, + "grad_norm": 1.1218370199203491, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 40569 + }, + { + "epoch": 0.4057, + "grad_norm": 1.076261281967163, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 40570 + }, + { + "epoch": 0.40571, + "grad_norm": 1.186962604522705, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 40571 + }, + { + "epoch": 0.40572, + "grad_norm": 0.7353163957595825, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 40572 + }, + { + "epoch": 0.40573, + "grad_norm": 0.9006279110908508, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 40573 + }, + { + "epoch": 0.40574, + "grad_norm": 1.3336507081985474, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 40574 + }, + { + "epoch": 0.40575, + "grad_norm": 0.8921779990196228, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 40575 + }, + { + "epoch": 0.40576, + "grad_norm": 0.6853928565979004, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 40576 + }, + { + "epoch": 0.40577, + "grad_norm": 0.6712722182273865, + "learning_rate": 0.003, + "loss": 3.969, + "step": 40577 + }, + { + "epoch": 0.40578, + "grad_norm": 0.7682621479034424, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 40578 + }, + { + "epoch": 0.40579, + "grad_norm": 0.9039539694786072, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 40579 + }, + { + "epoch": 0.4058, + "grad_norm": 1.2011961936950684, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 40580 + }, + { + "epoch": 0.40581, + "grad_norm": 0.8920822739601135, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 40581 + }, + { + "epoch": 0.40582, + "grad_norm": 1.114769458770752, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 40582 + }, + { + "epoch": 0.40583, + "grad_norm": 1.0505695343017578, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 40583 + }, + { + "epoch": 0.40584, + "grad_norm": 1.0692052841186523, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 40584 + }, + { + "epoch": 0.40585, + "grad_norm": 1.1419274806976318, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 40585 + }, + { + "epoch": 0.40586, + "grad_norm": 0.9682794213294983, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 40586 + }, + { + "epoch": 0.40587, + "grad_norm": 0.872249960899353, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 40587 + }, + { + "epoch": 0.40588, + "grad_norm": 0.7632177472114563, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 40588 + }, + { + "epoch": 0.40589, + "grad_norm": 0.8114426136016846, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 40589 + }, + { + "epoch": 0.4059, + "grad_norm": 0.7977334260940552, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 40590 + }, + { + "epoch": 0.40591, + "grad_norm": 0.9085190296173096, + "learning_rate": 0.003, + "loss": 3.982, + "step": 40591 + }, + { + "epoch": 0.40592, + "grad_norm": 1.04975163936615, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 40592 + }, + { + "epoch": 0.40593, + "grad_norm": 0.9830321669578552, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 40593 + }, + { + "epoch": 0.40594, + "grad_norm": 1.0746186971664429, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 40594 + }, + { + "epoch": 0.40595, + "grad_norm": 0.9834139347076416, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 40595 + }, + { + "epoch": 0.40596, + "grad_norm": 0.947290301322937, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 40596 + }, + { + "epoch": 0.40597, + "grad_norm": 1.1197022199630737, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 40597 + }, + { + "epoch": 0.40598, + "grad_norm": 1.334824562072754, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 40598 + }, + { + "epoch": 0.40599, + "grad_norm": 0.9308537244796753, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 40599 + }, + { + "epoch": 0.406, + "grad_norm": 0.753114640712738, + "learning_rate": 0.003, + "loss": 3.984, + "step": 40600 + }, + { + "epoch": 0.40601, + "grad_norm": 0.7552365660667419, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 40601 + }, + { + "epoch": 0.40602, + "grad_norm": 0.9112823605537415, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 40602 + }, + { + "epoch": 0.40603, + "grad_norm": 1.11295747756958, + "learning_rate": 0.003, + "loss": 3.978, + "step": 40603 + }, + { + "epoch": 0.40604, + "grad_norm": 0.9530515074729919, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 40604 + }, + { + "epoch": 0.40605, + "grad_norm": 1.033362865447998, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 40605 + }, + { + "epoch": 0.40606, + "grad_norm": 0.9096786379814148, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 40606 + }, + { + "epoch": 0.40607, + "grad_norm": 0.8042378425598145, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 40607 + }, + { + "epoch": 0.40608, + "grad_norm": 0.8151567578315735, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 40608 + }, + { + "epoch": 0.40609, + "grad_norm": 0.8850197196006775, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 40609 + }, + { + "epoch": 0.4061, + "grad_norm": 0.9074773192405701, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 40610 + }, + { + "epoch": 0.40611, + "grad_norm": 0.9433488845825195, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 40611 + }, + { + "epoch": 0.40612, + "grad_norm": 1.1273646354675293, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 40612 + }, + { + "epoch": 0.40613, + "grad_norm": 0.9880017042160034, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 40613 + }, + { + "epoch": 0.40614, + "grad_norm": 0.9216626286506653, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 40614 + }, + { + "epoch": 0.40615, + "grad_norm": 0.971742570400238, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 40615 + }, + { + "epoch": 0.40616, + "grad_norm": 1.0811877250671387, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 40616 + }, + { + "epoch": 0.40617, + "grad_norm": 0.9110305309295654, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 40617 + }, + { + "epoch": 0.40618, + "grad_norm": 0.9752118587493896, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 40618 + }, + { + "epoch": 0.40619, + "grad_norm": 1.1966983079910278, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 40619 + }, + { + "epoch": 0.4062, + "grad_norm": 0.9248772263526917, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 40620 + }, + { + "epoch": 0.40621, + "grad_norm": 0.9710018038749695, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 40621 + }, + { + "epoch": 0.40622, + "grad_norm": 1.0262972116470337, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 40622 + }, + { + "epoch": 0.40623, + "grad_norm": 1.1983858346939087, + "learning_rate": 0.003, + "loss": 3.981, + "step": 40623 + }, + { + "epoch": 0.40624, + "grad_norm": 0.7193641662597656, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 40624 + }, + { + "epoch": 0.40625, + "grad_norm": 0.7046363949775696, + "learning_rate": 0.003, + "loss": 3.981, + "step": 40625 + }, + { + "epoch": 0.40626, + "grad_norm": 0.8052520751953125, + "learning_rate": 0.003, + "loss": 3.974, + "step": 40626 + }, + { + "epoch": 0.40627, + "grad_norm": 1.1137114763259888, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 40627 + }, + { + "epoch": 0.40628, + "grad_norm": 0.8871970176696777, + "learning_rate": 0.003, + "loss": 3.961, + "step": 40628 + }, + { + "epoch": 0.40629, + "grad_norm": 0.658207893371582, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 40629 + }, + { + "epoch": 0.4063, + "grad_norm": 0.5658740401268005, + "learning_rate": 0.003, + "loss": 3.9282, + "step": 40630 + }, + { + "epoch": 0.40631, + "grad_norm": 0.4920901358127594, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 40631 + }, + { + "epoch": 0.40632, + "grad_norm": 0.6238511800765991, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 40632 + }, + { + "epoch": 0.40633, + "grad_norm": 0.6914446353912354, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 40633 + }, + { + "epoch": 0.40634, + "grad_norm": 0.7080362439155579, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 40634 + }, + { + "epoch": 0.40635, + "grad_norm": 0.7573118209838867, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 40635 + }, + { + "epoch": 0.40636, + "grad_norm": 0.9552774429321289, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 40636 + }, + { + "epoch": 0.40637, + "grad_norm": 1.1575126647949219, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 40637 + }, + { + "epoch": 0.40638, + "grad_norm": 0.7938718199729919, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 40638 + }, + { + "epoch": 0.40639, + "grad_norm": 0.7617126107215881, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 40639 + }, + { + "epoch": 0.4064, + "grad_norm": 0.7527756690979004, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 40640 + }, + { + "epoch": 0.40641, + "grad_norm": 0.8047425150871277, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 40641 + }, + { + "epoch": 0.40642, + "grad_norm": 0.8068206310272217, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 40642 + }, + { + "epoch": 0.40643, + "grad_norm": 0.814594030380249, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 40643 + }, + { + "epoch": 0.40644, + "grad_norm": 0.9461479783058167, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 40644 + }, + { + "epoch": 0.40645, + "grad_norm": 1.0269418954849243, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 40645 + }, + { + "epoch": 0.40646, + "grad_norm": 1.0866096019744873, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 40646 + }, + { + "epoch": 0.40647, + "grad_norm": 0.9655125141143799, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 40647 + }, + { + "epoch": 0.40648, + "grad_norm": 0.9753226637840271, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 40648 + }, + { + "epoch": 0.40649, + "grad_norm": 1.0880461931228638, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 40649 + }, + { + "epoch": 0.4065, + "grad_norm": 0.884387731552124, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 40650 + }, + { + "epoch": 0.40651, + "grad_norm": 0.8614346385002136, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 40651 + }, + { + "epoch": 0.40652, + "grad_norm": 0.9958211779594421, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 40652 + }, + { + "epoch": 0.40653, + "grad_norm": 1.1269540786743164, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 40653 + }, + { + "epoch": 0.40654, + "grad_norm": 0.9550330638885498, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 40654 + }, + { + "epoch": 0.40655, + "grad_norm": 0.8044222593307495, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 40655 + }, + { + "epoch": 0.40656, + "grad_norm": 0.7064229846000671, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 40656 + }, + { + "epoch": 0.40657, + "grad_norm": 0.827189028263092, + "learning_rate": 0.003, + "loss": 3.952, + "step": 40657 + }, + { + "epoch": 0.40658, + "grad_norm": 0.9303664565086365, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 40658 + }, + { + "epoch": 0.40659, + "grad_norm": 1.2311854362487793, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 40659 + }, + { + "epoch": 0.4066, + "grad_norm": 1.191450834274292, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 40660 + }, + { + "epoch": 0.40661, + "grad_norm": 0.7889836430549622, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 40661 + }, + { + "epoch": 0.40662, + "grad_norm": 0.6837198138237, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 40662 + }, + { + "epoch": 0.40663, + "grad_norm": 0.7072145938873291, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 40663 + }, + { + "epoch": 0.40664, + "grad_norm": 0.8084099888801575, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 40664 + }, + { + "epoch": 0.40665, + "grad_norm": 1.005152702331543, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 40665 + }, + { + "epoch": 0.40666, + "grad_norm": 1.1131055355072021, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 40666 + }, + { + "epoch": 0.40667, + "grad_norm": 0.8621871471405029, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 40667 + }, + { + "epoch": 0.40668, + "grad_norm": 0.8840975761413574, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 40668 + }, + { + "epoch": 0.40669, + "grad_norm": 0.9786555171012878, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 40669 + }, + { + "epoch": 0.4067, + "grad_norm": 1.29120934009552, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 40670 + }, + { + "epoch": 0.40671, + "grad_norm": 1.014501929283142, + "learning_rate": 0.003, + "loss": 3.972, + "step": 40671 + }, + { + "epoch": 0.40672, + "grad_norm": 0.9314291477203369, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 40672 + }, + { + "epoch": 0.40673, + "grad_norm": 0.9642642140388489, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 40673 + }, + { + "epoch": 0.40674, + "grad_norm": 1.279295563697815, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 40674 + }, + { + "epoch": 0.40675, + "grad_norm": 0.7905142307281494, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 40675 + }, + { + "epoch": 0.40676, + "grad_norm": 0.5804925560951233, + "learning_rate": 0.003, + "loss": 3.962, + "step": 40676 + }, + { + "epoch": 0.40677, + "grad_norm": 0.7279995679855347, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 40677 + }, + { + "epoch": 0.40678, + "grad_norm": 0.8159498572349548, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 40678 + }, + { + "epoch": 0.40679, + "grad_norm": 1.254531741142273, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 40679 + }, + { + "epoch": 0.4068, + "grad_norm": 0.9299194812774658, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 40680 + }, + { + "epoch": 0.40681, + "grad_norm": 0.8584794402122498, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 40681 + }, + { + "epoch": 0.40682, + "grad_norm": 0.8898563981056213, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 40682 + }, + { + "epoch": 0.40683, + "grad_norm": 1.051718831062317, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 40683 + }, + { + "epoch": 0.40684, + "grad_norm": 0.9178484678268433, + "learning_rate": 0.003, + "loss": 4.0, + "step": 40684 + }, + { + "epoch": 0.40685, + "grad_norm": 1.0427982807159424, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 40685 + }, + { + "epoch": 0.40686, + "grad_norm": 0.9594438672065735, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 40686 + }, + { + "epoch": 0.40687, + "grad_norm": 0.9409937262535095, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 40687 + }, + { + "epoch": 0.40688, + "grad_norm": 1.0408449172973633, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 40688 + }, + { + "epoch": 0.40689, + "grad_norm": 1.130816102027893, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 40689 + }, + { + "epoch": 0.4069, + "grad_norm": 0.9142009615898132, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 40690 + }, + { + "epoch": 0.40691, + "grad_norm": 1.0030320882797241, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 40691 + }, + { + "epoch": 0.40692, + "grad_norm": 1.0361065864562988, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 40692 + }, + { + "epoch": 0.40693, + "grad_norm": 1.0273336172103882, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 40693 + }, + { + "epoch": 0.40694, + "grad_norm": 0.8904470205307007, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 40694 + }, + { + "epoch": 0.40695, + "grad_norm": 0.8230336308479309, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 40695 + }, + { + "epoch": 0.40696, + "grad_norm": 0.7477582693099976, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 40696 + }, + { + "epoch": 0.40697, + "grad_norm": 0.7491962313652039, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 40697 + }, + { + "epoch": 0.40698, + "grad_norm": 0.8390926718711853, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 40698 + }, + { + "epoch": 0.40699, + "grad_norm": 0.9644114971160889, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 40699 + }, + { + "epoch": 0.407, + "grad_norm": 1.1737322807312012, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 40700 + }, + { + "epoch": 0.40701, + "grad_norm": 0.7937082052230835, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 40701 + }, + { + "epoch": 0.40702, + "grad_norm": 0.7549707889556885, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 40702 + }, + { + "epoch": 0.40703, + "grad_norm": 0.7242718935012817, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 40703 + }, + { + "epoch": 0.40704, + "grad_norm": 0.8940736651420593, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 40704 + }, + { + "epoch": 0.40705, + "grad_norm": 1.1118800640106201, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 40705 + }, + { + "epoch": 0.40706, + "grad_norm": 0.7349778413772583, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 40706 + }, + { + "epoch": 0.40707, + "grad_norm": 0.56712406873703, + "learning_rate": 0.003, + "loss": 3.9405, + "step": 40707 + }, + { + "epoch": 0.40708, + "grad_norm": 0.5879752039909363, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 40708 + }, + { + "epoch": 0.40709, + "grad_norm": 0.5923415422439575, + "learning_rate": 0.003, + "loss": 3.9363, + "step": 40709 + }, + { + "epoch": 0.4071, + "grad_norm": 0.5857088565826416, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 40710 + }, + { + "epoch": 0.40711, + "grad_norm": 0.7285358309745789, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 40711 + }, + { + "epoch": 0.40712, + "grad_norm": 1.030369758605957, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 40712 + }, + { + "epoch": 0.40713, + "grad_norm": 1.2901616096496582, + "learning_rate": 0.003, + "loss": 4.012, + "step": 40713 + }, + { + "epoch": 0.40714, + "grad_norm": 0.7955361008644104, + "learning_rate": 0.003, + "loss": 3.961, + "step": 40714 + }, + { + "epoch": 0.40715, + "grad_norm": 0.7298277020454407, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 40715 + }, + { + "epoch": 0.40716, + "grad_norm": 0.739479124546051, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 40716 + }, + { + "epoch": 0.40717, + "grad_norm": 0.8929224610328674, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 40717 + }, + { + "epoch": 0.40718, + "grad_norm": 0.9673499464988708, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 40718 + }, + { + "epoch": 0.40719, + "grad_norm": 0.9740890264511108, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 40719 + }, + { + "epoch": 0.4072, + "grad_norm": 0.9284461140632629, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 40720 + }, + { + "epoch": 0.40721, + "grad_norm": 0.9937244653701782, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 40721 + }, + { + "epoch": 0.40722, + "grad_norm": 1.1125558614730835, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 40722 + }, + { + "epoch": 0.40723, + "grad_norm": 0.9131464958190918, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 40723 + }, + { + "epoch": 0.40724, + "grad_norm": 0.9098989963531494, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 40724 + }, + { + "epoch": 0.40725, + "grad_norm": 1.0346808433532715, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 40725 + }, + { + "epoch": 0.40726, + "grad_norm": 0.9720780849456787, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 40726 + }, + { + "epoch": 0.40727, + "grad_norm": 1.0737711191177368, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 40727 + }, + { + "epoch": 0.40728, + "grad_norm": 1.0870130062103271, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 40728 + }, + { + "epoch": 0.40729, + "grad_norm": 1.1367831230163574, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 40729 + }, + { + "epoch": 0.4073, + "grad_norm": 0.9501994252204895, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 40730 + }, + { + "epoch": 0.40731, + "grad_norm": 0.8835151195526123, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 40731 + }, + { + "epoch": 0.40732, + "grad_norm": 0.9669712781906128, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 40732 + }, + { + "epoch": 0.40733, + "grad_norm": 1.0284016132354736, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 40733 + }, + { + "epoch": 0.40734, + "grad_norm": 1.1635572910308838, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 40734 + }, + { + "epoch": 0.40735, + "grad_norm": 0.780872106552124, + "learning_rate": 0.003, + "loss": 4.006, + "step": 40735 + }, + { + "epoch": 0.40736, + "grad_norm": 0.7749646306037903, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 40736 + }, + { + "epoch": 0.40737, + "grad_norm": 0.8011495471000671, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 40737 + }, + { + "epoch": 0.40738, + "grad_norm": 0.8934308290481567, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 40738 + }, + { + "epoch": 0.40739, + "grad_norm": 0.8674924969673157, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 40739 + }, + { + "epoch": 0.4074, + "grad_norm": 0.8078537583351135, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 40740 + }, + { + "epoch": 0.40741, + "grad_norm": 1.000654697418213, + "learning_rate": 0.003, + "loss": 4.001, + "step": 40741 + }, + { + "epoch": 0.40742, + "grad_norm": 0.9998019337654114, + "learning_rate": 0.003, + "loss": 3.976, + "step": 40742 + }, + { + "epoch": 0.40743, + "grad_norm": 1.0989378690719604, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 40743 + }, + { + "epoch": 0.40744, + "grad_norm": 1.0030533075332642, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 40744 + }, + { + "epoch": 0.40745, + "grad_norm": 0.9952508807182312, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 40745 + }, + { + "epoch": 0.40746, + "grad_norm": 1.0192081928253174, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 40746 + }, + { + "epoch": 0.40747, + "grad_norm": 1.1732152700424194, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 40747 + }, + { + "epoch": 0.40748, + "grad_norm": 0.9443259239196777, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 40748 + }, + { + "epoch": 0.40749, + "grad_norm": 0.8178516626358032, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 40749 + }, + { + "epoch": 0.4075, + "grad_norm": 0.8650404214859009, + "learning_rate": 0.003, + "loss": 3.98, + "step": 40750 + }, + { + "epoch": 0.40751, + "grad_norm": 1.1173689365386963, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 40751 + }, + { + "epoch": 0.40752, + "grad_norm": 1.0409622192382812, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 40752 + }, + { + "epoch": 0.40753, + "grad_norm": 0.9266519546508789, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 40753 + }, + { + "epoch": 0.40754, + "grad_norm": 0.8146383762359619, + "learning_rate": 0.003, + "loss": 4.025, + "step": 40754 + }, + { + "epoch": 0.40755, + "grad_norm": 0.7987688183784485, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 40755 + }, + { + "epoch": 0.40756, + "grad_norm": 0.7664963006973267, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 40756 + }, + { + "epoch": 0.40757, + "grad_norm": 0.6948964595794678, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 40757 + }, + { + "epoch": 0.40758, + "grad_norm": 0.561759352684021, + "learning_rate": 0.003, + "loss": 3.958, + "step": 40758 + }, + { + "epoch": 0.40759, + "grad_norm": 0.638824462890625, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 40759 + }, + { + "epoch": 0.4076, + "grad_norm": 0.6987271308898926, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 40760 + }, + { + "epoch": 0.40761, + "grad_norm": 0.7175271511077881, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 40761 + }, + { + "epoch": 0.40762, + "grad_norm": 0.991740345954895, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 40762 + }, + { + "epoch": 0.40763, + "grad_norm": 1.3831579685211182, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 40763 + }, + { + "epoch": 0.40764, + "grad_norm": 0.6828516721725464, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 40764 + }, + { + "epoch": 0.40765, + "grad_norm": 0.7284600734710693, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 40765 + }, + { + "epoch": 0.40766, + "grad_norm": 0.8955966234207153, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 40766 + }, + { + "epoch": 0.40767, + "grad_norm": 1.0794618129730225, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 40767 + }, + { + "epoch": 0.40768, + "grad_norm": 1.0150130987167358, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 40768 + }, + { + "epoch": 0.40769, + "grad_norm": 1.0333993434906006, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 40769 + }, + { + "epoch": 0.4077, + "grad_norm": 1.1559016704559326, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 40770 + }, + { + "epoch": 0.40771, + "grad_norm": 1.168816089630127, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 40771 + }, + { + "epoch": 0.40772, + "grad_norm": 0.8324388265609741, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 40772 + }, + { + "epoch": 0.40773, + "grad_norm": 0.6950848698616028, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 40773 + }, + { + "epoch": 0.40774, + "grad_norm": 0.6962714791297913, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 40774 + }, + { + "epoch": 0.40775, + "grad_norm": 0.7501577138900757, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 40775 + }, + { + "epoch": 0.40776, + "grad_norm": 0.892704963684082, + "learning_rate": 0.003, + "loss": 3.96, + "step": 40776 + }, + { + "epoch": 0.40777, + "grad_norm": 0.9707507491111755, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 40777 + }, + { + "epoch": 0.40778, + "grad_norm": 0.8015662431716919, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 40778 + }, + { + "epoch": 0.40779, + "grad_norm": 0.8446529507637024, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 40779 + }, + { + "epoch": 0.4078, + "grad_norm": 0.9322056174278259, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 40780 + }, + { + "epoch": 0.40781, + "grad_norm": 0.9191645383834839, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 40781 + }, + { + "epoch": 0.40782, + "grad_norm": 1.111959457397461, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 40782 + }, + { + "epoch": 0.40783, + "grad_norm": 1.1406352519989014, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 40783 + }, + { + "epoch": 0.40784, + "grad_norm": 0.8230763077735901, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 40784 + }, + { + "epoch": 0.40785, + "grad_norm": 0.8626942038536072, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 40785 + }, + { + "epoch": 0.40786, + "grad_norm": 0.9657783508300781, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 40786 + }, + { + "epoch": 0.40787, + "grad_norm": 0.9858313798904419, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 40787 + }, + { + "epoch": 0.40788, + "grad_norm": 1.0069444179534912, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 40788 + }, + { + "epoch": 0.40789, + "grad_norm": 1.0012260675430298, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 40789 + }, + { + "epoch": 0.4079, + "grad_norm": 1.075278401374817, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 40790 + }, + { + "epoch": 0.40791, + "grad_norm": 1.00630784034729, + "learning_rate": 0.003, + "loss": 3.984, + "step": 40791 + }, + { + "epoch": 0.40792, + "grad_norm": 1.009219765663147, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 40792 + }, + { + "epoch": 0.40793, + "grad_norm": 0.8339144587516785, + "learning_rate": 0.003, + "loss": 3.998, + "step": 40793 + }, + { + "epoch": 0.40794, + "grad_norm": 0.7790324091911316, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 40794 + }, + { + "epoch": 0.40795, + "grad_norm": 0.8313042521476746, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 40795 + }, + { + "epoch": 0.40796, + "grad_norm": 0.8563942313194275, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 40796 + }, + { + "epoch": 0.40797, + "grad_norm": 0.8141913414001465, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 40797 + }, + { + "epoch": 0.40798, + "grad_norm": 0.7680476903915405, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 40798 + }, + { + "epoch": 0.40799, + "grad_norm": 0.7659276723861694, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 40799 + }, + { + "epoch": 0.408, + "grad_norm": 0.8074519038200378, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 40800 + }, + { + "epoch": 0.40801, + "grad_norm": 0.9567832946777344, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 40801 + }, + { + "epoch": 0.40802, + "grad_norm": 1.2067852020263672, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 40802 + }, + { + "epoch": 0.40803, + "grad_norm": 0.8021277189254761, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 40803 + }, + { + "epoch": 0.40804, + "grad_norm": 0.7477496266365051, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 40804 + }, + { + "epoch": 0.40805, + "grad_norm": 0.9465550780296326, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 40805 + }, + { + "epoch": 0.40806, + "grad_norm": 1.0767254829406738, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 40806 + }, + { + "epoch": 0.40807, + "grad_norm": 0.9296196699142456, + "learning_rate": 0.003, + "loss": 3.9222, + "step": 40807 + }, + { + "epoch": 0.40808, + "grad_norm": 0.9159217476844788, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 40808 + }, + { + "epoch": 0.40809, + "grad_norm": 0.9154067039489746, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 40809 + }, + { + "epoch": 0.4081, + "grad_norm": 0.9379838109016418, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 40810 + }, + { + "epoch": 0.40811, + "grad_norm": 0.9782621264457703, + "learning_rate": 0.003, + "loss": 4.012, + "step": 40811 + }, + { + "epoch": 0.40812, + "grad_norm": 0.8636937141418457, + "learning_rate": 0.003, + "loss": 3.9451, + "step": 40812 + }, + { + "epoch": 0.40813, + "grad_norm": 0.6745209097862244, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 40813 + }, + { + "epoch": 0.40814, + "grad_norm": 0.6848140358924866, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 40814 + }, + { + "epoch": 0.40815, + "grad_norm": 0.6359965205192566, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 40815 + }, + { + "epoch": 0.40816, + "grad_norm": 0.789768636226654, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 40816 + }, + { + "epoch": 0.40817, + "grad_norm": 0.7950697541236877, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 40817 + }, + { + "epoch": 0.40818, + "grad_norm": 0.8104300498962402, + "learning_rate": 0.003, + "loss": 4.011, + "step": 40818 + }, + { + "epoch": 0.40819, + "grad_norm": 1.0085222721099854, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 40819 + }, + { + "epoch": 0.4082, + "grad_norm": 0.9716081619262695, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 40820 + }, + { + "epoch": 0.40821, + "grad_norm": 0.8557377457618713, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 40821 + }, + { + "epoch": 0.40822, + "grad_norm": 0.9374205470085144, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 40822 + }, + { + "epoch": 0.40823, + "grad_norm": 0.9950436949729919, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 40823 + }, + { + "epoch": 0.40824, + "grad_norm": 1.1519795656204224, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 40824 + }, + { + "epoch": 0.40825, + "grad_norm": 0.8011178374290466, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 40825 + }, + { + "epoch": 0.40826, + "grad_norm": 0.7916989922523499, + "learning_rate": 0.003, + "loss": 3.995, + "step": 40826 + }, + { + "epoch": 0.40827, + "grad_norm": 0.8753398656845093, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 40827 + }, + { + "epoch": 0.40828, + "grad_norm": 1.0374583005905151, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 40828 + }, + { + "epoch": 0.40829, + "grad_norm": 1.0875568389892578, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 40829 + }, + { + "epoch": 0.4083, + "grad_norm": 0.9556023478507996, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 40830 + }, + { + "epoch": 0.40831, + "grad_norm": 1.2384952306747437, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 40831 + }, + { + "epoch": 0.40832, + "grad_norm": 0.8316524028778076, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 40832 + }, + { + "epoch": 0.40833, + "grad_norm": 0.7606905698776245, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 40833 + }, + { + "epoch": 0.40834, + "grad_norm": 0.7899162173271179, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 40834 + }, + { + "epoch": 0.40835, + "grad_norm": 0.9629352688789368, + "learning_rate": 0.003, + "loss": 3.971, + "step": 40835 + }, + { + "epoch": 0.40836, + "grad_norm": 1.2013258934020996, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 40836 + }, + { + "epoch": 0.40837, + "grad_norm": 0.8586453795433044, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 40837 + }, + { + "epoch": 0.40838, + "grad_norm": 0.8349957466125488, + "learning_rate": 0.003, + "loss": 3.9478, + "step": 40838 + }, + { + "epoch": 0.40839, + "grad_norm": 0.8769649863243103, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 40839 + }, + { + "epoch": 0.4084, + "grad_norm": 0.8904590010643005, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 40840 + }, + { + "epoch": 0.40841, + "grad_norm": 1.0045552253723145, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 40841 + }, + { + "epoch": 0.40842, + "grad_norm": 1.0911177396774292, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 40842 + }, + { + "epoch": 0.40843, + "grad_norm": 0.8820469379425049, + "learning_rate": 0.003, + "loss": 3.961, + "step": 40843 + }, + { + "epoch": 0.40844, + "grad_norm": 0.8882225751876831, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 40844 + }, + { + "epoch": 0.40845, + "grad_norm": 1.0665992498397827, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 40845 + }, + { + "epoch": 0.40846, + "grad_norm": 1.0806416273117065, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 40846 + }, + { + "epoch": 0.40847, + "grad_norm": 0.8570404052734375, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 40847 + }, + { + "epoch": 0.40848, + "grad_norm": 0.6869903206825256, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 40848 + }, + { + "epoch": 0.40849, + "grad_norm": 0.6493359804153442, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 40849 + }, + { + "epoch": 0.4085, + "grad_norm": 0.7526227235794067, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 40850 + }, + { + "epoch": 0.40851, + "grad_norm": 1.0045596361160278, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 40851 + }, + { + "epoch": 0.40852, + "grad_norm": 1.3123384714126587, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 40852 + }, + { + "epoch": 0.40853, + "grad_norm": 0.7678728103637695, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 40853 + }, + { + "epoch": 0.40854, + "grad_norm": 0.6948805451393127, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 40854 + }, + { + "epoch": 0.40855, + "grad_norm": 0.7629196047782898, + "learning_rate": 0.003, + "loss": 3.983, + "step": 40855 + }, + { + "epoch": 0.40856, + "grad_norm": 0.9239124059677124, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 40856 + }, + { + "epoch": 0.40857, + "grad_norm": 1.0183539390563965, + "learning_rate": 0.003, + "loss": 4.007, + "step": 40857 + }, + { + "epoch": 0.40858, + "grad_norm": 1.0354870557785034, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 40858 + }, + { + "epoch": 0.40859, + "grad_norm": 1.0436315536499023, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 40859 + }, + { + "epoch": 0.4086, + "grad_norm": 1.039427638053894, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 40860 + }, + { + "epoch": 0.40861, + "grad_norm": 0.7412423491477966, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 40861 + }, + { + "epoch": 0.40862, + "grad_norm": 0.6156047582626343, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 40862 + }, + { + "epoch": 0.40863, + "grad_norm": 0.6433762311935425, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 40863 + }, + { + "epoch": 0.40864, + "grad_norm": 0.7440754771232605, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 40864 + }, + { + "epoch": 0.40865, + "grad_norm": 0.8317424058914185, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 40865 + }, + { + "epoch": 0.40866, + "grad_norm": 0.8699129819869995, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 40866 + }, + { + "epoch": 0.40867, + "grad_norm": 0.9801725745201111, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 40867 + }, + { + "epoch": 0.40868, + "grad_norm": 1.036360502243042, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 40868 + }, + { + "epoch": 0.40869, + "grad_norm": 1.0564167499542236, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 40869 + }, + { + "epoch": 0.4087, + "grad_norm": 0.939910352230072, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 40870 + }, + { + "epoch": 0.40871, + "grad_norm": 0.8225774765014648, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 40871 + }, + { + "epoch": 0.40872, + "grad_norm": 0.6832706928253174, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 40872 + }, + { + "epoch": 0.40873, + "grad_norm": 0.7122870683670044, + "learning_rate": 0.003, + "loss": 3.979, + "step": 40873 + }, + { + "epoch": 0.40874, + "grad_norm": 0.6858627796173096, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 40874 + }, + { + "epoch": 0.40875, + "grad_norm": 0.8362959623336792, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 40875 + }, + { + "epoch": 0.40876, + "grad_norm": 0.9104514718055725, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 40876 + }, + { + "epoch": 0.40877, + "grad_norm": 1.0251480340957642, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 40877 + }, + { + "epoch": 0.40878, + "grad_norm": 1.2523393630981445, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 40878 + }, + { + "epoch": 0.40879, + "grad_norm": 0.8138148188591003, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 40879 + }, + { + "epoch": 0.4088, + "grad_norm": 0.7899556756019592, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 40880 + }, + { + "epoch": 0.40881, + "grad_norm": 0.8247260451316833, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 40881 + }, + { + "epoch": 0.40882, + "grad_norm": 0.8856601119041443, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 40882 + }, + { + "epoch": 0.40883, + "grad_norm": 0.9919642806053162, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 40883 + }, + { + "epoch": 0.40884, + "grad_norm": 1.127158761024475, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 40884 + }, + { + "epoch": 0.40885, + "grad_norm": 0.9768333435058594, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 40885 + }, + { + "epoch": 0.40886, + "grad_norm": 0.9103394746780396, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 40886 + }, + { + "epoch": 0.40887, + "grad_norm": 1.0047166347503662, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 40887 + }, + { + "epoch": 0.40888, + "grad_norm": 1.1789143085479736, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 40888 + }, + { + "epoch": 0.40889, + "grad_norm": 0.9115755558013916, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 40889 + }, + { + "epoch": 0.4089, + "grad_norm": 0.9318013787269592, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 40890 + }, + { + "epoch": 0.40891, + "grad_norm": 0.8244155645370483, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 40891 + }, + { + "epoch": 0.40892, + "grad_norm": 0.8268870115280151, + "learning_rate": 0.003, + "loss": 3.946, + "step": 40892 + }, + { + "epoch": 0.40893, + "grad_norm": 1.0104153156280518, + "learning_rate": 0.003, + "loss": 3.9397, + "step": 40893 + }, + { + "epoch": 0.40894, + "grad_norm": 1.199453353881836, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 40894 + }, + { + "epoch": 0.40895, + "grad_norm": 0.9178307056427002, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 40895 + }, + { + "epoch": 0.40896, + "grad_norm": 1.0207449197769165, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 40896 + }, + { + "epoch": 0.40897, + "grad_norm": 1.1652621030807495, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 40897 + }, + { + "epoch": 0.40898, + "grad_norm": 0.7499950528144836, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 40898 + }, + { + "epoch": 0.40899, + "grad_norm": 0.70881587266922, + "learning_rate": 0.003, + "loss": 3.969, + "step": 40899 + }, + { + "epoch": 0.409, + "grad_norm": 0.786399245262146, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 40900 + }, + { + "epoch": 0.40901, + "grad_norm": 0.7904636263847351, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 40901 + }, + { + "epoch": 0.40902, + "grad_norm": 0.8236324787139893, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 40902 + }, + { + "epoch": 0.40903, + "grad_norm": 0.8834637403488159, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 40903 + }, + { + "epoch": 0.40904, + "grad_norm": 0.9515606760978699, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 40904 + }, + { + "epoch": 0.40905, + "grad_norm": 1.2621674537658691, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 40905 + }, + { + "epoch": 0.40906, + "grad_norm": 0.7426397800445557, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 40906 + }, + { + "epoch": 0.40907, + "grad_norm": 0.8090494871139526, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 40907 + }, + { + "epoch": 0.40908, + "grad_norm": 0.740097165107727, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 40908 + }, + { + "epoch": 0.40909, + "grad_norm": 0.7349540591239929, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 40909 + }, + { + "epoch": 0.4091, + "grad_norm": 0.7884804010391235, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 40910 + }, + { + "epoch": 0.40911, + "grad_norm": 0.709953248500824, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 40911 + }, + { + "epoch": 0.40912, + "grad_norm": 0.8267479538917542, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 40912 + }, + { + "epoch": 0.40913, + "grad_norm": 1.001508116722107, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 40913 + }, + { + "epoch": 0.40914, + "grad_norm": 1.1120959520339966, + "learning_rate": 0.003, + "loss": 3.952, + "step": 40914 + }, + { + "epoch": 0.40915, + "grad_norm": 0.9742152094841003, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 40915 + }, + { + "epoch": 0.40916, + "grad_norm": 1.1300910711288452, + "learning_rate": 0.003, + "loss": 3.981, + "step": 40916 + }, + { + "epoch": 0.40917, + "grad_norm": 0.9679609537124634, + "learning_rate": 0.003, + "loss": 3.96, + "step": 40917 + }, + { + "epoch": 0.40918, + "grad_norm": 0.9527541995048523, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 40918 + }, + { + "epoch": 0.40919, + "grad_norm": 1.0760650634765625, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 40919 + }, + { + "epoch": 0.4092, + "grad_norm": 1.012467861175537, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 40920 + }, + { + "epoch": 0.40921, + "grad_norm": 1.0822663307189941, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 40921 + }, + { + "epoch": 0.40922, + "grad_norm": 0.8059542179107666, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 40922 + }, + { + "epoch": 0.40923, + "grad_norm": 0.847276508808136, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 40923 + }, + { + "epoch": 0.40924, + "grad_norm": 0.9260768294334412, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 40924 + }, + { + "epoch": 0.40925, + "grad_norm": 1.1033390760421753, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 40925 + }, + { + "epoch": 0.40926, + "grad_norm": 1.0084326267242432, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 40926 + }, + { + "epoch": 0.40927, + "grad_norm": 1.0046119689941406, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 40927 + }, + { + "epoch": 0.40928, + "grad_norm": 0.7132079601287842, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 40928 + }, + { + "epoch": 0.40929, + "grad_norm": 0.5795491933822632, + "learning_rate": 0.003, + "loss": 3.9278, + "step": 40929 + }, + { + "epoch": 0.4093, + "grad_norm": 0.7017325758934021, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 40930 + }, + { + "epoch": 0.40931, + "grad_norm": 0.8203991055488586, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 40931 + }, + { + "epoch": 0.40932, + "grad_norm": 0.7983422875404358, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 40932 + }, + { + "epoch": 0.40933, + "grad_norm": 0.860371470451355, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 40933 + }, + { + "epoch": 0.40934, + "grad_norm": 0.9465568661689758, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 40934 + }, + { + "epoch": 0.40935, + "grad_norm": 0.9633901119232178, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 40935 + }, + { + "epoch": 0.40936, + "grad_norm": 0.9662157893180847, + "learning_rate": 0.003, + "loss": 4.001, + "step": 40936 + }, + { + "epoch": 0.40937, + "grad_norm": 0.9485762715339661, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 40937 + }, + { + "epoch": 0.40938, + "grad_norm": 0.9649510979652405, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 40938 + }, + { + "epoch": 0.40939, + "grad_norm": 1.179437279701233, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 40939 + }, + { + "epoch": 0.4094, + "grad_norm": 0.8942108750343323, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 40940 + }, + { + "epoch": 0.40941, + "grad_norm": 0.9638151526451111, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 40941 + }, + { + "epoch": 0.40942, + "grad_norm": 1.154526948928833, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 40942 + }, + { + "epoch": 0.40943, + "grad_norm": 0.8608864545822144, + "learning_rate": 0.003, + "loss": 3.995, + "step": 40943 + }, + { + "epoch": 0.40944, + "grad_norm": 0.7797127366065979, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 40944 + }, + { + "epoch": 0.40945, + "grad_norm": 0.8702636361122131, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 40945 + }, + { + "epoch": 0.40946, + "grad_norm": 0.8265849351882935, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 40946 + }, + { + "epoch": 0.40947, + "grad_norm": 0.799950122833252, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 40947 + }, + { + "epoch": 0.40948, + "grad_norm": 0.8026185631752014, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 40948 + }, + { + "epoch": 0.40949, + "grad_norm": 0.74305659532547, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 40949 + }, + { + "epoch": 0.4095, + "grad_norm": 0.8802561163902283, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 40950 + }, + { + "epoch": 0.40951, + "grad_norm": 0.9187690019607544, + "learning_rate": 0.003, + "loss": 3.968, + "step": 40951 + }, + { + "epoch": 0.40952, + "grad_norm": 0.9885168671607971, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 40952 + }, + { + "epoch": 0.40953, + "grad_norm": 1.1398817300796509, + "learning_rate": 0.003, + "loss": 3.992, + "step": 40953 + }, + { + "epoch": 0.40954, + "grad_norm": 0.8609106540679932, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 40954 + }, + { + "epoch": 0.40955, + "grad_norm": 0.8211296200752258, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 40955 + }, + { + "epoch": 0.40956, + "grad_norm": 0.7952517867088318, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 40956 + }, + { + "epoch": 0.40957, + "grad_norm": 0.9362047910690308, + "learning_rate": 0.003, + "loss": 3.987, + "step": 40957 + }, + { + "epoch": 0.40958, + "grad_norm": 1.1679637432098389, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 40958 + }, + { + "epoch": 0.40959, + "grad_norm": 1.2036867141723633, + "learning_rate": 0.003, + "loss": 3.964, + "step": 40959 + }, + { + "epoch": 0.4096, + "grad_norm": 0.9740427732467651, + "learning_rate": 0.003, + "loss": 3.985, + "step": 40960 + }, + { + "epoch": 0.40961, + "grad_norm": 0.9731001257896423, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 40961 + }, + { + "epoch": 0.40962, + "grad_norm": 0.9480272531509399, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 40962 + }, + { + "epoch": 0.40963, + "grad_norm": 0.783382773399353, + "learning_rate": 0.003, + "loss": 3.993, + "step": 40963 + }, + { + "epoch": 0.40964, + "grad_norm": 0.7426978349685669, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 40964 + }, + { + "epoch": 0.40965, + "grad_norm": 0.8028404116630554, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 40965 + }, + { + "epoch": 0.40966, + "grad_norm": 0.764695405960083, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 40966 + }, + { + "epoch": 0.40967, + "grad_norm": 0.7150115966796875, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 40967 + }, + { + "epoch": 0.40968, + "grad_norm": 0.823394775390625, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 40968 + }, + { + "epoch": 0.40969, + "grad_norm": 0.9625790119171143, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 40969 + }, + { + "epoch": 0.4097, + "grad_norm": 1.23568594455719, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 40970 + }, + { + "epoch": 0.40971, + "grad_norm": 0.9657372832298279, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 40971 + }, + { + "epoch": 0.40972, + "grad_norm": 0.8318502306938171, + "learning_rate": 0.003, + "loss": 3.986, + "step": 40972 + }, + { + "epoch": 0.40973, + "grad_norm": 0.7347708940505981, + "learning_rate": 0.003, + "loss": 3.984, + "step": 40973 + }, + { + "epoch": 0.40974, + "grad_norm": 0.7501605749130249, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 40974 + }, + { + "epoch": 0.40975, + "grad_norm": 0.7222383618354797, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 40975 + }, + { + "epoch": 0.40976, + "grad_norm": 0.7152555584907532, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 40976 + }, + { + "epoch": 0.40977, + "grad_norm": 0.7773206233978271, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 40977 + }, + { + "epoch": 0.40978, + "grad_norm": 0.7563687562942505, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 40978 + }, + { + "epoch": 0.40979, + "grad_norm": 0.6967066526412964, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 40979 + }, + { + "epoch": 0.4098, + "grad_norm": 0.7805966138839722, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 40980 + }, + { + "epoch": 0.40981, + "grad_norm": 1.0098085403442383, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 40981 + }, + { + "epoch": 0.40982, + "grad_norm": 1.2490935325622559, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 40982 + }, + { + "epoch": 0.40983, + "grad_norm": 0.8167494535446167, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 40983 + }, + { + "epoch": 0.40984, + "grad_norm": 0.7285228371620178, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 40984 + }, + { + "epoch": 0.40985, + "grad_norm": 0.7961179614067078, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 40985 + }, + { + "epoch": 0.40986, + "grad_norm": 0.9165686964988708, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 40986 + }, + { + "epoch": 0.40987, + "grad_norm": 1.0356218814849854, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 40987 + }, + { + "epoch": 0.40988, + "grad_norm": 1.1166633367538452, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 40988 + }, + { + "epoch": 0.40989, + "grad_norm": 0.8471337556838989, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 40989 + }, + { + "epoch": 0.4099, + "grad_norm": 0.7945396304130554, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 40990 + }, + { + "epoch": 0.40991, + "grad_norm": 0.7985069155693054, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 40991 + }, + { + "epoch": 0.40992, + "grad_norm": 0.76108717918396, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 40992 + }, + { + "epoch": 0.40993, + "grad_norm": 0.7510248422622681, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 40993 + }, + { + "epoch": 0.40994, + "grad_norm": 1.0964752435684204, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 40994 + }, + { + "epoch": 0.40995, + "grad_norm": 1.0831420421600342, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 40995 + }, + { + "epoch": 0.40996, + "grad_norm": 0.8908584117889404, + "learning_rate": 0.003, + "loss": 3.9459, + "step": 40996 + }, + { + "epoch": 0.40997, + "grad_norm": 1.0539026260375977, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 40997 + }, + { + "epoch": 0.40998, + "grad_norm": 1.2535185813903809, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 40998 + }, + { + "epoch": 0.40999, + "grad_norm": 0.5906004309654236, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 40999 + }, + { + "epoch": 0.41, + "grad_norm": 0.7677070498466492, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 41000 + }, + { + "epoch": 0.41001, + "grad_norm": 1.0350778102874756, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 41001 + }, + { + "epoch": 0.41002, + "grad_norm": 1.1204155683517456, + "learning_rate": 0.003, + "loss": 3.9399, + "step": 41002 + }, + { + "epoch": 0.41003, + "grad_norm": 0.86000657081604, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 41003 + }, + { + "epoch": 0.41004, + "grad_norm": 0.7019170522689819, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 41004 + }, + { + "epoch": 0.41005, + "grad_norm": 0.6616955399513245, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 41005 + }, + { + "epoch": 0.41006, + "grad_norm": 0.7530520558357239, + "learning_rate": 0.003, + "loss": 3.953, + "step": 41006 + }, + { + "epoch": 0.41007, + "grad_norm": 0.927556574344635, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 41007 + }, + { + "epoch": 0.41008, + "grad_norm": 0.9524747133255005, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 41008 + }, + { + "epoch": 0.41009, + "grad_norm": 0.9121966361999512, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 41009 + }, + { + "epoch": 0.4101, + "grad_norm": 0.9673008322715759, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 41010 + }, + { + "epoch": 0.41011, + "grad_norm": 0.9802336096763611, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 41011 + }, + { + "epoch": 0.41012, + "grad_norm": 0.9950745701789856, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 41012 + }, + { + "epoch": 0.41013, + "grad_norm": 0.9271800518035889, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 41013 + }, + { + "epoch": 0.41014, + "grad_norm": 0.9279474020004272, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 41014 + }, + { + "epoch": 0.41015, + "grad_norm": 1.029233455657959, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 41015 + }, + { + "epoch": 0.41016, + "grad_norm": 1.2986161708831787, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 41016 + }, + { + "epoch": 0.41017, + "grad_norm": 0.8768953084945679, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 41017 + }, + { + "epoch": 0.41018, + "grad_norm": 0.8493993282318115, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 41018 + }, + { + "epoch": 0.41019, + "grad_norm": 0.9512099623680115, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 41019 + }, + { + "epoch": 0.4102, + "grad_norm": 1.1506476402282715, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 41020 + }, + { + "epoch": 0.41021, + "grad_norm": 0.9206907749176025, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 41021 + }, + { + "epoch": 0.41022, + "grad_norm": 0.8127718567848206, + "learning_rate": 0.003, + "loss": 3.991, + "step": 41022 + }, + { + "epoch": 0.41023, + "grad_norm": 0.8436741828918457, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 41023 + }, + { + "epoch": 0.41024, + "grad_norm": 0.8165512681007385, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 41024 + }, + { + "epoch": 0.41025, + "grad_norm": 0.8907931447029114, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 41025 + }, + { + "epoch": 0.41026, + "grad_norm": 0.9810680747032166, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 41026 + }, + { + "epoch": 0.41027, + "grad_norm": 1.1285746097564697, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 41027 + }, + { + "epoch": 0.41028, + "grad_norm": 1.064990520477295, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 41028 + }, + { + "epoch": 0.41029, + "grad_norm": 1.1045668125152588, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 41029 + }, + { + "epoch": 0.4103, + "grad_norm": 0.8883934020996094, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 41030 + }, + { + "epoch": 0.41031, + "grad_norm": 0.916499137878418, + "learning_rate": 0.003, + "loss": 3.985, + "step": 41031 + }, + { + "epoch": 0.41032, + "grad_norm": 0.9281208515167236, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 41032 + }, + { + "epoch": 0.41033, + "grad_norm": 0.771556556224823, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 41033 + }, + { + "epoch": 0.41034, + "grad_norm": 0.6876526474952698, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 41034 + }, + { + "epoch": 0.41035, + "grad_norm": 0.7795351147651672, + "learning_rate": 0.003, + "loss": 4.005, + "step": 41035 + }, + { + "epoch": 0.41036, + "grad_norm": 1.0421662330627441, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 41036 + }, + { + "epoch": 0.41037, + "grad_norm": 1.1203205585479736, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 41037 + }, + { + "epoch": 0.41038, + "grad_norm": 0.948587954044342, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 41038 + }, + { + "epoch": 0.41039, + "grad_norm": 0.9405484199523926, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 41039 + }, + { + "epoch": 0.4104, + "grad_norm": 1.004388451576233, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 41040 + }, + { + "epoch": 0.41041, + "grad_norm": 1.030547857284546, + "learning_rate": 0.003, + "loss": 3.966, + "step": 41041 + }, + { + "epoch": 0.41042, + "grad_norm": 0.9344191551208496, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 41042 + }, + { + "epoch": 0.41043, + "grad_norm": 1.015516757965088, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 41043 + }, + { + "epoch": 0.41044, + "grad_norm": 0.9386818408966064, + "learning_rate": 0.003, + "loss": 3.957, + "step": 41044 + }, + { + "epoch": 0.41045, + "grad_norm": 0.983392596244812, + "learning_rate": 0.003, + "loss": 3.99, + "step": 41045 + }, + { + "epoch": 0.41046, + "grad_norm": 1.3320902585983276, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 41046 + }, + { + "epoch": 0.41047, + "grad_norm": 0.6506377458572388, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 41047 + }, + { + "epoch": 0.41048, + "grad_norm": 0.7939757704734802, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 41048 + }, + { + "epoch": 0.41049, + "grad_norm": 0.9418172836303711, + "learning_rate": 0.003, + "loss": 3.967, + "step": 41049 + }, + { + "epoch": 0.4105, + "grad_norm": 1.0498263835906982, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 41050 + }, + { + "epoch": 0.41051, + "grad_norm": 1.0197076797485352, + "learning_rate": 0.003, + "loss": 3.94, + "step": 41051 + }, + { + "epoch": 0.41052, + "grad_norm": 0.7907038927078247, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 41052 + }, + { + "epoch": 0.41053, + "grad_norm": 0.6310120820999146, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 41053 + }, + { + "epoch": 0.41054, + "grad_norm": 0.5357524752616882, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 41054 + }, + { + "epoch": 0.41055, + "grad_norm": 0.578028678894043, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 41055 + }, + { + "epoch": 0.41056, + "grad_norm": 0.5783238410949707, + "learning_rate": 0.003, + "loss": 3.9434, + "step": 41056 + }, + { + "epoch": 0.41057, + "grad_norm": 0.5835348963737488, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 41057 + }, + { + "epoch": 0.41058, + "grad_norm": 0.6169410347938538, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 41058 + }, + { + "epoch": 0.41059, + "grad_norm": 0.628527820110321, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 41059 + }, + { + "epoch": 0.4106, + "grad_norm": 0.9000558257102966, + "learning_rate": 0.003, + "loss": 3.9377, + "step": 41060 + }, + { + "epoch": 0.41061, + "grad_norm": 1.3379788398742676, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 41061 + }, + { + "epoch": 0.41062, + "grad_norm": 0.6064035892486572, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 41062 + }, + { + "epoch": 0.41063, + "grad_norm": 0.6273430585861206, + "learning_rate": 0.003, + "loss": 3.9442, + "step": 41063 + }, + { + "epoch": 0.41064, + "grad_norm": 0.9445578455924988, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 41064 + }, + { + "epoch": 0.41065, + "grad_norm": 1.2995905876159668, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 41065 + }, + { + "epoch": 0.41066, + "grad_norm": 0.7478456497192383, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 41066 + }, + { + "epoch": 0.41067, + "grad_norm": 0.6672636270523071, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 41067 + }, + { + "epoch": 0.41068, + "grad_norm": 0.6917980313301086, + "learning_rate": 0.003, + "loss": 3.989, + "step": 41068 + }, + { + "epoch": 0.41069, + "grad_norm": 0.7708619236946106, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 41069 + }, + { + "epoch": 0.4107, + "grad_norm": 0.93669193983078, + "learning_rate": 0.003, + "loss": 3.96, + "step": 41070 + }, + { + "epoch": 0.41071, + "grad_norm": 1.1146413087844849, + "learning_rate": 0.003, + "loss": 3.989, + "step": 41071 + }, + { + "epoch": 0.41072, + "grad_norm": 0.7844797968864441, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 41072 + }, + { + "epoch": 0.41073, + "grad_norm": 0.797376811504364, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 41073 + }, + { + "epoch": 0.41074, + "grad_norm": 0.799288272857666, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 41074 + }, + { + "epoch": 0.41075, + "grad_norm": 0.7864819765090942, + "learning_rate": 0.003, + "loss": 3.9458, + "step": 41075 + }, + { + "epoch": 0.41076, + "grad_norm": 0.8108957409858704, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 41076 + }, + { + "epoch": 0.41077, + "grad_norm": 0.9158667922019958, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 41077 + }, + { + "epoch": 0.41078, + "grad_norm": 1.3343125581741333, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 41078 + }, + { + "epoch": 0.41079, + "grad_norm": 0.8417260050773621, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 41079 + }, + { + "epoch": 0.4108, + "grad_norm": 0.8803749084472656, + "learning_rate": 0.003, + "loss": 3.969, + "step": 41080 + }, + { + "epoch": 0.41081, + "grad_norm": 1.1219199895858765, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 41081 + }, + { + "epoch": 0.41082, + "grad_norm": 0.9759713411331177, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 41082 + }, + { + "epoch": 0.41083, + "grad_norm": 1.1783573627471924, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 41083 + }, + { + "epoch": 0.41084, + "grad_norm": 1.0381158590316772, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 41084 + }, + { + "epoch": 0.41085, + "grad_norm": 1.11492121219635, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 41085 + }, + { + "epoch": 0.41086, + "grad_norm": 0.9815148115158081, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 41086 + }, + { + "epoch": 0.41087, + "grad_norm": 1.055314302444458, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 41087 + }, + { + "epoch": 0.41088, + "grad_norm": 1.0857040882110596, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 41088 + }, + { + "epoch": 0.41089, + "grad_norm": 1.0090471506118774, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 41089 + }, + { + "epoch": 0.4109, + "grad_norm": 1.1143465042114258, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 41090 + }, + { + "epoch": 0.41091, + "grad_norm": 1.163328766822815, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 41091 + }, + { + "epoch": 0.41092, + "grad_norm": 1.153816819190979, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 41092 + }, + { + "epoch": 0.41093, + "grad_norm": 0.962293267250061, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 41093 + }, + { + "epoch": 0.41094, + "grad_norm": 1.017983078956604, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 41094 + }, + { + "epoch": 0.41095, + "grad_norm": 1.3104294538497925, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 41095 + }, + { + "epoch": 0.41096, + "grad_norm": 0.8975563049316406, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 41096 + }, + { + "epoch": 0.41097, + "grad_norm": 0.8122672438621521, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 41097 + }, + { + "epoch": 0.41098, + "grad_norm": 0.9327647686004639, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 41098 + }, + { + "epoch": 0.41099, + "grad_norm": 1.04341721534729, + "learning_rate": 0.003, + "loss": 3.989, + "step": 41099 + }, + { + "epoch": 0.411, + "grad_norm": 0.9932865500450134, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 41100 + }, + { + "epoch": 0.41101, + "grad_norm": 0.8977075815200806, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 41101 + }, + { + "epoch": 0.41102, + "grad_norm": 0.8448808193206787, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 41102 + }, + { + "epoch": 0.41103, + "grad_norm": 0.8664383888244629, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 41103 + }, + { + "epoch": 0.41104, + "grad_norm": 0.980309009552002, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 41104 + }, + { + "epoch": 0.41105, + "grad_norm": 1.145534873008728, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 41105 + }, + { + "epoch": 0.41106, + "grad_norm": 0.8900923132896423, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 41106 + }, + { + "epoch": 0.41107, + "grad_norm": 1.0380444526672363, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 41107 + }, + { + "epoch": 0.41108, + "grad_norm": 1.1508980989456177, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 41108 + }, + { + "epoch": 0.41109, + "grad_norm": 0.9052751660346985, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 41109 + }, + { + "epoch": 0.4111, + "grad_norm": 1.1173584461212158, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 41110 + }, + { + "epoch": 0.41111, + "grad_norm": 1.247944712638855, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 41111 + }, + { + "epoch": 0.41112, + "grad_norm": 0.8205608129501343, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 41112 + }, + { + "epoch": 0.41113, + "grad_norm": 0.6765358448028564, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 41113 + }, + { + "epoch": 0.41114, + "grad_norm": 0.7063273787498474, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 41114 + }, + { + "epoch": 0.41115, + "grad_norm": 0.7972437739372253, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 41115 + }, + { + "epoch": 0.41116, + "grad_norm": 0.8736926317214966, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 41116 + }, + { + "epoch": 0.41117, + "grad_norm": 1.0099862813949585, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 41117 + }, + { + "epoch": 0.41118, + "grad_norm": 1.039422869682312, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 41118 + }, + { + "epoch": 0.41119, + "grad_norm": 0.9393274784088135, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 41119 + }, + { + "epoch": 0.4112, + "grad_norm": 0.9710643291473389, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 41120 + }, + { + "epoch": 0.41121, + "grad_norm": 1.2211799621582031, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 41121 + }, + { + "epoch": 0.41122, + "grad_norm": 0.8242526054382324, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 41122 + }, + { + "epoch": 0.41123, + "grad_norm": 0.8389295339584351, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 41123 + }, + { + "epoch": 0.41124, + "grad_norm": 0.8703290820121765, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 41124 + }, + { + "epoch": 0.41125, + "grad_norm": 0.9874466061592102, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 41125 + }, + { + "epoch": 0.41126, + "grad_norm": 1.0958625078201294, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 41126 + }, + { + "epoch": 0.41127, + "grad_norm": 0.8221285939216614, + "learning_rate": 0.003, + "loss": 4.004, + "step": 41127 + }, + { + "epoch": 0.41128, + "grad_norm": 0.6879100203514099, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 41128 + }, + { + "epoch": 0.41129, + "grad_norm": 0.628807544708252, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 41129 + }, + { + "epoch": 0.4113, + "grad_norm": 0.5563502311706543, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 41130 + }, + { + "epoch": 0.41131, + "grad_norm": 0.5559758543968201, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 41131 + }, + { + "epoch": 0.41132, + "grad_norm": 0.6768811941146851, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 41132 + }, + { + "epoch": 0.41133, + "grad_norm": 0.9010196924209595, + "learning_rate": 0.003, + "loss": 3.9431, + "step": 41133 + }, + { + "epoch": 0.41134, + "grad_norm": 1.1422616243362427, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 41134 + }, + { + "epoch": 0.41135, + "grad_norm": 1.0706051588058472, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 41135 + }, + { + "epoch": 0.41136, + "grad_norm": 1.1235431432724, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 41136 + }, + { + "epoch": 0.41137, + "grad_norm": 0.8864747881889343, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 41137 + }, + { + "epoch": 0.41138, + "grad_norm": 0.5609537363052368, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 41138 + }, + { + "epoch": 0.41139, + "grad_norm": 0.7350475192070007, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 41139 + }, + { + "epoch": 0.4114, + "grad_norm": 0.8697900772094727, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 41140 + }, + { + "epoch": 0.41141, + "grad_norm": 0.999229907989502, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 41141 + }, + { + "epoch": 0.41142, + "grad_norm": 1.1721735000610352, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 41142 + }, + { + "epoch": 0.41143, + "grad_norm": 0.8820127844810486, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 41143 + }, + { + "epoch": 0.41144, + "grad_norm": 0.8752126097679138, + "learning_rate": 0.003, + "loss": 3.9379, + "step": 41144 + }, + { + "epoch": 0.41145, + "grad_norm": 0.8261533975601196, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 41145 + }, + { + "epoch": 0.41146, + "grad_norm": 0.6669045090675354, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 41146 + }, + { + "epoch": 0.41147, + "grad_norm": 0.7135475873947144, + "learning_rate": 0.003, + "loss": 3.987, + "step": 41147 + }, + { + "epoch": 0.41148, + "grad_norm": 0.78183513879776, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 41148 + }, + { + "epoch": 0.41149, + "grad_norm": 0.7535926103591919, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 41149 + }, + { + "epoch": 0.4115, + "grad_norm": 0.8986193537712097, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 41150 + }, + { + "epoch": 0.41151, + "grad_norm": 1.1912401914596558, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 41151 + }, + { + "epoch": 0.41152, + "grad_norm": 1.0614322423934937, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 41152 + }, + { + "epoch": 0.41153, + "grad_norm": 0.8848280310630798, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 41153 + }, + { + "epoch": 0.41154, + "grad_norm": 1.0173283815383911, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 41154 + }, + { + "epoch": 0.41155, + "grad_norm": 0.929611325263977, + "learning_rate": 0.003, + "loss": 3.9438, + "step": 41155 + }, + { + "epoch": 0.41156, + "grad_norm": 0.8661308288574219, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 41156 + }, + { + "epoch": 0.41157, + "grad_norm": 1.0551953315734863, + "learning_rate": 0.003, + "loss": 4.006, + "step": 41157 + }, + { + "epoch": 0.41158, + "grad_norm": 1.495051383972168, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 41158 + }, + { + "epoch": 0.41159, + "grad_norm": 0.9146769046783447, + "learning_rate": 0.003, + "loss": 4.005, + "step": 41159 + }, + { + "epoch": 0.4116, + "grad_norm": 0.7872540354728699, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 41160 + }, + { + "epoch": 0.41161, + "grad_norm": 0.7123496532440186, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 41161 + }, + { + "epoch": 0.41162, + "grad_norm": 0.6748815774917603, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 41162 + }, + { + "epoch": 0.41163, + "grad_norm": 0.8522852659225464, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 41163 + }, + { + "epoch": 0.41164, + "grad_norm": 0.9836320877075195, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 41164 + }, + { + "epoch": 0.41165, + "grad_norm": 1.1360325813293457, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 41165 + }, + { + "epoch": 0.41166, + "grad_norm": 0.8793897032737732, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 41166 + }, + { + "epoch": 0.41167, + "grad_norm": 0.8827987909317017, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 41167 + }, + { + "epoch": 0.41168, + "grad_norm": 0.9457340240478516, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 41168 + }, + { + "epoch": 0.41169, + "grad_norm": 0.9730079770088196, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 41169 + }, + { + "epoch": 0.4117, + "grad_norm": 0.9816914796829224, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 41170 + }, + { + "epoch": 0.41171, + "grad_norm": 1.008500337600708, + "learning_rate": 0.003, + "loss": 3.95, + "step": 41171 + }, + { + "epoch": 0.41172, + "grad_norm": 1.0347192287445068, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 41172 + }, + { + "epoch": 0.41173, + "grad_norm": 0.89842689037323, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 41173 + }, + { + "epoch": 0.41174, + "grad_norm": 0.9469254016876221, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 41174 + }, + { + "epoch": 0.41175, + "grad_norm": 1.149324893951416, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 41175 + }, + { + "epoch": 0.41176, + "grad_norm": 1.0429409742355347, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 41176 + }, + { + "epoch": 0.41177, + "grad_norm": 0.8470731973648071, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 41177 + }, + { + "epoch": 0.41178, + "grad_norm": 0.789577305316925, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 41178 + }, + { + "epoch": 0.41179, + "grad_norm": 0.8535863161087036, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 41179 + }, + { + "epoch": 0.4118, + "grad_norm": 1.0148999691009521, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 41180 + }, + { + "epoch": 0.41181, + "grad_norm": 1.013362169265747, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 41181 + }, + { + "epoch": 0.41182, + "grad_norm": 0.7987837195396423, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 41182 + }, + { + "epoch": 0.41183, + "grad_norm": 0.7590625882148743, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 41183 + }, + { + "epoch": 0.41184, + "grad_norm": 0.6750516295433044, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 41184 + }, + { + "epoch": 0.41185, + "grad_norm": 0.6140913367271423, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 41185 + }, + { + "epoch": 0.41186, + "grad_norm": 0.5938458442687988, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 41186 + }, + { + "epoch": 0.41187, + "grad_norm": 0.6752866506576538, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 41187 + }, + { + "epoch": 0.41188, + "grad_norm": 0.6822843551635742, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 41188 + }, + { + "epoch": 0.41189, + "grad_norm": 0.7658665180206299, + "learning_rate": 0.003, + "loss": 3.997, + "step": 41189 + }, + { + "epoch": 0.4119, + "grad_norm": 0.8707765340805054, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 41190 + }, + { + "epoch": 0.41191, + "grad_norm": 0.9419683218002319, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 41191 + }, + { + "epoch": 0.41192, + "grad_norm": 1.0786751508712769, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 41192 + }, + { + "epoch": 0.41193, + "grad_norm": 1.0110807418823242, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 41193 + }, + { + "epoch": 0.41194, + "grad_norm": 0.9862281084060669, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 41194 + }, + { + "epoch": 0.41195, + "grad_norm": 1.1171320676803589, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 41195 + }, + { + "epoch": 0.41196, + "grad_norm": 0.9377790093421936, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 41196 + }, + { + "epoch": 0.41197, + "grad_norm": 0.8717906475067139, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 41197 + }, + { + "epoch": 0.41198, + "grad_norm": 0.839555025100708, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 41198 + }, + { + "epoch": 0.41199, + "grad_norm": 1.1131622791290283, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 41199 + }, + { + "epoch": 0.412, + "grad_norm": 0.9518706798553467, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 41200 + }, + { + "epoch": 0.41201, + "grad_norm": 0.9351011514663696, + "learning_rate": 0.003, + "loss": 3.994, + "step": 41201 + }, + { + "epoch": 0.41202, + "grad_norm": 0.9827096462249756, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 41202 + }, + { + "epoch": 0.41203, + "grad_norm": 1.074039340019226, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 41203 + }, + { + "epoch": 0.41204, + "grad_norm": 0.789680004119873, + "learning_rate": 0.003, + "loss": 3.978, + "step": 41204 + }, + { + "epoch": 0.41205, + "grad_norm": 0.7798881530761719, + "learning_rate": 0.003, + "loss": 3.94, + "step": 41205 + }, + { + "epoch": 0.41206, + "grad_norm": 0.9540780186653137, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 41206 + }, + { + "epoch": 0.41207, + "grad_norm": 1.1977087259292603, + "learning_rate": 0.003, + "loss": 4.007, + "step": 41207 + }, + { + "epoch": 0.41208, + "grad_norm": 1.022961974143982, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 41208 + }, + { + "epoch": 0.41209, + "grad_norm": 1.4040374755859375, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 41209 + }, + { + "epoch": 0.4121, + "grad_norm": 0.5737341642379761, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 41210 + }, + { + "epoch": 0.41211, + "grad_norm": 0.7975238561630249, + "learning_rate": 0.003, + "loss": 3.9407, + "step": 41211 + }, + { + "epoch": 0.41212, + "grad_norm": 1.3096200227737427, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 41212 + }, + { + "epoch": 0.41213, + "grad_norm": 0.7294857501983643, + "learning_rate": 0.003, + "loss": 3.956, + "step": 41213 + }, + { + "epoch": 0.41214, + "grad_norm": 0.7035763263702393, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 41214 + }, + { + "epoch": 0.41215, + "grad_norm": 0.9778889417648315, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 41215 + }, + { + "epoch": 0.41216, + "grad_norm": 1.148123025894165, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 41216 + }, + { + "epoch": 0.41217, + "grad_norm": 0.7110883593559265, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 41217 + }, + { + "epoch": 0.41218, + "grad_norm": 0.767838716506958, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 41218 + }, + { + "epoch": 0.41219, + "grad_norm": 0.8654146790504456, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 41219 + }, + { + "epoch": 0.4122, + "grad_norm": 1.0761165618896484, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 41220 + }, + { + "epoch": 0.41221, + "grad_norm": 1.2309768199920654, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 41221 + }, + { + "epoch": 0.41222, + "grad_norm": 0.6474784016609192, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 41222 + }, + { + "epoch": 0.41223, + "grad_norm": 0.6681444644927979, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 41223 + }, + { + "epoch": 0.41224, + "grad_norm": 0.9039144515991211, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 41224 + }, + { + "epoch": 0.41225, + "grad_norm": 1.083840250968933, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 41225 + }, + { + "epoch": 0.41226, + "grad_norm": 0.9629534482955933, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 41226 + }, + { + "epoch": 0.41227, + "grad_norm": 0.7553152441978455, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 41227 + }, + { + "epoch": 0.41228, + "grad_norm": 0.5986404418945312, + "learning_rate": 0.003, + "loss": 3.976, + "step": 41228 + }, + { + "epoch": 0.41229, + "grad_norm": 0.7322625517845154, + "learning_rate": 0.003, + "loss": 3.994, + "step": 41229 + }, + { + "epoch": 0.4123, + "grad_norm": 0.8027428984642029, + "learning_rate": 0.003, + "loss": 3.956, + "step": 41230 + }, + { + "epoch": 0.41231, + "grad_norm": 0.7786287069320679, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 41231 + }, + { + "epoch": 0.41232, + "grad_norm": 0.7762612700462341, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 41232 + }, + { + "epoch": 0.41233, + "grad_norm": 0.8646380305290222, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 41233 + }, + { + "epoch": 0.41234, + "grad_norm": 0.8564216494560242, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 41234 + }, + { + "epoch": 0.41235, + "grad_norm": 0.792558491230011, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 41235 + }, + { + "epoch": 0.41236, + "grad_norm": 0.7459580898284912, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 41236 + }, + { + "epoch": 0.41237, + "grad_norm": 0.7958836555480957, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 41237 + }, + { + "epoch": 0.41238, + "grad_norm": 0.9576721787452698, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 41238 + }, + { + "epoch": 0.41239, + "grad_norm": 1.5110573768615723, + "learning_rate": 0.003, + "loss": 3.97, + "step": 41239 + }, + { + "epoch": 0.4124, + "grad_norm": 0.4812391698360443, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 41240 + }, + { + "epoch": 0.41241, + "grad_norm": 0.9702315330505371, + "learning_rate": 0.003, + "loss": 3.97, + "step": 41241 + }, + { + "epoch": 0.41242, + "grad_norm": 1.5433423519134521, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 41242 + }, + { + "epoch": 0.41243, + "grad_norm": 0.5733510255813599, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 41243 + }, + { + "epoch": 0.41244, + "grad_norm": 0.7870966196060181, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 41244 + }, + { + "epoch": 0.41245, + "grad_norm": 0.8211700916290283, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 41245 + }, + { + "epoch": 0.41246, + "grad_norm": 0.8374504446983337, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 41246 + }, + { + "epoch": 0.41247, + "grad_norm": 0.9674665331840515, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 41247 + }, + { + "epoch": 0.41248, + "grad_norm": 0.9995651245117188, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 41248 + }, + { + "epoch": 0.41249, + "grad_norm": 1.1207813024520874, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 41249 + }, + { + "epoch": 0.4125, + "grad_norm": 1.1006609201431274, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 41250 + }, + { + "epoch": 0.41251, + "grad_norm": 0.7733182907104492, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 41251 + }, + { + "epoch": 0.41252, + "grad_norm": 0.8031817078590393, + "learning_rate": 0.003, + "loss": 3.978, + "step": 41252 + }, + { + "epoch": 0.41253, + "grad_norm": 0.8541613221168518, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 41253 + }, + { + "epoch": 0.41254, + "grad_norm": 0.9947472810745239, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 41254 + }, + { + "epoch": 0.41255, + "grad_norm": 1.062720537185669, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 41255 + }, + { + "epoch": 0.41256, + "grad_norm": 0.8616950511932373, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 41256 + }, + { + "epoch": 0.41257, + "grad_norm": 0.8281533718109131, + "learning_rate": 0.003, + "loss": 3.9465, + "step": 41257 + }, + { + "epoch": 0.41258, + "grad_norm": 0.8191540837287903, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 41258 + }, + { + "epoch": 0.41259, + "grad_norm": 1.0651952028274536, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 41259 + }, + { + "epoch": 0.4126, + "grad_norm": 1.0906858444213867, + "learning_rate": 0.003, + "loss": 3.982, + "step": 41260 + }, + { + "epoch": 0.41261, + "grad_norm": 0.915855884552002, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 41261 + }, + { + "epoch": 0.41262, + "grad_norm": 0.8238592147827148, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 41262 + }, + { + "epoch": 0.41263, + "grad_norm": 0.8416628837585449, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 41263 + }, + { + "epoch": 0.41264, + "grad_norm": 0.9615981578826904, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 41264 + }, + { + "epoch": 0.41265, + "grad_norm": 1.1905570030212402, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 41265 + }, + { + "epoch": 0.41266, + "grad_norm": 0.8771082162857056, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 41266 + }, + { + "epoch": 0.41267, + "grad_norm": 0.9114190936088562, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 41267 + }, + { + "epoch": 0.41268, + "grad_norm": 0.8746061325073242, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 41268 + }, + { + "epoch": 0.41269, + "grad_norm": 0.8482958674430847, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 41269 + }, + { + "epoch": 0.4127, + "grad_norm": 0.9962107539176941, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 41270 + }, + { + "epoch": 0.41271, + "grad_norm": 1.122338891029358, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 41271 + }, + { + "epoch": 0.41272, + "grad_norm": 0.81988525390625, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 41272 + }, + { + "epoch": 0.41273, + "grad_norm": 0.744938850402832, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 41273 + }, + { + "epoch": 0.41274, + "grad_norm": 0.8078799247741699, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 41274 + }, + { + "epoch": 0.41275, + "grad_norm": 0.909911036491394, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 41275 + }, + { + "epoch": 0.41276, + "grad_norm": 1.0167089700698853, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 41276 + }, + { + "epoch": 0.41277, + "grad_norm": 0.9696038961410522, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 41277 + }, + { + "epoch": 0.41278, + "grad_norm": 0.8540838956832886, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 41278 + }, + { + "epoch": 0.41279, + "grad_norm": 0.8563671708106995, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 41279 + }, + { + "epoch": 0.4128, + "grad_norm": 0.9517725110054016, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 41280 + }, + { + "epoch": 0.41281, + "grad_norm": 1.2616362571716309, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 41281 + }, + { + "epoch": 0.41282, + "grad_norm": 0.919472336769104, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 41282 + }, + { + "epoch": 0.41283, + "grad_norm": 0.9346612095832825, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 41283 + }, + { + "epoch": 0.41284, + "grad_norm": 0.9987502694129944, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 41284 + }, + { + "epoch": 0.41285, + "grad_norm": 0.9873175024986267, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 41285 + }, + { + "epoch": 0.41286, + "grad_norm": 0.9692069292068481, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 41286 + }, + { + "epoch": 0.41287, + "grad_norm": 1.129156231880188, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 41287 + }, + { + "epoch": 0.41288, + "grad_norm": 0.8904610276222229, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 41288 + }, + { + "epoch": 0.41289, + "grad_norm": 0.7899903059005737, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 41289 + }, + { + "epoch": 0.4129, + "grad_norm": 0.6676017642021179, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 41290 + }, + { + "epoch": 0.41291, + "grad_norm": 0.6912239193916321, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 41291 + }, + { + "epoch": 0.41292, + "grad_norm": 0.885860025882721, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 41292 + }, + { + "epoch": 0.41293, + "grad_norm": 1.0374398231506348, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 41293 + }, + { + "epoch": 0.41294, + "grad_norm": 1.0804933309555054, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 41294 + }, + { + "epoch": 0.41295, + "grad_norm": 0.8052554130554199, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 41295 + }, + { + "epoch": 0.41296, + "grad_norm": 0.6908950805664062, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 41296 + }, + { + "epoch": 0.41297, + "grad_norm": 0.7351971864700317, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 41297 + }, + { + "epoch": 0.41298, + "grad_norm": 0.7216674089431763, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 41298 + }, + { + "epoch": 0.41299, + "grad_norm": 0.6545940637588501, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 41299 + }, + { + "epoch": 0.413, + "grad_norm": 0.6149832010269165, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 41300 + }, + { + "epoch": 0.41301, + "grad_norm": 0.6930363774299622, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 41301 + }, + { + "epoch": 0.41302, + "grad_norm": 0.8770902156829834, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 41302 + }, + { + "epoch": 0.41303, + "grad_norm": 0.8959327340126038, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 41303 + }, + { + "epoch": 0.41304, + "grad_norm": 0.9942192435264587, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 41304 + }, + { + "epoch": 0.41305, + "grad_norm": 1.2125627994537354, + "learning_rate": 0.003, + "loss": 3.975, + "step": 41305 + }, + { + "epoch": 0.41306, + "grad_norm": 0.6157641410827637, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 41306 + }, + { + "epoch": 0.41307, + "grad_norm": 0.6972652077674866, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 41307 + }, + { + "epoch": 0.41308, + "grad_norm": 0.9573046565055847, + "learning_rate": 0.003, + "loss": 3.9379, + "step": 41308 + }, + { + "epoch": 0.41309, + "grad_norm": 1.018226146697998, + "learning_rate": 0.003, + "loss": 3.96, + "step": 41309 + }, + { + "epoch": 0.4131, + "grad_norm": 1.093867540359497, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 41310 + }, + { + "epoch": 0.41311, + "grad_norm": 1.0073182582855225, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 41311 + }, + { + "epoch": 0.41312, + "grad_norm": 0.9279995560646057, + "learning_rate": 0.003, + "loss": 3.984, + "step": 41312 + }, + { + "epoch": 0.41313, + "grad_norm": 0.9839658141136169, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 41313 + }, + { + "epoch": 0.41314, + "grad_norm": 1.0209439992904663, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 41314 + }, + { + "epoch": 0.41315, + "grad_norm": 1.1240357160568237, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 41315 + }, + { + "epoch": 0.41316, + "grad_norm": 0.8516234755516052, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 41316 + }, + { + "epoch": 0.41317, + "grad_norm": 0.8166253566741943, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 41317 + }, + { + "epoch": 0.41318, + "grad_norm": 1.0586974620819092, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 41318 + }, + { + "epoch": 0.41319, + "grad_norm": 1.1207467317581177, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 41319 + }, + { + "epoch": 0.4132, + "grad_norm": 0.8127905130386353, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 41320 + }, + { + "epoch": 0.41321, + "grad_norm": 0.8626341223716736, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 41321 + }, + { + "epoch": 0.41322, + "grad_norm": 1.1260058879852295, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 41322 + }, + { + "epoch": 0.41323, + "grad_norm": 0.9458111524581909, + "learning_rate": 0.003, + "loss": 3.98, + "step": 41323 + }, + { + "epoch": 0.41324, + "grad_norm": 0.7057847380638123, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 41324 + }, + { + "epoch": 0.41325, + "grad_norm": 0.6988601684570312, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 41325 + }, + { + "epoch": 0.41326, + "grad_norm": 0.7371112704277039, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 41326 + }, + { + "epoch": 0.41327, + "grad_norm": 0.8271274566650391, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 41327 + }, + { + "epoch": 0.41328, + "grad_norm": 1.0688356161117554, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 41328 + }, + { + "epoch": 0.41329, + "grad_norm": 0.9096440076828003, + "learning_rate": 0.003, + "loss": 3.996, + "step": 41329 + }, + { + "epoch": 0.4133, + "grad_norm": 0.787532389163971, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 41330 + }, + { + "epoch": 0.41331, + "grad_norm": 0.7028806805610657, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 41331 + }, + { + "epoch": 0.41332, + "grad_norm": 0.7617402076721191, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 41332 + }, + { + "epoch": 0.41333, + "grad_norm": 0.9727951288223267, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 41333 + }, + { + "epoch": 0.41334, + "grad_norm": 1.3112189769744873, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 41334 + }, + { + "epoch": 0.41335, + "grad_norm": 0.9965007305145264, + "learning_rate": 0.003, + "loss": 3.975, + "step": 41335 + }, + { + "epoch": 0.41336, + "grad_norm": 1.0708988904953003, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 41336 + }, + { + "epoch": 0.41337, + "grad_norm": 0.763308048248291, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 41337 + }, + { + "epoch": 0.41338, + "grad_norm": 0.671310544013977, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 41338 + }, + { + "epoch": 0.41339, + "grad_norm": 0.608788251876831, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 41339 + }, + { + "epoch": 0.4134, + "grad_norm": 0.6028150916099548, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 41340 + }, + { + "epoch": 0.41341, + "grad_norm": 0.6726163625717163, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 41341 + }, + { + "epoch": 0.41342, + "grad_norm": 0.7364725470542908, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 41342 + }, + { + "epoch": 0.41343, + "grad_norm": 0.9235562086105347, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 41343 + }, + { + "epoch": 0.41344, + "grad_norm": 1.1893846988677979, + "learning_rate": 0.003, + "loss": 3.973, + "step": 41344 + }, + { + "epoch": 0.41345, + "grad_norm": 1.0483876466751099, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 41345 + }, + { + "epoch": 0.41346, + "grad_norm": 1.031944751739502, + "learning_rate": 0.003, + "loss": 3.9422, + "step": 41346 + }, + { + "epoch": 0.41347, + "grad_norm": 0.9840798377990723, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 41347 + }, + { + "epoch": 0.41348, + "grad_norm": 1.1253345012664795, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 41348 + }, + { + "epoch": 0.41349, + "grad_norm": 0.9466031193733215, + "learning_rate": 0.003, + "loss": 4.002, + "step": 41349 + }, + { + "epoch": 0.4135, + "grad_norm": 0.8361152410507202, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 41350 + }, + { + "epoch": 0.41351, + "grad_norm": 0.8625922203063965, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 41351 + }, + { + "epoch": 0.41352, + "grad_norm": 0.9526265263557434, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 41352 + }, + { + "epoch": 0.41353, + "grad_norm": 1.0642346143722534, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 41353 + }, + { + "epoch": 0.41354, + "grad_norm": 1.0474615097045898, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 41354 + }, + { + "epoch": 0.41355, + "grad_norm": 1.0191493034362793, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 41355 + }, + { + "epoch": 0.41356, + "grad_norm": 0.9607821702957153, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 41356 + }, + { + "epoch": 0.41357, + "grad_norm": 0.7748322486877441, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 41357 + }, + { + "epoch": 0.41358, + "grad_norm": 0.7973750829696655, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 41358 + }, + { + "epoch": 0.41359, + "grad_norm": 0.8451427817344666, + "learning_rate": 0.003, + "loss": 3.9479, + "step": 41359 + }, + { + "epoch": 0.4136, + "grad_norm": 1.0403165817260742, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 41360 + }, + { + "epoch": 0.41361, + "grad_norm": 1.0367035865783691, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 41361 + }, + { + "epoch": 0.41362, + "grad_norm": 0.9754197001457214, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 41362 + }, + { + "epoch": 0.41363, + "grad_norm": 0.8864544630050659, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 41363 + }, + { + "epoch": 0.41364, + "grad_norm": 0.7808422446250916, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 41364 + }, + { + "epoch": 0.41365, + "grad_norm": 0.7492804527282715, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 41365 + }, + { + "epoch": 0.41366, + "grad_norm": 0.7419714331626892, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 41366 + }, + { + "epoch": 0.41367, + "grad_norm": 0.8335008025169373, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 41367 + }, + { + "epoch": 0.41368, + "grad_norm": 0.8666634559631348, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 41368 + }, + { + "epoch": 0.41369, + "grad_norm": 0.7992943525314331, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 41369 + }, + { + "epoch": 0.4137, + "grad_norm": 0.822175920009613, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 41370 + }, + { + "epoch": 0.41371, + "grad_norm": 0.93186354637146, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 41371 + }, + { + "epoch": 0.41372, + "grad_norm": 0.9314267039299011, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 41372 + }, + { + "epoch": 0.41373, + "grad_norm": 0.9726598262786865, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 41373 + }, + { + "epoch": 0.41374, + "grad_norm": 0.8385924696922302, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 41374 + }, + { + "epoch": 0.41375, + "grad_norm": 0.9178421497344971, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 41375 + }, + { + "epoch": 0.41376, + "grad_norm": 1.0565296411514282, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 41376 + }, + { + "epoch": 0.41377, + "grad_norm": 1.0391570329666138, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 41377 + }, + { + "epoch": 0.41378, + "grad_norm": 1.0480468273162842, + "learning_rate": 0.003, + "loss": 3.9428, + "step": 41378 + }, + { + "epoch": 0.41379, + "grad_norm": 1.2126390933990479, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 41379 + }, + { + "epoch": 0.4138, + "grad_norm": 0.9497050642967224, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 41380 + }, + { + "epoch": 0.41381, + "grad_norm": 0.9054000973701477, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 41381 + }, + { + "epoch": 0.41382, + "grad_norm": 0.9818316698074341, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 41382 + }, + { + "epoch": 0.41383, + "grad_norm": 1.0570635795593262, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 41383 + }, + { + "epoch": 0.41384, + "grad_norm": 0.8641547560691833, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 41384 + }, + { + "epoch": 0.41385, + "grad_norm": 0.9634584784507751, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 41385 + }, + { + "epoch": 0.41386, + "grad_norm": 1.1352579593658447, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 41386 + }, + { + "epoch": 0.41387, + "grad_norm": 0.8468684554100037, + "learning_rate": 0.003, + "loss": 3.996, + "step": 41387 + }, + { + "epoch": 0.41388, + "grad_norm": 0.7578032612800598, + "learning_rate": 0.003, + "loss": 3.9444, + "step": 41388 + }, + { + "epoch": 0.41389, + "grad_norm": 0.786304771900177, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 41389 + }, + { + "epoch": 0.4139, + "grad_norm": 0.9355652928352356, + "learning_rate": 0.003, + "loss": 3.9455, + "step": 41390 + }, + { + "epoch": 0.41391, + "grad_norm": 1.0333236455917358, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 41391 + }, + { + "epoch": 0.41392, + "grad_norm": 1.0084583759307861, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 41392 + }, + { + "epoch": 0.41393, + "grad_norm": 1.2327288389205933, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 41393 + }, + { + "epoch": 0.41394, + "grad_norm": 0.8156059980392456, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 41394 + }, + { + "epoch": 0.41395, + "grad_norm": 0.7481407523155212, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 41395 + }, + { + "epoch": 0.41396, + "grad_norm": 0.7371836304664612, + "learning_rate": 0.003, + "loss": 3.988, + "step": 41396 + }, + { + "epoch": 0.41397, + "grad_norm": 0.9362191557884216, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 41397 + }, + { + "epoch": 0.41398, + "grad_norm": 1.0973403453826904, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 41398 + }, + { + "epoch": 0.41399, + "grad_norm": 0.8086583018302917, + "learning_rate": 0.003, + "loss": 3.958, + "step": 41399 + }, + { + "epoch": 0.414, + "grad_norm": 0.6175271272659302, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 41400 + }, + { + "epoch": 0.41401, + "grad_norm": 0.6450279355049133, + "learning_rate": 0.003, + "loss": 3.99, + "step": 41401 + }, + { + "epoch": 0.41402, + "grad_norm": 0.6808347105979919, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 41402 + }, + { + "epoch": 0.41403, + "grad_norm": 0.7559964656829834, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 41403 + }, + { + "epoch": 0.41404, + "grad_norm": 1.051424264907837, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 41404 + }, + { + "epoch": 0.41405, + "grad_norm": 1.1953181028366089, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 41405 + }, + { + "epoch": 0.41406, + "grad_norm": 0.8915272951126099, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 41406 + }, + { + "epoch": 0.41407, + "grad_norm": 0.9341866374015808, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 41407 + }, + { + "epoch": 0.41408, + "grad_norm": 0.9800822734832764, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 41408 + }, + { + "epoch": 0.41409, + "grad_norm": 0.9437863826751709, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 41409 + }, + { + "epoch": 0.4141, + "grad_norm": 0.8921228647232056, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 41410 + }, + { + "epoch": 0.41411, + "grad_norm": 0.9464541673660278, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 41411 + }, + { + "epoch": 0.41412, + "grad_norm": 0.9667695760726929, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 41412 + }, + { + "epoch": 0.41413, + "grad_norm": 0.9481440782546997, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 41413 + }, + { + "epoch": 0.41414, + "grad_norm": 0.9223023653030396, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 41414 + }, + { + "epoch": 0.41415, + "grad_norm": 1.0697579383850098, + "learning_rate": 0.003, + "loss": 3.97, + "step": 41415 + }, + { + "epoch": 0.41416, + "grad_norm": 0.9125940203666687, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 41416 + }, + { + "epoch": 0.41417, + "grad_norm": 0.9982422590255737, + "learning_rate": 0.003, + "loss": 3.9457, + "step": 41417 + }, + { + "epoch": 0.41418, + "grad_norm": 1.4373911619186401, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 41418 + }, + { + "epoch": 0.41419, + "grad_norm": 0.8151742219924927, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 41419 + }, + { + "epoch": 0.4142, + "grad_norm": 0.844042956829071, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 41420 + }, + { + "epoch": 0.41421, + "grad_norm": 0.8173197507858276, + "learning_rate": 0.003, + "loss": 4.001, + "step": 41421 + }, + { + "epoch": 0.41422, + "grad_norm": 0.913108766078949, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 41422 + }, + { + "epoch": 0.41423, + "grad_norm": 1.0914417505264282, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 41423 + }, + { + "epoch": 0.41424, + "grad_norm": 0.9980021715164185, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 41424 + }, + { + "epoch": 0.41425, + "grad_norm": 0.8656253814697266, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 41425 + }, + { + "epoch": 0.41426, + "grad_norm": 0.9775957465171814, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 41426 + }, + { + "epoch": 0.41427, + "grad_norm": 1.1653683185577393, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 41427 + }, + { + "epoch": 0.41428, + "grad_norm": 0.8563591241836548, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 41428 + }, + { + "epoch": 0.41429, + "grad_norm": 0.7942968606948853, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 41429 + }, + { + "epoch": 0.4143, + "grad_norm": 0.7489545941352844, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 41430 + }, + { + "epoch": 0.41431, + "grad_norm": 0.6664227843284607, + "learning_rate": 0.003, + "loss": 3.97, + "step": 41431 + }, + { + "epoch": 0.41432, + "grad_norm": 0.6862231492996216, + "learning_rate": 0.003, + "loss": 3.9459, + "step": 41432 + }, + { + "epoch": 0.41433, + "grad_norm": 0.6336501240730286, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 41433 + }, + { + "epoch": 0.41434, + "grad_norm": 0.6197352409362793, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 41434 + }, + { + "epoch": 0.41435, + "grad_norm": 0.753503143787384, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 41435 + }, + { + "epoch": 0.41436, + "grad_norm": 1.0431426763534546, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 41436 + }, + { + "epoch": 0.41437, + "grad_norm": 1.2280124425888062, + "learning_rate": 0.003, + "loss": 3.9239, + "step": 41437 + }, + { + "epoch": 0.41438, + "grad_norm": 0.5626353621482849, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 41438 + }, + { + "epoch": 0.41439, + "grad_norm": 0.6316342949867249, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 41439 + }, + { + "epoch": 0.4144, + "grad_norm": 0.8875305652618408, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 41440 + }, + { + "epoch": 0.41441, + "grad_norm": 1.1504231691360474, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 41441 + }, + { + "epoch": 0.41442, + "grad_norm": 0.9439363479614258, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 41442 + }, + { + "epoch": 0.41443, + "grad_norm": 0.8687266707420349, + "learning_rate": 0.003, + "loss": 3.958, + "step": 41443 + }, + { + "epoch": 0.41444, + "grad_norm": 0.88819420337677, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 41444 + }, + { + "epoch": 0.41445, + "grad_norm": 1.0000874996185303, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 41445 + }, + { + "epoch": 0.41446, + "grad_norm": 1.179417610168457, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 41446 + }, + { + "epoch": 0.41447, + "grad_norm": 0.8538842797279358, + "learning_rate": 0.003, + "loss": 3.9439, + "step": 41447 + }, + { + "epoch": 0.41448, + "grad_norm": 0.7808514833450317, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 41448 + }, + { + "epoch": 0.41449, + "grad_norm": 0.856380820274353, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 41449 + }, + { + "epoch": 0.4145, + "grad_norm": 1.0157564878463745, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 41450 + }, + { + "epoch": 0.41451, + "grad_norm": 1.068506121635437, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 41451 + }, + { + "epoch": 0.41452, + "grad_norm": 1.0058115720748901, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 41452 + }, + { + "epoch": 0.41453, + "grad_norm": 1.0148507356643677, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 41453 + }, + { + "epoch": 0.41454, + "grad_norm": 0.9529242515563965, + "learning_rate": 0.003, + "loss": 4.009, + "step": 41454 + }, + { + "epoch": 0.41455, + "grad_norm": 0.8789528012275696, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 41455 + }, + { + "epoch": 0.41456, + "grad_norm": 0.8309822082519531, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 41456 + }, + { + "epoch": 0.41457, + "grad_norm": 1.0033763647079468, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 41457 + }, + { + "epoch": 0.41458, + "grad_norm": 1.0486961603164673, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 41458 + }, + { + "epoch": 0.41459, + "grad_norm": 1.0697826147079468, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 41459 + }, + { + "epoch": 0.4146, + "grad_norm": 0.9804463982582092, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 41460 + }, + { + "epoch": 0.41461, + "grad_norm": 1.0196162462234497, + "learning_rate": 0.003, + "loss": 4.008, + "step": 41461 + }, + { + "epoch": 0.41462, + "grad_norm": 1.0426400899887085, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 41462 + }, + { + "epoch": 0.41463, + "grad_norm": 0.7819979190826416, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 41463 + }, + { + "epoch": 0.41464, + "grad_norm": 0.7773728966712952, + "learning_rate": 0.003, + "loss": 3.985, + "step": 41464 + }, + { + "epoch": 0.41465, + "grad_norm": 0.7727846503257751, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 41465 + }, + { + "epoch": 0.41466, + "grad_norm": 0.8579714298248291, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 41466 + }, + { + "epoch": 0.41467, + "grad_norm": 0.8685242533683777, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 41467 + }, + { + "epoch": 0.41468, + "grad_norm": 0.9692903161048889, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 41468 + }, + { + "epoch": 0.41469, + "grad_norm": 1.078296184539795, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 41469 + }, + { + "epoch": 0.4147, + "grad_norm": 0.989795446395874, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 41470 + }, + { + "epoch": 0.41471, + "grad_norm": 0.9869727492332458, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 41471 + }, + { + "epoch": 0.41472, + "grad_norm": 0.8242354393005371, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 41472 + }, + { + "epoch": 0.41473, + "grad_norm": 0.6932294368743896, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 41473 + }, + { + "epoch": 0.41474, + "grad_norm": 0.6906421184539795, + "learning_rate": 0.003, + "loss": 3.987, + "step": 41474 + }, + { + "epoch": 0.41475, + "grad_norm": 0.7281799912452698, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 41475 + }, + { + "epoch": 0.41476, + "grad_norm": 0.6983469128608704, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 41476 + }, + { + "epoch": 0.41477, + "grad_norm": 0.7688484191894531, + "learning_rate": 0.003, + "loss": 3.9475, + "step": 41477 + }, + { + "epoch": 0.41478, + "grad_norm": 0.8687107563018799, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 41478 + }, + { + "epoch": 0.41479, + "grad_norm": 0.9300827383995056, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 41479 + }, + { + "epoch": 0.4148, + "grad_norm": 1.1162364482879639, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 41480 + }, + { + "epoch": 0.41481, + "grad_norm": 1.009658932685852, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 41481 + }, + { + "epoch": 0.41482, + "grad_norm": 0.9189461469650269, + "learning_rate": 0.003, + "loss": 3.989, + "step": 41482 + }, + { + "epoch": 0.41483, + "grad_norm": 0.899902880191803, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 41483 + }, + { + "epoch": 0.41484, + "grad_norm": 1.106929898262024, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 41484 + }, + { + "epoch": 0.41485, + "grad_norm": 1.2174115180969238, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 41485 + }, + { + "epoch": 0.41486, + "grad_norm": 0.9782124161720276, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 41486 + }, + { + "epoch": 0.41487, + "grad_norm": 0.9038683176040649, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 41487 + }, + { + "epoch": 0.41488, + "grad_norm": 0.7427064180374146, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 41488 + }, + { + "epoch": 0.41489, + "grad_norm": 0.6828621029853821, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 41489 + }, + { + "epoch": 0.4149, + "grad_norm": 0.7655100226402283, + "learning_rate": 0.003, + "loss": 3.9369, + "step": 41490 + }, + { + "epoch": 0.41491, + "grad_norm": 0.7214524745941162, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 41491 + }, + { + "epoch": 0.41492, + "grad_norm": 0.6558262705802917, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 41492 + }, + { + "epoch": 0.41493, + "grad_norm": 0.6669369339942932, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 41493 + }, + { + "epoch": 0.41494, + "grad_norm": 0.7217309474945068, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 41494 + }, + { + "epoch": 0.41495, + "grad_norm": 0.7943300008773804, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 41495 + }, + { + "epoch": 0.41496, + "grad_norm": 0.9653634428977966, + "learning_rate": 0.003, + "loss": 3.991, + "step": 41496 + }, + { + "epoch": 0.41497, + "grad_norm": 1.1073806285858154, + "learning_rate": 0.003, + "loss": 3.974, + "step": 41497 + }, + { + "epoch": 0.41498, + "grad_norm": 0.9554896354675293, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 41498 + }, + { + "epoch": 0.41499, + "grad_norm": 1.0757646560668945, + "learning_rate": 0.003, + "loss": 3.9439, + "step": 41499 + }, + { + "epoch": 0.415, + "grad_norm": 1.0066146850585938, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 41500 + }, + { + "epoch": 0.41501, + "grad_norm": 0.8463053107261658, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 41501 + }, + { + "epoch": 0.41502, + "grad_norm": 0.8590133190155029, + "learning_rate": 0.003, + "loss": 3.97, + "step": 41502 + }, + { + "epoch": 0.41503, + "grad_norm": 0.8684083819389343, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 41503 + }, + { + "epoch": 0.41504, + "grad_norm": 0.8005397915840149, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 41504 + }, + { + "epoch": 0.41505, + "grad_norm": 0.8624087572097778, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 41505 + }, + { + "epoch": 0.41506, + "grad_norm": 0.9647194743156433, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 41506 + }, + { + "epoch": 0.41507, + "grad_norm": 0.9976023435592651, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 41507 + }, + { + "epoch": 0.41508, + "grad_norm": 1.1899064779281616, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 41508 + }, + { + "epoch": 0.41509, + "grad_norm": 0.83128821849823, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 41509 + }, + { + "epoch": 0.4151, + "grad_norm": 0.7331534624099731, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 41510 + }, + { + "epoch": 0.41511, + "grad_norm": 0.8792011141777039, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 41511 + }, + { + "epoch": 0.41512, + "grad_norm": 0.9883187413215637, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 41512 + }, + { + "epoch": 0.41513, + "grad_norm": 0.9860872030258179, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 41513 + }, + { + "epoch": 0.41514, + "grad_norm": 1.1599760055541992, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 41514 + }, + { + "epoch": 0.41515, + "grad_norm": 0.9738403558731079, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 41515 + }, + { + "epoch": 0.41516, + "grad_norm": 0.9653417468070984, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 41516 + }, + { + "epoch": 0.41517, + "grad_norm": 1.1127129793167114, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 41517 + }, + { + "epoch": 0.41518, + "grad_norm": 1.105210781097412, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 41518 + }, + { + "epoch": 0.41519, + "grad_norm": 0.8764727115631104, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 41519 + }, + { + "epoch": 0.4152, + "grad_norm": 1.000046968460083, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 41520 + }, + { + "epoch": 0.41521, + "grad_norm": 1.0969732999801636, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 41521 + }, + { + "epoch": 0.41522, + "grad_norm": 0.9351027011871338, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 41522 + }, + { + "epoch": 0.41523, + "grad_norm": 0.7419541478157043, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 41523 + }, + { + "epoch": 0.41524, + "grad_norm": 0.7308628559112549, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 41524 + }, + { + "epoch": 0.41525, + "grad_norm": 0.8802032470703125, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 41525 + }, + { + "epoch": 0.41526, + "grad_norm": 1.0390472412109375, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 41526 + }, + { + "epoch": 0.41527, + "grad_norm": 0.9434019327163696, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 41527 + }, + { + "epoch": 0.41528, + "grad_norm": 1.0293512344360352, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 41528 + }, + { + "epoch": 0.41529, + "grad_norm": 1.032213568687439, + "learning_rate": 0.003, + "loss": 3.9397, + "step": 41529 + }, + { + "epoch": 0.4153, + "grad_norm": 0.7340732216835022, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 41530 + }, + { + "epoch": 0.41531, + "grad_norm": 0.7099645137786865, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 41531 + }, + { + "epoch": 0.41532, + "grad_norm": 0.7674568891525269, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 41532 + }, + { + "epoch": 0.41533, + "grad_norm": 0.8024232983589172, + "learning_rate": 0.003, + "loss": 3.974, + "step": 41533 + }, + { + "epoch": 0.41534, + "grad_norm": 0.8969841599464417, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 41534 + }, + { + "epoch": 0.41535, + "grad_norm": 0.8801705241203308, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 41535 + }, + { + "epoch": 0.41536, + "grad_norm": 0.8516685366630554, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 41536 + }, + { + "epoch": 0.41537, + "grad_norm": 0.7631872892379761, + "learning_rate": 0.003, + "loss": 3.98, + "step": 41537 + }, + { + "epoch": 0.41538, + "grad_norm": 0.7227178812026978, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 41538 + }, + { + "epoch": 0.41539, + "grad_norm": 0.9066436886787415, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 41539 + }, + { + "epoch": 0.4154, + "grad_norm": 1.3605661392211914, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 41540 + }, + { + "epoch": 0.41541, + "grad_norm": 0.9231537580490112, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 41541 + }, + { + "epoch": 0.41542, + "grad_norm": 0.8429957628250122, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 41542 + }, + { + "epoch": 0.41543, + "grad_norm": 0.7613388895988464, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 41543 + }, + { + "epoch": 0.41544, + "grad_norm": 0.7574102878570557, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 41544 + }, + { + "epoch": 0.41545, + "grad_norm": 0.8124598264694214, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 41545 + }, + { + "epoch": 0.41546, + "grad_norm": 0.8148533701896667, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 41546 + }, + { + "epoch": 0.41547, + "grad_norm": 0.9257217645645142, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 41547 + }, + { + "epoch": 0.41548, + "grad_norm": 1.0159965753555298, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 41548 + }, + { + "epoch": 0.41549, + "grad_norm": 1.216745138168335, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 41549 + }, + { + "epoch": 0.4155, + "grad_norm": 0.6919053792953491, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 41550 + }, + { + "epoch": 0.41551, + "grad_norm": 0.692706286907196, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 41551 + }, + { + "epoch": 0.41552, + "grad_norm": 0.7689783573150635, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 41552 + }, + { + "epoch": 0.41553, + "grad_norm": 0.7560137510299683, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 41553 + }, + { + "epoch": 0.41554, + "grad_norm": 0.7136902213096619, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 41554 + }, + { + "epoch": 0.41555, + "grad_norm": 0.8351882696151733, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 41555 + }, + { + "epoch": 0.41556, + "grad_norm": 0.914222776889801, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 41556 + }, + { + "epoch": 0.41557, + "grad_norm": 0.8916282653808594, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 41557 + }, + { + "epoch": 0.41558, + "grad_norm": 0.8457095623016357, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 41558 + }, + { + "epoch": 0.41559, + "grad_norm": 0.7422141432762146, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 41559 + }, + { + "epoch": 0.4156, + "grad_norm": 0.7111146450042725, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 41560 + }, + { + "epoch": 0.41561, + "grad_norm": 0.7317217588424683, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 41561 + }, + { + "epoch": 0.41562, + "grad_norm": 0.8696567416191101, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 41562 + }, + { + "epoch": 0.41563, + "grad_norm": 1.4063847064971924, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 41563 + }, + { + "epoch": 0.41564, + "grad_norm": 0.6960455775260925, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 41564 + }, + { + "epoch": 0.41565, + "grad_norm": 0.6317731738090515, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 41565 + }, + { + "epoch": 0.41566, + "grad_norm": 0.7795658111572266, + "learning_rate": 0.003, + "loss": 3.969, + "step": 41566 + }, + { + "epoch": 0.41567, + "grad_norm": 0.9489884376525879, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 41567 + }, + { + "epoch": 0.41568, + "grad_norm": 1.087045431137085, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 41568 + }, + { + "epoch": 0.41569, + "grad_norm": 0.7999508380889893, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 41569 + }, + { + "epoch": 0.4157, + "grad_norm": 0.9675031304359436, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 41570 + }, + { + "epoch": 0.41571, + "grad_norm": 1.3748787641525269, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 41571 + }, + { + "epoch": 0.41572, + "grad_norm": 0.8614315986633301, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 41572 + }, + { + "epoch": 0.41573, + "grad_norm": 0.689304530620575, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 41573 + }, + { + "epoch": 0.41574, + "grad_norm": 0.6677052974700928, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 41574 + }, + { + "epoch": 0.41575, + "grad_norm": 0.7778767943382263, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 41575 + }, + { + "epoch": 0.41576, + "grad_norm": 0.8863914012908936, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 41576 + }, + { + "epoch": 0.41577, + "grad_norm": 1.0288077592849731, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 41577 + }, + { + "epoch": 0.41578, + "grad_norm": 0.9325556755065918, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 41578 + }, + { + "epoch": 0.41579, + "grad_norm": 0.8735014200210571, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 41579 + }, + { + "epoch": 0.4158, + "grad_norm": 0.7294093370437622, + "learning_rate": 0.003, + "loss": 3.99, + "step": 41580 + }, + { + "epoch": 0.41581, + "grad_norm": 0.8391856551170349, + "learning_rate": 0.003, + "loss": 3.972, + "step": 41581 + }, + { + "epoch": 0.41582, + "grad_norm": 0.9935675859451294, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 41582 + }, + { + "epoch": 0.41583, + "grad_norm": 1.3188183307647705, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 41583 + }, + { + "epoch": 0.41584, + "grad_norm": 0.8848585486412048, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 41584 + }, + { + "epoch": 0.41585, + "grad_norm": 0.827663242816925, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 41585 + }, + { + "epoch": 0.41586, + "grad_norm": 0.8195993304252625, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 41586 + }, + { + "epoch": 0.41587, + "grad_norm": 0.9614384174346924, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 41587 + }, + { + "epoch": 0.41588, + "grad_norm": 0.963782548904419, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 41588 + }, + { + "epoch": 0.41589, + "grad_norm": 0.9769009351730347, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 41589 + }, + { + "epoch": 0.4159, + "grad_norm": 1.239612340927124, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 41590 + }, + { + "epoch": 0.41591, + "grad_norm": 1.1390702724456787, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 41591 + }, + { + "epoch": 0.41592, + "grad_norm": 0.896447479724884, + "learning_rate": 0.003, + "loss": 3.959, + "step": 41592 + }, + { + "epoch": 0.41593, + "grad_norm": 0.8249761462211609, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 41593 + }, + { + "epoch": 0.41594, + "grad_norm": 1.0043280124664307, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 41594 + }, + { + "epoch": 0.41595, + "grad_norm": 1.0467348098754883, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 41595 + }, + { + "epoch": 0.41596, + "grad_norm": 1.2589244842529297, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 41596 + }, + { + "epoch": 0.41597, + "grad_norm": 0.827523946762085, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 41597 + }, + { + "epoch": 0.41598, + "grad_norm": 0.7215304374694824, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 41598 + }, + { + "epoch": 0.41599, + "grad_norm": 0.7887710928916931, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 41599 + }, + { + "epoch": 0.416, + "grad_norm": 0.8515225052833557, + "learning_rate": 0.003, + "loss": 3.952, + "step": 41600 + }, + { + "epoch": 0.41601, + "grad_norm": 1.0676403045654297, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 41601 + }, + { + "epoch": 0.41602, + "grad_norm": 1.1142418384552002, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 41602 + }, + { + "epoch": 0.41603, + "grad_norm": 0.7484758496284485, + "learning_rate": 0.003, + "loss": 3.995, + "step": 41603 + }, + { + "epoch": 0.41604, + "grad_norm": 0.6082140207290649, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 41604 + }, + { + "epoch": 0.41605, + "grad_norm": 0.6484959721565247, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 41605 + }, + { + "epoch": 0.41606, + "grad_norm": 0.7103104591369629, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 41606 + }, + { + "epoch": 0.41607, + "grad_norm": 0.8700833320617676, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 41607 + }, + { + "epoch": 0.41608, + "grad_norm": 1.1049628257751465, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 41608 + }, + { + "epoch": 0.41609, + "grad_norm": 0.9532009959220886, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 41609 + }, + { + "epoch": 0.4161, + "grad_norm": 0.903940737247467, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 41610 + }, + { + "epoch": 0.41611, + "grad_norm": 0.8733935356140137, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 41611 + }, + { + "epoch": 0.41612, + "grad_norm": 0.825504720211029, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 41612 + }, + { + "epoch": 0.41613, + "grad_norm": 0.8816273808479309, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 41613 + }, + { + "epoch": 0.41614, + "grad_norm": 0.9223227500915527, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 41614 + }, + { + "epoch": 0.41615, + "grad_norm": 0.7288163900375366, + "learning_rate": 0.003, + "loss": 3.9469, + "step": 41615 + }, + { + "epoch": 0.41616, + "grad_norm": 0.6733263731002808, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 41616 + }, + { + "epoch": 0.41617, + "grad_norm": 0.767016589641571, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 41617 + }, + { + "epoch": 0.41618, + "grad_norm": 1.144839882850647, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 41618 + }, + { + "epoch": 0.41619, + "grad_norm": 0.978800892829895, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 41619 + }, + { + "epoch": 0.4162, + "grad_norm": 1.180605411529541, + "learning_rate": 0.003, + "loss": 3.9467, + "step": 41620 + }, + { + "epoch": 0.41621, + "grad_norm": 0.8049744367599487, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 41621 + }, + { + "epoch": 0.41622, + "grad_norm": 0.5855854749679565, + "learning_rate": 0.003, + "loss": 3.9468, + "step": 41622 + }, + { + "epoch": 0.41623, + "grad_norm": 0.7513129115104675, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 41623 + }, + { + "epoch": 0.41624, + "grad_norm": 1.1035584211349487, + "learning_rate": 0.003, + "loss": 3.964, + "step": 41624 + }, + { + "epoch": 0.41625, + "grad_norm": 1.166159749031067, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 41625 + }, + { + "epoch": 0.41626, + "grad_norm": 0.6335336565971375, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 41626 + }, + { + "epoch": 0.41627, + "grad_norm": 0.6184056997299194, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 41627 + }, + { + "epoch": 0.41628, + "grad_norm": 0.6912200450897217, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 41628 + }, + { + "epoch": 0.41629, + "grad_norm": 0.7253939509391785, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 41629 + }, + { + "epoch": 0.4163, + "grad_norm": 0.8184320330619812, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 41630 + }, + { + "epoch": 0.41631, + "grad_norm": 0.8491071462631226, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 41631 + }, + { + "epoch": 0.41632, + "grad_norm": 0.8360460996627808, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 41632 + }, + { + "epoch": 0.41633, + "grad_norm": 0.8722298741340637, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 41633 + }, + { + "epoch": 0.41634, + "grad_norm": 0.962294340133667, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 41634 + }, + { + "epoch": 0.41635, + "grad_norm": 1.0598711967468262, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 41635 + }, + { + "epoch": 0.41636, + "grad_norm": 0.9537031650543213, + "learning_rate": 0.003, + "loss": 4.0, + "step": 41636 + }, + { + "epoch": 0.41637, + "grad_norm": 0.9055473208427429, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 41637 + }, + { + "epoch": 0.41638, + "grad_norm": 1.0533101558685303, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 41638 + }, + { + "epoch": 0.41639, + "grad_norm": 0.9610040187835693, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 41639 + }, + { + "epoch": 0.4164, + "grad_norm": 1.036647081375122, + "learning_rate": 0.003, + "loss": 3.99, + "step": 41640 + }, + { + "epoch": 0.41641, + "grad_norm": 1.1246002912521362, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 41641 + }, + { + "epoch": 0.41642, + "grad_norm": 1.1510398387908936, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 41642 + }, + { + "epoch": 0.41643, + "grad_norm": 1.32384192943573, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 41643 + }, + { + "epoch": 0.41644, + "grad_norm": 0.8114997744560242, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 41644 + }, + { + "epoch": 0.41645, + "grad_norm": 0.6741251945495605, + "learning_rate": 0.003, + "loss": 3.985, + "step": 41645 + }, + { + "epoch": 0.41646, + "grad_norm": 0.7450345754623413, + "learning_rate": 0.003, + "loss": 3.95, + "step": 41646 + }, + { + "epoch": 0.41647, + "grad_norm": 0.847010612487793, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 41647 + }, + { + "epoch": 0.41648, + "grad_norm": 1.1017754077911377, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 41648 + }, + { + "epoch": 0.41649, + "grad_norm": 0.9185215830802917, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 41649 + }, + { + "epoch": 0.4165, + "grad_norm": 0.847628116607666, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 41650 + }, + { + "epoch": 0.41651, + "grad_norm": 0.8056631088256836, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 41651 + }, + { + "epoch": 0.41652, + "grad_norm": 0.8833078742027283, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 41652 + }, + { + "epoch": 0.41653, + "grad_norm": 0.9519262909889221, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 41653 + }, + { + "epoch": 0.41654, + "grad_norm": 0.9917106032371521, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 41654 + }, + { + "epoch": 0.41655, + "grad_norm": 1.0821012258529663, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 41655 + }, + { + "epoch": 0.41656, + "grad_norm": 0.949783444404602, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 41656 + }, + { + "epoch": 0.41657, + "grad_norm": 0.9790940880775452, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 41657 + }, + { + "epoch": 0.41658, + "grad_norm": 1.2080028057098389, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 41658 + }, + { + "epoch": 0.41659, + "grad_norm": 0.8530771136283875, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 41659 + }, + { + "epoch": 0.4166, + "grad_norm": 0.8036303520202637, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 41660 + }, + { + "epoch": 0.41661, + "grad_norm": 0.9824885725975037, + "learning_rate": 0.003, + "loss": 3.9241, + "step": 41661 + }, + { + "epoch": 0.41662, + "grad_norm": 1.2652992010116577, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 41662 + }, + { + "epoch": 0.41663, + "grad_norm": 0.7287797331809998, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 41663 + }, + { + "epoch": 0.41664, + "grad_norm": 0.8245384097099304, + "learning_rate": 0.003, + "loss": 3.97, + "step": 41664 + }, + { + "epoch": 0.41665, + "grad_norm": 1.0319124460220337, + "learning_rate": 0.003, + "loss": 3.9388, + "step": 41665 + }, + { + "epoch": 0.41666, + "grad_norm": 1.2663922309875488, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 41666 + }, + { + "epoch": 0.41667, + "grad_norm": 0.5734889507293701, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 41667 + }, + { + "epoch": 0.41668, + "grad_norm": 0.6697536110877991, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 41668 + }, + { + "epoch": 0.41669, + "grad_norm": 0.7140712141990662, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 41669 + }, + { + "epoch": 0.4167, + "grad_norm": 0.9072823524475098, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 41670 + }, + { + "epoch": 0.41671, + "grad_norm": 1.088985562324524, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 41671 + }, + { + "epoch": 0.41672, + "grad_norm": 1.2300211191177368, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 41672 + }, + { + "epoch": 0.41673, + "grad_norm": 0.7889381647109985, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 41673 + }, + { + "epoch": 0.41674, + "grad_norm": 0.7079100608825684, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 41674 + }, + { + "epoch": 0.41675, + "grad_norm": 0.6032965779304504, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 41675 + }, + { + "epoch": 0.41676, + "grad_norm": 0.5461152791976929, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 41676 + }, + { + "epoch": 0.41677, + "grad_norm": 0.7161197066307068, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 41677 + }, + { + "epoch": 0.41678, + "grad_norm": 0.8755851984024048, + "learning_rate": 0.003, + "loss": 3.972, + "step": 41678 + }, + { + "epoch": 0.41679, + "grad_norm": 1.24301278591156, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 41679 + }, + { + "epoch": 0.4168, + "grad_norm": 0.8471684455871582, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 41680 + }, + { + "epoch": 0.41681, + "grad_norm": 0.7893093824386597, + "learning_rate": 0.003, + "loss": 3.964, + "step": 41681 + }, + { + "epoch": 0.41682, + "grad_norm": 0.9106740951538086, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 41682 + }, + { + "epoch": 0.41683, + "grad_norm": 0.8583029508590698, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 41683 + }, + { + "epoch": 0.41684, + "grad_norm": 0.7469415664672852, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 41684 + }, + { + "epoch": 0.41685, + "grad_norm": 0.7232999801635742, + "learning_rate": 0.003, + "loss": 3.957, + "step": 41685 + }, + { + "epoch": 0.41686, + "grad_norm": 0.7682600617408752, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 41686 + }, + { + "epoch": 0.41687, + "grad_norm": 0.8734975457191467, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 41687 + }, + { + "epoch": 0.41688, + "grad_norm": 0.9602192640304565, + "learning_rate": 0.003, + "loss": 3.944, + "step": 41688 + }, + { + "epoch": 0.41689, + "grad_norm": 1.1559133529663086, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 41689 + }, + { + "epoch": 0.4169, + "grad_norm": 0.9583809971809387, + "learning_rate": 0.003, + "loss": 3.978, + "step": 41690 + }, + { + "epoch": 0.41691, + "grad_norm": 0.9036871790885925, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 41691 + }, + { + "epoch": 0.41692, + "grad_norm": 0.9130793213844299, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 41692 + }, + { + "epoch": 0.41693, + "grad_norm": 1.1319290399551392, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 41693 + }, + { + "epoch": 0.41694, + "grad_norm": 0.8708423972129822, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 41694 + }, + { + "epoch": 0.41695, + "grad_norm": 0.6977680921554565, + "learning_rate": 0.003, + "loss": 3.992, + "step": 41695 + }, + { + "epoch": 0.41696, + "grad_norm": 0.7747987508773804, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 41696 + }, + { + "epoch": 0.41697, + "grad_norm": 1.0271227359771729, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 41697 + }, + { + "epoch": 0.41698, + "grad_norm": 1.0880508422851562, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 41698 + }, + { + "epoch": 0.41699, + "grad_norm": 0.8265838623046875, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 41699 + }, + { + "epoch": 0.417, + "grad_norm": 0.7839170694351196, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 41700 + }, + { + "epoch": 0.41701, + "grad_norm": 0.9540314674377441, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 41701 + }, + { + "epoch": 0.41702, + "grad_norm": 1.0734306573867798, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 41702 + }, + { + "epoch": 0.41703, + "grad_norm": 1.1094743013381958, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 41703 + }, + { + "epoch": 0.41704, + "grad_norm": 1.144701361656189, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 41704 + }, + { + "epoch": 0.41705, + "grad_norm": 1.2458248138427734, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 41705 + }, + { + "epoch": 0.41706, + "grad_norm": 0.830872118473053, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 41706 + }, + { + "epoch": 0.41707, + "grad_norm": 0.7718713283538818, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 41707 + }, + { + "epoch": 0.41708, + "grad_norm": 0.948427677154541, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 41708 + }, + { + "epoch": 0.41709, + "grad_norm": 1.229819655418396, + "learning_rate": 0.003, + "loss": 3.973, + "step": 41709 + }, + { + "epoch": 0.4171, + "grad_norm": 0.7437449097633362, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 41710 + }, + { + "epoch": 0.41711, + "grad_norm": 0.6619100570678711, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 41711 + }, + { + "epoch": 0.41712, + "grad_norm": 0.6908870339393616, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 41712 + }, + { + "epoch": 0.41713, + "grad_norm": 0.7600235939025879, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 41713 + }, + { + "epoch": 0.41714, + "grad_norm": 0.7855256795883179, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 41714 + }, + { + "epoch": 0.41715, + "grad_norm": 0.8629507422447205, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 41715 + }, + { + "epoch": 0.41716, + "grad_norm": 0.9126588702201843, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 41716 + }, + { + "epoch": 0.41717, + "grad_norm": 1.0155349969863892, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 41717 + }, + { + "epoch": 0.41718, + "grad_norm": 1.086417555809021, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 41718 + }, + { + "epoch": 0.41719, + "grad_norm": 1.0906994342803955, + "learning_rate": 0.003, + "loss": 4.002, + "step": 41719 + }, + { + "epoch": 0.4172, + "grad_norm": 1.0651140213012695, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 41720 + }, + { + "epoch": 0.41721, + "grad_norm": 0.8993005752563477, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 41721 + }, + { + "epoch": 0.41722, + "grad_norm": 0.8964649438858032, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 41722 + }, + { + "epoch": 0.41723, + "grad_norm": 0.972926914691925, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 41723 + }, + { + "epoch": 0.41724, + "grad_norm": 0.9439943432807922, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 41724 + }, + { + "epoch": 0.41725, + "grad_norm": 0.8709940910339355, + "learning_rate": 0.003, + "loss": 3.976, + "step": 41725 + }, + { + "epoch": 0.41726, + "grad_norm": 0.9176397919654846, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 41726 + }, + { + "epoch": 0.41727, + "grad_norm": 0.9563674926757812, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 41727 + }, + { + "epoch": 0.41728, + "grad_norm": 0.9565734267234802, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 41728 + }, + { + "epoch": 0.41729, + "grad_norm": 1.0375947952270508, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 41729 + }, + { + "epoch": 0.4173, + "grad_norm": 1.202088475227356, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 41730 + }, + { + "epoch": 0.41731, + "grad_norm": 1.038665771484375, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 41731 + }, + { + "epoch": 0.41732, + "grad_norm": 0.9778182506561279, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 41732 + }, + { + "epoch": 0.41733, + "grad_norm": 0.9887055158615112, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 41733 + }, + { + "epoch": 0.41734, + "grad_norm": 0.9139209389686584, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 41734 + }, + { + "epoch": 0.41735, + "grad_norm": 0.9293668270111084, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 41735 + }, + { + "epoch": 0.41736, + "grad_norm": 0.9712164402008057, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 41736 + }, + { + "epoch": 0.41737, + "grad_norm": 0.9010220766067505, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 41737 + }, + { + "epoch": 0.41738, + "grad_norm": 0.7923110723495483, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 41738 + }, + { + "epoch": 0.41739, + "grad_norm": 0.7939367294311523, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 41739 + }, + { + "epoch": 0.4174, + "grad_norm": 0.8022342324256897, + "learning_rate": 0.003, + "loss": 3.996, + "step": 41740 + }, + { + "epoch": 0.41741, + "grad_norm": 0.9487379193305969, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 41741 + }, + { + "epoch": 0.41742, + "grad_norm": 1.1058557033538818, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 41742 + }, + { + "epoch": 0.41743, + "grad_norm": 0.8953562378883362, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 41743 + }, + { + "epoch": 0.41744, + "grad_norm": 0.9824709296226501, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 41744 + }, + { + "epoch": 0.41745, + "grad_norm": 0.9875802397727966, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 41745 + }, + { + "epoch": 0.41746, + "grad_norm": 0.9123646020889282, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 41746 + }, + { + "epoch": 0.41747, + "grad_norm": 0.866552472114563, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 41747 + }, + { + "epoch": 0.41748, + "grad_norm": 0.7149774432182312, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 41748 + }, + { + "epoch": 0.41749, + "grad_norm": 0.7719079852104187, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 41749 + }, + { + "epoch": 0.4175, + "grad_norm": 0.7360686659812927, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 41750 + }, + { + "epoch": 0.41751, + "grad_norm": 0.7795864939689636, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 41751 + }, + { + "epoch": 0.41752, + "grad_norm": 0.8839349746704102, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 41752 + }, + { + "epoch": 0.41753, + "grad_norm": 0.9949274063110352, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 41753 + }, + { + "epoch": 0.41754, + "grad_norm": 0.9918501377105713, + "learning_rate": 0.003, + "loss": 3.992, + "step": 41754 + }, + { + "epoch": 0.41755, + "grad_norm": 1.0083290338516235, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 41755 + }, + { + "epoch": 0.41756, + "grad_norm": 1.1221567392349243, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 41756 + }, + { + "epoch": 0.41757, + "grad_norm": 0.9177605509757996, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 41757 + }, + { + "epoch": 0.41758, + "grad_norm": 0.8349646925926208, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 41758 + }, + { + "epoch": 0.41759, + "grad_norm": 0.7328534126281738, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 41759 + }, + { + "epoch": 0.4176, + "grad_norm": 0.7615523338317871, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 41760 + }, + { + "epoch": 0.41761, + "grad_norm": 0.8182579874992371, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 41761 + }, + { + "epoch": 0.41762, + "grad_norm": 0.8854471445083618, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 41762 + }, + { + "epoch": 0.41763, + "grad_norm": 0.9611912369728088, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 41763 + }, + { + "epoch": 0.41764, + "grad_norm": 0.9705116152763367, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 41764 + }, + { + "epoch": 0.41765, + "grad_norm": 0.9875006079673767, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 41765 + }, + { + "epoch": 0.41766, + "grad_norm": 1.279151439666748, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 41766 + }, + { + "epoch": 0.41767, + "grad_norm": 0.7238859534263611, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 41767 + }, + { + "epoch": 0.41768, + "grad_norm": 0.6314805746078491, + "learning_rate": 0.003, + "loss": 3.9401, + "step": 41768 + }, + { + "epoch": 0.41769, + "grad_norm": 0.6100450158119202, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 41769 + }, + { + "epoch": 0.4177, + "grad_norm": 0.6545692682266235, + "learning_rate": 0.003, + "loss": 3.9453, + "step": 41770 + }, + { + "epoch": 0.41771, + "grad_norm": 0.7457104921340942, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 41771 + }, + { + "epoch": 0.41772, + "grad_norm": 1.002542495727539, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 41772 + }, + { + "epoch": 0.41773, + "grad_norm": 1.5946714878082275, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 41773 + }, + { + "epoch": 0.41774, + "grad_norm": 0.5755379796028137, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 41774 + }, + { + "epoch": 0.41775, + "grad_norm": 1.0154838562011719, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 41775 + }, + { + "epoch": 0.41776, + "grad_norm": 1.2768278121948242, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 41776 + }, + { + "epoch": 0.41777, + "grad_norm": 0.7058563232421875, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 41777 + }, + { + "epoch": 0.41778, + "grad_norm": 0.7031910419464111, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 41778 + }, + { + "epoch": 0.41779, + "grad_norm": 0.6815803050994873, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 41779 + }, + { + "epoch": 0.4178, + "grad_norm": 0.7305906414985657, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 41780 + }, + { + "epoch": 0.41781, + "grad_norm": 0.9788250923156738, + "learning_rate": 0.003, + "loss": 3.99, + "step": 41781 + }, + { + "epoch": 0.41782, + "grad_norm": 1.2322819232940674, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 41782 + }, + { + "epoch": 0.41783, + "grad_norm": 0.8681236505508423, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 41783 + }, + { + "epoch": 0.41784, + "grad_norm": 1.0019859075546265, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 41784 + }, + { + "epoch": 0.41785, + "grad_norm": 0.8944970965385437, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 41785 + }, + { + "epoch": 0.41786, + "grad_norm": 0.827441394329071, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 41786 + }, + { + "epoch": 0.41787, + "grad_norm": 0.8986883759498596, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 41787 + }, + { + "epoch": 0.41788, + "grad_norm": 1.1026761531829834, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 41788 + }, + { + "epoch": 0.41789, + "grad_norm": 1.031598687171936, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 41789 + }, + { + "epoch": 0.4179, + "grad_norm": 0.9933899641036987, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 41790 + }, + { + "epoch": 0.41791, + "grad_norm": 1.010242223739624, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 41791 + }, + { + "epoch": 0.41792, + "grad_norm": 1.0950888395309448, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 41792 + }, + { + "epoch": 0.41793, + "grad_norm": 1.0209726095199585, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 41793 + }, + { + "epoch": 0.41794, + "grad_norm": 0.9103503227233887, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 41794 + }, + { + "epoch": 0.41795, + "grad_norm": 0.81130051612854, + "learning_rate": 0.003, + "loss": 3.948, + "step": 41795 + }, + { + "epoch": 0.41796, + "grad_norm": 0.8050071597099304, + "learning_rate": 0.003, + "loss": 3.949, + "step": 41796 + }, + { + "epoch": 0.41797, + "grad_norm": 0.9741461873054504, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 41797 + }, + { + "epoch": 0.41798, + "grad_norm": 1.2312473058700562, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 41798 + }, + { + "epoch": 0.41799, + "grad_norm": 0.8213366270065308, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 41799 + }, + { + "epoch": 0.418, + "grad_norm": 0.6328867673873901, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 41800 + }, + { + "epoch": 0.41801, + "grad_norm": 0.6383018493652344, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 41801 + }, + { + "epoch": 0.41802, + "grad_norm": 0.665459394454956, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 41802 + }, + { + "epoch": 0.41803, + "grad_norm": 0.7507277727127075, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 41803 + }, + { + "epoch": 0.41804, + "grad_norm": 0.8322629928588867, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 41804 + }, + { + "epoch": 0.41805, + "grad_norm": 0.8362572193145752, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 41805 + }, + { + "epoch": 0.41806, + "grad_norm": 0.8413284420967102, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 41806 + }, + { + "epoch": 0.41807, + "grad_norm": 0.8335764408111572, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 41807 + }, + { + "epoch": 0.41808, + "grad_norm": 0.7999314665794373, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 41808 + }, + { + "epoch": 0.41809, + "grad_norm": 0.794805109500885, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 41809 + }, + { + "epoch": 0.4181, + "grad_norm": 0.8388710021972656, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 41810 + }, + { + "epoch": 0.41811, + "grad_norm": 0.8670327067375183, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 41811 + }, + { + "epoch": 0.41812, + "grad_norm": 0.8084828853607178, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 41812 + }, + { + "epoch": 0.41813, + "grad_norm": 0.8893064856529236, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 41813 + }, + { + "epoch": 0.41814, + "grad_norm": 0.8815672993659973, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 41814 + }, + { + "epoch": 0.41815, + "grad_norm": 1.1821085214614868, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 41815 + }, + { + "epoch": 0.41816, + "grad_norm": 0.9460654854774475, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 41816 + }, + { + "epoch": 0.41817, + "grad_norm": 0.8363175392150879, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 41817 + }, + { + "epoch": 0.41818, + "grad_norm": 0.8504256010055542, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 41818 + }, + { + "epoch": 0.41819, + "grad_norm": 0.7437569499015808, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 41819 + }, + { + "epoch": 0.4182, + "grad_norm": 0.7027272582054138, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 41820 + }, + { + "epoch": 0.41821, + "grad_norm": 0.6872142553329468, + "learning_rate": 0.003, + "loss": 3.981, + "step": 41821 + }, + { + "epoch": 0.41822, + "grad_norm": 0.8769466876983643, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 41822 + }, + { + "epoch": 0.41823, + "grad_norm": 1.017802357673645, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 41823 + }, + { + "epoch": 0.41824, + "grad_norm": 1.2054425477981567, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 41824 + }, + { + "epoch": 0.41825, + "grad_norm": 0.9761111736297607, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 41825 + }, + { + "epoch": 0.41826, + "grad_norm": 1.001716136932373, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 41826 + }, + { + "epoch": 0.41827, + "grad_norm": 1.1464390754699707, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 41827 + }, + { + "epoch": 0.41828, + "grad_norm": 0.9819540977478027, + "learning_rate": 0.003, + "loss": 4.001, + "step": 41828 + }, + { + "epoch": 0.41829, + "grad_norm": 1.0795775651931763, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 41829 + }, + { + "epoch": 0.4183, + "grad_norm": 1.0464235544204712, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 41830 + }, + { + "epoch": 0.41831, + "grad_norm": 0.7601978778839111, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 41831 + }, + { + "epoch": 0.41832, + "grad_norm": 0.6663917899131775, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 41832 + }, + { + "epoch": 0.41833, + "grad_norm": 0.6814583539962769, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 41833 + }, + { + "epoch": 0.41834, + "grad_norm": 0.7032614946365356, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 41834 + }, + { + "epoch": 0.41835, + "grad_norm": 0.6771630644798279, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 41835 + }, + { + "epoch": 0.41836, + "grad_norm": 0.7683953642845154, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 41836 + }, + { + "epoch": 0.41837, + "grad_norm": 0.8663979172706604, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 41837 + }, + { + "epoch": 0.41838, + "grad_norm": 0.9078354835510254, + "learning_rate": 0.003, + "loss": 3.9388, + "step": 41838 + }, + { + "epoch": 0.41839, + "grad_norm": 0.9302870035171509, + "learning_rate": 0.003, + "loss": 3.9475, + "step": 41839 + }, + { + "epoch": 0.4184, + "grad_norm": 0.9199128746986389, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 41840 + }, + { + "epoch": 0.41841, + "grad_norm": 0.9874418377876282, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 41841 + }, + { + "epoch": 0.41842, + "grad_norm": 1.083472490310669, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 41842 + }, + { + "epoch": 0.41843, + "grad_norm": 1.0819084644317627, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 41843 + }, + { + "epoch": 0.41844, + "grad_norm": 1.0080339908599854, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 41844 + }, + { + "epoch": 0.41845, + "grad_norm": 1.0395694971084595, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 41845 + }, + { + "epoch": 0.41846, + "grad_norm": 0.9116122722625732, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 41846 + }, + { + "epoch": 0.41847, + "grad_norm": 0.9038614630699158, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 41847 + }, + { + "epoch": 0.41848, + "grad_norm": 1.026509165763855, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 41848 + }, + { + "epoch": 0.41849, + "grad_norm": 1.167220115661621, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 41849 + }, + { + "epoch": 0.4185, + "grad_norm": 0.9190277457237244, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 41850 + }, + { + "epoch": 0.41851, + "grad_norm": 0.8294478058815002, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 41851 + }, + { + "epoch": 0.41852, + "grad_norm": 0.7751627564430237, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 41852 + }, + { + "epoch": 0.41853, + "grad_norm": 0.7261764407157898, + "learning_rate": 0.003, + "loss": 3.996, + "step": 41853 + }, + { + "epoch": 0.41854, + "grad_norm": 0.7530628442764282, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 41854 + }, + { + "epoch": 0.41855, + "grad_norm": 0.8085359334945679, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 41855 + }, + { + "epoch": 0.41856, + "grad_norm": 1.0342845916748047, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 41856 + }, + { + "epoch": 0.41857, + "grad_norm": 1.142037272453308, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 41857 + }, + { + "epoch": 0.41858, + "grad_norm": 1.036226749420166, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 41858 + }, + { + "epoch": 0.41859, + "grad_norm": 0.9947273135185242, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 41859 + }, + { + "epoch": 0.4186, + "grad_norm": 1.1832904815673828, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 41860 + }, + { + "epoch": 0.41861, + "grad_norm": 0.8898239731788635, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 41861 + }, + { + "epoch": 0.41862, + "grad_norm": 0.8029105067253113, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 41862 + }, + { + "epoch": 0.41863, + "grad_norm": 0.6919757127761841, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 41863 + }, + { + "epoch": 0.41864, + "grad_norm": 0.6282244324684143, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 41864 + }, + { + "epoch": 0.41865, + "grad_norm": 0.6127671003341675, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 41865 + }, + { + "epoch": 0.41866, + "grad_norm": 0.6235458850860596, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 41866 + }, + { + "epoch": 0.41867, + "grad_norm": 0.6634382605552673, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 41867 + }, + { + "epoch": 0.41868, + "grad_norm": 0.76779705286026, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 41868 + }, + { + "epoch": 0.41869, + "grad_norm": 0.8864414095878601, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 41869 + }, + { + "epoch": 0.4187, + "grad_norm": 1.1413171291351318, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 41870 + }, + { + "epoch": 0.41871, + "grad_norm": 0.9209337830543518, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 41871 + }, + { + "epoch": 0.41872, + "grad_norm": 0.8029301762580872, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 41872 + }, + { + "epoch": 0.41873, + "grad_norm": 0.7037258148193359, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 41873 + }, + { + "epoch": 0.41874, + "grad_norm": 0.6573392748832703, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 41874 + }, + { + "epoch": 0.41875, + "grad_norm": 0.7151067852973938, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 41875 + }, + { + "epoch": 0.41876, + "grad_norm": 0.8832536339759827, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 41876 + }, + { + "epoch": 0.41877, + "grad_norm": 1.2666099071502686, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 41877 + }, + { + "epoch": 0.41878, + "grad_norm": 0.7817795872688293, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 41878 + }, + { + "epoch": 0.41879, + "grad_norm": 0.7296764254570007, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 41879 + }, + { + "epoch": 0.4188, + "grad_norm": 0.6964972019195557, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 41880 + }, + { + "epoch": 0.41881, + "grad_norm": 0.6686973571777344, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 41881 + }, + { + "epoch": 0.41882, + "grad_norm": 0.834729790687561, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 41882 + }, + { + "epoch": 0.41883, + "grad_norm": 1.123180627822876, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 41883 + }, + { + "epoch": 0.41884, + "grad_norm": 1.0952889919281006, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 41884 + }, + { + "epoch": 0.41885, + "grad_norm": 0.9497065544128418, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 41885 + }, + { + "epoch": 0.41886, + "grad_norm": 0.8330485820770264, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 41886 + }, + { + "epoch": 0.41887, + "grad_norm": 0.8272863626480103, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 41887 + }, + { + "epoch": 0.41888, + "grad_norm": 1.0877801179885864, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 41888 + }, + { + "epoch": 0.41889, + "grad_norm": 1.2403656244277954, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 41889 + }, + { + "epoch": 0.4189, + "grad_norm": 0.7847372889518738, + "learning_rate": 0.003, + "loss": 3.9385, + "step": 41890 + }, + { + "epoch": 0.41891, + "grad_norm": 0.7844259738922119, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 41891 + }, + { + "epoch": 0.41892, + "grad_norm": 0.9241584539413452, + "learning_rate": 0.003, + "loss": 3.999, + "step": 41892 + }, + { + "epoch": 0.41893, + "grad_norm": 1.0244722366333008, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 41893 + }, + { + "epoch": 0.41894, + "grad_norm": 1.0129095315933228, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 41894 + }, + { + "epoch": 0.41895, + "grad_norm": 1.3103803396224976, + "learning_rate": 0.003, + "loss": 3.965, + "step": 41895 + }, + { + "epoch": 0.41896, + "grad_norm": 0.8340649604797363, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 41896 + }, + { + "epoch": 0.41897, + "grad_norm": 1.0273176431655884, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 41897 + }, + { + "epoch": 0.41898, + "grad_norm": 1.1558300256729126, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 41898 + }, + { + "epoch": 0.41899, + "grad_norm": 0.807949960231781, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 41899 + }, + { + "epoch": 0.419, + "grad_norm": 0.8302352428436279, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 41900 + }, + { + "epoch": 0.41901, + "grad_norm": 1.1273624897003174, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 41901 + }, + { + "epoch": 0.41902, + "grad_norm": 1.0178266763687134, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 41902 + }, + { + "epoch": 0.41903, + "grad_norm": 1.0628437995910645, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 41903 + }, + { + "epoch": 0.41904, + "grad_norm": 0.9855813980102539, + "learning_rate": 0.003, + "loss": 3.975, + "step": 41904 + }, + { + "epoch": 0.41905, + "grad_norm": 0.9670136570930481, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 41905 + }, + { + "epoch": 0.41906, + "grad_norm": 0.9209439754486084, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 41906 + }, + { + "epoch": 0.41907, + "grad_norm": 0.966050922870636, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 41907 + }, + { + "epoch": 0.41908, + "grad_norm": 0.9417829513549805, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 41908 + }, + { + "epoch": 0.41909, + "grad_norm": 0.9293376207351685, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 41909 + }, + { + "epoch": 0.4191, + "grad_norm": 1.0297374725341797, + "learning_rate": 0.003, + "loss": 3.965, + "step": 41910 + }, + { + "epoch": 0.41911, + "grad_norm": 0.9152306914329529, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 41911 + }, + { + "epoch": 0.41912, + "grad_norm": 0.9139352440834045, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 41912 + }, + { + "epoch": 0.41913, + "grad_norm": 0.9473594427108765, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 41913 + }, + { + "epoch": 0.41914, + "grad_norm": 1.0521279573440552, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 41914 + }, + { + "epoch": 0.41915, + "grad_norm": 1.0277146100997925, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 41915 + }, + { + "epoch": 0.41916, + "grad_norm": 0.8869934678077698, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 41916 + }, + { + "epoch": 0.41917, + "grad_norm": 0.8457193970680237, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 41917 + }, + { + "epoch": 0.41918, + "grad_norm": 1.0053911209106445, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 41918 + }, + { + "epoch": 0.41919, + "grad_norm": 1.2274410724639893, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 41919 + }, + { + "epoch": 0.4192, + "grad_norm": 0.7981455326080322, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 41920 + }, + { + "epoch": 0.41921, + "grad_norm": 0.7881396412849426, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 41921 + }, + { + "epoch": 0.41922, + "grad_norm": 0.8435888290405273, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 41922 + }, + { + "epoch": 0.41923, + "grad_norm": 0.8897566199302673, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 41923 + }, + { + "epoch": 0.41924, + "grad_norm": 0.9450140595436096, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 41924 + }, + { + "epoch": 0.41925, + "grad_norm": 1.0492905378341675, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 41925 + }, + { + "epoch": 0.41926, + "grad_norm": 0.9207710027694702, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 41926 + }, + { + "epoch": 0.41927, + "grad_norm": 0.9980649948120117, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 41927 + }, + { + "epoch": 0.41928, + "grad_norm": 1.2946600914001465, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 41928 + }, + { + "epoch": 0.41929, + "grad_norm": 0.8544475436210632, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 41929 + }, + { + "epoch": 0.4193, + "grad_norm": 0.8394744992256165, + "learning_rate": 0.003, + "loss": 3.968, + "step": 41930 + }, + { + "epoch": 0.41931, + "grad_norm": 0.870150089263916, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 41931 + }, + { + "epoch": 0.41932, + "grad_norm": 0.9464729428291321, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 41932 + }, + { + "epoch": 0.41933, + "grad_norm": 0.8237200379371643, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 41933 + }, + { + "epoch": 0.41934, + "grad_norm": 0.6951212286949158, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 41934 + }, + { + "epoch": 0.41935, + "grad_norm": 0.5024895668029785, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 41935 + }, + { + "epoch": 0.41936, + "grad_norm": 0.5435690879821777, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 41936 + }, + { + "epoch": 0.41937, + "grad_norm": 0.6623722314834595, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 41937 + }, + { + "epoch": 0.41938, + "grad_norm": 0.8402624130249023, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 41938 + }, + { + "epoch": 0.41939, + "grad_norm": 1.2453358173370361, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 41939 + }, + { + "epoch": 0.4194, + "grad_norm": 0.8191980123519897, + "learning_rate": 0.003, + "loss": 3.961, + "step": 41940 + }, + { + "epoch": 0.41941, + "grad_norm": 0.6836709380149841, + "learning_rate": 0.003, + "loss": 3.97, + "step": 41941 + }, + { + "epoch": 0.41942, + "grad_norm": 0.7664406895637512, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 41942 + }, + { + "epoch": 0.41943, + "grad_norm": 0.8701636791229248, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 41943 + }, + { + "epoch": 0.41944, + "grad_norm": 0.8948732018470764, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 41944 + }, + { + "epoch": 0.41945, + "grad_norm": 1.1276600360870361, + "learning_rate": 0.003, + "loss": 3.9413, + "step": 41945 + }, + { + "epoch": 0.41946, + "grad_norm": 1.2992409467697144, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 41946 + }, + { + "epoch": 0.41947, + "grad_norm": 0.6025580763816833, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 41947 + }, + { + "epoch": 0.41948, + "grad_norm": 0.7258050441741943, + "learning_rate": 0.003, + "loss": 3.977, + "step": 41948 + }, + { + "epoch": 0.41949, + "grad_norm": 0.75006502866745, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 41949 + }, + { + "epoch": 0.4195, + "grad_norm": 0.7915489673614502, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 41950 + }, + { + "epoch": 0.41951, + "grad_norm": 1.0197511911392212, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 41951 + }, + { + "epoch": 0.41952, + "grad_norm": 1.261157512664795, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 41952 + }, + { + "epoch": 0.41953, + "grad_norm": 0.6605160236358643, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 41953 + }, + { + "epoch": 0.41954, + "grad_norm": 0.7386797070503235, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 41954 + }, + { + "epoch": 0.41955, + "grad_norm": 0.9491297006607056, + "learning_rate": 0.003, + "loss": 3.9478, + "step": 41955 + }, + { + "epoch": 0.41956, + "grad_norm": 1.157981276512146, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 41956 + }, + { + "epoch": 0.41957, + "grad_norm": 0.9997947216033936, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 41957 + }, + { + "epoch": 0.41958, + "grad_norm": 1.13455331325531, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 41958 + }, + { + "epoch": 0.41959, + "grad_norm": 0.9380020499229431, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 41959 + }, + { + "epoch": 0.4196, + "grad_norm": 0.8647164702415466, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 41960 + }, + { + "epoch": 0.41961, + "grad_norm": 1.0094149112701416, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 41961 + }, + { + "epoch": 0.41962, + "grad_norm": 1.0822159051895142, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 41962 + }, + { + "epoch": 0.41963, + "grad_norm": 1.1199344396591187, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 41963 + }, + { + "epoch": 0.41964, + "grad_norm": 0.9775411486625671, + "learning_rate": 0.003, + "loss": 3.982, + "step": 41964 + }, + { + "epoch": 0.41965, + "grad_norm": 1.0544674396514893, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 41965 + }, + { + "epoch": 0.41966, + "grad_norm": 0.8476653695106506, + "learning_rate": 0.003, + "loss": 3.991, + "step": 41966 + }, + { + "epoch": 0.41967, + "grad_norm": 0.9201408624649048, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 41967 + }, + { + "epoch": 0.41968, + "grad_norm": 1.117361068725586, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 41968 + }, + { + "epoch": 0.41969, + "grad_norm": 0.9379578232765198, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 41969 + }, + { + "epoch": 0.4197, + "grad_norm": 0.9865742921829224, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 41970 + }, + { + "epoch": 0.41971, + "grad_norm": 1.3055201768875122, + "learning_rate": 0.003, + "loss": 3.989, + "step": 41971 + }, + { + "epoch": 0.41972, + "grad_norm": 0.9812478423118591, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 41972 + }, + { + "epoch": 0.41973, + "grad_norm": 0.8831146955490112, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 41973 + }, + { + "epoch": 0.41974, + "grad_norm": 0.869031548500061, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 41974 + }, + { + "epoch": 0.41975, + "grad_norm": 1.0522338151931763, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 41975 + }, + { + "epoch": 0.41976, + "grad_norm": 1.0991499423980713, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 41976 + }, + { + "epoch": 0.41977, + "grad_norm": 0.7598135471343994, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 41977 + }, + { + "epoch": 0.41978, + "grad_norm": 0.6961091756820679, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 41978 + }, + { + "epoch": 0.41979, + "grad_norm": 0.8173364400863647, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 41979 + }, + { + "epoch": 0.4198, + "grad_norm": 1.2655153274536133, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 41980 + }, + { + "epoch": 0.41981, + "grad_norm": 0.9381774067878723, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 41981 + }, + { + "epoch": 0.41982, + "grad_norm": 0.92786705493927, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 41982 + }, + { + "epoch": 0.41983, + "grad_norm": 1.0648120641708374, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 41983 + }, + { + "epoch": 0.41984, + "grad_norm": 1.2239007949829102, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 41984 + }, + { + "epoch": 0.41985, + "grad_norm": 0.6307240128517151, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 41985 + }, + { + "epoch": 0.41986, + "grad_norm": 0.6600730419158936, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 41986 + }, + { + "epoch": 0.41987, + "grad_norm": 0.8481485247612, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 41987 + }, + { + "epoch": 0.41988, + "grad_norm": 0.8041019439697266, + "learning_rate": 0.003, + "loss": 3.9393, + "step": 41988 + }, + { + "epoch": 0.41989, + "grad_norm": 0.9070835709571838, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 41989 + }, + { + "epoch": 0.4199, + "grad_norm": 1.0746164321899414, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 41990 + }, + { + "epoch": 0.41991, + "grad_norm": 0.8502286672592163, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 41991 + }, + { + "epoch": 0.41992, + "grad_norm": 0.8337586522102356, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 41992 + }, + { + "epoch": 0.41993, + "grad_norm": 1.01089346408844, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 41993 + }, + { + "epoch": 0.41994, + "grad_norm": 0.9783124327659607, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 41994 + }, + { + "epoch": 0.41995, + "grad_norm": 0.8523246049880981, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 41995 + }, + { + "epoch": 0.41996, + "grad_norm": 0.7789532542228699, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 41996 + }, + { + "epoch": 0.41997, + "grad_norm": 0.824758768081665, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 41997 + }, + { + "epoch": 0.41998, + "grad_norm": 1.0299389362335205, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 41998 + }, + { + "epoch": 0.41999, + "grad_norm": 1.2097073793411255, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 41999 + }, + { + "epoch": 0.42, + "grad_norm": 0.8553361296653748, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 42000 + }, + { + "epoch": 0.42001, + "grad_norm": 0.8478964567184448, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 42001 + }, + { + "epoch": 0.42002, + "grad_norm": 0.7487559914588928, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 42002 + }, + { + "epoch": 0.42003, + "grad_norm": 0.9500738978385925, + "learning_rate": 0.003, + "loss": 3.994, + "step": 42003 + }, + { + "epoch": 0.42004, + "grad_norm": 1.353739857673645, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 42004 + }, + { + "epoch": 0.42005, + "grad_norm": 0.8474625945091248, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 42005 + }, + { + "epoch": 0.42006, + "grad_norm": 0.7681830525398254, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 42006 + }, + { + "epoch": 0.42007, + "grad_norm": 0.8978322744369507, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 42007 + }, + { + "epoch": 0.42008, + "grad_norm": 0.8604069948196411, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 42008 + }, + { + "epoch": 0.42009, + "grad_norm": 0.7012848258018494, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 42009 + }, + { + "epoch": 0.4201, + "grad_norm": 0.8304893374443054, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 42010 + }, + { + "epoch": 0.42011, + "grad_norm": 1.1029130220413208, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 42011 + }, + { + "epoch": 0.42012, + "grad_norm": 1.0835511684417725, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 42012 + }, + { + "epoch": 0.42013, + "grad_norm": 0.9599417448043823, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 42013 + }, + { + "epoch": 0.42014, + "grad_norm": 0.8840417861938477, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 42014 + }, + { + "epoch": 0.42015, + "grad_norm": 0.9611601829528809, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 42015 + }, + { + "epoch": 0.42016, + "grad_norm": 1.15231454372406, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 42016 + }, + { + "epoch": 0.42017, + "grad_norm": 1.002186894416809, + "learning_rate": 0.003, + "loss": 3.9479, + "step": 42017 + }, + { + "epoch": 0.42018, + "grad_norm": 0.9933626055717468, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 42018 + }, + { + "epoch": 0.42019, + "grad_norm": 0.9985564351081848, + "learning_rate": 0.003, + "loss": 3.998, + "step": 42019 + }, + { + "epoch": 0.4202, + "grad_norm": 1.0626851320266724, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 42020 + }, + { + "epoch": 0.42021, + "grad_norm": 0.9577644467353821, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 42021 + }, + { + "epoch": 0.42022, + "grad_norm": 1.0668082237243652, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 42022 + }, + { + "epoch": 0.42023, + "grad_norm": 0.8769780397415161, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 42023 + }, + { + "epoch": 0.42024, + "grad_norm": 0.9133299589157104, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 42024 + }, + { + "epoch": 0.42025, + "grad_norm": 1.0875298976898193, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 42025 + }, + { + "epoch": 0.42026, + "grad_norm": 0.892898678779602, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 42026 + }, + { + "epoch": 0.42027, + "grad_norm": 0.7875732183456421, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 42027 + }, + { + "epoch": 0.42028, + "grad_norm": 0.8242537379264832, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 42028 + }, + { + "epoch": 0.42029, + "grad_norm": 0.8425043225288391, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 42029 + }, + { + "epoch": 0.4203, + "grad_norm": 0.9549620151519775, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 42030 + }, + { + "epoch": 0.42031, + "grad_norm": 1.176029920578003, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 42031 + }, + { + "epoch": 0.42032, + "grad_norm": 1.1294593811035156, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 42032 + }, + { + "epoch": 0.42033, + "grad_norm": 0.9061161279678345, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 42033 + }, + { + "epoch": 0.42034, + "grad_norm": 0.7598848342895508, + "learning_rate": 0.003, + "loss": 3.9449, + "step": 42034 + }, + { + "epoch": 0.42035, + "grad_norm": 0.843571662902832, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 42035 + }, + { + "epoch": 0.42036, + "grad_norm": 0.9222847819328308, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 42036 + }, + { + "epoch": 0.42037, + "grad_norm": 0.7750633955001831, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 42037 + }, + { + "epoch": 0.42038, + "grad_norm": 0.7704171538352966, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 42038 + }, + { + "epoch": 0.42039, + "grad_norm": 0.997468113899231, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 42039 + }, + { + "epoch": 0.4204, + "grad_norm": 1.114558458328247, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 42040 + }, + { + "epoch": 0.42041, + "grad_norm": 1.022201418876648, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 42041 + }, + { + "epoch": 0.42042, + "grad_norm": 1.067073106765747, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 42042 + }, + { + "epoch": 0.42043, + "grad_norm": 0.7695409655570984, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 42043 + }, + { + "epoch": 0.42044, + "grad_norm": 0.6757507920265198, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 42044 + }, + { + "epoch": 0.42045, + "grad_norm": 0.7241597771644592, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 42045 + }, + { + "epoch": 0.42046, + "grad_norm": 0.7094652056694031, + "learning_rate": 0.003, + "loss": 3.9378, + "step": 42046 + }, + { + "epoch": 0.42047, + "grad_norm": 0.7655900716781616, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 42047 + }, + { + "epoch": 0.42048, + "grad_norm": 0.812701404094696, + "learning_rate": 0.003, + "loss": 3.9354, + "step": 42048 + }, + { + "epoch": 0.42049, + "grad_norm": 0.6606895923614502, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 42049 + }, + { + "epoch": 0.4205, + "grad_norm": 0.6952167749404907, + "learning_rate": 0.003, + "loss": 3.97, + "step": 42050 + }, + { + "epoch": 0.42051, + "grad_norm": 0.8852167129516602, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 42051 + }, + { + "epoch": 0.42052, + "grad_norm": 1.2744057178497314, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 42052 + }, + { + "epoch": 0.42053, + "grad_norm": 0.7731963396072388, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 42053 + }, + { + "epoch": 0.42054, + "grad_norm": 0.7000293135643005, + "learning_rate": 0.003, + "loss": 3.987, + "step": 42054 + }, + { + "epoch": 0.42055, + "grad_norm": 0.7190661430358887, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 42055 + }, + { + "epoch": 0.42056, + "grad_norm": 0.7548413276672363, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 42056 + }, + { + "epoch": 0.42057, + "grad_norm": 0.7900766134262085, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 42057 + }, + { + "epoch": 0.42058, + "grad_norm": 0.757445752620697, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 42058 + }, + { + "epoch": 0.42059, + "grad_norm": 0.6769258975982666, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 42059 + }, + { + "epoch": 0.4206, + "grad_norm": 0.6873776912689209, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 42060 + }, + { + "epoch": 0.42061, + "grad_norm": 0.6729027032852173, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 42061 + }, + { + "epoch": 0.42062, + "grad_norm": 0.6245238184928894, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 42062 + }, + { + "epoch": 0.42063, + "grad_norm": 0.6508016586303711, + "learning_rate": 0.003, + "loss": 3.946, + "step": 42063 + }, + { + "epoch": 0.42064, + "grad_norm": 0.6462544202804565, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 42064 + }, + { + "epoch": 0.42065, + "grad_norm": 0.6147368550300598, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 42065 + }, + { + "epoch": 0.42066, + "grad_norm": 0.6498724818229675, + "learning_rate": 0.003, + "loss": 3.938, + "step": 42066 + }, + { + "epoch": 0.42067, + "grad_norm": 0.8277249932289124, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 42067 + }, + { + "epoch": 0.42068, + "grad_norm": 1.2632845640182495, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 42068 + }, + { + "epoch": 0.42069, + "grad_norm": 1.0988842248916626, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 42069 + }, + { + "epoch": 0.4207, + "grad_norm": 1.1335283517837524, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 42070 + }, + { + "epoch": 0.42071, + "grad_norm": 0.8481789231300354, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 42071 + }, + { + "epoch": 0.42072, + "grad_norm": 0.8510912656784058, + "learning_rate": 0.003, + "loss": 3.9438, + "step": 42072 + }, + { + "epoch": 0.42073, + "grad_norm": 0.951499342918396, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 42073 + }, + { + "epoch": 0.42074, + "grad_norm": 1.226568341255188, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 42074 + }, + { + "epoch": 0.42075, + "grad_norm": 0.9445271492004395, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 42075 + }, + { + "epoch": 0.42076, + "grad_norm": 0.8454564809799194, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 42076 + }, + { + "epoch": 0.42077, + "grad_norm": 0.7480685710906982, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 42077 + }, + { + "epoch": 0.42078, + "grad_norm": 0.8525599837303162, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 42078 + }, + { + "epoch": 0.42079, + "grad_norm": 0.9426072835922241, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 42079 + }, + { + "epoch": 0.4208, + "grad_norm": 1.165714144706726, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 42080 + }, + { + "epoch": 0.42081, + "grad_norm": 0.9288713335990906, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 42081 + }, + { + "epoch": 0.42082, + "grad_norm": 1.0195281505584717, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 42082 + }, + { + "epoch": 0.42083, + "grad_norm": 1.2514866590499878, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 42083 + }, + { + "epoch": 0.42084, + "grad_norm": 0.8701124787330627, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 42084 + }, + { + "epoch": 0.42085, + "grad_norm": 0.8780737519264221, + "learning_rate": 0.003, + "loss": 3.9371, + "step": 42085 + }, + { + "epoch": 0.42086, + "grad_norm": 1.0095691680908203, + "learning_rate": 0.003, + "loss": 3.986, + "step": 42086 + }, + { + "epoch": 0.42087, + "grad_norm": 0.9783424735069275, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 42087 + }, + { + "epoch": 0.42088, + "grad_norm": 1.0357651710510254, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 42088 + }, + { + "epoch": 0.42089, + "grad_norm": 0.8889274597167969, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 42089 + }, + { + "epoch": 0.4209, + "grad_norm": 0.9274566769599915, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 42090 + }, + { + "epoch": 0.42091, + "grad_norm": 0.9179766774177551, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 42091 + }, + { + "epoch": 0.42092, + "grad_norm": 0.9766592979431152, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 42092 + }, + { + "epoch": 0.42093, + "grad_norm": 1.0631519556045532, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 42093 + }, + { + "epoch": 0.42094, + "grad_norm": 1.0688247680664062, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 42094 + }, + { + "epoch": 0.42095, + "grad_norm": 0.9726391434669495, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 42095 + }, + { + "epoch": 0.42096, + "grad_norm": 1.1506797075271606, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 42096 + }, + { + "epoch": 0.42097, + "grad_norm": 0.7072702050209045, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 42097 + }, + { + "epoch": 0.42098, + "grad_norm": 0.636873722076416, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 42098 + }, + { + "epoch": 0.42099, + "grad_norm": 0.5981071591377258, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 42099 + }, + { + "epoch": 0.421, + "grad_norm": 0.6107245087623596, + "learning_rate": 0.003, + "loss": 3.955, + "step": 42100 + }, + { + "epoch": 0.42101, + "grad_norm": 0.8832762241363525, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 42101 + }, + { + "epoch": 0.42102, + "grad_norm": 1.1617532968521118, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 42102 + }, + { + "epoch": 0.42103, + "grad_norm": 0.8212234377861023, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 42103 + }, + { + "epoch": 0.42104, + "grad_norm": 0.9080137014389038, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 42104 + }, + { + "epoch": 0.42105, + "grad_norm": 1.2774862051010132, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 42105 + }, + { + "epoch": 0.42106, + "grad_norm": 1.1678370237350464, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 42106 + }, + { + "epoch": 0.42107, + "grad_norm": 1.0050138235092163, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 42107 + }, + { + "epoch": 0.42108, + "grad_norm": 1.0765464305877686, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 42108 + }, + { + "epoch": 0.42109, + "grad_norm": 0.9982749819755554, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 42109 + }, + { + "epoch": 0.4211, + "grad_norm": 1.269875407218933, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 42110 + }, + { + "epoch": 0.42111, + "grad_norm": 1.0128581523895264, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 42111 + }, + { + "epoch": 0.42112, + "grad_norm": 1.0893125534057617, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 42112 + }, + { + "epoch": 0.42113, + "grad_norm": 0.9050706028938293, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 42113 + }, + { + "epoch": 0.42114, + "grad_norm": 1.041722297668457, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 42114 + }, + { + "epoch": 0.42115, + "grad_norm": 1.017285943031311, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 42115 + }, + { + "epoch": 0.42116, + "grad_norm": 1.050196886062622, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 42116 + }, + { + "epoch": 0.42117, + "grad_norm": 0.7247793078422546, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 42117 + }, + { + "epoch": 0.42118, + "grad_norm": 0.6994543075561523, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 42118 + }, + { + "epoch": 0.42119, + "grad_norm": 0.7968862652778625, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 42119 + }, + { + "epoch": 0.4212, + "grad_norm": 0.9543164372444153, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 42120 + }, + { + "epoch": 0.42121, + "grad_norm": 1.116675853729248, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 42121 + }, + { + "epoch": 0.42122, + "grad_norm": 0.8315114974975586, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 42122 + }, + { + "epoch": 0.42123, + "grad_norm": 0.7331727147102356, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 42123 + }, + { + "epoch": 0.42124, + "grad_norm": 0.8072373270988464, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 42124 + }, + { + "epoch": 0.42125, + "grad_norm": 1.0173819065093994, + "learning_rate": 0.003, + "loss": 3.984, + "step": 42125 + }, + { + "epoch": 0.42126, + "grad_norm": 1.1540014743804932, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 42126 + }, + { + "epoch": 0.42127, + "grad_norm": 0.8211039304733276, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 42127 + }, + { + "epoch": 0.42128, + "grad_norm": 0.9367123246192932, + "learning_rate": 0.003, + "loss": 3.98, + "step": 42128 + }, + { + "epoch": 0.42129, + "grad_norm": 0.9667950868606567, + "learning_rate": 0.003, + "loss": 3.978, + "step": 42129 + }, + { + "epoch": 0.4213, + "grad_norm": 1.000628113746643, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 42130 + }, + { + "epoch": 0.42131, + "grad_norm": 0.93967205286026, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 42131 + }, + { + "epoch": 0.42132, + "grad_norm": 0.8921983242034912, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 42132 + }, + { + "epoch": 0.42133, + "grad_norm": 0.802221417427063, + "learning_rate": 0.003, + "loss": 3.9488, + "step": 42133 + }, + { + "epoch": 0.42134, + "grad_norm": 0.757866621017456, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 42134 + }, + { + "epoch": 0.42135, + "grad_norm": 0.6910175681114197, + "learning_rate": 0.003, + "loss": 3.981, + "step": 42135 + }, + { + "epoch": 0.42136, + "grad_norm": 0.6598626971244812, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 42136 + }, + { + "epoch": 0.42137, + "grad_norm": 0.7793872952461243, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 42137 + }, + { + "epoch": 0.42138, + "grad_norm": 1.0425297021865845, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 42138 + }, + { + "epoch": 0.42139, + "grad_norm": 1.3100628852844238, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 42139 + }, + { + "epoch": 0.4214, + "grad_norm": 0.8499155640602112, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 42140 + }, + { + "epoch": 0.42141, + "grad_norm": 1.0519535541534424, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 42141 + }, + { + "epoch": 0.42142, + "grad_norm": 1.198704719543457, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 42142 + }, + { + "epoch": 0.42143, + "grad_norm": 0.8828746676445007, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 42143 + }, + { + "epoch": 0.42144, + "grad_norm": 0.8089920282363892, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 42144 + }, + { + "epoch": 0.42145, + "grad_norm": 0.7958637475967407, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 42145 + }, + { + "epoch": 0.42146, + "grad_norm": 0.8022576570510864, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 42146 + }, + { + "epoch": 0.42147, + "grad_norm": 0.7965492010116577, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 42147 + }, + { + "epoch": 0.42148, + "grad_norm": 0.9260627031326294, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 42148 + }, + { + "epoch": 0.42149, + "grad_norm": 1.1746397018432617, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 42149 + }, + { + "epoch": 0.4215, + "grad_norm": 1.0714595317840576, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 42150 + }, + { + "epoch": 0.42151, + "grad_norm": 1.020939588546753, + "learning_rate": 0.003, + "loss": 3.9422, + "step": 42151 + }, + { + "epoch": 0.42152, + "grad_norm": 0.9052588939666748, + "learning_rate": 0.003, + "loss": 3.943, + "step": 42152 + }, + { + "epoch": 0.42153, + "grad_norm": 0.855438768863678, + "learning_rate": 0.003, + "loss": 3.951, + "step": 42153 + }, + { + "epoch": 0.42154, + "grad_norm": 0.8144277930259705, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 42154 + }, + { + "epoch": 0.42155, + "grad_norm": 0.7369678616523743, + "learning_rate": 0.003, + "loss": 3.944, + "step": 42155 + }, + { + "epoch": 0.42156, + "grad_norm": 0.7680621147155762, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 42156 + }, + { + "epoch": 0.42157, + "grad_norm": 1.0080255270004272, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 42157 + }, + { + "epoch": 0.42158, + "grad_norm": 1.176808476448059, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 42158 + }, + { + "epoch": 0.42159, + "grad_norm": 0.9600800275802612, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 42159 + }, + { + "epoch": 0.4216, + "grad_norm": 1.011674404144287, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 42160 + }, + { + "epoch": 0.42161, + "grad_norm": 0.8736683130264282, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 42161 + }, + { + "epoch": 0.42162, + "grad_norm": 0.7304939031600952, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 42162 + }, + { + "epoch": 0.42163, + "grad_norm": 0.6337166428565979, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 42163 + }, + { + "epoch": 0.42164, + "grad_norm": 0.6555385589599609, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 42164 + }, + { + "epoch": 0.42165, + "grad_norm": 0.6858662366867065, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 42165 + }, + { + "epoch": 0.42166, + "grad_norm": 0.8321378231048584, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 42166 + }, + { + "epoch": 0.42167, + "grad_norm": 0.9731445908546448, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 42167 + }, + { + "epoch": 0.42168, + "grad_norm": 1.0350679159164429, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 42168 + }, + { + "epoch": 0.42169, + "grad_norm": 1.0281016826629639, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 42169 + }, + { + "epoch": 0.4217, + "grad_norm": 1.0293238162994385, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 42170 + }, + { + "epoch": 0.42171, + "grad_norm": 0.9705419540405273, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 42171 + }, + { + "epoch": 0.42172, + "grad_norm": 0.7628311514854431, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 42172 + }, + { + "epoch": 0.42173, + "grad_norm": 0.6929522752761841, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 42173 + }, + { + "epoch": 0.42174, + "grad_norm": 0.8888788819313049, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 42174 + }, + { + "epoch": 0.42175, + "grad_norm": 1.0699025392532349, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 42175 + }, + { + "epoch": 0.42176, + "grad_norm": 1.1590551137924194, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 42176 + }, + { + "epoch": 0.42177, + "grad_norm": 1.0900390148162842, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 42177 + }, + { + "epoch": 0.42178, + "grad_norm": 0.9458721280097961, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 42178 + }, + { + "epoch": 0.42179, + "grad_norm": 0.7924609184265137, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 42179 + }, + { + "epoch": 0.4218, + "grad_norm": 0.7753890752792358, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 42180 + }, + { + "epoch": 0.42181, + "grad_norm": 0.8370930552482605, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 42181 + }, + { + "epoch": 0.42182, + "grad_norm": 0.8798096179962158, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 42182 + }, + { + "epoch": 0.42183, + "grad_norm": 0.7772161364555359, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 42183 + }, + { + "epoch": 0.42184, + "grad_norm": 0.7968723773956299, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 42184 + }, + { + "epoch": 0.42185, + "grad_norm": 0.9307878613471985, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 42185 + }, + { + "epoch": 0.42186, + "grad_norm": 1.1758692264556885, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 42186 + }, + { + "epoch": 0.42187, + "grad_norm": 0.9195138216018677, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 42187 + }, + { + "epoch": 0.42188, + "grad_norm": 0.907511293888092, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 42188 + }, + { + "epoch": 0.42189, + "grad_norm": 0.9915151000022888, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 42189 + }, + { + "epoch": 0.4219, + "grad_norm": 1.0755826234817505, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 42190 + }, + { + "epoch": 0.42191, + "grad_norm": 0.9209916591644287, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 42191 + }, + { + "epoch": 0.42192, + "grad_norm": 0.8337661623954773, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 42192 + }, + { + "epoch": 0.42193, + "grad_norm": 0.7988234162330627, + "learning_rate": 0.003, + "loss": 3.9429, + "step": 42193 + }, + { + "epoch": 0.42194, + "grad_norm": 0.6748762130737305, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 42194 + }, + { + "epoch": 0.42195, + "grad_norm": 0.7309790253639221, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 42195 + }, + { + "epoch": 0.42196, + "grad_norm": 0.8322020769119263, + "learning_rate": 0.003, + "loss": 3.957, + "step": 42196 + }, + { + "epoch": 0.42197, + "grad_norm": 1.0242667198181152, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 42197 + }, + { + "epoch": 0.42198, + "grad_norm": 1.4329049587249756, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 42198 + }, + { + "epoch": 0.42199, + "grad_norm": 0.538928747177124, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 42199 + }, + { + "epoch": 0.422, + "grad_norm": 0.9254873394966125, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 42200 + }, + { + "epoch": 0.42201, + "grad_norm": 1.4786734580993652, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 42201 + }, + { + "epoch": 0.42202, + "grad_norm": 0.4303119480609894, + "learning_rate": 0.003, + "loss": 3.992, + "step": 42202 + }, + { + "epoch": 0.42203, + "grad_norm": 0.8972743153572083, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 42203 + }, + { + "epoch": 0.42204, + "grad_norm": 1.0091710090637207, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 42204 + }, + { + "epoch": 0.42205, + "grad_norm": 0.8755699396133423, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 42205 + }, + { + "epoch": 0.42206, + "grad_norm": 0.7254472970962524, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 42206 + }, + { + "epoch": 0.42207, + "grad_norm": 0.7757805585861206, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 42207 + }, + { + "epoch": 0.42208, + "grad_norm": 1.0596511363983154, + "learning_rate": 0.003, + "loss": 3.985, + "step": 42208 + }, + { + "epoch": 0.42209, + "grad_norm": 1.0401813983917236, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 42209 + }, + { + "epoch": 0.4221, + "grad_norm": 0.8921640515327454, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 42210 + }, + { + "epoch": 0.42211, + "grad_norm": 1.0626922845840454, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 42211 + }, + { + "epoch": 0.42212, + "grad_norm": 1.0968167781829834, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 42212 + }, + { + "epoch": 0.42213, + "grad_norm": 0.90384840965271, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 42213 + }, + { + "epoch": 0.42214, + "grad_norm": 0.8501759767532349, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 42214 + }, + { + "epoch": 0.42215, + "grad_norm": 0.8421027660369873, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 42215 + }, + { + "epoch": 0.42216, + "grad_norm": 0.8280860185623169, + "learning_rate": 0.003, + "loss": 3.938, + "step": 42216 + }, + { + "epoch": 0.42217, + "grad_norm": 0.7945264577865601, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 42217 + }, + { + "epoch": 0.42218, + "grad_norm": 0.8470200896263123, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 42218 + }, + { + "epoch": 0.42219, + "grad_norm": 0.8418667912483215, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 42219 + }, + { + "epoch": 0.4222, + "grad_norm": 0.9157212972640991, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 42220 + }, + { + "epoch": 0.42221, + "grad_norm": 1.0813641548156738, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 42221 + }, + { + "epoch": 0.42222, + "grad_norm": 0.9749755859375, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 42222 + }, + { + "epoch": 0.42223, + "grad_norm": 0.897399365901947, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 42223 + }, + { + "epoch": 0.42224, + "grad_norm": 1.1123144626617432, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 42224 + }, + { + "epoch": 0.42225, + "grad_norm": 1.0209912061691284, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 42225 + }, + { + "epoch": 0.42226, + "grad_norm": 0.9417264461517334, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 42226 + }, + { + "epoch": 0.42227, + "grad_norm": 1.1644960641860962, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 42227 + }, + { + "epoch": 0.42228, + "grad_norm": 1.1384679079055786, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 42228 + }, + { + "epoch": 0.42229, + "grad_norm": 0.9512472748756409, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 42229 + }, + { + "epoch": 0.4223, + "grad_norm": 0.9098550081253052, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 42230 + }, + { + "epoch": 0.42231, + "grad_norm": 1.0044158697128296, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 42231 + }, + { + "epoch": 0.42232, + "grad_norm": 1.1180957555770874, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 42232 + }, + { + "epoch": 0.42233, + "grad_norm": 0.8137633800506592, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 42233 + }, + { + "epoch": 0.42234, + "grad_norm": 0.7708417177200317, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 42234 + }, + { + "epoch": 0.42235, + "grad_norm": 0.8223283886909485, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 42235 + }, + { + "epoch": 0.42236, + "grad_norm": 0.832150936126709, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 42236 + }, + { + "epoch": 0.42237, + "grad_norm": 0.9179397821426392, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 42237 + }, + { + "epoch": 0.42238, + "grad_norm": 1.0318247079849243, + "learning_rate": 0.003, + "loss": 3.971, + "step": 42238 + }, + { + "epoch": 0.42239, + "grad_norm": 0.924018144607544, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 42239 + }, + { + "epoch": 0.4224, + "grad_norm": 0.8152686953544617, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 42240 + }, + { + "epoch": 0.42241, + "grad_norm": 0.6895917057991028, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 42241 + }, + { + "epoch": 0.42242, + "grad_norm": 0.7180904746055603, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 42242 + }, + { + "epoch": 0.42243, + "grad_norm": 0.735775351524353, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 42243 + }, + { + "epoch": 0.42244, + "grad_norm": 0.6763343214988708, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 42244 + }, + { + "epoch": 0.42245, + "grad_norm": 0.5667502880096436, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 42245 + }, + { + "epoch": 0.42246, + "grad_norm": 0.5419217944145203, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 42246 + }, + { + "epoch": 0.42247, + "grad_norm": 0.6040149331092834, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 42247 + }, + { + "epoch": 0.42248, + "grad_norm": 0.7310181260108948, + "learning_rate": 0.003, + "loss": 3.968, + "step": 42248 + }, + { + "epoch": 0.42249, + "grad_norm": 0.9922952055931091, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 42249 + }, + { + "epoch": 0.4225, + "grad_norm": 1.456612229347229, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 42250 + }, + { + "epoch": 0.42251, + "grad_norm": 0.811983585357666, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 42251 + }, + { + "epoch": 0.42252, + "grad_norm": 0.6964582800865173, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 42252 + }, + { + "epoch": 0.42253, + "grad_norm": 0.6373400688171387, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 42253 + }, + { + "epoch": 0.42254, + "grad_norm": 0.8124823570251465, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 42254 + }, + { + "epoch": 0.42255, + "grad_norm": 0.8743725419044495, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 42255 + }, + { + "epoch": 0.42256, + "grad_norm": 1.1834858655929565, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 42256 + }, + { + "epoch": 0.42257, + "grad_norm": 0.9429344534873962, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 42257 + }, + { + "epoch": 0.42258, + "grad_norm": 0.8897679448127747, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 42258 + }, + { + "epoch": 0.42259, + "grad_norm": 1.006256103515625, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 42259 + }, + { + "epoch": 0.4226, + "grad_norm": 1.045028567314148, + "learning_rate": 0.003, + "loss": 3.9449, + "step": 42260 + }, + { + "epoch": 0.42261, + "grad_norm": 0.8663146495819092, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 42261 + }, + { + "epoch": 0.42262, + "grad_norm": 0.9585965871810913, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 42262 + }, + { + "epoch": 0.42263, + "grad_norm": 0.921161949634552, + "learning_rate": 0.003, + "loss": 3.999, + "step": 42263 + }, + { + "epoch": 0.42264, + "grad_norm": 0.8235267996788025, + "learning_rate": 0.003, + "loss": 3.993, + "step": 42264 + }, + { + "epoch": 0.42265, + "grad_norm": 0.9038822650909424, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 42265 + }, + { + "epoch": 0.42266, + "grad_norm": 0.9448302388191223, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 42266 + }, + { + "epoch": 0.42267, + "grad_norm": 0.9293007254600525, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 42267 + }, + { + "epoch": 0.42268, + "grad_norm": 0.9735960364341736, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 42268 + }, + { + "epoch": 0.42269, + "grad_norm": 1.051993489265442, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 42269 + }, + { + "epoch": 0.4227, + "grad_norm": 1.3815503120422363, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 42270 + }, + { + "epoch": 0.42271, + "grad_norm": 0.9671689867973328, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 42271 + }, + { + "epoch": 0.42272, + "grad_norm": 0.9966925978660583, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 42272 + }, + { + "epoch": 0.42273, + "grad_norm": 1.2597086429595947, + "learning_rate": 0.003, + "loss": 4.002, + "step": 42273 + }, + { + "epoch": 0.42274, + "grad_norm": 0.8165028095245361, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 42274 + }, + { + "epoch": 0.42275, + "grad_norm": 0.8265511393547058, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 42275 + }, + { + "epoch": 0.42276, + "grad_norm": 0.9298005104064941, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 42276 + }, + { + "epoch": 0.42277, + "grad_norm": 1.015591025352478, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 42277 + }, + { + "epoch": 0.42278, + "grad_norm": 1.0115445852279663, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 42278 + }, + { + "epoch": 0.42279, + "grad_norm": 1.0032286643981934, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 42279 + }, + { + "epoch": 0.4228, + "grad_norm": 0.7531154751777649, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 42280 + }, + { + "epoch": 0.42281, + "grad_norm": 0.689769446849823, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 42281 + }, + { + "epoch": 0.42282, + "grad_norm": 0.8072424530982971, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 42282 + }, + { + "epoch": 0.42283, + "grad_norm": 0.8517639636993408, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 42283 + }, + { + "epoch": 0.42284, + "grad_norm": 0.8800426721572876, + "learning_rate": 0.003, + "loss": 3.947, + "step": 42284 + }, + { + "epoch": 0.42285, + "grad_norm": 0.8093209266662598, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 42285 + }, + { + "epoch": 0.42286, + "grad_norm": 0.717248260974884, + "learning_rate": 0.003, + "loss": 3.976, + "step": 42286 + }, + { + "epoch": 0.42287, + "grad_norm": 0.8465209007263184, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 42287 + }, + { + "epoch": 0.42288, + "grad_norm": 0.9425678253173828, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 42288 + }, + { + "epoch": 0.42289, + "grad_norm": 1.1492911577224731, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 42289 + }, + { + "epoch": 0.4229, + "grad_norm": 1.0931999683380127, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 42290 + }, + { + "epoch": 0.42291, + "grad_norm": 0.8781285285949707, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 42291 + }, + { + "epoch": 0.42292, + "grad_norm": 0.9057723879814148, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 42292 + }, + { + "epoch": 0.42293, + "grad_norm": 0.908826470375061, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 42293 + }, + { + "epoch": 0.42294, + "grad_norm": 0.8360826969146729, + "learning_rate": 0.003, + "loss": 3.969, + "step": 42294 + }, + { + "epoch": 0.42295, + "grad_norm": 0.9688440561294556, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 42295 + }, + { + "epoch": 0.42296, + "grad_norm": 0.9240896105766296, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 42296 + }, + { + "epoch": 0.42297, + "grad_norm": 0.8592174053192139, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 42297 + }, + { + "epoch": 0.42298, + "grad_norm": 1.2000943422317505, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 42298 + }, + { + "epoch": 0.42299, + "grad_norm": 0.8934785723686218, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 42299 + }, + { + "epoch": 0.423, + "grad_norm": 0.8475959300994873, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 42300 + }, + { + "epoch": 0.42301, + "grad_norm": 0.9050998091697693, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 42301 + }, + { + "epoch": 0.42302, + "grad_norm": 1.218196988105774, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 42302 + }, + { + "epoch": 0.42303, + "grad_norm": 0.6818972826004028, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 42303 + }, + { + "epoch": 0.42304, + "grad_norm": 0.5688571929931641, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 42304 + }, + { + "epoch": 0.42305, + "grad_norm": 0.6594070792198181, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 42305 + }, + { + "epoch": 0.42306, + "grad_norm": 0.682380735874176, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 42306 + }, + { + "epoch": 0.42307, + "grad_norm": 0.72989821434021, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 42307 + }, + { + "epoch": 0.42308, + "grad_norm": 0.7634022235870361, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 42308 + }, + { + "epoch": 0.42309, + "grad_norm": 0.8539128303527832, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 42309 + }, + { + "epoch": 0.4231, + "grad_norm": 0.9720349311828613, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 42310 + }, + { + "epoch": 0.42311, + "grad_norm": 0.9479154348373413, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 42311 + }, + { + "epoch": 0.42312, + "grad_norm": 0.9847483038902283, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 42312 + }, + { + "epoch": 0.42313, + "grad_norm": 0.9214261174201965, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 42313 + }, + { + "epoch": 0.42314, + "grad_norm": 0.952075183391571, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 42314 + }, + { + "epoch": 0.42315, + "grad_norm": 1.0641480684280396, + "learning_rate": 0.003, + "loss": 3.984, + "step": 42315 + }, + { + "epoch": 0.42316, + "grad_norm": 0.9447535872459412, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 42316 + }, + { + "epoch": 0.42317, + "grad_norm": 0.932114839553833, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 42317 + }, + { + "epoch": 0.42318, + "grad_norm": 0.9862046241760254, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 42318 + }, + { + "epoch": 0.42319, + "grad_norm": 1.3459359407424927, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 42319 + }, + { + "epoch": 0.4232, + "grad_norm": 1.1201785802841187, + "learning_rate": 0.003, + "loss": 3.958, + "step": 42320 + }, + { + "epoch": 0.42321, + "grad_norm": 1.004422664642334, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 42321 + }, + { + "epoch": 0.42322, + "grad_norm": 1.043755054473877, + "learning_rate": 0.003, + "loss": 4.011, + "step": 42322 + }, + { + "epoch": 0.42323, + "grad_norm": 1.080659031867981, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 42323 + }, + { + "epoch": 0.42324, + "grad_norm": 0.8309747576713562, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 42324 + }, + { + "epoch": 0.42325, + "grad_norm": 0.7515255808830261, + "learning_rate": 0.003, + "loss": 3.973, + "step": 42325 + }, + { + "epoch": 0.42326, + "grad_norm": 0.8421974778175354, + "learning_rate": 0.003, + "loss": 3.9458, + "step": 42326 + }, + { + "epoch": 0.42327, + "grad_norm": 0.9276965856552124, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 42327 + }, + { + "epoch": 0.42328, + "grad_norm": 1.167865514755249, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 42328 + }, + { + "epoch": 0.42329, + "grad_norm": 0.9170730710029602, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 42329 + }, + { + "epoch": 0.4233, + "grad_norm": 1.0224519968032837, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 42330 + }, + { + "epoch": 0.42331, + "grad_norm": 1.1010164022445679, + "learning_rate": 0.003, + "loss": 3.977, + "step": 42331 + }, + { + "epoch": 0.42332, + "grad_norm": 0.8804015517234802, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 42332 + }, + { + "epoch": 0.42333, + "grad_norm": 0.8042630553245544, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 42333 + }, + { + "epoch": 0.42334, + "grad_norm": 0.7424907684326172, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 42334 + }, + { + "epoch": 0.42335, + "grad_norm": 0.8085160255432129, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 42335 + }, + { + "epoch": 0.42336, + "grad_norm": 0.8425464630126953, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 42336 + }, + { + "epoch": 0.42337, + "grad_norm": 0.9465603232383728, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 42337 + }, + { + "epoch": 0.42338, + "grad_norm": 1.1862263679504395, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 42338 + }, + { + "epoch": 0.42339, + "grad_norm": 0.9766764044761658, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 42339 + }, + { + "epoch": 0.4234, + "grad_norm": 0.8985991477966309, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 42340 + }, + { + "epoch": 0.42341, + "grad_norm": 0.7668522000312805, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 42341 + }, + { + "epoch": 0.42342, + "grad_norm": 0.6892523765563965, + "learning_rate": 0.003, + "loss": 3.962, + "step": 42342 + }, + { + "epoch": 0.42343, + "grad_norm": 0.770885169506073, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 42343 + }, + { + "epoch": 0.42344, + "grad_norm": 0.8675910830497742, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 42344 + }, + { + "epoch": 0.42345, + "grad_norm": 0.8391100764274597, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 42345 + }, + { + "epoch": 0.42346, + "grad_norm": 0.8025185465812683, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 42346 + }, + { + "epoch": 0.42347, + "grad_norm": 0.7664393782615662, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 42347 + }, + { + "epoch": 0.42348, + "grad_norm": 0.7784303426742554, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 42348 + }, + { + "epoch": 0.42349, + "grad_norm": 0.8904587626457214, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 42349 + }, + { + "epoch": 0.4235, + "grad_norm": 0.8966587781906128, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 42350 + }, + { + "epoch": 0.42351, + "grad_norm": 0.9417306780815125, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 42351 + }, + { + "epoch": 0.42352, + "grad_norm": 1.2374125719070435, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 42352 + }, + { + "epoch": 0.42353, + "grad_norm": 1.094472050666809, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 42353 + }, + { + "epoch": 0.42354, + "grad_norm": 1.1741513013839722, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 42354 + }, + { + "epoch": 0.42355, + "grad_norm": 0.8202704787254333, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 42355 + }, + { + "epoch": 0.42356, + "grad_norm": 0.7587077021598816, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 42356 + }, + { + "epoch": 0.42357, + "grad_norm": 0.8632664680480957, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 42357 + }, + { + "epoch": 0.42358, + "grad_norm": 0.8470784425735474, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 42358 + }, + { + "epoch": 0.42359, + "grad_norm": 0.7853544354438782, + "learning_rate": 0.003, + "loss": 3.998, + "step": 42359 + }, + { + "epoch": 0.4236, + "grad_norm": 0.8185623288154602, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 42360 + }, + { + "epoch": 0.42361, + "grad_norm": 0.8938801288604736, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 42361 + }, + { + "epoch": 0.42362, + "grad_norm": 0.8323565721511841, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 42362 + }, + { + "epoch": 0.42363, + "grad_norm": 0.9913350343704224, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 42363 + }, + { + "epoch": 0.42364, + "grad_norm": 1.339945673942566, + "learning_rate": 0.003, + "loss": 3.9493, + "step": 42364 + }, + { + "epoch": 0.42365, + "grad_norm": 0.8915945291519165, + "learning_rate": 0.003, + "loss": 3.967, + "step": 42365 + }, + { + "epoch": 0.42366, + "grad_norm": 0.7482735514640808, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 42366 + }, + { + "epoch": 0.42367, + "grad_norm": 0.7807809114456177, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 42367 + }, + { + "epoch": 0.42368, + "grad_norm": 0.8227871060371399, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 42368 + }, + { + "epoch": 0.42369, + "grad_norm": 0.9306555986404419, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 42369 + }, + { + "epoch": 0.4237, + "grad_norm": 1.0461686849594116, + "learning_rate": 0.003, + "loss": 3.9424, + "step": 42370 + }, + { + "epoch": 0.42371, + "grad_norm": 0.9909291863441467, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 42371 + }, + { + "epoch": 0.42372, + "grad_norm": 1.1293613910675049, + "learning_rate": 0.003, + "loss": 3.98, + "step": 42372 + }, + { + "epoch": 0.42373, + "grad_norm": 1.0315203666687012, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 42373 + }, + { + "epoch": 0.42374, + "grad_norm": 1.2634459733963013, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 42374 + }, + { + "epoch": 0.42375, + "grad_norm": 0.8617721199989319, + "learning_rate": 0.003, + "loss": 3.962, + "step": 42375 + }, + { + "epoch": 0.42376, + "grad_norm": 0.7531991004943848, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 42376 + }, + { + "epoch": 0.42377, + "grad_norm": 0.9094658493995667, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 42377 + }, + { + "epoch": 0.42378, + "grad_norm": 1.042626142501831, + "learning_rate": 0.003, + "loss": 3.966, + "step": 42378 + }, + { + "epoch": 0.42379, + "grad_norm": 1.0038442611694336, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 42379 + }, + { + "epoch": 0.4238, + "grad_norm": 1.0407429933547974, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 42380 + }, + { + "epoch": 0.42381, + "grad_norm": 1.1857755184173584, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 42381 + }, + { + "epoch": 0.42382, + "grad_norm": 0.8804183602333069, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 42382 + }, + { + "epoch": 0.42383, + "grad_norm": 0.6715772747993469, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 42383 + }, + { + "epoch": 0.42384, + "grad_norm": 0.678072988986969, + "learning_rate": 0.003, + "loss": 4.012, + "step": 42384 + }, + { + "epoch": 0.42385, + "grad_norm": 0.6478498578071594, + "learning_rate": 0.003, + "loss": 3.9387, + "step": 42385 + }, + { + "epoch": 0.42386, + "grad_norm": 0.608386218547821, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 42386 + }, + { + "epoch": 0.42387, + "grad_norm": 0.5998275876045227, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 42387 + }, + { + "epoch": 0.42388, + "grad_norm": 0.6548370718955994, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 42388 + }, + { + "epoch": 0.42389, + "grad_norm": 0.8541951775550842, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 42389 + }, + { + "epoch": 0.4239, + "grad_norm": 1.2063632011413574, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 42390 + }, + { + "epoch": 0.42391, + "grad_norm": 0.9830247163772583, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 42391 + }, + { + "epoch": 0.42392, + "grad_norm": 1.0146644115447998, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 42392 + }, + { + "epoch": 0.42393, + "grad_norm": 1.0345818996429443, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 42393 + }, + { + "epoch": 0.42394, + "grad_norm": 1.0778769254684448, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 42394 + }, + { + "epoch": 0.42395, + "grad_norm": 0.9699714183807373, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 42395 + }, + { + "epoch": 0.42396, + "grad_norm": 1.0434333086013794, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 42396 + }, + { + "epoch": 0.42397, + "grad_norm": 0.9414575695991516, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 42397 + }, + { + "epoch": 0.42398, + "grad_norm": 1.0913318395614624, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 42398 + }, + { + "epoch": 0.42399, + "grad_norm": 1.0126910209655762, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 42399 + }, + { + "epoch": 0.424, + "grad_norm": 0.9927183985710144, + "learning_rate": 0.003, + "loss": 3.986, + "step": 42400 + }, + { + "epoch": 0.42401, + "grad_norm": 0.9343582987785339, + "learning_rate": 0.003, + "loss": 3.982, + "step": 42401 + }, + { + "epoch": 0.42402, + "grad_norm": 0.9037278890609741, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 42402 + }, + { + "epoch": 0.42403, + "grad_norm": 0.8152856230735779, + "learning_rate": 0.003, + "loss": 3.9445, + "step": 42403 + }, + { + "epoch": 0.42404, + "grad_norm": 0.8999679088592529, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 42404 + }, + { + "epoch": 0.42405, + "grad_norm": 1.0117456912994385, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 42405 + }, + { + "epoch": 0.42406, + "grad_norm": 1.2089875936508179, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 42406 + }, + { + "epoch": 0.42407, + "grad_norm": 0.7993122339248657, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 42407 + }, + { + "epoch": 0.42408, + "grad_norm": 0.736745297908783, + "learning_rate": 0.003, + "loss": 3.945, + "step": 42408 + }, + { + "epoch": 0.42409, + "grad_norm": 0.8009971380233765, + "learning_rate": 0.003, + "loss": 3.994, + "step": 42409 + }, + { + "epoch": 0.4241, + "grad_norm": 0.8610479235649109, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 42410 + }, + { + "epoch": 0.42411, + "grad_norm": 0.7657590508460999, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 42411 + }, + { + "epoch": 0.42412, + "grad_norm": 0.7728025913238525, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 42412 + }, + { + "epoch": 0.42413, + "grad_norm": 0.9443517327308655, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 42413 + }, + { + "epoch": 0.42414, + "grad_norm": 1.2204989194869995, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 42414 + }, + { + "epoch": 0.42415, + "grad_norm": 0.7181851863861084, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 42415 + }, + { + "epoch": 0.42416, + "grad_norm": 0.597356915473938, + "learning_rate": 0.003, + "loss": 3.9299, + "step": 42416 + }, + { + "epoch": 0.42417, + "grad_norm": 0.692503035068512, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 42417 + }, + { + "epoch": 0.42418, + "grad_norm": 0.9295587539672852, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 42418 + }, + { + "epoch": 0.42419, + "grad_norm": 0.7607778906822205, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 42419 + }, + { + "epoch": 0.4242, + "grad_norm": 0.7597779035568237, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 42420 + }, + { + "epoch": 0.42421, + "grad_norm": 0.9616101980209351, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 42421 + }, + { + "epoch": 0.42422, + "grad_norm": 1.0976629257202148, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 42422 + }, + { + "epoch": 0.42423, + "grad_norm": 0.9426258206367493, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 42423 + }, + { + "epoch": 0.42424, + "grad_norm": 1.156848430633545, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 42424 + }, + { + "epoch": 0.42425, + "grad_norm": 1.0068848133087158, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 42425 + }, + { + "epoch": 0.42426, + "grad_norm": 0.9687142968177795, + "learning_rate": 0.003, + "loss": 3.962, + "step": 42426 + }, + { + "epoch": 0.42427, + "grad_norm": 1.0031180381774902, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 42427 + }, + { + "epoch": 0.42428, + "grad_norm": 1.1196359395980835, + "learning_rate": 0.003, + "loss": 3.978, + "step": 42428 + }, + { + "epoch": 0.42429, + "grad_norm": 0.8677242398262024, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 42429 + }, + { + "epoch": 0.4243, + "grad_norm": 0.8749775290489197, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 42430 + }, + { + "epoch": 0.42431, + "grad_norm": 1.3229771852493286, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 42431 + }, + { + "epoch": 0.42432, + "grad_norm": 1.0483543872833252, + "learning_rate": 0.003, + "loss": 3.976, + "step": 42432 + }, + { + "epoch": 0.42433, + "grad_norm": 1.2433435916900635, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 42433 + }, + { + "epoch": 0.42434, + "grad_norm": 0.7156078815460205, + "learning_rate": 0.003, + "loss": 3.98, + "step": 42434 + }, + { + "epoch": 0.42435, + "grad_norm": 0.6325322389602661, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 42435 + }, + { + "epoch": 0.42436, + "grad_norm": 0.7991132736206055, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 42436 + }, + { + "epoch": 0.42437, + "grad_norm": 1.0825568437576294, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 42437 + }, + { + "epoch": 0.42438, + "grad_norm": 1.240855097770691, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 42438 + }, + { + "epoch": 0.42439, + "grad_norm": 0.5675491094589233, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 42439 + }, + { + "epoch": 0.4244, + "grad_norm": 0.777411699295044, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 42440 + }, + { + "epoch": 0.42441, + "grad_norm": 0.9686617851257324, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 42441 + }, + { + "epoch": 0.42442, + "grad_norm": 0.989738941192627, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 42442 + }, + { + "epoch": 0.42443, + "grad_norm": 0.8371227979660034, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 42443 + }, + { + "epoch": 0.42444, + "grad_norm": 0.697815477848053, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 42444 + }, + { + "epoch": 0.42445, + "grad_norm": 0.7431186437606812, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 42445 + }, + { + "epoch": 0.42446, + "grad_norm": 0.8334270715713501, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 42446 + }, + { + "epoch": 0.42447, + "grad_norm": 1.0193778276443481, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 42447 + }, + { + "epoch": 0.42448, + "grad_norm": 1.0224868059158325, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 42448 + }, + { + "epoch": 0.42449, + "grad_norm": 1.0896302461624146, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 42449 + }, + { + "epoch": 0.4245, + "grad_norm": 1.1238951683044434, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 42450 + }, + { + "epoch": 0.42451, + "grad_norm": 0.9760202169418335, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 42451 + }, + { + "epoch": 0.42452, + "grad_norm": 1.0947158336639404, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 42452 + }, + { + "epoch": 0.42453, + "grad_norm": 1.0832711458206177, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 42453 + }, + { + "epoch": 0.42454, + "grad_norm": 0.8536838293075562, + "learning_rate": 0.003, + "loss": 3.989, + "step": 42454 + }, + { + "epoch": 0.42455, + "grad_norm": 0.8616016507148743, + "learning_rate": 0.003, + "loss": 3.975, + "step": 42455 + }, + { + "epoch": 0.42456, + "grad_norm": 1.002132773399353, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 42456 + }, + { + "epoch": 0.42457, + "grad_norm": 1.0922783613204956, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 42457 + }, + { + "epoch": 0.42458, + "grad_norm": 0.9097080826759338, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 42458 + }, + { + "epoch": 0.42459, + "grad_norm": 0.9215481281280518, + "learning_rate": 0.003, + "loss": 3.947, + "step": 42459 + }, + { + "epoch": 0.4246, + "grad_norm": 1.142515778541565, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 42460 + }, + { + "epoch": 0.42461, + "grad_norm": 1.0249218940734863, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 42461 + }, + { + "epoch": 0.42462, + "grad_norm": 1.062357783317566, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 42462 + }, + { + "epoch": 0.42463, + "grad_norm": 0.9361176490783691, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 42463 + }, + { + "epoch": 0.42464, + "grad_norm": 0.9326622486114502, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 42464 + }, + { + "epoch": 0.42465, + "grad_norm": 1.1431374549865723, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 42465 + }, + { + "epoch": 0.42466, + "grad_norm": 0.8079103827476501, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 42466 + }, + { + "epoch": 0.42467, + "grad_norm": 0.6932924389839172, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 42467 + }, + { + "epoch": 0.42468, + "grad_norm": 0.895732581615448, + "learning_rate": 0.003, + "loss": 4.001, + "step": 42468 + }, + { + "epoch": 0.42469, + "grad_norm": 1.155074954032898, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 42469 + }, + { + "epoch": 0.4247, + "grad_norm": 0.6802091002464294, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 42470 + }, + { + "epoch": 0.42471, + "grad_norm": 0.6318747997283936, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 42471 + }, + { + "epoch": 0.42472, + "grad_norm": 0.689579963684082, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 42472 + }, + { + "epoch": 0.42473, + "grad_norm": 0.6817358732223511, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 42473 + }, + { + "epoch": 0.42474, + "grad_norm": 0.7320082783699036, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 42474 + }, + { + "epoch": 0.42475, + "grad_norm": 0.7198011875152588, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 42475 + }, + { + "epoch": 0.42476, + "grad_norm": 0.7824849486351013, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 42476 + }, + { + "epoch": 0.42477, + "grad_norm": 0.8839243650436401, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 42477 + }, + { + "epoch": 0.42478, + "grad_norm": 1.0819629430770874, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 42478 + }, + { + "epoch": 0.42479, + "grad_norm": 0.9480621814727783, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 42479 + }, + { + "epoch": 0.4248, + "grad_norm": 0.7403396368026733, + "learning_rate": 0.003, + "loss": 3.972, + "step": 42480 + }, + { + "epoch": 0.42481, + "grad_norm": 0.7214800715446472, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 42481 + }, + { + "epoch": 0.42482, + "grad_norm": 0.6483220458030701, + "learning_rate": 0.003, + "loss": 3.9288, + "step": 42482 + }, + { + "epoch": 0.42483, + "grad_norm": 0.6435839533805847, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 42483 + }, + { + "epoch": 0.42484, + "grad_norm": 0.6922389268875122, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 42484 + }, + { + "epoch": 0.42485, + "grad_norm": 0.8091126084327698, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 42485 + }, + { + "epoch": 0.42486, + "grad_norm": 1.029704213142395, + "learning_rate": 0.003, + "loss": 3.9334, + "step": 42486 + }, + { + "epoch": 0.42487, + "grad_norm": 1.3151726722717285, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 42487 + }, + { + "epoch": 0.42488, + "grad_norm": 0.6562666893005371, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 42488 + }, + { + "epoch": 0.42489, + "grad_norm": 0.8303307294845581, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 42489 + }, + { + "epoch": 0.4249, + "grad_norm": 1.1985912322998047, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 42490 + }, + { + "epoch": 0.42491, + "grad_norm": 0.8155670166015625, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 42491 + }, + { + "epoch": 0.42492, + "grad_norm": 0.8177903294563293, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 42492 + }, + { + "epoch": 0.42493, + "grad_norm": 1.010221242904663, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 42493 + }, + { + "epoch": 0.42494, + "grad_norm": 1.178423523902893, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 42494 + }, + { + "epoch": 0.42495, + "grad_norm": 0.6871355175971985, + "learning_rate": 0.003, + "loss": 3.9457, + "step": 42495 + }, + { + "epoch": 0.42496, + "grad_norm": 0.705180823802948, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 42496 + }, + { + "epoch": 0.42497, + "grad_norm": 0.7558739185333252, + "learning_rate": 0.003, + "loss": 3.9386, + "step": 42497 + }, + { + "epoch": 0.42498, + "grad_norm": 0.7276865839958191, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 42498 + }, + { + "epoch": 0.42499, + "grad_norm": 0.8936841487884521, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 42499 + }, + { + "epoch": 0.425, + "grad_norm": 0.9958357810974121, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 42500 + }, + { + "epoch": 0.42501, + "grad_norm": 1.0780150890350342, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 42501 + }, + { + "epoch": 0.42502, + "grad_norm": 1.1324177980422974, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 42502 + }, + { + "epoch": 0.42503, + "grad_norm": 1.1520516872406006, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 42503 + }, + { + "epoch": 0.42504, + "grad_norm": 0.9394820928573608, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 42504 + }, + { + "epoch": 0.42505, + "grad_norm": 0.9498576521873474, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 42505 + }, + { + "epoch": 0.42506, + "grad_norm": 1.069806694984436, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 42506 + }, + { + "epoch": 0.42507, + "grad_norm": 0.9648048281669617, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 42507 + }, + { + "epoch": 0.42508, + "grad_norm": 0.8569185733795166, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 42508 + }, + { + "epoch": 0.42509, + "grad_norm": 0.9196972846984863, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 42509 + }, + { + "epoch": 0.4251, + "grad_norm": 1.059780478477478, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 42510 + }, + { + "epoch": 0.42511, + "grad_norm": 1.0859334468841553, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 42511 + }, + { + "epoch": 0.42512, + "grad_norm": 1.011539340019226, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 42512 + }, + { + "epoch": 0.42513, + "grad_norm": 1.0840022563934326, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 42513 + }, + { + "epoch": 0.42514, + "grad_norm": 1.1894949674606323, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 42514 + }, + { + "epoch": 0.42515, + "grad_norm": 0.690557599067688, + "learning_rate": 0.003, + "loss": 3.966, + "step": 42515 + }, + { + "epoch": 0.42516, + "grad_norm": 0.6730562448501587, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 42516 + }, + { + "epoch": 0.42517, + "grad_norm": 0.7962388396263123, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 42517 + }, + { + "epoch": 0.42518, + "grad_norm": 0.9593138694763184, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 42518 + }, + { + "epoch": 0.42519, + "grad_norm": 1.3467674255371094, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 42519 + }, + { + "epoch": 0.4252, + "grad_norm": 0.635230302810669, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 42520 + }, + { + "epoch": 0.42521, + "grad_norm": 0.8679112195968628, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 42521 + }, + { + "epoch": 0.42522, + "grad_norm": 1.0958163738250732, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 42522 + }, + { + "epoch": 0.42523, + "grad_norm": 0.9457026720046997, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 42523 + }, + { + "epoch": 0.42524, + "grad_norm": 0.9533184170722961, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 42524 + }, + { + "epoch": 0.42525, + "grad_norm": 0.7930358052253723, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 42525 + }, + { + "epoch": 0.42526, + "grad_norm": 0.67387855052948, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 42526 + }, + { + "epoch": 0.42527, + "grad_norm": 0.7378572821617126, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 42527 + }, + { + "epoch": 0.42528, + "grad_norm": 0.6942078471183777, + "learning_rate": 0.003, + "loss": 3.981, + "step": 42528 + }, + { + "epoch": 0.42529, + "grad_norm": 0.7453786134719849, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 42529 + }, + { + "epoch": 0.4253, + "grad_norm": 0.7573012113571167, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 42530 + }, + { + "epoch": 0.42531, + "grad_norm": 0.7358159422874451, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 42531 + }, + { + "epoch": 0.42532, + "grad_norm": 0.7098086476325989, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 42532 + }, + { + "epoch": 0.42533, + "grad_norm": 0.7602662444114685, + "learning_rate": 0.003, + "loss": 4.017, + "step": 42533 + }, + { + "epoch": 0.42534, + "grad_norm": 0.7759242057800293, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 42534 + }, + { + "epoch": 0.42535, + "grad_norm": 0.8965311646461487, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 42535 + }, + { + "epoch": 0.42536, + "grad_norm": 0.9778637290000916, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 42536 + }, + { + "epoch": 0.42537, + "grad_norm": 0.9672253131866455, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 42537 + }, + { + "epoch": 0.42538, + "grad_norm": 1.1633026599884033, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 42538 + }, + { + "epoch": 0.42539, + "grad_norm": 1.1250579357147217, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 42539 + }, + { + "epoch": 0.4254, + "grad_norm": 0.9347428679466248, + "learning_rate": 0.003, + "loss": 3.974, + "step": 42540 + }, + { + "epoch": 0.42541, + "grad_norm": 0.9891582727432251, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 42541 + }, + { + "epoch": 0.42542, + "grad_norm": 1.145808219909668, + "learning_rate": 0.003, + "loss": 3.983, + "step": 42542 + }, + { + "epoch": 0.42543, + "grad_norm": 0.899291455745697, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 42543 + }, + { + "epoch": 0.42544, + "grad_norm": 0.7621572613716125, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 42544 + }, + { + "epoch": 0.42545, + "grad_norm": 0.7527821660041809, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 42545 + }, + { + "epoch": 0.42546, + "grad_norm": 0.8863541483879089, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 42546 + }, + { + "epoch": 0.42547, + "grad_norm": 0.9335974454879761, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 42547 + }, + { + "epoch": 0.42548, + "grad_norm": 0.8853792548179626, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 42548 + }, + { + "epoch": 0.42549, + "grad_norm": 0.9828547835350037, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 42549 + }, + { + "epoch": 0.4255, + "grad_norm": 1.0410630702972412, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 42550 + }, + { + "epoch": 0.42551, + "grad_norm": 1.318141222000122, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 42551 + }, + { + "epoch": 0.42552, + "grad_norm": 0.8459156155586243, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 42552 + }, + { + "epoch": 0.42553, + "grad_norm": 0.7766560912132263, + "learning_rate": 0.003, + "loss": 4.03, + "step": 42553 + }, + { + "epoch": 0.42554, + "grad_norm": 0.7829617261886597, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 42554 + }, + { + "epoch": 0.42555, + "grad_norm": 0.8263078331947327, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 42555 + }, + { + "epoch": 0.42556, + "grad_norm": 0.8058593273162842, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 42556 + }, + { + "epoch": 0.42557, + "grad_norm": 0.8611670732498169, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 42557 + }, + { + "epoch": 0.42558, + "grad_norm": 1.0804979801177979, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 42558 + }, + { + "epoch": 0.42559, + "grad_norm": 1.2670241594314575, + "learning_rate": 0.003, + "loss": 3.989, + "step": 42559 + }, + { + "epoch": 0.4256, + "grad_norm": 0.8786394596099854, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 42560 + }, + { + "epoch": 0.42561, + "grad_norm": 0.970331609249115, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 42561 + }, + { + "epoch": 0.42562, + "grad_norm": 1.2443227767944336, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 42562 + }, + { + "epoch": 0.42563, + "grad_norm": 0.8051514029502869, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 42563 + }, + { + "epoch": 0.42564, + "grad_norm": 0.6590875387191772, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 42564 + }, + { + "epoch": 0.42565, + "grad_norm": 0.781002402305603, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 42565 + }, + { + "epoch": 0.42566, + "grad_norm": 0.8007655143737793, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 42566 + }, + { + "epoch": 0.42567, + "grad_norm": 0.9613795876502991, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 42567 + }, + { + "epoch": 0.42568, + "grad_norm": 0.9732794761657715, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 42568 + }, + { + "epoch": 0.42569, + "grad_norm": 1.0279511213302612, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 42569 + }, + { + "epoch": 0.4257, + "grad_norm": 0.9284290075302124, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 42570 + }, + { + "epoch": 0.42571, + "grad_norm": 0.9155558347702026, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 42571 + }, + { + "epoch": 0.42572, + "grad_norm": 0.8585010766983032, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 42572 + }, + { + "epoch": 0.42573, + "grad_norm": 0.9103490114212036, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 42573 + }, + { + "epoch": 0.42574, + "grad_norm": 0.8969002962112427, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 42574 + }, + { + "epoch": 0.42575, + "grad_norm": 0.72723388671875, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 42575 + }, + { + "epoch": 0.42576, + "grad_norm": 0.7621877789497375, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 42576 + }, + { + "epoch": 0.42577, + "grad_norm": 0.8924505114555359, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 42577 + }, + { + "epoch": 0.42578, + "grad_norm": 1.0455421209335327, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 42578 + }, + { + "epoch": 0.42579, + "grad_norm": 1.0663450956344604, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 42579 + }, + { + "epoch": 0.4258, + "grad_norm": 1.1798473596572876, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 42580 + }, + { + "epoch": 0.42581, + "grad_norm": 0.8656402230262756, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 42581 + }, + { + "epoch": 0.42582, + "grad_norm": 0.741845965385437, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 42582 + }, + { + "epoch": 0.42583, + "grad_norm": 0.708532989025116, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 42583 + }, + { + "epoch": 0.42584, + "grad_norm": 0.7629429697990417, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 42584 + }, + { + "epoch": 0.42585, + "grad_norm": 0.991066575050354, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 42585 + }, + { + "epoch": 0.42586, + "grad_norm": 1.2058278322219849, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 42586 + }, + { + "epoch": 0.42587, + "grad_norm": 0.8455241322517395, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 42587 + }, + { + "epoch": 0.42588, + "grad_norm": 0.9564096331596375, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 42588 + }, + { + "epoch": 0.42589, + "grad_norm": 1.0377750396728516, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 42589 + }, + { + "epoch": 0.4259, + "grad_norm": 1.0576908588409424, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 42590 + }, + { + "epoch": 0.42591, + "grad_norm": 1.0134620666503906, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 42591 + }, + { + "epoch": 0.42592, + "grad_norm": 1.0506970882415771, + "learning_rate": 0.003, + "loss": 4.007, + "step": 42592 + }, + { + "epoch": 0.42593, + "grad_norm": 0.8943461179733276, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 42593 + }, + { + "epoch": 0.42594, + "grad_norm": 0.7470794916152954, + "learning_rate": 0.003, + "loss": 3.971, + "step": 42594 + }, + { + "epoch": 0.42595, + "grad_norm": 0.758034884929657, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 42595 + }, + { + "epoch": 0.42596, + "grad_norm": 0.8003572225570679, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 42596 + }, + { + "epoch": 0.42597, + "grad_norm": 0.8149428963661194, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 42597 + }, + { + "epoch": 0.42598, + "grad_norm": 0.9698595404624939, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 42598 + }, + { + "epoch": 0.42599, + "grad_norm": 1.105096459388733, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 42599 + }, + { + "epoch": 0.426, + "grad_norm": 0.8537248969078064, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 42600 + }, + { + "epoch": 0.42601, + "grad_norm": 0.9133638739585876, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 42601 + }, + { + "epoch": 0.42602, + "grad_norm": 0.9534811973571777, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 42602 + }, + { + "epoch": 0.42603, + "grad_norm": 0.9383721947669983, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 42603 + }, + { + "epoch": 0.42604, + "grad_norm": 1.3316328525543213, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 42604 + }, + { + "epoch": 0.42605, + "grad_norm": 0.9449316263198853, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 42605 + }, + { + "epoch": 0.42606, + "grad_norm": 1.0728358030319214, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 42606 + }, + { + "epoch": 0.42607, + "grad_norm": 0.979304313659668, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 42607 + }, + { + "epoch": 0.42608, + "grad_norm": 1.0371681451797485, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 42608 + }, + { + "epoch": 0.42609, + "grad_norm": 0.9126209020614624, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 42609 + }, + { + "epoch": 0.4261, + "grad_norm": 1.0708163976669312, + "learning_rate": 0.003, + "loss": 3.988, + "step": 42610 + }, + { + "epoch": 0.42611, + "grad_norm": 1.1999666690826416, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 42611 + }, + { + "epoch": 0.42612, + "grad_norm": 0.9229190945625305, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 42612 + }, + { + "epoch": 0.42613, + "grad_norm": 0.8609812259674072, + "learning_rate": 0.003, + "loss": 3.977, + "step": 42613 + }, + { + "epoch": 0.42614, + "grad_norm": 0.8047899603843689, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 42614 + }, + { + "epoch": 0.42615, + "grad_norm": 0.7297985553741455, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 42615 + }, + { + "epoch": 0.42616, + "grad_norm": 0.7573359608650208, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 42616 + }, + { + "epoch": 0.42617, + "grad_norm": 0.7693973779678345, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 42617 + }, + { + "epoch": 0.42618, + "grad_norm": 0.7862663269042969, + "learning_rate": 0.003, + "loss": 3.972, + "step": 42618 + }, + { + "epoch": 0.42619, + "grad_norm": 0.8019823431968689, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 42619 + }, + { + "epoch": 0.4262, + "grad_norm": 0.9120521545410156, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 42620 + }, + { + "epoch": 0.42621, + "grad_norm": 1.237778663635254, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 42621 + }, + { + "epoch": 0.42622, + "grad_norm": 1.0605236291885376, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 42622 + }, + { + "epoch": 0.42623, + "grad_norm": 0.9376776814460754, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 42623 + }, + { + "epoch": 0.42624, + "grad_norm": 0.991249680519104, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 42624 + }, + { + "epoch": 0.42625, + "grad_norm": 1.1089606285095215, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 42625 + }, + { + "epoch": 0.42626, + "grad_norm": 0.811248242855072, + "learning_rate": 0.003, + "loss": 3.996, + "step": 42626 + }, + { + "epoch": 0.42627, + "grad_norm": 0.6328935027122498, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 42627 + }, + { + "epoch": 0.42628, + "grad_norm": 0.6288989782333374, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 42628 + }, + { + "epoch": 0.42629, + "grad_norm": 0.6905278563499451, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 42629 + }, + { + "epoch": 0.4263, + "grad_norm": 0.8167542815208435, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 42630 + }, + { + "epoch": 0.42631, + "grad_norm": 0.9800211787223816, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 42631 + }, + { + "epoch": 0.42632, + "grad_norm": 1.0531001091003418, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 42632 + }, + { + "epoch": 0.42633, + "grad_norm": 0.9165847301483154, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 42633 + }, + { + "epoch": 0.42634, + "grad_norm": 0.8586801886558533, + "learning_rate": 0.003, + "loss": 4.004, + "step": 42634 + }, + { + "epoch": 0.42635, + "grad_norm": 0.8474493026733398, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 42635 + }, + { + "epoch": 0.42636, + "grad_norm": 0.7963594198226929, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 42636 + }, + { + "epoch": 0.42637, + "grad_norm": 0.7636728286743164, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 42637 + }, + { + "epoch": 0.42638, + "grad_norm": 0.9238458275794983, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 42638 + }, + { + "epoch": 0.42639, + "grad_norm": 1.1383813619613647, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 42639 + }, + { + "epoch": 0.4264, + "grad_norm": 0.8203203678131104, + "learning_rate": 0.003, + "loss": 3.947, + "step": 42640 + }, + { + "epoch": 0.42641, + "grad_norm": 0.8571634888648987, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 42641 + }, + { + "epoch": 0.42642, + "grad_norm": 0.9383385181427002, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 42642 + }, + { + "epoch": 0.42643, + "grad_norm": 0.788755476474762, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 42643 + }, + { + "epoch": 0.42644, + "grad_norm": 0.7494588494300842, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 42644 + }, + { + "epoch": 0.42645, + "grad_norm": 0.9543783664703369, + "learning_rate": 0.003, + "loss": 3.981, + "step": 42645 + }, + { + "epoch": 0.42646, + "grad_norm": 1.1474571228027344, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 42646 + }, + { + "epoch": 0.42647, + "grad_norm": 1.0948692560195923, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 42647 + }, + { + "epoch": 0.42648, + "grad_norm": 0.8064208030700684, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 42648 + }, + { + "epoch": 0.42649, + "grad_norm": 0.7223473191261292, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 42649 + }, + { + "epoch": 0.4265, + "grad_norm": 0.8390978574752808, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 42650 + }, + { + "epoch": 0.42651, + "grad_norm": 0.8128206133842468, + "learning_rate": 0.003, + "loss": 3.986, + "step": 42651 + }, + { + "epoch": 0.42652, + "grad_norm": 0.8763503432273865, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 42652 + }, + { + "epoch": 0.42653, + "grad_norm": 0.9101155996322632, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 42653 + }, + { + "epoch": 0.42654, + "grad_norm": 0.817190945148468, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 42654 + }, + { + "epoch": 0.42655, + "grad_norm": 0.8662226796150208, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 42655 + }, + { + "epoch": 0.42656, + "grad_norm": 0.8938532471656799, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 42656 + }, + { + "epoch": 0.42657, + "grad_norm": 1.1843137741088867, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 42657 + }, + { + "epoch": 0.42658, + "grad_norm": 0.9842067360877991, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 42658 + }, + { + "epoch": 0.42659, + "grad_norm": 1.049143671989441, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 42659 + }, + { + "epoch": 0.4266, + "grad_norm": 1.2795131206512451, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 42660 + }, + { + "epoch": 0.42661, + "grad_norm": 0.802749514579773, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 42661 + }, + { + "epoch": 0.42662, + "grad_norm": 0.7008389830589294, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 42662 + }, + { + "epoch": 0.42663, + "grad_norm": 0.7619183659553528, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 42663 + }, + { + "epoch": 0.42664, + "grad_norm": 0.8671061992645264, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 42664 + }, + { + "epoch": 0.42665, + "grad_norm": 0.8434338569641113, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 42665 + }, + { + "epoch": 0.42666, + "grad_norm": 0.8012887239456177, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 42666 + }, + { + "epoch": 0.42667, + "grad_norm": 0.9131534695625305, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 42667 + }, + { + "epoch": 0.42668, + "grad_norm": 0.8736546635627747, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 42668 + }, + { + "epoch": 0.42669, + "grad_norm": 0.9090237021446228, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 42669 + }, + { + "epoch": 0.4267, + "grad_norm": 0.9519315361976624, + "learning_rate": 0.003, + "loss": 3.993, + "step": 42670 + }, + { + "epoch": 0.42671, + "grad_norm": 1.0367543697357178, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 42671 + }, + { + "epoch": 0.42672, + "grad_norm": 1.1327769756317139, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 42672 + }, + { + "epoch": 0.42673, + "grad_norm": 1.0484645366668701, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 42673 + }, + { + "epoch": 0.42674, + "grad_norm": 1.1327457427978516, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 42674 + }, + { + "epoch": 0.42675, + "grad_norm": 0.8827289938926697, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 42675 + }, + { + "epoch": 0.42676, + "grad_norm": 0.8946552276611328, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 42676 + }, + { + "epoch": 0.42677, + "grad_norm": 1.0204626321792603, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 42677 + }, + { + "epoch": 0.42678, + "grad_norm": 0.8555961847305298, + "learning_rate": 0.003, + "loss": 3.945, + "step": 42678 + }, + { + "epoch": 0.42679, + "grad_norm": 0.7664663195610046, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 42679 + }, + { + "epoch": 0.4268, + "grad_norm": 0.748386800289154, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 42680 + }, + { + "epoch": 0.42681, + "grad_norm": 0.8126998543739319, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 42681 + }, + { + "epoch": 0.42682, + "grad_norm": 0.9217205047607422, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 42682 + }, + { + "epoch": 0.42683, + "grad_norm": 1.0361597537994385, + "learning_rate": 0.003, + "loss": 3.961, + "step": 42683 + }, + { + "epoch": 0.42684, + "grad_norm": 1.180876612663269, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 42684 + }, + { + "epoch": 0.42685, + "grad_norm": 0.9134324193000793, + "learning_rate": 0.003, + "loss": 4.003, + "step": 42685 + }, + { + "epoch": 0.42686, + "grad_norm": 0.9044310450553894, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 42686 + }, + { + "epoch": 0.42687, + "grad_norm": 0.9829003214836121, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 42687 + }, + { + "epoch": 0.42688, + "grad_norm": 1.242799997329712, + "learning_rate": 0.003, + "loss": 3.982, + "step": 42688 + }, + { + "epoch": 0.42689, + "grad_norm": 0.8552564978599548, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 42689 + }, + { + "epoch": 0.4269, + "grad_norm": 0.8247222900390625, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 42690 + }, + { + "epoch": 0.42691, + "grad_norm": 0.7937273383140564, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 42691 + }, + { + "epoch": 0.42692, + "grad_norm": 1.0168514251708984, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 42692 + }, + { + "epoch": 0.42693, + "grad_norm": 1.153249740600586, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 42693 + }, + { + "epoch": 0.42694, + "grad_norm": 0.6567758917808533, + "learning_rate": 0.003, + "loss": 3.99, + "step": 42694 + }, + { + "epoch": 0.42695, + "grad_norm": 0.8094883561134338, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 42695 + }, + { + "epoch": 0.42696, + "grad_norm": 0.8533717393875122, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 42696 + }, + { + "epoch": 0.42697, + "grad_norm": 1.067064642906189, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 42697 + }, + { + "epoch": 0.42698, + "grad_norm": 1.1390033960342407, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 42698 + }, + { + "epoch": 0.42699, + "grad_norm": 0.8704254031181335, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 42699 + }, + { + "epoch": 0.427, + "grad_norm": 0.8549358248710632, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 42700 + }, + { + "epoch": 0.42701, + "grad_norm": 0.7930123805999756, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 42701 + }, + { + "epoch": 0.42702, + "grad_norm": 0.9048150777816772, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 42702 + }, + { + "epoch": 0.42703, + "grad_norm": 1.3045216798782349, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 42703 + }, + { + "epoch": 0.42704, + "grad_norm": 1.0568931102752686, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 42704 + }, + { + "epoch": 0.42705, + "grad_norm": 0.8876510858535767, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 42705 + }, + { + "epoch": 0.42706, + "grad_norm": 0.8500170111656189, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 42706 + }, + { + "epoch": 0.42707, + "grad_norm": 0.8413448333740234, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 42707 + }, + { + "epoch": 0.42708, + "grad_norm": 1.0165917873382568, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 42708 + }, + { + "epoch": 0.42709, + "grad_norm": 1.0455173254013062, + "learning_rate": 0.003, + "loss": 3.96, + "step": 42709 + }, + { + "epoch": 0.4271, + "grad_norm": 0.884128212928772, + "learning_rate": 0.003, + "loss": 3.986, + "step": 42710 + }, + { + "epoch": 0.42711, + "grad_norm": 0.7755604386329651, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 42711 + }, + { + "epoch": 0.42712, + "grad_norm": 0.648620069026947, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 42712 + }, + { + "epoch": 0.42713, + "grad_norm": 0.6133002638816833, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 42713 + }, + { + "epoch": 0.42714, + "grad_norm": 0.7076454758644104, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 42714 + }, + { + "epoch": 0.42715, + "grad_norm": 0.9578649997711182, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 42715 + }, + { + "epoch": 0.42716, + "grad_norm": 1.0379573106765747, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 42716 + }, + { + "epoch": 0.42717, + "grad_norm": 1.028477668762207, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 42717 + }, + { + "epoch": 0.42718, + "grad_norm": 1.0945875644683838, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 42718 + }, + { + "epoch": 0.42719, + "grad_norm": 0.9171095490455627, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 42719 + }, + { + "epoch": 0.4272, + "grad_norm": 1.0712909698486328, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 42720 + }, + { + "epoch": 0.42721, + "grad_norm": 1.0632615089416504, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 42721 + }, + { + "epoch": 0.42722, + "grad_norm": 0.8855400085449219, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 42722 + }, + { + "epoch": 0.42723, + "grad_norm": 0.8973804116249084, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 42723 + }, + { + "epoch": 0.42724, + "grad_norm": 0.9940391778945923, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 42724 + }, + { + "epoch": 0.42725, + "grad_norm": 0.9403582811355591, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 42725 + }, + { + "epoch": 0.42726, + "grad_norm": 0.655571460723877, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 42726 + }, + { + "epoch": 0.42727, + "grad_norm": 0.6276904344558716, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 42727 + }, + { + "epoch": 0.42728, + "grad_norm": 0.644092321395874, + "learning_rate": 0.003, + "loss": 3.964, + "step": 42728 + }, + { + "epoch": 0.42729, + "grad_norm": 0.7910151481628418, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 42729 + }, + { + "epoch": 0.4273, + "grad_norm": 0.8137441873550415, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 42730 + }, + { + "epoch": 0.42731, + "grad_norm": 1.0150963068008423, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 42731 + }, + { + "epoch": 0.42732, + "grad_norm": 1.1786881685256958, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 42732 + }, + { + "epoch": 0.42733, + "grad_norm": 0.8066445589065552, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 42733 + }, + { + "epoch": 0.42734, + "grad_norm": 0.7813840508460999, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 42734 + }, + { + "epoch": 0.42735, + "grad_norm": 0.7920549511909485, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 42735 + }, + { + "epoch": 0.42736, + "grad_norm": 0.9183588624000549, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 42736 + }, + { + "epoch": 0.42737, + "grad_norm": 1.0172197818756104, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 42737 + }, + { + "epoch": 0.42738, + "grad_norm": 0.9882456660270691, + "learning_rate": 0.003, + "loss": 3.977, + "step": 42738 + }, + { + "epoch": 0.42739, + "grad_norm": 0.9400572776794434, + "learning_rate": 0.003, + "loss": 3.972, + "step": 42739 + }, + { + "epoch": 0.4274, + "grad_norm": 0.9751688838005066, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 42740 + }, + { + "epoch": 0.42741, + "grad_norm": 1.0333571434020996, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 42741 + }, + { + "epoch": 0.42742, + "grad_norm": 0.9920343160629272, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 42742 + }, + { + "epoch": 0.42743, + "grad_norm": 1.057868480682373, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 42743 + }, + { + "epoch": 0.42744, + "grad_norm": 0.979604959487915, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 42744 + }, + { + "epoch": 0.42745, + "grad_norm": 0.9749791622161865, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 42745 + }, + { + "epoch": 0.42746, + "grad_norm": 1.005642294883728, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 42746 + }, + { + "epoch": 0.42747, + "grad_norm": 1.0545861721038818, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 42747 + }, + { + "epoch": 0.42748, + "grad_norm": 1.0359629392623901, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 42748 + }, + { + "epoch": 0.42749, + "grad_norm": 0.9174767732620239, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 42749 + }, + { + "epoch": 0.4275, + "grad_norm": 0.8363479375839233, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 42750 + }, + { + "epoch": 0.42751, + "grad_norm": 0.9250129461288452, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 42751 + }, + { + "epoch": 0.42752, + "grad_norm": 1.1131441593170166, + "learning_rate": 0.003, + "loss": 3.999, + "step": 42752 + }, + { + "epoch": 0.42753, + "grad_norm": 1.0350860357284546, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 42753 + }, + { + "epoch": 0.42754, + "grad_norm": 1.230672001838684, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 42754 + }, + { + "epoch": 0.42755, + "grad_norm": 0.9186279773712158, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 42755 + }, + { + "epoch": 0.42756, + "grad_norm": 0.8871832489967346, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 42756 + }, + { + "epoch": 0.42757, + "grad_norm": 0.9250750541687012, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 42757 + }, + { + "epoch": 0.42758, + "grad_norm": 1.1044042110443115, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 42758 + }, + { + "epoch": 0.42759, + "grad_norm": 1.0277057886123657, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 42759 + }, + { + "epoch": 0.4276, + "grad_norm": 1.0191489458084106, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 42760 + }, + { + "epoch": 0.42761, + "grad_norm": 1.086053729057312, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 42761 + }, + { + "epoch": 0.42762, + "grad_norm": 0.8981273174285889, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 42762 + }, + { + "epoch": 0.42763, + "grad_norm": 0.8681610822677612, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 42763 + }, + { + "epoch": 0.42764, + "grad_norm": 0.8719543218612671, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 42764 + }, + { + "epoch": 0.42765, + "grad_norm": 0.9252362847328186, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 42765 + }, + { + "epoch": 0.42766, + "grad_norm": 0.9408925175666809, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 42766 + }, + { + "epoch": 0.42767, + "grad_norm": 0.9267804026603699, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 42767 + }, + { + "epoch": 0.42768, + "grad_norm": 0.8924443125724792, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 42768 + }, + { + "epoch": 0.42769, + "grad_norm": 1.0106596946716309, + "learning_rate": 0.003, + "loss": 3.973, + "step": 42769 + }, + { + "epoch": 0.4277, + "grad_norm": 0.9544715881347656, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 42770 + }, + { + "epoch": 0.42771, + "grad_norm": 0.870242178440094, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 42771 + }, + { + "epoch": 0.42772, + "grad_norm": 1.0865778923034668, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 42772 + }, + { + "epoch": 0.42773, + "grad_norm": 1.0892771482467651, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 42773 + }, + { + "epoch": 0.42774, + "grad_norm": 0.7496567368507385, + "learning_rate": 0.003, + "loss": 4.007, + "step": 42774 + }, + { + "epoch": 0.42775, + "grad_norm": 0.7472078204154968, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 42775 + }, + { + "epoch": 0.42776, + "grad_norm": 0.828458309173584, + "learning_rate": 0.003, + "loss": 3.9446, + "step": 42776 + }, + { + "epoch": 0.42777, + "grad_norm": 0.9254755973815918, + "learning_rate": 0.003, + "loss": 3.97, + "step": 42777 + }, + { + "epoch": 0.42778, + "grad_norm": 1.0418459177017212, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 42778 + }, + { + "epoch": 0.42779, + "grad_norm": 0.881507396697998, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 42779 + }, + { + "epoch": 0.4278, + "grad_norm": 0.7226645946502686, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 42780 + }, + { + "epoch": 0.42781, + "grad_norm": 0.7094212174415588, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 42781 + }, + { + "epoch": 0.42782, + "grad_norm": 0.768419086933136, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 42782 + }, + { + "epoch": 0.42783, + "grad_norm": 1.0426427125930786, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 42783 + }, + { + "epoch": 0.42784, + "grad_norm": 1.2789899110794067, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 42784 + }, + { + "epoch": 0.42785, + "grad_norm": 0.6689742207527161, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 42785 + }, + { + "epoch": 0.42786, + "grad_norm": 0.6364821791648865, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 42786 + }, + { + "epoch": 0.42787, + "grad_norm": 0.7405193448066711, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 42787 + }, + { + "epoch": 0.42788, + "grad_norm": 0.9352587461471558, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 42788 + }, + { + "epoch": 0.42789, + "grad_norm": 0.9054877161979675, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 42789 + }, + { + "epoch": 0.4279, + "grad_norm": 0.7344231009483337, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 42790 + }, + { + "epoch": 0.42791, + "grad_norm": 0.6831894516944885, + "learning_rate": 0.003, + "loss": 3.9337, + "step": 42791 + }, + { + "epoch": 0.42792, + "grad_norm": 0.7342104315757751, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 42792 + }, + { + "epoch": 0.42793, + "grad_norm": 0.6549692153930664, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 42793 + }, + { + "epoch": 0.42794, + "grad_norm": 0.6308777928352356, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 42794 + }, + { + "epoch": 0.42795, + "grad_norm": 0.6366708874702454, + "learning_rate": 0.003, + "loss": 3.954, + "step": 42795 + }, + { + "epoch": 0.42796, + "grad_norm": 0.6932421326637268, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 42796 + }, + { + "epoch": 0.42797, + "grad_norm": 0.7610704302787781, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 42797 + }, + { + "epoch": 0.42798, + "grad_norm": 0.8336536884307861, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 42798 + }, + { + "epoch": 0.42799, + "grad_norm": 0.9354425668716431, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 42799 + }, + { + "epoch": 0.428, + "grad_norm": 1.2273281812667847, + "learning_rate": 0.003, + "loss": 3.978, + "step": 42800 + }, + { + "epoch": 0.42801, + "grad_norm": 0.9372303485870361, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 42801 + }, + { + "epoch": 0.42802, + "grad_norm": 1.1229761838912964, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 42802 + }, + { + "epoch": 0.42803, + "grad_norm": 1.0307849645614624, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 42803 + }, + { + "epoch": 0.42804, + "grad_norm": 1.0384758710861206, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 42804 + }, + { + "epoch": 0.42805, + "grad_norm": 0.9492842555046082, + "learning_rate": 0.003, + "loss": 3.973, + "step": 42805 + }, + { + "epoch": 0.42806, + "grad_norm": 0.8667718172073364, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 42806 + }, + { + "epoch": 0.42807, + "grad_norm": 0.7426442503929138, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 42807 + }, + { + "epoch": 0.42808, + "grad_norm": 0.7232751846313477, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 42808 + }, + { + "epoch": 0.42809, + "grad_norm": 0.6891670823097229, + "learning_rate": 0.003, + "loss": 3.9397, + "step": 42809 + }, + { + "epoch": 0.4281, + "grad_norm": 0.7875507473945618, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 42810 + }, + { + "epoch": 0.42811, + "grad_norm": 0.8311426639556885, + "learning_rate": 0.003, + "loss": 3.962, + "step": 42811 + }, + { + "epoch": 0.42812, + "grad_norm": 0.8785718679428101, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 42812 + }, + { + "epoch": 0.42813, + "grad_norm": 0.7350001931190491, + "learning_rate": 0.003, + "loss": 3.967, + "step": 42813 + }, + { + "epoch": 0.42814, + "grad_norm": 0.8833147287368774, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 42814 + }, + { + "epoch": 0.42815, + "grad_norm": 1.2389202117919922, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 42815 + }, + { + "epoch": 0.42816, + "grad_norm": 1.19158136844635, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 42816 + }, + { + "epoch": 0.42817, + "grad_norm": 0.7728590965270996, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 42817 + }, + { + "epoch": 0.42818, + "grad_norm": 0.570640504360199, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 42818 + }, + { + "epoch": 0.42819, + "grad_norm": 0.798224687576294, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 42819 + }, + { + "epoch": 0.4282, + "grad_norm": 1.03883957862854, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 42820 + }, + { + "epoch": 0.42821, + "grad_norm": 1.3861538171768188, + "learning_rate": 0.003, + "loss": 3.9323, + "step": 42821 + }, + { + "epoch": 0.42822, + "grad_norm": 0.6123051047325134, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 42822 + }, + { + "epoch": 0.42823, + "grad_norm": 0.6987836956977844, + "learning_rate": 0.003, + "loss": 3.974, + "step": 42823 + }, + { + "epoch": 0.42824, + "grad_norm": 0.9415832757949829, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 42824 + }, + { + "epoch": 0.42825, + "grad_norm": 0.9781697988510132, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 42825 + }, + { + "epoch": 0.42826, + "grad_norm": 1.024505615234375, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 42826 + }, + { + "epoch": 0.42827, + "grad_norm": 0.9480403065681458, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 42827 + }, + { + "epoch": 0.42828, + "grad_norm": 0.8423345685005188, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 42828 + }, + { + "epoch": 0.42829, + "grad_norm": 0.7339491844177246, + "learning_rate": 0.003, + "loss": 3.99, + "step": 42829 + }, + { + "epoch": 0.4283, + "grad_norm": 0.8445951342582703, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 42830 + }, + { + "epoch": 0.42831, + "grad_norm": 0.9229695200920105, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 42831 + }, + { + "epoch": 0.42832, + "grad_norm": 1.0199999809265137, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 42832 + }, + { + "epoch": 0.42833, + "grad_norm": 1.3436918258666992, + "learning_rate": 0.003, + "loss": 3.971, + "step": 42833 + }, + { + "epoch": 0.42834, + "grad_norm": 0.8019916415214539, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 42834 + }, + { + "epoch": 0.42835, + "grad_norm": 0.7830618023872375, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 42835 + }, + { + "epoch": 0.42836, + "grad_norm": 0.8399112224578857, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 42836 + }, + { + "epoch": 0.42837, + "grad_norm": 0.886987566947937, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 42837 + }, + { + "epoch": 0.42838, + "grad_norm": 1.1567564010620117, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 42838 + }, + { + "epoch": 0.42839, + "grad_norm": 1.1115834712982178, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 42839 + }, + { + "epoch": 0.4284, + "grad_norm": 0.9050938487052917, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 42840 + }, + { + "epoch": 0.42841, + "grad_norm": 0.8800444006919861, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 42841 + }, + { + "epoch": 0.42842, + "grad_norm": 0.8969903588294983, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 42842 + }, + { + "epoch": 0.42843, + "grad_norm": 0.8944269418716431, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 42843 + }, + { + "epoch": 0.42844, + "grad_norm": 0.7955301403999329, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 42844 + }, + { + "epoch": 0.42845, + "grad_norm": 0.8500604033470154, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 42845 + }, + { + "epoch": 0.42846, + "grad_norm": 1.0393000841140747, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 42846 + }, + { + "epoch": 0.42847, + "grad_norm": 1.1867990493774414, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 42847 + }, + { + "epoch": 0.42848, + "grad_norm": 0.8879284858703613, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 42848 + }, + { + "epoch": 0.42849, + "grad_norm": 0.8505696058273315, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 42849 + }, + { + "epoch": 0.4285, + "grad_norm": 0.7265779376029968, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 42850 + }, + { + "epoch": 0.42851, + "grad_norm": 0.7709575295448303, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 42851 + }, + { + "epoch": 0.42852, + "grad_norm": 0.8515470623970032, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 42852 + }, + { + "epoch": 0.42853, + "grad_norm": 1.0956546068191528, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 42853 + }, + { + "epoch": 0.42854, + "grad_norm": 1.1079360246658325, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 42854 + }, + { + "epoch": 0.42855, + "grad_norm": 0.7954277992248535, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 42855 + }, + { + "epoch": 0.42856, + "grad_norm": 0.7925911545753479, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 42856 + }, + { + "epoch": 0.42857, + "grad_norm": 0.8824431300163269, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 42857 + }, + { + "epoch": 0.42858, + "grad_norm": 1.0897104740142822, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 42858 + }, + { + "epoch": 0.42859, + "grad_norm": 0.9900853037834167, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 42859 + }, + { + "epoch": 0.4286, + "grad_norm": 0.9246540069580078, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 42860 + }, + { + "epoch": 0.42861, + "grad_norm": 0.8997389078140259, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 42861 + }, + { + "epoch": 0.42862, + "grad_norm": 0.8008478283882141, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 42862 + }, + { + "epoch": 0.42863, + "grad_norm": 0.8262541890144348, + "learning_rate": 0.003, + "loss": 3.9392, + "step": 42863 + }, + { + "epoch": 0.42864, + "grad_norm": 0.8398754000663757, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 42864 + }, + { + "epoch": 0.42865, + "grad_norm": 0.8331605792045593, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 42865 + }, + { + "epoch": 0.42866, + "grad_norm": 0.8348963856697083, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 42866 + }, + { + "epoch": 0.42867, + "grad_norm": 0.9182835221290588, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 42867 + }, + { + "epoch": 0.42868, + "grad_norm": 1.0515598058700562, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 42868 + }, + { + "epoch": 0.42869, + "grad_norm": 0.9119529128074646, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 42869 + }, + { + "epoch": 0.4287, + "grad_norm": 1.3700321912765503, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 42870 + }, + { + "epoch": 0.42871, + "grad_norm": 0.8714045882225037, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 42871 + }, + { + "epoch": 0.42872, + "grad_norm": 0.6850482225418091, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 42872 + }, + { + "epoch": 0.42873, + "grad_norm": 0.7792858481407166, + "learning_rate": 0.003, + "loss": 3.9483, + "step": 42873 + }, + { + "epoch": 0.42874, + "grad_norm": 0.7572187185287476, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 42874 + }, + { + "epoch": 0.42875, + "grad_norm": 0.7861167192459106, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 42875 + }, + { + "epoch": 0.42876, + "grad_norm": 0.8321620225906372, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 42876 + }, + { + "epoch": 0.42877, + "grad_norm": 0.941750705242157, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 42877 + }, + { + "epoch": 0.42878, + "grad_norm": 1.0692968368530273, + "learning_rate": 0.003, + "loss": 3.986, + "step": 42878 + }, + { + "epoch": 0.42879, + "grad_norm": 1.1590458154678345, + "learning_rate": 0.003, + "loss": 3.977, + "step": 42879 + }, + { + "epoch": 0.4288, + "grad_norm": 0.9808570742607117, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 42880 + }, + { + "epoch": 0.42881, + "grad_norm": 1.0925498008728027, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 42881 + }, + { + "epoch": 0.42882, + "grad_norm": 0.9340634346008301, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 42882 + }, + { + "epoch": 0.42883, + "grad_norm": 0.8078262805938721, + "learning_rate": 0.003, + "loss": 3.9463, + "step": 42883 + }, + { + "epoch": 0.42884, + "grad_norm": 0.7004592418670654, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 42884 + }, + { + "epoch": 0.42885, + "grad_norm": 0.6973031163215637, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 42885 + }, + { + "epoch": 0.42886, + "grad_norm": 0.6615633964538574, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 42886 + }, + { + "epoch": 0.42887, + "grad_norm": 0.7695558071136475, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 42887 + }, + { + "epoch": 0.42888, + "grad_norm": 0.8176336288452148, + "learning_rate": 0.003, + "loss": 3.993, + "step": 42888 + }, + { + "epoch": 0.42889, + "grad_norm": 0.9139233827590942, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 42889 + }, + { + "epoch": 0.4289, + "grad_norm": 1.200202226638794, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 42890 + }, + { + "epoch": 0.42891, + "grad_norm": 1.1157437562942505, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 42891 + }, + { + "epoch": 0.42892, + "grad_norm": 0.7827249765396118, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 42892 + }, + { + "epoch": 0.42893, + "grad_norm": 0.8004035353660583, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 42893 + }, + { + "epoch": 0.42894, + "grad_norm": 0.9431825280189514, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 42894 + }, + { + "epoch": 0.42895, + "grad_norm": 0.9859488010406494, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 42895 + }, + { + "epoch": 0.42896, + "grad_norm": 0.9556178450584412, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 42896 + }, + { + "epoch": 0.42897, + "grad_norm": 0.885049045085907, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 42897 + }, + { + "epoch": 0.42898, + "grad_norm": 0.9460813403129578, + "learning_rate": 0.003, + "loss": 3.993, + "step": 42898 + }, + { + "epoch": 0.42899, + "grad_norm": 1.0817636251449585, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 42899 + }, + { + "epoch": 0.429, + "grad_norm": 0.8888627290725708, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 42900 + }, + { + "epoch": 0.42901, + "grad_norm": 0.8172315359115601, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 42901 + }, + { + "epoch": 0.42902, + "grad_norm": 1.0283621549606323, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 42902 + }, + { + "epoch": 0.42903, + "grad_norm": 1.3750190734863281, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 42903 + }, + { + "epoch": 0.42904, + "grad_norm": 0.7093423008918762, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 42904 + }, + { + "epoch": 0.42905, + "grad_norm": 0.7318344712257385, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 42905 + }, + { + "epoch": 0.42906, + "grad_norm": 0.7892559170722961, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 42906 + }, + { + "epoch": 0.42907, + "grad_norm": 0.8076654672622681, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 42907 + }, + { + "epoch": 0.42908, + "grad_norm": 0.7715041637420654, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 42908 + }, + { + "epoch": 0.42909, + "grad_norm": 0.7613096833229065, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 42909 + }, + { + "epoch": 0.4291, + "grad_norm": 0.7441433668136597, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 42910 + }, + { + "epoch": 0.42911, + "grad_norm": 0.9087297320365906, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 42911 + }, + { + "epoch": 0.42912, + "grad_norm": 0.9851729273796082, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 42912 + }, + { + "epoch": 0.42913, + "grad_norm": 1.0069856643676758, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 42913 + }, + { + "epoch": 0.42914, + "grad_norm": 1.3483715057373047, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 42914 + }, + { + "epoch": 0.42915, + "grad_norm": 0.8308852910995483, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 42915 + }, + { + "epoch": 0.42916, + "grad_norm": 0.6939243078231812, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 42916 + }, + { + "epoch": 0.42917, + "grad_norm": 0.7647532224655151, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 42917 + }, + { + "epoch": 0.42918, + "grad_norm": 0.7734504342079163, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 42918 + }, + { + "epoch": 0.42919, + "grad_norm": 1.186203122138977, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 42919 + }, + { + "epoch": 0.4292, + "grad_norm": 1.1764198541641235, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 42920 + }, + { + "epoch": 0.42921, + "grad_norm": 0.9807078242301941, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 42921 + }, + { + "epoch": 0.42922, + "grad_norm": 1.1432580947875977, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 42922 + }, + { + "epoch": 0.42923, + "grad_norm": 1.104835867881775, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 42923 + }, + { + "epoch": 0.42924, + "grad_norm": 0.9218307137489319, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 42924 + }, + { + "epoch": 0.42925, + "grad_norm": 0.8591558337211609, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 42925 + }, + { + "epoch": 0.42926, + "grad_norm": 0.8720154762268066, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 42926 + }, + { + "epoch": 0.42927, + "grad_norm": 1.146032452583313, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 42927 + }, + { + "epoch": 0.42928, + "grad_norm": 0.8604015707969666, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 42928 + }, + { + "epoch": 0.42929, + "grad_norm": 0.7265124917030334, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 42929 + }, + { + "epoch": 0.4293, + "grad_norm": 0.8232010006904602, + "learning_rate": 0.003, + "loss": 3.9336, + "step": 42930 + }, + { + "epoch": 0.42931, + "grad_norm": 0.8383179306983948, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 42931 + }, + { + "epoch": 0.42932, + "grad_norm": 0.9076027274131775, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 42932 + }, + { + "epoch": 0.42933, + "grad_norm": 0.9660912752151489, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 42933 + }, + { + "epoch": 0.42934, + "grad_norm": 0.9247288703918457, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 42934 + }, + { + "epoch": 0.42935, + "grad_norm": 0.9016752243041992, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 42935 + }, + { + "epoch": 0.42936, + "grad_norm": 1.0076736211776733, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 42936 + }, + { + "epoch": 0.42937, + "grad_norm": 1.371353268623352, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 42937 + }, + { + "epoch": 0.42938, + "grad_norm": 1.0220386981964111, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 42938 + }, + { + "epoch": 0.42939, + "grad_norm": 0.9317317008972168, + "learning_rate": 0.003, + "loss": 4.029, + "step": 42939 + }, + { + "epoch": 0.4294, + "grad_norm": 0.8876065015792847, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 42940 + }, + { + "epoch": 0.42941, + "grad_norm": 1.0300836563110352, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 42941 + }, + { + "epoch": 0.42942, + "grad_norm": 1.33454167842865, + "learning_rate": 0.003, + "loss": 3.998, + "step": 42942 + }, + { + "epoch": 0.42943, + "grad_norm": 0.7289376258850098, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 42943 + }, + { + "epoch": 0.42944, + "grad_norm": 0.7823198437690735, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 42944 + }, + { + "epoch": 0.42945, + "grad_norm": 1.0396506786346436, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 42945 + }, + { + "epoch": 0.42946, + "grad_norm": 0.9550511240959167, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 42946 + }, + { + "epoch": 0.42947, + "grad_norm": 1.0120439529418945, + "learning_rate": 0.003, + "loss": 4.002, + "step": 42947 + }, + { + "epoch": 0.42948, + "grad_norm": 1.1693744659423828, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 42948 + }, + { + "epoch": 0.42949, + "grad_norm": 0.9021157026290894, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 42949 + }, + { + "epoch": 0.4295, + "grad_norm": 0.7398384809494019, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 42950 + }, + { + "epoch": 0.42951, + "grad_norm": 0.7167650461196899, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 42951 + }, + { + "epoch": 0.42952, + "grad_norm": 0.7771020531654358, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 42952 + }, + { + "epoch": 0.42953, + "grad_norm": 0.6869805455207825, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 42953 + }, + { + "epoch": 0.42954, + "grad_norm": 0.8396334052085876, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 42954 + }, + { + "epoch": 0.42955, + "grad_norm": 1.0697258710861206, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 42955 + }, + { + "epoch": 0.42956, + "grad_norm": 0.9000361561775208, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 42956 + }, + { + "epoch": 0.42957, + "grad_norm": 0.7861975431442261, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 42957 + }, + { + "epoch": 0.42958, + "grad_norm": 0.7109542489051819, + "learning_rate": 0.003, + "loss": 4.006, + "step": 42958 + }, + { + "epoch": 0.42959, + "grad_norm": 0.8314705491065979, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 42959 + }, + { + "epoch": 0.4296, + "grad_norm": 0.8267878293991089, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 42960 + }, + { + "epoch": 0.42961, + "grad_norm": 0.9508774876594543, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 42961 + }, + { + "epoch": 0.42962, + "grad_norm": 1.2103098630905151, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 42962 + }, + { + "epoch": 0.42963, + "grad_norm": 1.03971529006958, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 42963 + }, + { + "epoch": 0.42964, + "grad_norm": 0.9793987274169922, + "learning_rate": 0.003, + "loss": 3.982, + "step": 42964 + }, + { + "epoch": 0.42965, + "grad_norm": 1.0248432159423828, + "learning_rate": 0.003, + "loss": 3.984, + "step": 42965 + }, + { + "epoch": 0.42966, + "grad_norm": 1.0578430891036987, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 42966 + }, + { + "epoch": 0.42967, + "grad_norm": 0.9783953428268433, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 42967 + }, + { + "epoch": 0.42968, + "grad_norm": 0.9451237320899963, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 42968 + }, + { + "epoch": 0.42969, + "grad_norm": 0.8640233874320984, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 42969 + }, + { + "epoch": 0.4297, + "grad_norm": 1.1436481475830078, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 42970 + }, + { + "epoch": 0.42971, + "grad_norm": 0.9517384767532349, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 42971 + }, + { + "epoch": 0.42972, + "grad_norm": 0.7539302706718445, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 42972 + }, + { + "epoch": 0.42973, + "grad_norm": 0.6774719953536987, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 42973 + }, + { + "epoch": 0.42974, + "grad_norm": 0.7509461641311646, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 42974 + }, + { + "epoch": 0.42975, + "grad_norm": 0.8038415908813477, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 42975 + }, + { + "epoch": 0.42976, + "grad_norm": 0.7824161052703857, + "learning_rate": 0.003, + "loss": 3.959, + "step": 42976 + }, + { + "epoch": 0.42977, + "grad_norm": 0.9113063216209412, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 42977 + }, + { + "epoch": 0.42978, + "grad_norm": 1.0800426006317139, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 42978 + }, + { + "epoch": 0.42979, + "grad_norm": 1.0067353248596191, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 42979 + }, + { + "epoch": 0.4298, + "grad_norm": 0.9371845126152039, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 42980 + }, + { + "epoch": 0.42981, + "grad_norm": 0.880688488483429, + "learning_rate": 0.003, + "loss": 4.002, + "step": 42981 + }, + { + "epoch": 0.42982, + "grad_norm": 0.9218140244483948, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 42982 + }, + { + "epoch": 0.42983, + "grad_norm": 0.9870793223381042, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 42983 + }, + { + "epoch": 0.42984, + "grad_norm": 0.9804760217666626, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 42984 + }, + { + "epoch": 0.42985, + "grad_norm": 0.957008421421051, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 42985 + }, + { + "epoch": 0.42986, + "grad_norm": 1.0567387342453003, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 42986 + }, + { + "epoch": 0.42987, + "grad_norm": 1.1388685703277588, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 42987 + }, + { + "epoch": 0.42988, + "grad_norm": 0.8656157851219177, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 42988 + }, + { + "epoch": 0.42989, + "grad_norm": 0.8266494274139404, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 42989 + }, + { + "epoch": 0.4299, + "grad_norm": 0.8853937387466431, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 42990 + }, + { + "epoch": 0.42991, + "grad_norm": 0.9952439069747925, + "learning_rate": 0.003, + "loss": 3.9325, + "step": 42991 + }, + { + "epoch": 0.42992, + "grad_norm": 0.9258648753166199, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 42992 + }, + { + "epoch": 0.42993, + "grad_norm": 0.7016871571540833, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 42993 + }, + { + "epoch": 0.42994, + "grad_norm": 0.6443409323692322, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 42994 + }, + { + "epoch": 0.42995, + "grad_norm": 0.7340396046638489, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 42995 + }, + { + "epoch": 0.42996, + "grad_norm": 0.7587727904319763, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 42996 + }, + { + "epoch": 0.42997, + "grad_norm": 0.715825080871582, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 42997 + }, + { + "epoch": 0.42998, + "grad_norm": 0.7427233457565308, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 42998 + }, + { + "epoch": 0.42999, + "grad_norm": 0.6894306540489197, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 42999 + }, + { + "epoch": 0.43, + "grad_norm": 0.7385161519050598, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 43000 + }, + { + "epoch": 0.43001, + "grad_norm": 0.801401674747467, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 43001 + }, + { + "epoch": 0.43002, + "grad_norm": 0.9759196043014526, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 43002 + }, + { + "epoch": 0.43003, + "grad_norm": 1.1423979997634888, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 43003 + }, + { + "epoch": 0.43004, + "grad_norm": 0.8599446415901184, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 43004 + }, + { + "epoch": 0.43005, + "grad_norm": 1.0777769088745117, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 43005 + }, + { + "epoch": 0.43006, + "grad_norm": 1.2421342134475708, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 43006 + }, + { + "epoch": 0.43007, + "grad_norm": 0.9339633584022522, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 43007 + }, + { + "epoch": 0.43008, + "grad_norm": 0.9560884833335876, + "learning_rate": 0.003, + "loss": 3.964, + "step": 43008 + }, + { + "epoch": 0.43009, + "grad_norm": 1.040961503982544, + "learning_rate": 0.003, + "loss": 3.994, + "step": 43009 + }, + { + "epoch": 0.4301, + "grad_norm": 1.2349660396575928, + "learning_rate": 0.003, + "loss": 3.95, + "step": 43010 + }, + { + "epoch": 0.43011, + "grad_norm": 0.618956983089447, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 43011 + }, + { + "epoch": 0.43012, + "grad_norm": 0.8013079762458801, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 43012 + }, + { + "epoch": 0.43013, + "grad_norm": 1.0664623975753784, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 43013 + }, + { + "epoch": 0.43014, + "grad_norm": 0.9780603051185608, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 43014 + }, + { + "epoch": 0.43015, + "grad_norm": 0.9937858581542969, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 43015 + }, + { + "epoch": 0.43016, + "grad_norm": 0.9768044352531433, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 43016 + }, + { + "epoch": 0.43017, + "grad_norm": 0.8545278906822205, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 43017 + }, + { + "epoch": 0.43018, + "grad_norm": 0.855994701385498, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 43018 + }, + { + "epoch": 0.43019, + "grad_norm": 1.0490118265151978, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 43019 + }, + { + "epoch": 0.4302, + "grad_norm": 1.4076948165893555, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 43020 + }, + { + "epoch": 0.43021, + "grad_norm": 0.8441290259361267, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 43021 + }, + { + "epoch": 0.43022, + "grad_norm": 1.0103797912597656, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 43022 + }, + { + "epoch": 0.43023, + "grad_norm": 1.1451553106307983, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 43023 + }, + { + "epoch": 0.43024, + "grad_norm": 0.7825626134872437, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 43024 + }, + { + "epoch": 0.43025, + "grad_norm": 0.7115554809570312, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 43025 + }, + { + "epoch": 0.43026, + "grad_norm": 0.7365210652351379, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 43026 + }, + { + "epoch": 0.43027, + "grad_norm": 0.7852792739868164, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 43027 + }, + { + "epoch": 0.43028, + "grad_norm": 0.7487519383430481, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 43028 + }, + { + "epoch": 0.43029, + "grad_norm": 0.7945826649665833, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 43029 + }, + { + "epoch": 0.4303, + "grad_norm": 0.7713397145271301, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 43030 + }, + { + "epoch": 0.43031, + "grad_norm": 0.8905251026153564, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 43031 + }, + { + "epoch": 0.43032, + "grad_norm": 1.0931071043014526, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 43032 + }, + { + "epoch": 0.43033, + "grad_norm": 0.8411900997161865, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 43033 + }, + { + "epoch": 0.43034, + "grad_norm": 0.635378897190094, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 43034 + }, + { + "epoch": 0.43035, + "grad_norm": 0.5680957436561584, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 43035 + }, + { + "epoch": 0.43036, + "grad_norm": 0.7255839705467224, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 43036 + }, + { + "epoch": 0.43037, + "grad_norm": 0.9715405702590942, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 43037 + }, + { + "epoch": 0.43038, + "grad_norm": 1.1617536544799805, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 43038 + }, + { + "epoch": 0.43039, + "grad_norm": 0.8776196241378784, + "learning_rate": 0.003, + "loss": 3.983, + "step": 43039 + }, + { + "epoch": 0.4304, + "grad_norm": 0.9060823917388916, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 43040 + }, + { + "epoch": 0.43041, + "grad_norm": 0.9559985995292664, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 43041 + }, + { + "epoch": 0.43042, + "grad_norm": 0.9378650188446045, + "learning_rate": 0.003, + "loss": 3.9412, + "step": 43042 + }, + { + "epoch": 0.43043, + "grad_norm": 0.7773640751838684, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 43043 + }, + { + "epoch": 0.43044, + "grad_norm": 0.8519057631492615, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 43044 + }, + { + "epoch": 0.43045, + "grad_norm": 0.9096124172210693, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 43045 + }, + { + "epoch": 0.43046, + "grad_norm": 1.0651240348815918, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 43046 + }, + { + "epoch": 0.43047, + "grad_norm": 1.197625756263733, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 43047 + }, + { + "epoch": 0.43048, + "grad_norm": 0.8073554635047913, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 43048 + }, + { + "epoch": 0.43049, + "grad_norm": 0.8114461898803711, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 43049 + }, + { + "epoch": 0.4305, + "grad_norm": 0.9109265208244324, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 43050 + }, + { + "epoch": 0.43051, + "grad_norm": 0.8701890110969543, + "learning_rate": 0.003, + "loss": 3.974, + "step": 43051 + }, + { + "epoch": 0.43052, + "grad_norm": 0.7030467987060547, + "learning_rate": 0.003, + "loss": 3.948, + "step": 43052 + }, + { + "epoch": 0.43053, + "grad_norm": 0.7265157699584961, + "learning_rate": 0.003, + "loss": 3.998, + "step": 43053 + }, + { + "epoch": 0.43054, + "grad_norm": 0.7822422981262207, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 43054 + }, + { + "epoch": 0.43055, + "grad_norm": 0.8003215789794922, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 43055 + }, + { + "epoch": 0.43056, + "grad_norm": 0.8989118933677673, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 43056 + }, + { + "epoch": 0.43057, + "grad_norm": 0.8646408319473267, + "learning_rate": 0.003, + "loss": 3.9418, + "step": 43057 + }, + { + "epoch": 0.43058, + "grad_norm": 0.9471102356910706, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 43058 + }, + { + "epoch": 0.43059, + "grad_norm": 0.9434589743614197, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 43059 + }, + { + "epoch": 0.4306, + "grad_norm": 1.0930839776992798, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 43060 + }, + { + "epoch": 0.43061, + "grad_norm": 0.9276874661445618, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 43061 + }, + { + "epoch": 0.43062, + "grad_norm": 1.1143460273742676, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 43062 + }, + { + "epoch": 0.43063, + "grad_norm": 1.104797601699829, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 43063 + }, + { + "epoch": 0.43064, + "grad_norm": 1.205322265625, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 43064 + }, + { + "epoch": 0.43065, + "grad_norm": 0.7864982485771179, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 43065 + }, + { + "epoch": 0.43066, + "grad_norm": 0.7364057898521423, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 43066 + }, + { + "epoch": 0.43067, + "grad_norm": 0.6836933493614197, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 43067 + }, + { + "epoch": 0.43068, + "grad_norm": 0.6665953397750854, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 43068 + }, + { + "epoch": 0.43069, + "grad_norm": 0.8208069205284119, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 43069 + }, + { + "epoch": 0.4307, + "grad_norm": 1.1210194826126099, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 43070 + }, + { + "epoch": 0.43071, + "grad_norm": 1.311453938484192, + "learning_rate": 0.003, + "loss": 4.003, + "step": 43071 + }, + { + "epoch": 0.43072, + "grad_norm": 0.5497981905937195, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 43072 + }, + { + "epoch": 0.43073, + "grad_norm": 0.6869296431541443, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 43073 + }, + { + "epoch": 0.43074, + "grad_norm": 1.0245558023452759, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 43074 + }, + { + "epoch": 0.43075, + "grad_norm": 1.1674927473068237, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 43075 + }, + { + "epoch": 0.43076, + "grad_norm": 0.8747367262840271, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 43076 + }, + { + "epoch": 0.43077, + "grad_norm": 0.6917994618415833, + "learning_rate": 0.003, + "loss": 3.985, + "step": 43077 + }, + { + "epoch": 0.43078, + "grad_norm": 0.6710734963417053, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 43078 + }, + { + "epoch": 0.43079, + "grad_norm": 0.8206791877746582, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 43079 + }, + { + "epoch": 0.4308, + "grad_norm": 0.9276152849197388, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 43080 + }, + { + "epoch": 0.43081, + "grad_norm": 1.1069265604019165, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 43081 + }, + { + "epoch": 0.43082, + "grad_norm": 0.9334841370582581, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 43082 + }, + { + "epoch": 0.43083, + "grad_norm": 0.8323163986206055, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 43083 + }, + { + "epoch": 0.43084, + "grad_norm": 0.8260614275932312, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 43084 + }, + { + "epoch": 0.43085, + "grad_norm": 0.8831632137298584, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 43085 + }, + { + "epoch": 0.43086, + "grad_norm": 0.8067678213119507, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 43086 + }, + { + "epoch": 0.43087, + "grad_norm": 0.766862154006958, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 43087 + }, + { + "epoch": 0.43088, + "grad_norm": 0.8626965284347534, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 43088 + }, + { + "epoch": 0.43089, + "grad_norm": 0.854381263256073, + "learning_rate": 0.003, + "loss": 3.9399, + "step": 43089 + }, + { + "epoch": 0.4309, + "grad_norm": 0.9265123009681702, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 43090 + }, + { + "epoch": 0.43091, + "grad_norm": 1.1156742572784424, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 43091 + }, + { + "epoch": 0.43092, + "grad_norm": 1.0897372961044312, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 43092 + }, + { + "epoch": 0.43093, + "grad_norm": 1.0792181491851807, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 43093 + }, + { + "epoch": 0.43094, + "grad_norm": 0.7979403734207153, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 43094 + }, + { + "epoch": 0.43095, + "grad_norm": 0.754623293876648, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 43095 + }, + { + "epoch": 0.43096, + "grad_norm": 0.7717350125312805, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 43096 + }, + { + "epoch": 0.43097, + "grad_norm": 0.7795308828353882, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 43097 + }, + { + "epoch": 0.43098, + "grad_norm": 0.624937891960144, + "learning_rate": 0.003, + "loss": 3.9456, + "step": 43098 + }, + { + "epoch": 0.43099, + "grad_norm": 0.5737281441688538, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 43099 + }, + { + "epoch": 0.431, + "grad_norm": 0.662065863609314, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 43100 + }, + { + "epoch": 0.43101, + "grad_norm": 0.8062628507614136, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 43101 + }, + { + "epoch": 0.43102, + "grad_norm": 1.020876169204712, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 43102 + }, + { + "epoch": 0.43103, + "grad_norm": 1.279839277267456, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 43103 + }, + { + "epoch": 0.43104, + "grad_norm": 0.8494821190834045, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 43104 + }, + { + "epoch": 0.43105, + "grad_norm": 0.989224910736084, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 43105 + }, + { + "epoch": 0.43106, + "grad_norm": 1.0644370317459106, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 43106 + }, + { + "epoch": 0.43107, + "grad_norm": 1.175068974494934, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 43107 + }, + { + "epoch": 0.43108, + "grad_norm": 0.7614088654518127, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 43108 + }, + { + "epoch": 0.43109, + "grad_norm": 0.6509212851524353, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 43109 + }, + { + "epoch": 0.4311, + "grad_norm": 0.6828076243400574, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 43110 + }, + { + "epoch": 0.43111, + "grad_norm": 0.6928569674491882, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 43111 + }, + { + "epoch": 0.43112, + "grad_norm": 0.7979059219360352, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 43112 + }, + { + "epoch": 0.43113, + "grad_norm": 0.9912794828414917, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 43113 + }, + { + "epoch": 0.43114, + "grad_norm": 1.2998300790786743, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 43114 + }, + { + "epoch": 0.43115, + "grad_norm": 0.6037266850471497, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 43115 + }, + { + "epoch": 0.43116, + "grad_norm": 0.7201408743858337, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 43116 + }, + { + "epoch": 0.43117, + "grad_norm": 0.9692350029945374, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 43117 + }, + { + "epoch": 0.43118, + "grad_norm": 1.047060489654541, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 43118 + }, + { + "epoch": 0.43119, + "grad_norm": 1.0360915660858154, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 43119 + }, + { + "epoch": 0.4312, + "grad_norm": 1.0613549947738647, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 43120 + }, + { + "epoch": 0.43121, + "grad_norm": 0.8515404462814331, + "learning_rate": 0.003, + "loss": 3.9422, + "step": 43121 + }, + { + "epoch": 0.43122, + "grad_norm": 0.9521795511245728, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 43122 + }, + { + "epoch": 0.43123, + "grad_norm": 1.0636720657348633, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 43123 + }, + { + "epoch": 0.43124, + "grad_norm": 0.960645854473114, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 43124 + }, + { + "epoch": 0.43125, + "grad_norm": 1.0102689266204834, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 43125 + }, + { + "epoch": 0.43126, + "grad_norm": 1.1547740697860718, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 43126 + }, + { + "epoch": 0.43127, + "grad_norm": 0.868390679359436, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 43127 + }, + { + "epoch": 0.43128, + "grad_norm": 0.9181337952613831, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 43128 + }, + { + "epoch": 0.43129, + "grad_norm": 1.0262963771820068, + "learning_rate": 0.003, + "loss": 4.013, + "step": 43129 + }, + { + "epoch": 0.4313, + "grad_norm": 1.0912381410598755, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 43130 + }, + { + "epoch": 0.43131, + "grad_norm": 0.8522229790687561, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 43131 + }, + { + "epoch": 0.43132, + "grad_norm": 0.9869105219841003, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 43132 + }, + { + "epoch": 0.43133, + "grad_norm": 1.0555247068405151, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 43133 + }, + { + "epoch": 0.43134, + "grad_norm": 1.237805962562561, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 43134 + }, + { + "epoch": 0.43135, + "grad_norm": 1.4784091711044312, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 43135 + }, + { + "epoch": 0.43136, + "grad_norm": 0.5960695743560791, + "learning_rate": 0.003, + "loss": 3.9408, + "step": 43136 + }, + { + "epoch": 0.43137, + "grad_norm": 0.7834169268608093, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 43137 + }, + { + "epoch": 0.43138, + "grad_norm": 0.9546573758125305, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 43138 + }, + { + "epoch": 0.43139, + "grad_norm": 1.0754239559173584, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 43139 + }, + { + "epoch": 0.4314, + "grad_norm": 1.1579278707504272, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 43140 + }, + { + "epoch": 0.43141, + "grad_norm": 0.8607880473136902, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 43141 + }, + { + "epoch": 0.43142, + "grad_norm": 0.6805599331855774, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 43142 + }, + { + "epoch": 0.43143, + "grad_norm": 0.6612477898597717, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 43143 + }, + { + "epoch": 0.43144, + "grad_norm": 0.6925220489501953, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 43144 + }, + { + "epoch": 0.43145, + "grad_norm": 0.7464480996131897, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 43145 + }, + { + "epoch": 0.43146, + "grad_norm": 0.8059023022651672, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 43146 + }, + { + "epoch": 0.43147, + "grad_norm": 1.0540283918380737, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 43147 + }, + { + "epoch": 0.43148, + "grad_norm": 1.155572533607483, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 43148 + }, + { + "epoch": 0.43149, + "grad_norm": 0.6611666679382324, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 43149 + }, + { + "epoch": 0.4315, + "grad_norm": 0.6868916749954224, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 43150 + }, + { + "epoch": 0.43151, + "grad_norm": 0.843339204788208, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 43151 + }, + { + "epoch": 0.43152, + "grad_norm": 0.9078029990196228, + "learning_rate": 0.003, + "loss": 3.9458, + "step": 43152 + }, + { + "epoch": 0.43153, + "grad_norm": 1.1082156896591187, + "learning_rate": 0.003, + "loss": 3.9488, + "step": 43153 + }, + { + "epoch": 0.43154, + "grad_norm": 0.8764249682426453, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 43154 + }, + { + "epoch": 0.43155, + "grad_norm": 0.6523086428642273, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 43155 + }, + { + "epoch": 0.43156, + "grad_norm": 0.8478659391403198, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 43156 + }, + { + "epoch": 0.43157, + "grad_norm": 1.1238958835601807, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 43157 + }, + { + "epoch": 0.43158, + "grad_norm": 0.9026181101799011, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 43158 + }, + { + "epoch": 0.43159, + "grad_norm": 0.817754328250885, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 43159 + }, + { + "epoch": 0.4316, + "grad_norm": 0.7723127007484436, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 43160 + }, + { + "epoch": 0.43161, + "grad_norm": 0.8320155143737793, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 43161 + }, + { + "epoch": 0.43162, + "grad_norm": 0.8062442541122437, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 43162 + }, + { + "epoch": 0.43163, + "grad_norm": 0.8798151612281799, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 43163 + }, + { + "epoch": 0.43164, + "grad_norm": 1.1648790836334229, + "learning_rate": 0.003, + "loss": 3.97, + "step": 43164 + }, + { + "epoch": 0.43165, + "grad_norm": 0.9210562109947205, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 43165 + }, + { + "epoch": 0.43166, + "grad_norm": 0.8328824043273926, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 43166 + }, + { + "epoch": 0.43167, + "grad_norm": 0.8015958666801453, + "learning_rate": 0.003, + "loss": 3.963, + "step": 43167 + }, + { + "epoch": 0.43168, + "grad_norm": 0.7692462205886841, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 43168 + }, + { + "epoch": 0.43169, + "grad_norm": 0.8901144862174988, + "learning_rate": 0.003, + "loss": 3.9468, + "step": 43169 + }, + { + "epoch": 0.4317, + "grad_norm": 1.1041419506072998, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 43170 + }, + { + "epoch": 0.43171, + "grad_norm": 0.8843094706535339, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 43171 + }, + { + "epoch": 0.43172, + "grad_norm": 0.8398434519767761, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 43172 + }, + { + "epoch": 0.43173, + "grad_norm": 0.7819473147392273, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 43173 + }, + { + "epoch": 0.43174, + "grad_norm": 0.7691479921340942, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 43174 + }, + { + "epoch": 0.43175, + "grad_norm": 0.9644784331321716, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 43175 + }, + { + "epoch": 0.43176, + "grad_norm": 1.4167671203613281, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 43176 + }, + { + "epoch": 0.43177, + "grad_norm": 0.9251941442489624, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 43177 + }, + { + "epoch": 0.43178, + "grad_norm": 0.8136219382286072, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 43178 + }, + { + "epoch": 0.43179, + "grad_norm": 0.6800025105476379, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 43179 + }, + { + "epoch": 0.4318, + "grad_norm": 0.7800778150558472, + "learning_rate": 0.003, + "loss": 3.9444, + "step": 43180 + }, + { + "epoch": 0.43181, + "grad_norm": 0.7998527884483337, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 43181 + }, + { + "epoch": 0.43182, + "grad_norm": 0.9024674296379089, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 43182 + }, + { + "epoch": 0.43183, + "grad_norm": 0.9537393450737, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 43183 + }, + { + "epoch": 0.43184, + "grad_norm": 0.9596313834190369, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 43184 + }, + { + "epoch": 0.43185, + "grad_norm": 1.0255006551742554, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 43185 + }, + { + "epoch": 0.43186, + "grad_norm": 1.0437458753585815, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 43186 + }, + { + "epoch": 0.43187, + "grad_norm": 0.9110845923423767, + "learning_rate": 0.003, + "loss": 3.97, + "step": 43187 + }, + { + "epoch": 0.43188, + "grad_norm": 0.9881011247634888, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 43188 + }, + { + "epoch": 0.43189, + "grad_norm": 0.9590907096862793, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 43189 + }, + { + "epoch": 0.4319, + "grad_norm": 0.746597170829773, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 43190 + }, + { + "epoch": 0.43191, + "grad_norm": 0.9406048655509949, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 43191 + }, + { + "epoch": 0.43192, + "grad_norm": 1.1266319751739502, + "learning_rate": 0.003, + "loss": 3.998, + "step": 43192 + }, + { + "epoch": 0.43193, + "grad_norm": 1.1119394302368164, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 43193 + }, + { + "epoch": 0.43194, + "grad_norm": 1.0614603757858276, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 43194 + }, + { + "epoch": 0.43195, + "grad_norm": 1.2403181791305542, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 43195 + }, + { + "epoch": 0.43196, + "grad_norm": 0.7348313927650452, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 43196 + }, + { + "epoch": 0.43197, + "grad_norm": 0.6742569804191589, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 43197 + }, + { + "epoch": 0.43198, + "grad_norm": 0.7142609357833862, + "learning_rate": 0.003, + "loss": 4.001, + "step": 43198 + }, + { + "epoch": 0.43199, + "grad_norm": 0.8317811489105225, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 43199 + }, + { + "epoch": 0.432, + "grad_norm": 0.9826295971870422, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 43200 + }, + { + "epoch": 0.43201, + "grad_norm": 1.348559021949768, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 43201 + }, + { + "epoch": 0.43202, + "grad_norm": 0.635410487651825, + "learning_rate": 0.003, + "loss": 3.9488, + "step": 43202 + }, + { + "epoch": 0.43203, + "grad_norm": 0.6785230040550232, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 43203 + }, + { + "epoch": 0.43204, + "grad_norm": 0.783306896686554, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 43204 + }, + { + "epoch": 0.43205, + "grad_norm": 0.8468449115753174, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 43205 + }, + { + "epoch": 0.43206, + "grad_norm": 0.8801425695419312, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 43206 + }, + { + "epoch": 0.43207, + "grad_norm": 0.8869838714599609, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 43207 + }, + { + "epoch": 0.43208, + "grad_norm": 0.8983903527259827, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 43208 + }, + { + "epoch": 0.43209, + "grad_norm": 1.014544129371643, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 43209 + }, + { + "epoch": 0.4321, + "grad_norm": 1.0235260725021362, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 43210 + }, + { + "epoch": 0.43211, + "grad_norm": 1.0699716806411743, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 43211 + }, + { + "epoch": 0.43212, + "grad_norm": 0.8676535487174988, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 43212 + }, + { + "epoch": 0.43213, + "grad_norm": 0.9454278945922852, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 43213 + }, + { + "epoch": 0.43214, + "grad_norm": 1.0690770149230957, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 43214 + }, + { + "epoch": 0.43215, + "grad_norm": 1.0528095960617065, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 43215 + }, + { + "epoch": 0.43216, + "grad_norm": 1.1003696918487549, + "learning_rate": 0.003, + "loss": 3.961, + "step": 43216 + }, + { + "epoch": 0.43217, + "grad_norm": 1.2295255661010742, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 43217 + }, + { + "epoch": 0.43218, + "grad_norm": 0.8166696429252625, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 43218 + }, + { + "epoch": 0.43219, + "grad_norm": 0.7105631232261658, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 43219 + }, + { + "epoch": 0.4322, + "grad_norm": 0.8997461199760437, + "learning_rate": 0.003, + "loss": 3.978, + "step": 43220 + }, + { + "epoch": 0.43221, + "grad_norm": 1.002463459968567, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 43221 + }, + { + "epoch": 0.43222, + "grad_norm": 1.1724334955215454, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 43222 + }, + { + "epoch": 0.43223, + "grad_norm": 0.8685101866722107, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 43223 + }, + { + "epoch": 0.43224, + "grad_norm": 0.7641156911849976, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 43224 + }, + { + "epoch": 0.43225, + "grad_norm": 0.7924295663833618, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 43225 + }, + { + "epoch": 0.43226, + "grad_norm": 0.8495567440986633, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 43226 + }, + { + "epoch": 0.43227, + "grad_norm": 0.927215039730072, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 43227 + }, + { + "epoch": 0.43228, + "grad_norm": 1.2276384830474854, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 43228 + }, + { + "epoch": 0.43229, + "grad_norm": 0.640519380569458, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 43229 + }, + { + "epoch": 0.4323, + "grad_norm": 0.5659250617027283, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 43230 + }, + { + "epoch": 0.43231, + "grad_norm": 0.7729545831680298, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 43231 + }, + { + "epoch": 0.43232, + "grad_norm": 0.9753103256225586, + "learning_rate": 0.003, + "loss": 3.971, + "step": 43232 + }, + { + "epoch": 0.43233, + "grad_norm": 1.2436472177505493, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 43233 + }, + { + "epoch": 0.43234, + "grad_norm": 0.6788517832756042, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 43234 + }, + { + "epoch": 0.43235, + "grad_norm": 0.6598168015480042, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 43235 + }, + { + "epoch": 0.43236, + "grad_norm": 0.6328692436218262, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 43236 + }, + { + "epoch": 0.43237, + "grad_norm": 0.639378011226654, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 43237 + }, + { + "epoch": 0.43238, + "grad_norm": 0.6391421556472778, + "learning_rate": 0.003, + "loss": 3.962, + "step": 43238 + }, + { + "epoch": 0.43239, + "grad_norm": 0.7611197233200073, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 43239 + }, + { + "epoch": 0.4324, + "grad_norm": 0.8288336396217346, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 43240 + }, + { + "epoch": 0.43241, + "grad_norm": 0.9285392165184021, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 43241 + }, + { + "epoch": 0.43242, + "grad_norm": 1.1640410423278809, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 43242 + }, + { + "epoch": 0.43243, + "grad_norm": 0.9283625483512878, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 43243 + }, + { + "epoch": 0.43244, + "grad_norm": 0.961089015007019, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 43244 + }, + { + "epoch": 0.43245, + "grad_norm": 0.9862825870513916, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 43245 + }, + { + "epoch": 0.43246, + "grad_norm": 0.9239320755004883, + "learning_rate": 0.003, + "loss": 3.981, + "step": 43246 + }, + { + "epoch": 0.43247, + "grad_norm": 0.7833354473114014, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 43247 + }, + { + "epoch": 0.43248, + "grad_norm": 0.9618902206420898, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 43248 + }, + { + "epoch": 0.43249, + "grad_norm": 1.3002092838287354, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 43249 + }, + { + "epoch": 0.4325, + "grad_norm": 0.8382505774497986, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 43250 + }, + { + "epoch": 0.43251, + "grad_norm": 0.7425589561462402, + "learning_rate": 0.003, + "loss": 3.96, + "step": 43251 + }, + { + "epoch": 0.43252, + "grad_norm": 0.8163500428199768, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 43252 + }, + { + "epoch": 0.43253, + "grad_norm": 0.9228964447975159, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 43253 + }, + { + "epoch": 0.43254, + "grad_norm": 1.1989749670028687, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 43254 + }, + { + "epoch": 0.43255, + "grad_norm": 1.1365097761154175, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 43255 + }, + { + "epoch": 0.43256, + "grad_norm": 0.998374342918396, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 43256 + }, + { + "epoch": 0.43257, + "grad_norm": 0.8477112650871277, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 43257 + }, + { + "epoch": 0.43258, + "grad_norm": 0.8200920820236206, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 43258 + }, + { + "epoch": 0.43259, + "grad_norm": 0.8084766268730164, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 43259 + }, + { + "epoch": 0.4326, + "grad_norm": 0.8415750861167908, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 43260 + }, + { + "epoch": 0.43261, + "grad_norm": 0.9971763491630554, + "learning_rate": 0.003, + "loss": 3.991, + "step": 43261 + }, + { + "epoch": 0.43262, + "grad_norm": 1.1852763891220093, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 43262 + }, + { + "epoch": 0.43263, + "grad_norm": 1.1274147033691406, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 43263 + }, + { + "epoch": 0.43264, + "grad_norm": 1.0181777477264404, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 43264 + }, + { + "epoch": 0.43265, + "grad_norm": 0.9492008686065674, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 43265 + }, + { + "epoch": 0.43266, + "grad_norm": 0.9012932777404785, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 43266 + }, + { + "epoch": 0.43267, + "grad_norm": 0.8024832010269165, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 43267 + }, + { + "epoch": 0.43268, + "grad_norm": 0.9755657315254211, + "learning_rate": 0.003, + "loss": 3.962, + "step": 43268 + }, + { + "epoch": 0.43269, + "grad_norm": 1.3207552433013916, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 43269 + }, + { + "epoch": 0.4327, + "grad_norm": 0.8784767389297485, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 43270 + }, + { + "epoch": 0.43271, + "grad_norm": 0.7594501972198486, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 43271 + }, + { + "epoch": 0.43272, + "grad_norm": 0.8633622527122498, + "learning_rate": 0.003, + "loss": 3.976, + "step": 43272 + }, + { + "epoch": 0.43273, + "grad_norm": 0.9495154619216919, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 43273 + }, + { + "epoch": 0.43274, + "grad_norm": 0.8387694954872131, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 43274 + }, + { + "epoch": 0.43275, + "grad_norm": 0.7330785393714905, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 43275 + }, + { + "epoch": 0.43276, + "grad_norm": 0.8005468845367432, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 43276 + }, + { + "epoch": 0.43277, + "grad_norm": 0.8583207130432129, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 43277 + }, + { + "epoch": 0.43278, + "grad_norm": 0.9652092456817627, + "learning_rate": 0.003, + "loss": 3.992, + "step": 43278 + }, + { + "epoch": 0.43279, + "grad_norm": 1.1695150136947632, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 43279 + }, + { + "epoch": 0.4328, + "grad_norm": 1.0696752071380615, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 43280 + }, + { + "epoch": 0.43281, + "grad_norm": 0.9447618722915649, + "learning_rate": 0.003, + "loss": 3.96, + "step": 43281 + }, + { + "epoch": 0.43282, + "grad_norm": 0.9214510321617126, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 43282 + }, + { + "epoch": 0.43283, + "grad_norm": 0.9527389407157898, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 43283 + }, + { + "epoch": 0.43284, + "grad_norm": 0.9236548542976379, + "learning_rate": 0.003, + "loss": 3.997, + "step": 43284 + }, + { + "epoch": 0.43285, + "grad_norm": 0.8136249780654907, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 43285 + }, + { + "epoch": 0.43286, + "grad_norm": 0.7904257774353027, + "learning_rate": 0.003, + "loss": 3.981, + "step": 43286 + }, + { + "epoch": 0.43287, + "grad_norm": 0.8699457049369812, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 43287 + }, + { + "epoch": 0.43288, + "grad_norm": 1.0448311567306519, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 43288 + }, + { + "epoch": 0.43289, + "grad_norm": 1.042641520500183, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 43289 + }, + { + "epoch": 0.4329, + "grad_norm": 0.9135768413543701, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 43290 + }, + { + "epoch": 0.43291, + "grad_norm": 0.939169704914093, + "learning_rate": 0.003, + "loss": 3.975, + "step": 43291 + }, + { + "epoch": 0.43292, + "grad_norm": 0.974663496017456, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 43292 + }, + { + "epoch": 0.43293, + "grad_norm": 0.7635183334350586, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 43293 + }, + { + "epoch": 0.43294, + "grad_norm": 0.769352912902832, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 43294 + }, + { + "epoch": 0.43295, + "grad_norm": 0.7844465374946594, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 43295 + }, + { + "epoch": 0.43296, + "grad_norm": 1.0897125005722046, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 43296 + }, + { + "epoch": 0.43297, + "grad_norm": 1.066344976425171, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 43297 + }, + { + "epoch": 0.43298, + "grad_norm": 0.9965066909790039, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 43298 + }, + { + "epoch": 0.43299, + "grad_norm": 1.106488585472107, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 43299 + }, + { + "epoch": 0.433, + "grad_norm": 0.9441744685173035, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 43300 + }, + { + "epoch": 0.43301, + "grad_norm": 0.8243303894996643, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 43301 + }, + { + "epoch": 0.43302, + "grad_norm": 0.6854305267333984, + "learning_rate": 0.003, + "loss": 3.985, + "step": 43302 + }, + { + "epoch": 0.43303, + "grad_norm": 0.7607675194740295, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 43303 + }, + { + "epoch": 0.43304, + "grad_norm": 0.8357384204864502, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 43304 + }, + { + "epoch": 0.43305, + "grad_norm": 0.9413255453109741, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 43305 + }, + { + "epoch": 0.43306, + "grad_norm": 1.1714799404144287, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 43306 + }, + { + "epoch": 0.43307, + "grad_norm": 0.8438315391540527, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 43307 + }, + { + "epoch": 0.43308, + "grad_norm": 0.9460282325744629, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 43308 + }, + { + "epoch": 0.43309, + "grad_norm": 0.9957338571548462, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 43309 + }, + { + "epoch": 0.4331, + "grad_norm": 1.1778128147125244, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 43310 + }, + { + "epoch": 0.43311, + "grad_norm": 0.9381806254386902, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 43311 + }, + { + "epoch": 0.43312, + "grad_norm": 0.9402326941490173, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 43312 + }, + { + "epoch": 0.43313, + "grad_norm": 1.0786805152893066, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 43313 + }, + { + "epoch": 0.43314, + "grad_norm": 0.9836130142211914, + "learning_rate": 0.003, + "loss": 3.992, + "step": 43314 + }, + { + "epoch": 0.43315, + "grad_norm": 0.7904523015022278, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 43315 + }, + { + "epoch": 0.43316, + "grad_norm": 0.7006754875183105, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 43316 + }, + { + "epoch": 0.43317, + "grad_norm": 0.9561275839805603, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 43317 + }, + { + "epoch": 0.43318, + "grad_norm": 1.2564386129379272, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 43318 + }, + { + "epoch": 0.43319, + "grad_norm": 0.6983397603034973, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 43319 + }, + { + "epoch": 0.4332, + "grad_norm": 0.6475402116775513, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 43320 + }, + { + "epoch": 0.43321, + "grad_norm": 0.7891172170639038, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 43321 + }, + { + "epoch": 0.43322, + "grad_norm": 0.8611558079719543, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 43322 + }, + { + "epoch": 0.43323, + "grad_norm": 0.8963682651519775, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 43323 + }, + { + "epoch": 0.43324, + "grad_norm": 0.9003245830535889, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 43324 + }, + { + "epoch": 0.43325, + "grad_norm": 1.0572872161865234, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 43325 + }, + { + "epoch": 0.43326, + "grad_norm": 1.2204471826553345, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 43326 + }, + { + "epoch": 0.43327, + "grad_norm": 0.7679536938667297, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 43327 + }, + { + "epoch": 0.43328, + "grad_norm": 0.6921120285987854, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 43328 + }, + { + "epoch": 0.43329, + "grad_norm": 0.751862108707428, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 43329 + }, + { + "epoch": 0.4333, + "grad_norm": 0.798917293548584, + "learning_rate": 0.003, + "loss": 3.97, + "step": 43330 + }, + { + "epoch": 0.43331, + "grad_norm": 0.995856761932373, + "learning_rate": 0.003, + "loss": 3.962, + "step": 43331 + }, + { + "epoch": 0.43332, + "grad_norm": 1.0949082374572754, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 43332 + }, + { + "epoch": 0.43333, + "grad_norm": 0.9507946968078613, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 43333 + }, + { + "epoch": 0.43334, + "grad_norm": 1.1986706256866455, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 43334 + }, + { + "epoch": 0.43335, + "grad_norm": 0.9810526967048645, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 43335 + }, + { + "epoch": 0.43336, + "grad_norm": 0.8047610521316528, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 43336 + }, + { + "epoch": 0.43337, + "grad_norm": 0.7398959994316101, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 43337 + }, + { + "epoch": 0.43338, + "grad_norm": 0.739040195941925, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 43338 + }, + { + "epoch": 0.43339, + "grad_norm": 0.7941365242004395, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 43339 + }, + { + "epoch": 0.4334, + "grad_norm": 0.731209933757782, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 43340 + }, + { + "epoch": 0.43341, + "grad_norm": 0.6507309675216675, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 43341 + }, + { + "epoch": 0.43342, + "grad_norm": 0.7693453431129456, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 43342 + }, + { + "epoch": 0.43343, + "grad_norm": 1.0721359252929688, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 43343 + }, + { + "epoch": 0.43344, + "grad_norm": 1.3143874406814575, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 43344 + }, + { + "epoch": 0.43345, + "grad_norm": 0.8911823630332947, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 43345 + }, + { + "epoch": 0.43346, + "grad_norm": 0.7822197079658508, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 43346 + }, + { + "epoch": 0.43347, + "grad_norm": 0.870945930480957, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 43347 + }, + { + "epoch": 0.43348, + "grad_norm": 0.8315239548683167, + "learning_rate": 0.003, + "loss": 3.9341, + "step": 43348 + }, + { + "epoch": 0.43349, + "grad_norm": 0.8067573308944702, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 43349 + }, + { + "epoch": 0.4335, + "grad_norm": 1.0197123289108276, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 43350 + }, + { + "epoch": 0.43351, + "grad_norm": 1.0073148012161255, + "learning_rate": 0.003, + "loss": 4.01, + "step": 43351 + }, + { + "epoch": 0.43352, + "grad_norm": 0.9558846354484558, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 43352 + }, + { + "epoch": 0.43353, + "grad_norm": 0.9351833462715149, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 43353 + }, + { + "epoch": 0.43354, + "grad_norm": 1.0010669231414795, + "learning_rate": 0.003, + "loss": 3.9328, + "step": 43354 + }, + { + "epoch": 0.43355, + "grad_norm": 1.1194156408309937, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 43355 + }, + { + "epoch": 0.43356, + "grad_norm": 0.9162479043006897, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 43356 + }, + { + "epoch": 0.43357, + "grad_norm": 0.9059404730796814, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 43357 + }, + { + "epoch": 0.43358, + "grad_norm": 1.0649869441986084, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 43358 + }, + { + "epoch": 0.43359, + "grad_norm": 1.1821097135543823, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 43359 + }, + { + "epoch": 0.4336, + "grad_norm": 1.151318907737732, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 43360 + }, + { + "epoch": 0.43361, + "grad_norm": 1.0891886949539185, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 43361 + }, + { + "epoch": 0.43362, + "grad_norm": 1.1244357824325562, + "learning_rate": 0.003, + "loss": 4.028, + "step": 43362 + }, + { + "epoch": 0.43363, + "grad_norm": 0.8955857753753662, + "learning_rate": 0.003, + "loss": 3.972, + "step": 43363 + }, + { + "epoch": 0.43364, + "grad_norm": 1.025452971458435, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 43364 + }, + { + "epoch": 0.43365, + "grad_norm": 1.1733025312423706, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 43365 + }, + { + "epoch": 0.43366, + "grad_norm": 0.6856675744056702, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 43366 + }, + { + "epoch": 0.43367, + "grad_norm": 0.6161444187164307, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 43367 + }, + { + "epoch": 0.43368, + "grad_norm": 0.7273669838905334, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 43368 + }, + { + "epoch": 0.43369, + "grad_norm": 0.8064104318618774, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 43369 + }, + { + "epoch": 0.4337, + "grad_norm": 0.8811393976211548, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 43370 + }, + { + "epoch": 0.43371, + "grad_norm": 0.9911395907402039, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 43371 + }, + { + "epoch": 0.43372, + "grad_norm": 1.2292531728744507, + "learning_rate": 0.003, + "loss": 3.998, + "step": 43372 + }, + { + "epoch": 0.43373, + "grad_norm": 0.7514126896858215, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 43373 + }, + { + "epoch": 0.43374, + "grad_norm": 0.6654276251792908, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 43374 + }, + { + "epoch": 0.43375, + "grad_norm": 0.5970287919044495, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 43375 + }, + { + "epoch": 0.43376, + "grad_norm": 0.6244418025016785, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 43376 + }, + { + "epoch": 0.43377, + "grad_norm": 0.8068926334381104, + "learning_rate": 0.003, + "loss": 3.9314, + "step": 43377 + }, + { + "epoch": 0.43378, + "grad_norm": 1.074257254600525, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 43378 + }, + { + "epoch": 0.43379, + "grad_norm": 0.9230802059173584, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 43379 + }, + { + "epoch": 0.4338, + "grad_norm": 0.6659383773803711, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 43380 + }, + { + "epoch": 0.43381, + "grad_norm": 0.7147281169891357, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 43381 + }, + { + "epoch": 0.43382, + "grad_norm": 0.841432511806488, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 43382 + }, + { + "epoch": 0.43383, + "grad_norm": 0.9677046537399292, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 43383 + }, + { + "epoch": 0.43384, + "grad_norm": 1.064004898071289, + "learning_rate": 0.003, + "loss": 3.9407, + "step": 43384 + }, + { + "epoch": 0.43385, + "grad_norm": 0.886688768863678, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 43385 + }, + { + "epoch": 0.43386, + "grad_norm": 0.8557284474372864, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 43386 + }, + { + "epoch": 0.43387, + "grad_norm": 0.8057675361633301, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 43387 + }, + { + "epoch": 0.43388, + "grad_norm": 0.8911383152008057, + "learning_rate": 0.003, + "loss": 3.9432, + "step": 43388 + }, + { + "epoch": 0.43389, + "grad_norm": 0.8690953254699707, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 43389 + }, + { + "epoch": 0.4339, + "grad_norm": 1.0048801898956299, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 43390 + }, + { + "epoch": 0.43391, + "grad_norm": 1.1388312578201294, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 43391 + }, + { + "epoch": 0.43392, + "grad_norm": 1.0149530172348022, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 43392 + }, + { + "epoch": 0.43393, + "grad_norm": 1.1870306730270386, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 43393 + }, + { + "epoch": 0.43394, + "grad_norm": 1.022458791732788, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 43394 + }, + { + "epoch": 0.43395, + "grad_norm": 1.0143436193466187, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 43395 + }, + { + "epoch": 0.43396, + "grad_norm": 1.00832200050354, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 43396 + }, + { + "epoch": 0.43397, + "grad_norm": 0.9012414813041687, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 43397 + }, + { + "epoch": 0.43398, + "grad_norm": 0.883553683757782, + "learning_rate": 0.003, + "loss": 3.997, + "step": 43398 + }, + { + "epoch": 0.43399, + "grad_norm": 0.9957208037376404, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 43399 + }, + { + "epoch": 0.434, + "grad_norm": 1.199958086013794, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 43400 + }, + { + "epoch": 0.43401, + "grad_norm": 1.0220558643341064, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 43401 + }, + { + "epoch": 0.43402, + "grad_norm": 1.0600517988204956, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 43402 + }, + { + "epoch": 0.43403, + "grad_norm": 0.9859919548034668, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 43403 + }, + { + "epoch": 0.43404, + "grad_norm": 1.0617424249649048, + "learning_rate": 0.003, + "loss": 3.994, + "step": 43404 + }, + { + "epoch": 0.43405, + "grad_norm": 0.936523973941803, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 43405 + }, + { + "epoch": 0.43406, + "grad_norm": 1.0819364786148071, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 43406 + }, + { + "epoch": 0.43407, + "grad_norm": 0.9030259251594543, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 43407 + }, + { + "epoch": 0.43408, + "grad_norm": 0.7574011087417603, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 43408 + }, + { + "epoch": 0.43409, + "grad_norm": 0.7524277567863464, + "learning_rate": 0.003, + "loss": 3.98, + "step": 43409 + }, + { + "epoch": 0.4341, + "grad_norm": 0.793065071105957, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 43410 + }, + { + "epoch": 0.43411, + "grad_norm": 0.8898347616195679, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 43411 + }, + { + "epoch": 0.43412, + "grad_norm": 1.0187656879425049, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 43412 + }, + { + "epoch": 0.43413, + "grad_norm": 0.9814345836639404, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 43413 + }, + { + "epoch": 0.43414, + "grad_norm": 0.9548584818840027, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 43414 + }, + { + "epoch": 0.43415, + "grad_norm": 1.0288424491882324, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 43415 + }, + { + "epoch": 0.43416, + "grad_norm": 1.0908069610595703, + "learning_rate": 0.003, + "loss": 4.002, + "step": 43416 + }, + { + "epoch": 0.43417, + "grad_norm": 0.9326081275939941, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 43417 + }, + { + "epoch": 0.43418, + "grad_norm": 0.9444082975387573, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 43418 + }, + { + "epoch": 0.43419, + "grad_norm": 0.8476058840751648, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 43419 + }, + { + "epoch": 0.4342, + "grad_norm": 0.7476775646209717, + "learning_rate": 0.003, + "loss": 3.97, + "step": 43420 + }, + { + "epoch": 0.43421, + "grad_norm": 0.734124481678009, + "learning_rate": 0.003, + "loss": 3.9396, + "step": 43421 + }, + { + "epoch": 0.43422, + "grad_norm": 0.8253819942474365, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 43422 + }, + { + "epoch": 0.43423, + "grad_norm": 1.027877688407898, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 43423 + }, + { + "epoch": 0.43424, + "grad_norm": 1.1739121675491333, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 43424 + }, + { + "epoch": 0.43425, + "grad_norm": 0.8806194067001343, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 43425 + }, + { + "epoch": 0.43426, + "grad_norm": 0.8322902321815491, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 43426 + }, + { + "epoch": 0.43427, + "grad_norm": 0.9655109643936157, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 43427 + }, + { + "epoch": 0.43428, + "grad_norm": 1.0997867584228516, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 43428 + }, + { + "epoch": 0.43429, + "grad_norm": 0.9620680212974548, + "learning_rate": 0.003, + "loss": 3.982, + "step": 43429 + }, + { + "epoch": 0.4343, + "grad_norm": 1.0214158296585083, + "learning_rate": 0.003, + "loss": 3.9336, + "step": 43430 + }, + { + "epoch": 0.43431, + "grad_norm": 1.020668387413025, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 43431 + }, + { + "epoch": 0.43432, + "grad_norm": 0.7699960470199585, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 43432 + }, + { + "epoch": 0.43433, + "grad_norm": 0.6669672131538391, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 43433 + }, + { + "epoch": 0.43434, + "grad_norm": 0.6738522052764893, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 43434 + }, + { + "epoch": 0.43435, + "grad_norm": 0.6677774786949158, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 43435 + }, + { + "epoch": 0.43436, + "grad_norm": 0.8238857984542847, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 43436 + }, + { + "epoch": 0.43437, + "grad_norm": 0.8700777292251587, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 43437 + }, + { + "epoch": 0.43438, + "grad_norm": 0.8301176428794861, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 43438 + }, + { + "epoch": 0.43439, + "grad_norm": 0.8997572660446167, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 43439 + }, + { + "epoch": 0.4344, + "grad_norm": 1.1603068113327026, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 43440 + }, + { + "epoch": 0.43441, + "grad_norm": 1.0792381763458252, + "learning_rate": 0.003, + "loss": 3.943, + "step": 43441 + }, + { + "epoch": 0.43442, + "grad_norm": 0.760966420173645, + "learning_rate": 0.003, + "loss": 3.953, + "step": 43442 + }, + { + "epoch": 0.43443, + "grad_norm": 0.8180087208747864, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 43443 + }, + { + "epoch": 0.43444, + "grad_norm": 0.9288723468780518, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 43444 + }, + { + "epoch": 0.43445, + "grad_norm": 1.02459716796875, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 43445 + }, + { + "epoch": 0.43446, + "grad_norm": 0.9722915887832642, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 43446 + }, + { + "epoch": 0.43447, + "grad_norm": 0.8779733180999756, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 43447 + }, + { + "epoch": 0.43448, + "grad_norm": 0.9753257632255554, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 43448 + }, + { + "epoch": 0.43449, + "grad_norm": 1.0531059503555298, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 43449 + }, + { + "epoch": 0.4345, + "grad_norm": 0.8642395734786987, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 43450 + }, + { + "epoch": 0.43451, + "grad_norm": 0.8558493852615356, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 43451 + }, + { + "epoch": 0.43452, + "grad_norm": 1.0192228555679321, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 43452 + }, + { + "epoch": 0.43453, + "grad_norm": 1.4058834314346313, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 43453 + }, + { + "epoch": 0.43454, + "grad_norm": 0.7170462012290955, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 43454 + }, + { + "epoch": 0.43455, + "grad_norm": 0.7499242424964905, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 43455 + }, + { + "epoch": 0.43456, + "grad_norm": 0.6678207516670227, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 43456 + }, + { + "epoch": 0.43457, + "grad_norm": 0.7165668606758118, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 43457 + }, + { + "epoch": 0.43458, + "grad_norm": 0.76201993227005, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 43458 + }, + { + "epoch": 0.43459, + "grad_norm": 0.7390522360801697, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 43459 + }, + { + "epoch": 0.4346, + "grad_norm": 0.7493404150009155, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 43460 + }, + { + "epoch": 0.43461, + "grad_norm": 0.8944758772850037, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 43461 + }, + { + "epoch": 0.43462, + "grad_norm": 1.1238030195236206, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 43462 + }, + { + "epoch": 0.43463, + "grad_norm": 1.4158042669296265, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 43463 + }, + { + "epoch": 0.43464, + "grad_norm": 0.6196944117546082, + "learning_rate": 0.003, + "loss": 3.982, + "step": 43464 + }, + { + "epoch": 0.43465, + "grad_norm": 0.847284734249115, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 43465 + }, + { + "epoch": 0.43466, + "grad_norm": 1.0691068172454834, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 43466 + }, + { + "epoch": 0.43467, + "grad_norm": 1.129833698272705, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 43467 + }, + { + "epoch": 0.43468, + "grad_norm": 0.8154141306877136, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 43468 + }, + { + "epoch": 0.43469, + "grad_norm": 0.7691035866737366, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 43469 + }, + { + "epoch": 0.4347, + "grad_norm": 0.8382527232170105, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 43470 + }, + { + "epoch": 0.43471, + "grad_norm": 0.9138081073760986, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 43471 + }, + { + "epoch": 0.43472, + "grad_norm": 1.1299561262130737, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 43472 + }, + { + "epoch": 0.43473, + "grad_norm": 0.8973124027252197, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 43473 + }, + { + "epoch": 0.43474, + "grad_norm": 0.9404584765434265, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 43474 + }, + { + "epoch": 0.43475, + "grad_norm": 0.9391273856163025, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 43475 + }, + { + "epoch": 0.43476, + "grad_norm": 0.8055273294448853, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 43476 + }, + { + "epoch": 0.43477, + "grad_norm": 0.975058376789093, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 43477 + }, + { + "epoch": 0.43478, + "grad_norm": 0.9735938906669617, + "learning_rate": 0.003, + "loss": 3.9349, + "step": 43478 + }, + { + "epoch": 0.43479, + "grad_norm": 1.0829440355300903, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 43479 + }, + { + "epoch": 0.4348, + "grad_norm": 1.0231925249099731, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 43480 + }, + { + "epoch": 0.43481, + "grad_norm": 1.0269163846969604, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 43481 + }, + { + "epoch": 0.43482, + "grad_norm": 0.9052227139472961, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 43482 + }, + { + "epoch": 0.43483, + "grad_norm": 0.9167878031730652, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 43483 + }, + { + "epoch": 0.43484, + "grad_norm": 0.8429620265960693, + "learning_rate": 0.003, + "loss": 3.966, + "step": 43484 + }, + { + "epoch": 0.43485, + "grad_norm": 0.9433333277702332, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 43485 + }, + { + "epoch": 0.43486, + "grad_norm": 1.0235872268676758, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 43486 + }, + { + "epoch": 0.43487, + "grad_norm": 1.169748067855835, + "learning_rate": 0.003, + "loss": 3.986, + "step": 43487 + }, + { + "epoch": 0.43488, + "grad_norm": 0.7608219981193542, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 43488 + }, + { + "epoch": 0.43489, + "grad_norm": 0.7175837159156799, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 43489 + }, + { + "epoch": 0.4349, + "grad_norm": 0.8188678622245789, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 43490 + }, + { + "epoch": 0.43491, + "grad_norm": 0.8980924487113953, + "learning_rate": 0.003, + "loss": 3.953, + "step": 43491 + }, + { + "epoch": 0.43492, + "grad_norm": 1.096123456954956, + "learning_rate": 0.003, + "loss": 3.957, + "step": 43492 + }, + { + "epoch": 0.43493, + "grad_norm": 1.0269941091537476, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 43493 + }, + { + "epoch": 0.43494, + "grad_norm": 0.9586839079856873, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 43494 + }, + { + "epoch": 0.43495, + "grad_norm": 1.0561095476150513, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 43495 + }, + { + "epoch": 0.43496, + "grad_norm": 1.2230600118637085, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 43496 + }, + { + "epoch": 0.43497, + "grad_norm": 0.895345151424408, + "learning_rate": 0.003, + "loss": 3.961, + "step": 43497 + }, + { + "epoch": 0.43498, + "grad_norm": 0.8139778971672058, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 43498 + }, + { + "epoch": 0.43499, + "grad_norm": 0.6771894097328186, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 43499 + }, + { + "epoch": 0.435, + "grad_norm": 0.7092623710632324, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 43500 + }, + { + "epoch": 0.43501, + "grad_norm": 0.7769561409950256, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 43501 + }, + { + "epoch": 0.43502, + "grad_norm": 0.919235110282898, + "learning_rate": 0.003, + "loss": 3.968, + "step": 43502 + }, + { + "epoch": 0.43503, + "grad_norm": 1.0782328844070435, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 43503 + }, + { + "epoch": 0.43504, + "grad_norm": 0.8020234107971191, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 43504 + }, + { + "epoch": 0.43505, + "grad_norm": 0.7429790496826172, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 43505 + }, + { + "epoch": 0.43506, + "grad_norm": 0.8031718730926514, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 43506 + }, + { + "epoch": 0.43507, + "grad_norm": 0.9842937588691711, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 43507 + }, + { + "epoch": 0.43508, + "grad_norm": 1.0327796936035156, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 43508 + }, + { + "epoch": 0.43509, + "grad_norm": 0.9425268173217773, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 43509 + }, + { + "epoch": 0.4351, + "grad_norm": 1.123896837234497, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 43510 + }, + { + "epoch": 0.43511, + "grad_norm": 0.9058920741081238, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 43511 + }, + { + "epoch": 0.43512, + "grad_norm": 0.8425044417381287, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 43512 + }, + { + "epoch": 0.43513, + "grad_norm": 0.8777697682380676, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 43513 + }, + { + "epoch": 0.43514, + "grad_norm": 1.2447125911712646, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 43514 + }, + { + "epoch": 0.43515, + "grad_norm": 0.9856958985328674, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 43515 + }, + { + "epoch": 0.43516, + "grad_norm": 1.015489935874939, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 43516 + }, + { + "epoch": 0.43517, + "grad_norm": 1.0643552541732788, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 43517 + }, + { + "epoch": 0.43518, + "grad_norm": 1.036040186882019, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 43518 + }, + { + "epoch": 0.43519, + "grad_norm": 0.9451100826263428, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 43519 + }, + { + "epoch": 0.4352, + "grad_norm": 0.9925513863563538, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 43520 + }, + { + "epoch": 0.43521, + "grad_norm": 1.1944355964660645, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 43521 + }, + { + "epoch": 0.43522, + "grad_norm": 0.8482833504676819, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 43522 + }, + { + "epoch": 0.43523, + "grad_norm": 0.7014367580413818, + "learning_rate": 0.003, + "loss": 3.999, + "step": 43523 + }, + { + "epoch": 0.43524, + "grad_norm": 0.9092677235603333, + "learning_rate": 0.003, + "loss": 3.964, + "step": 43524 + }, + { + "epoch": 0.43525, + "grad_norm": 0.8851945400238037, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 43525 + }, + { + "epoch": 0.43526, + "grad_norm": 0.7917414307594299, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 43526 + }, + { + "epoch": 0.43527, + "grad_norm": 0.7420490980148315, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 43527 + }, + { + "epoch": 0.43528, + "grad_norm": 0.8755151033401489, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 43528 + }, + { + "epoch": 0.43529, + "grad_norm": 0.8704920411109924, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 43529 + }, + { + "epoch": 0.4353, + "grad_norm": 1.0887759923934937, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 43530 + }, + { + "epoch": 0.43531, + "grad_norm": 1.2369868755340576, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 43531 + }, + { + "epoch": 0.43532, + "grad_norm": 0.7212787866592407, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 43532 + }, + { + "epoch": 0.43533, + "grad_norm": 0.7327218055725098, + "learning_rate": 0.003, + "loss": 3.985, + "step": 43533 + }, + { + "epoch": 0.43534, + "grad_norm": 0.7404852509498596, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 43534 + }, + { + "epoch": 0.43535, + "grad_norm": 0.7814499139785767, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 43535 + }, + { + "epoch": 0.43536, + "grad_norm": 0.8329561352729797, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 43536 + }, + { + "epoch": 0.43537, + "grad_norm": 0.8832117319107056, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 43537 + }, + { + "epoch": 0.43538, + "grad_norm": 0.91550213098526, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 43538 + }, + { + "epoch": 0.43539, + "grad_norm": 0.9661002159118652, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 43539 + }, + { + "epoch": 0.4354, + "grad_norm": 1.1010866165161133, + "learning_rate": 0.003, + "loss": 3.962, + "step": 43540 + }, + { + "epoch": 0.43541, + "grad_norm": 0.8932246565818787, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 43541 + }, + { + "epoch": 0.43542, + "grad_norm": 0.75054931640625, + "learning_rate": 0.003, + "loss": 3.968, + "step": 43542 + }, + { + "epoch": 0.43543, + "grad_norm": 0.769917905330658, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 43543 + }, + { + "epoch": 0.43544, + "grad_norm": 0.7558873891830444, + "learning_rate": 0.003, + "loss": 3.9409, + "step": 43544 + }, + { + "epoch": 0.43545, + "grad_norm": 0.7397926449775696, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 43545 + }, + { + "epoch": 0.43546, + "grad_norm": 0.678499162197113, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 43546 + }, + { + "epoch": 0.43547, + "grad_norm": 0.7044616341590881, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 43547 + }, + { + "epoch": 0.43548, + "grad_norm": 0.6665639281272888, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 43548 + }, + { + "epoch": 0.43549, + "grad_norm": 0.7124735116958618, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 43549 + }, + { + "epoch": 0.4355, + "grad_norm": 0.8420284986495972, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 43550 + }, + { + "epoch": 0.43551, + "grad_norm": 1.1895496845245361, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 43551 + }, + { + "epoch": 0.43552, + "grad_norm": 1.2510111331939697, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 43552 + }, + { + "epoch": 0.43553, + "grad_norm": 0.8310280442237854, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 43553 + }, + { + "epoch": 0.43554, + "grad_norm": 0.8116994500160217, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 43554 + }, + { + "epoch": 0.43555, + "grad_norm": 0.8981159925460815, + "learning_rate": 0.003, + "loss": 3.985, + "step": 43555 + }, + { + "epoch": 0.43556, + "grad_norm": 0.9867517948150635, + "learning_rate": 0.003, + "loss": 3.9509, + "step": 43556 + }, + { + "epoch": 0.43557, + "grad_norm": 1.0880591869354248, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 43557 + }, + { + "epoch": 0.43558, + "grad_norm": 0.96437668800354, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 43558 + }, + { + "epoch": 0.43559, + "grad_norm": 0.9189323782920837, + "learning_rate": 0.003, + "loss": 4.002, + "step": 43559 + }, + { + "epoch": 0.4356, + "grad_norm": 0.9286249876022339, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 43560 + }, + { + "epoch": 0.43561, + "grad_norm": 0.9087772369384766, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 43561 + }, + { + "epoch": 0.43562, + "grad_norm": 0.9409539699554443, + "learning_rate": 0.003, + "loss": 3.9338, + "step": 43562 + }, + { + "epoch": 0.43563, + "grad_norm": 1.0100553035736084, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 43563 + }, + { + "epoch": 0.43564, + "grad_norm": 0.9577314257621765, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 43564 + }, + { + "epoch": 0.43565, + "grad_norm": 1.0612252950668335, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 43565 + }, + { + "epoch": 0.43566, + "grad_norm": 0.9572547674179077, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 43566 + }, + { + "epoch": 0.43567, + "grad_norm": 0.9701207876205444, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 43567 + }, + { + "epoch": 0.43568, + "grad_norm": 1.1310855150222778, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 43568 + }, + { + "epoch": 0.43569, + "grad_norm": 1.0611244440078735, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 43569 + }, + { + "epoch": 0.4357, + "grad_norm": 0.9167026281356812, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 43570 + }, + { + "epoch": 0.43571, + "grad_norm": 0.9181137084960938, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 43571 + }, + { + "epoch": 0.43572, + "grad_norm": 0.9520986080169678, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 43572 + }, + { + "epoch": 0.43573, + "grad_norm": 0.8424277305603027, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 43573 + }, + { + "epoch": 0.43574, + "grad_norm": 1.0409603118896484, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 43574 + }, + { + "epoch": 0.43575, + "grad_norm": 1.206460952758789, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 43575 + }, + { + "epoch": 0.43576, + "grad_norm": 0.6865056753158569, + "learning_rate": 0.003, + "loss": 4.002, + "step": 43576 + }, + { + "epoch": 0.43577, + "grad_norm": 0.6526920795440674, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 43577 + }, + { + "epoch": 0.43578, + "grad_norm": 0.7465384602546692, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 43578 + }, + { + "epoch": 0.43579, + "grad_norm": 0.8255109190940857, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 43579 + }, + { + "epoch": 0.4358, + "grad_norm": 1.05483877658844, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 43580 + }, + { + "epoch": 0.43581, + "grad_norm": 1.0970609188079834, + "learning_rate": 0.003, + "loss": 3.971, + "step": 43581 + }, + { + "epoch": 0.43582, + "grad_norm": 0.7773786783218384, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 43582 + }, + { + "epoch": 0.43583, + "grad_norm": 0.6809654235839844, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 43583 + }, + { + "epoch": 0.43584, + "grad_norm": 0.5985221862792969, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 43584 + }, + { + "epoch": 0.43585, + "grad_norm": 0.6232450008392334, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 43585 + }, + { + "epoch": 0.43586, + "grad_norm": 0.6105892062187195, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 43586 + }, + { + "epoch": 0.43587, + "grad_norm": 0.6544758081436157, + "learning_rate": 0.003, + "loss": 3.994, + "step": 43587 + }, + { + "epoch": 0.43588, + "grad_norm": 0.774728000164032, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 43588 + }, + { + "epoch": 0.43589, + "grad_norm": 0.9457888007164001, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 43589 + }, + { + "epoch": 0.4359, + "grad_norm": 1.1818336248397827, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 43590 + }, + { + "epoch": 0.43591, + "grad_norm": 0.814972460269928, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 43591 + }, + { + "epoch": 0.43592, + "grad_norm": 0.6832465529441833, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 43592 + }, + { + "epoch": 0.43593, + "grad_norm": 0.6608511209487915, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 43593 + }, + { + "epoch": 0.43594, + "grad_norm": 0.7893688082695007, + "learning_rate": 0.003, + "loss": 3.967, + "step": 43594 + }, + { + "epoch": 0.43595, + "grad_norm": 0.9752978682518005, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 43595 + }, + { + "epoch": 0.43596, + "grad_norm": 1.2371100187301636, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 43596 + }, + { + "epoch": 0.43597, + "grad_norm": 0.9059889316558838, + "learning_rate": 0.003, + "loss": 3.981, + "step": 43597 + }, + { + "epoch": 0.43598, + "grad_norm": 0.9844261407852173, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 43598 + }, + { + "epoch": 0.43599, + "grad_norm": 1.047602653503418, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 43599 + }, + { + "epoch": 0.436, + "grad_norm": 0.9209626913070679, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 43600 + }, + { + "epoch": 0.43601, + "grad_norm": 0.7785860896110535, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 43601 + }, + { + "epoch": 0.43602, + "grad_norm": 0.8499147891998291, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 43602 + }, + { + "epoch": 0.43603, + "grad_norm": 0.9878125786781311, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 43603 + }, + { + "epoch": 0.43604, + "grad_norm": 0.9353342652320862, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 43604 + }, + { + "epoch": 0.43605, + "grad_norm": 0.887472927570343, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 43605 + }, + { + "epoch": 0.43606, + "grad_norm": 1.0312345027923584, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 43606 + }, + { + "epoch": 0.43607, + "grad_norm": 0.9915590882301331, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 43607 + }, + { + "epoch": 0.43608, + "grad_norm": 1.0996589660644531, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 43608 + }, + { + "epoch": 0.43609, + "grad_norm": 0.875842809677124, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 43609 + }, + { + "epoch": 0.4361, + "grad_norm": 0.8313008546829224, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 43610 + }, + { + "epoch": 0.43611, + "grad_norm": 0.9179646372795105, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 43611 + }, + { + "epoch": 0.43612, + "grad_norm": 0.8955705165863037, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 43612 + }, + { + "epoch": 0.43613, + "grad_norm": 1.036848545074463, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 43613 + }, + { + "epoch": 0.43614, + "grad_norm": 1.2940220832824707, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 43614 + }, + { + "epoch": 0.43615, + "grad_norm": 0.9264659881591797, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 43615 + }, + { + "epoch": 0.43616, + "grad_norm": 1.049825668334961, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 43616 + }, + { + "epoch": 0.43617, + "grad_norm": 1.083762288093567, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 43617 + }, + { + "epoch": 0.43618, + "grad_norm": 0.8488559722900391, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 43618 + }, + { + "epoch": 0.43619, + "grad_norm": 0.9693797826766968, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 43619 + }, + { + "epoch": 0.4362, + "grad_norm": 1.2023260593414307, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 43620 + }, + { + "epoch": 0.43621, + "grad_norm": 0.8525391221046448, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 43621 + }, + { + "epoch": 0.43622, + "grad_norm": 0.974899172782898, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 43622 + }, + { + "epoch": 0.43623, + "grad_norm": 1.0255224704742432, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 43623 + }, + { + "epoch": 0.43624, + "grad_norm": 0.9932893514633179, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 43624 + }, + { + "epoch": 0.43625, + "grad_norm": 1.1371527910232544, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 43625 + }, + { + "epoch": 0.43626, + "grad_norm": 0.9321824908256531, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 43626 + }, + { + "epoch": 0.43627, + "grad_norm": 0.8198844790458679, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 43627 + }, + { + "epoch": 0.43628, + "grad_norm": 0.8486245274543762, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 43628 + }, + { + "epoch": 0.43629, + "grad_norm": 0.9531255960464478, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 43629 + }, + { + "epoch": 0.4363, + "grad_norm": 1.0875784158706665, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 43630 + }, + { + "epoch": 0.43631, + "grad_norm": 0.9811896085739136, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 43631 + }, + { + "epoch": 0.43632, + "grad_norm": 0.9008752107620239, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 43632 + }, + { + "epoch": 0.43633, + "grad_norm": 0.7646161913871765, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 43633 + }, + { + "epoch": 0.43634, + "grad_norm": 0.7149195671081543, + "learning_rate": 0.003, + "loss": 3.9442, + "step": 43634 + }, + { + "epoch": 0.43635, + "grad_norm": 0.674821674823761, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 43635 + }, + { + "epoch": 0.43636, + "grad_norm": 0.7219679355621338, + "learning_rate": 0.003, + "loss": 3.98, + "step": 43636 + }, + { + "epoch": 0.43637, + "grad_norm": 0.7036224007606506, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 43637 + }, + { + "epoch": 0.43638, + "grad_norm": 0.6848741769790649, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 43638 + }, + { + "epoch": 0.43639, + "grad_norm": 0.742648184299469, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 43639 + }, + { + "epoch": 0.4364, + "grad_norm": 0.8267311453819275, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 43640 + }, + { + "epoch": 0.43641, + "grad_norm": 0.9978892803192139, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 43641 + }, + { + "epoch": 0.43642, + "grad_norm": 1.2340093851089478, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 43642 + }, + { + "epoch": 0.43643, + "grad_norm": 0.7496351599693298, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 43643 + }, + { + "epoch": 0.43644, + "grad_norm": 0.702610969543457, + "learning_rate": 0.003, + "loss": 3.968, + "step": 43644 + }, + { + "epoch": 0.43645, + "grad_norm": 0.7815605401992798, + "learning_rate": 0.003, + "loss": 3.94, + "step": 43645 + }, + { + "epoch": 0.43646, + "grad_norm": 0.8689611554145813, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 43646 + }, + { + "epoch": 0.43647, + "grad_norm": 1.0410096645355225, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 43647 + }, + { + "epoch": 0.43648, + "grad_norm": 1.2308846712112427, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 43648 + }, + { + "epoch": 0.43649, + "grad_norm": 1.0512937307357788, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 43649 + }, + { + "epoch": 0.4365, + "grad_norm": 1.0649811029434204, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 43650 + }, + { + "epoch": 0.43651, + "grad_norm": 0.9705869555473328, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 43651 + }, + { + "epoch": 0.43652, + "grad_norm": 0.8493708968162537, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 43652 + }, + { + "epoch": 0.43653, + "grad_norm": 0.782290518283844, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 43653 + }, + { + "epoch": 0.43654, + "grad_norm": 0.7883424162864685, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 43654 + }, + { + "epoch": 0.43655, + "grad_norm": 0.7345767021179199, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 43655 + }, + { + "epoch": 0.43656, + "grad_norm": 0.7287741303443909, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 43656 + }, + { + "epoch": 0.43657, + "grad_norm": 0.7449452877044678, + "learning_rate": 0.003, + "loss": 3.9337, + "step": 43657 + }, + { + "epoch": 0.43658, + "grad_norm": 0.7208945751190186, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 43658 + }, + { + "epoch": 0.43659, + "grad_norm": 0.6441061496734619, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 43659 + }, + { + "epoch": 0.4366, + "grad_norm": 0.8389315009117126, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 43660 + }, + { + "epoch": 0.43661, + "grad_norm": 1.1005491018295288, + "learning_rate": 0.003, + "loss": 3.983, + "step": 43661 + }, + { + "epoch": 0.43662, + "grad_norm": 1.1988263130187988, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 43662 + }, + { + "epoch": 0.43663, + "grad_norm": 0.9571794271469116, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 43663 + }, + { + "epoch": 0.43664, + "grad_norm": 1.0371191501617432, + "learning_rate": 0.003, + "loss": 3.982, + "step": 43664 + }, + { + "epoch": 0.43665, + "grad_norm": 1.1400812864303589, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 43665 + }, + { + "epoch": 0.43666, + "grad_norm": 0.9347547888755798, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 43666 + }, + { + "epoch": 0.43667, + "grad_norm": 1.0791339874267578, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 43667 + }, + { + "epoch": 0.43668, + "grad_norm": 0.9409129023551941, + "learning_rate": 0.003, + "loss": 3.963, + "step": 43668 + }, + { + "epoch": 0.43669, + "grad_norm": 0.8973995447158813, + "learning_rate": 0.003, + "loss": 3.995, + "step": 43669 + }, + { + "epoch": 0.4367, + "grad_norm": 0.9992116689682007, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 43670 + }, + { + "epoch": 0.43671, + "grad_norm": 1.1592429876327515, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 43671 + }, + { + "epoch": 0.43672, + "grad_norm": 0.9885715246200562, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 43672 + }, + { + "epoch": 0.43673, + "grad_norm": 0.9878379106521606, + "learning_rate": 0.003, + "loss": 3.971, + "step": 43673 + }, + { + "epoch": 0.43674, + "grad_norm": 1.0755702257156372, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 43674 + }, + { + "epoch": 0.43675, + "grad_norm": 0.9059950709342957, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 43675 + }, + { + "epoch": 0.43676, + "grad_norm": 0.872459352016449, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 43676 + }, + { + "epoch": 0.43677, + "grad_norm": 0.947107195854187, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 43677 + }, + { + "epoch": 0.43678, + "grad_norm": 0.9355326294898987, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 43678 + }, + { + "epoch": 0.43679, + "grad_norm": 0.8730417490005493, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 43679 + }, + { + "epoch": 0.4368, + "grad_norm": 1.0054808855056763, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 43680 + }, + { + "epoch": 0.43681, + "grad_norm": 1.0677565336227417, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 43681 + }, + { + "epoch": 0.43682, + "grad_norm": 1.0154727697372437, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 43682 + }, + { + "epoch": 0.43683, + "grad_norm": 0.9208515882492065, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 43683 + }, + { + "epoch": 0.43684, + "grad_norm": 1.0409356355667114, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 43684 + }, + { + "epoch": 0.43685, + "grad_norm": 1.19400155544281, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 43685 + }, + { + "epoch": 0.43686, + "grad_norm": 0.916235625743866, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 43686 + }, + { + "epoch": 0.43687, + "grad_norm": 1.0562410354614258, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 43687 + }, + { + "epoch": 0.43688, + "grad_norm": 0.9372755885124207, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 43688 + }, + { + "epoch": 0.43689, + "grad_norm": 0.9001801013946533, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 43689 + }, + { + "epoch": 0.4369, + "grad_norm": 0.8642283082008362, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 43690 + }, + { + "epoch": 0.43691, + "grad_norm": 0.8274567723274231, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 43691 + }, + { + "epoch": 0.43692, + "grad_norm": 1.0065159797668457, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 43692 + }, + { + "epoch": 0.43693, + "grad_norm": 1.2604143619537354, + "learning_rate": 0.003, + "loss": 4.013, + "step": 43693 + }, + { + "epoch": 0.43694, + "grad_norm": 0.7884988784790039, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 43694 + }, + { + "epoch": 0.43695, + "grad_norm": 0.9949652552604675, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 43695 + }, + { + "epoch": 0.43696, + "grad_norm": 1.1338305473327637, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 43696 + }, + { + "epoch": 0.43697, + "grad_norm": 0.7876393795013428, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 43697 + }, + { + "epoch": 0.43698, + "grad_norm": 0.6667061448097229, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 43698 + }, + { + "epoch": 0.43699, + "grad_norm": 0.7170143127441406, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 43699 + }, + { + "epoch": 0.437, + "grad_norm": 0.712083101272583, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 43700 + }, + { + "epoch": 0.43701, + "grad_norm": 0.629951536655426, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 43701 + }, + { + "epoch": 0.43702, + "grad_norm": 0.584871768951416, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 43702 + }, + { + "epoch": 0.43703, + "grad_norm": 0.6408160328865051, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 43703 + }, + { + "epoch": 0.43704, + "grad_norm": 0.7220133543014526, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 43704 + }, + { + "epoch": 0.43705, + "grad_norm": 0.981572151184082, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 43705 + }, + { + "epoch": 0.43706, + "grad_norm": 1.3442035913467407, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 43706 + }, + { + "epoch": 0.43707, + "grad_norm": 0.6638705134391785, + "learning_rate": 0.003, + "loss": 3.9469, + "step": 43707 + }, + { + "epoch": 0.43708, + "grad_norm": 0.8234612345695496, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 43708 + }, + { + "epoch": 0.43709, + "grad_norm": 0.9034761190414429, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 43709 + }, + { + "epoch": 0.4371, + "grad_norm": 0.9362307786941528, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 43710 + }, + { + "epoch": 0.43711, + "grad_norm": 0.9088234901428223, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 43711 + }, + { + "epoch": 0.43712, + "grad_norm": 1.0040932893753052, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 43712 + }, + { + "epoch": 0.43713, + "grad_norm": 1.189695954322815, + "learning_rate": 0.003, + "loss": 3.984, + "step": 43713 + }, + { + "epoch": 0.43714, + "grad_norm": 1.190625548362732, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 43714 + }, + { + "epoch": 0.43715, + "grad_norm": 1.0611317157745361, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 43715 + }, + { + "epoch": 0.43716, + "grad_norm": 0.986787736415863, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 43716 + }, + { + "epoch": 0.43717, + "grad_norm": 1.0952762365341187, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 43717 + }, + { + "epoch": 0.43718, + "grad_norm": 0.711463212966919, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 43718 + }, + { + "epoch": 0.43719, + "grad_norm": 0.7378372550010681, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 43719 + }, + { + "epoch": 0.4372, + "grad_norm": 0.7515296339988708, + "learning_rate": 0.003, + "loss": 3.975, + "step": 43720 + }, + { + "epoch": 0.43721, + "grad_norm": 0.7376033663749695, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 43721 + }, + { + "epoch": 0.43722, + "grad_norm": 0.9304897785186768, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 43722 + }, + { + "epoch": 0.43723, + "grad_norm": 0.9940231442451477, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 43723 + }, + { + "epoch": 0.43724, + "grad_norm": 1.105017066001892, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 43724 + }, + { + "epoch": 0.43725, + "grad_norm": 0.9709047079086304, + "learning_rate": 0.003, + "loss": 3.987, + "step": 43725 + }, + { + "epoch": 0.43726, + "grad_norm": 1.0899534225463867, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 43726 + }, + { + "epoch": 0.43727, + "grad_norm": 0.9245927333831787, + "learning_rate": 0.003, + "loss": 3.968, + "step": 43727 + }, + { + "epoch": 0.43728, + "grad_norm": 0.9005722403526306, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 43728 + }, + { + "epoch": 0.43729, + "grad_norm": 0.8778570294380188, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 43729 + }, + { + "epoch": 0.4373, + "grad_norm": 0.9347235560417175, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 43730 + }, + { + "epoch": 0.43731, + "grad_norm": 0.8935981392860413, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 43731 + }, + { + "epoch": 0.43732, + "grad_norm": 0.9749335050582886, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 43732 + }, + { + "epoch": 0.43733, + "grad_norm": 0.9926341772079468, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 43733 + }, + { + "epoch": 0.43734, + "grad_norm": 1.0316630601882935, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 43734 + }, + { + "epoch": 0.43735, + "grad_norm": 1.1687713861465454, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 43735 + }, + { + "epoch": 0.43736, + "grad_norm": 1.1353765726089478, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 43736 + }, + { + "epoch": 0.43737, + "grad_norm": 1.0632370710372925, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 43737 + }, + { + "epoch": 0.43738, + "grad_norm": 0.7834498882293701, + "learning_rate": 0.003, + "loss": 3.957, + "step": 43738 + }, + { + "epoch": 0.43739, + "grad_norm": 0.8349553942680359, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 43739 + }, + { + "epoch": 0.4374, + "grad_norm": 0.9406440258026123, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 43740 + }, + { + "epoch": 0.43741, + "grad_norm": 0.8197376728057861, + "learning_rate": 0.003, + "loss": 3.9395, + "step": 43741 + }, + { + "epoch": 0.43742, + "grad_norm": 0.8115569949150085, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 43742 + }, + { + "epoch": 0.43743, + "grad_norm": 0.7588692307472229, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 43743 + }, + { + "epoch": 0.43744, + "grad_norm": 0.7255585789680481, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 43744 + }, + { + "epoch": 0.43745, + "grad_norm": 0.6874576807022095, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 43745 + }, + { + "epoch": 0.43746, + "grad_norm": 0.685957670211792, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 43746 + }, + { + "epoch": 0.43747, + "grad_norm": 0.7490964531898499, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 43747 + }, + { + "epoch": 0.43748, + "grad_norm": 0.877931535243988, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 43748 + }, + { + "epoch": 0.43749, + "grad_norm": 1.3580825328826904, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 43749 + }, + { + "epoch": 0.4375, + "grad_norm": 1.057317852973938, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 43750 + }, + { + "epoch": 0.43751, + "grad_norm": 0.8883966207504272, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 43751 + }, + { + "epoch": 0.43752, + "grad_norm": 0.8146699666976929, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 43752 + }, + { + "epoch": 0.43753, + "grad_norm": 0.919983983039856, + "learning_rate": 0.003, + "loss": 3.969, + "step": 43753 + }, + { + "epoch": 0.43754, + "grad_norm": 1.0019479990005493, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 43754 + }, + { + "epoch": 0.43755, + "grad_norm": 1.0217090845108032, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 43755 + }, + { + "epoch": 0.43756, + "grad_norm": 1.2483396530151367, + "learning_rate": 0.003, + "loss": 3.975, + "step": 43756 + }, + { + "epoch": 0.43757, + "grad_norm": 0.9081950783729553, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 43757 + }, + { + "epoch": 0.43758, + "grad_norm": 0.8132002353668213, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 43758 + }, + { + "epoch": 0.43759, + "grad_norm": 0.8529813885688782, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 43759 + }, + { + "epoch": 0.4376, + "grad_norm": 0.9862698912620544, + "learning_rate": 0.003, + "loss": 3.977, + "step": 43760 + }, + { + "epoch": 0.43761, + "grad_norm": 1.0967769622802734, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 43761 + }, + { + "epoch": 0.43762, + "grad_norm": 0.7708786725997925, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 43762 + }, + { + "epoch": 0.43763, + "grad_norm": 0.6659170389175415, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 43763 + }, + { + "epoch": 0.43764, + "grad_norm": 0.6221954226493835, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 43764 + }, + { + "epoch": 0.43765, + "grad_norm": 0.6447919607162476, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 43765 + }, + { + "epoch": 0.43766, + "grad_norm": 0.7060909271240234, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 43766 + }, + { + "epoch": 0.43767, + "grad_norm": 0.8766050338745117, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 43767 + }, + { + "epoch": 0.43768, + "grad_norm": 1.202480435371399, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 43768 + }, + { + "epoch": 0.43769, + "grad_norm": 0.8707833886146545, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 43769 + }, + { + "epoch": 0.4377, + "grad_norm": 0.7913832068443298, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 43770 + }, + { + "epoch": 0.43771, + "grad_norm": 0.7858709096908569, + "learning_rate": 0.003, + "loss": 3.9324, + "step": 43771 + }, + { + "epoch": 0.43772, + "grad_norm": 0.8725440502166748, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 43772 + }, + { + "epoch": 0.43773, + "grad_norm": 0.9355215430259705, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 43773 + }, + { + "epoch": 0.43774, + "grad_norm": 0.9317418932914734, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 43774 + }, + { + "epoch": 0.43775, + "grad_norm": 0.8668287396430969, + "learning_rate": 0.003, + "loss": 3.9453, + "step": 43775 + }, + { + "epoch": 0.43776, + "grad_norm": 0.8360666632652283, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 43776 + }, + { + "epoch": 0.43777, + "grad_norm": 0.9455897212028503, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 43777 + }, + { + "epoch": 0.43778, + "grad_norm": 1.0530463457107544, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 43778 + }, + { + "epoch": 0.43779, + "grad_norm": 0.9249354600906372, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 43779 + }, + { + "epoch": 0.4378, + "grad_norm": 0.7276681661605835, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 43780 + }, + { + "epoch": 0.43781, + "grad_norm": 0.6267253160476685, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 43781 + }, + { + "epoch": 0.43782, + "grad_norm": 0.6599195599555969, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 43782 + }, + { + "epoch": 0.43783, + "grad_norm": 0.6600732803344727, + "learning_rate": 0.003, + "loss": 3.9405, + "step": 43783 + }, + { + "epoch": 0.43784, + "grad_norm": 0.6864059567451477, + "learning_rate": 0.003, + "loss": 3.975, + "step": 43784 + }, + { + "epoch": 0.43785, + "grad_norm": 0.8908969759941101, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 43785 + }, + { + "epoch": 0.43786, + "grad_norm": 1.2826062440872192, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 43786 + }, + { + "epoch": 0.43787, + "grad_norm": 0.8429520726203918, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 43787 + }, + { + "epoch": 0.43788, + "grad_norm": 0.7211974859237671, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 43788 + }, + { + "epoch": 0.43789, + "grad_norm": 0.833355724811554, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 43789 + }, + { + "epoch": 0.4379, + "grad_norm": 1.0305733680725098, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 43790 + }, + { + "epoch": 0.43791, + "grad_norm": 1.0733524560928345, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 43791 + }, + { + "epoch": 0.43792, + "grad_norm": 0.9050900340080261, + "learning_rate": 0.003, + "loss": 3.984, + "step": 43792 + }, + { + "epoch": 0.43793, + "grad_norm": 0.9738329648971558, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 43793 + }, + { + "epoch": 0.43794, + "grad_norm": 1.1710737943649292, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 43794 + }, + { + "epoch": 0.43795, + "grad_norm": 1.0718960762023926, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 43795 + }, + { + "epoch": 0.43796, + "grad_norm": 0.9007288217544556, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 43796 + }, + { + "epoch": 0.43797, + "grad_norm": 0.7601912021636963, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 43797 + }, + { + "epoch": 0.43798, + "grad_norm": 0.815015435218811, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 43798 + }, + { + "epoch": 0.43799, + "grad_norm": 0.9829233884811401, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 43799 + }, + { + "epoch": 0.438, + "grad_norm": 1.0131752490997314, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 43800 + }, + { + "epoch": 0.43801, + "grad_norm": 1.0762615203857422, + "learning_rate": 0.003, + "loss": 3.998, + "step": 43801 + }, + { + "epoch": 0.43802, + "grad_norm": 0.9007741212844849, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 43802 + }, + { + "epoch": 0.43803, + "grad_norm": 0.9636232852935791, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 43803 + }, + { + "epoch": 0.43804, + "grad_norm": 0.8701421022415161, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 43804 + }, + { + "epoch": 0.43805, + "grad_norm": 0.8508917689323425, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 43805 + }, + { + "epoch": 0.43806, + "grad_norm": 0.9987356066703796, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 43806 + }, + { + "epoch": 0.43807, + "grad_norm": 1.0941914319992065, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 43807 + }, + { + "epoch": 0.43808, + "grad_norm": 0.9500325322151184, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 43808 + }, + { + "epoch": 0.43809, + "grad_norm": 0.939785897731781, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 43809 + }, + { + "epoch": 0.4381, + "grad_norm": 1.3064115047454834, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 43810 + }, + { + "epoch": 0.43811, + "grad_norm": 1.1191155910491943, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 43811 + }, + { + "epoch": 0.43812, + "grad_norm": 1.0446248054504395, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 43812 + }, + { + "epoch": 0.43813, + "grad_norm": 0.9379903674125671, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 43813 + }, + { + "epoch": 0.43814, + "grad_norm": 0.9009407758712769, + "learning_rate": 0.003, + "loss": 3.9351, + "step": 43814 + }, + { + "epoch": 0.43815, + "grad_norm": 0.8297748565673828, + "learning_rate": 0.003, + "loss": 3.994, + "step": 43815 + }, + { + "epoch": 0.43816, + "grad_norm": 0.8739208579063416, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 43816 + }, + { + "epoch": 0.43817, + "grad_norm": 0.9992815256118774, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 43817 + }, + { + "epoch": 0.43818, + "grad_norm": 1.0437798500061035, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 43818 + }, + { + "epoch": 0.43819, + "grad_norm": 0.8663788437843323, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 43819 + }, + { + "epoch": 0.4382, + "grad_norm": 0.8471039533615112, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 43820 + }, + { + "epoch": 0.43821, + "grad_norm": 1.0863088369369507, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 43821 + }, + { + "epoch": 0.43822, + "grad_norm": 1.046347975730896, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 43822 + }, + { + "epoch": 0.43823, + "grad_norm": 0.8217136859893799, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 43823 + }, + { + "epoch": 0.43824, + "grad_norm": 0.7901987433433533, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 43824 + }, + { + "epoch": 0.43825, + "grad_norm": 0.8162271976470947, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 43825 + }, + { + "epoch": 0.43826, + "grad_norm": 0.8499007225036621, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 43826 + }, + { + "epoch": 0.43827, + "grad_norm": 0.8921765685081482, + "learning_rate": 0.003, + "loss": 3.974, + "step": 43827 + }, + { + "epoch": 0.43828, + "grad_norm": 1.1382064819335938, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 43828 + }, + { + "epoch": 0.43829, + "grad_norm": 1.093032956123352, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 43829 + }, + { + "epoch": 0.4383, + "grad_norm": 0.8524508476257324, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 43830 + }, + { + "epoch": 0.43831, + "grad_norm": 0.7521675229072571, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 43831 + }, + { + "epoch": 0.43832, + "grad_norm": 0.6631840467453003, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 43832 + }, + { + "epoch": 0.43833, + "grad_norm": 0.6999641060829163, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 43833 + }, + { + "epoch": 0.43834, + "grad_norm": 0.7456563711166382, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 43834 + }, + { + "epoch": 0.43835, + "grad_norm": 0.7278878092765808, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 43835 + }, + { + "epoch": 0.43836, + "grad_norm": 0.6575174331665039, + "learning_rate": 0.003, + "loss": 3.9451, + "step": 43836 + }, + { + "epoch": 0.43837, + "grad_norm": 0.7991610765457153, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 43837 + }, + { + "epoch": 0.43838, + "grad_norm": 1.08770751953125, + "learning_rate": 0.003, + "loss": 3.9445, + "step": 43838 + }, + { + "epoch": 0.43839, + "grad_norm": 1.0958013534545898, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 43839 + }, + { + "epoch": 0.4384, + "grad_norm": 0.9163428544998169, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 43840 + }, + { + "epoch": 0.43841, + "grad_norm": 0.8585946559906006, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 43841 + }, + { + "epoch": 0.43842, + "grad_norm": 0.8036940693855286, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 43842 + }, + { + "epoch": 0.43843, + "grad_norm": 0.7894436120986938, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 43843 + }, + { + "epoch": 0.43844, + "grad_norm": 1.0152026414871216, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 43844 + }, + { + "epoch": 0.43845, + "grad_norm": 1.1481125354766846, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 43845 + }, + { + "epoch": 0.43846, + "grad_norm": 0.701307475566864, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 43846 + }, + { + "epoch": 0.43847, + "grad_norm": 0.739874005317688, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 43847 + }, + { + "epoch": 0.43848, + "grad_norm": 0.8406406044960022, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 43848 + }, + { + "epoch": 0.43849, + "grad_norm": 0.9256648421287537, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 43849 + }, + { + "epoch": 0.4385, + "grad_norm": 0.8816402554512024, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 43850 + }, + { + "epoch": 0.43851, + "grad_norm": 0.9405737519264221, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 43851 + }, + { + "epoch": 0.43852, + "grad_norm": 1.0126049518585205, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 43852 + }, + { + "epoch": 0.43853, + "grad_norm": 1.268293023109436, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 43853 + }, + { + "epoch": 0.43854, + "grad_norm": 1.0085912942886353, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 43854 + }, + { + "epoch": 0.43855, + "grad_norm": 0.999880850315094, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 43855 + }, + { + "epoch": 0.43856, + "grad_norm": 0.8976179361343384, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 43856 + }, + { + "epoch": 0.43857, + "grad_norm": 0.8176790475845337, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 43857 + }, + { + "epoch": 0.43858, + "grad_norm": 0.774459719657898, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 43858 + }, + { + "epoch": 0.43859, + "grad_norm": 0.864695131778717, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 43859 + }, + { + "epoch": 0.4386, + "grad_norm": 1.0084553956985474, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 43860 + }, + { + "epoch": 0.43861, + "grad_norm": 1.140192985534668, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 43861 + }, + { + "epoch": 0.43862, + "grad_norm": 0.8311940431594849, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 43862 + }, + { + "epoch": 0.43863, + "grad_norm": 0.8315536975860596, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 43863 + }, + { + "epoch": 0.43864, + "grad_norm": 0.8687194585800171, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 43864 + }, + { + "epoch": 0.43865, + "grad_norm": 0.9480254650115967, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 43865 + }, + { + "epoch": 0.43866, + "grad_norm": 1.2653461694717407, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 43866 + }, + { + "epoch": 0.43867, + "grad_norm": 0.7255833148956299, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 43867 + }, + { + "epoch": 0.43868, + "grad_norm": 0.631828784942627, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 43868 + }, + { + "epoch": 0.43869, + "grad_norm": 0.6446248888969421, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 43869 + }, + { + "epoch": 0.4387, + "grad_norm": 0.7134871482849121, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 43870 + }, + { + "epoch": 0.43871, + "grad_norm": 0.8111317157745361, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 43871 + }, + { + "epoch": 0.43872, + "grad_norm": 0.975958526134491, + "learning_rate": 0.003, + "loss": 3.9337, + "step": 43872 + }, + { + "epoch": 0.43873, + "grad_norm": 1.260738730430603, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 43873 + }, + { + "epoch": 0.43874, + "grad_norm": 0.7942665815353394, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 43874 + }, + { + "epoch": 0.43875, + "grad_norm": 0.817103385925293, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 43875 + }, + { + "epoch": 0.43876, + "grad_norm": 0.9712016582489014, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 43876 + }, + { + "epoch": 0.43877, + "grad_norm": 1.1320942640304565, + "learning_rate": 0.003, + "loss": 4.025, + "step": 43877 + }, + { + "epoch": 0.43878, + "grad_norm": 0.8139370083808899, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 43878 + }, + { + "epoch": 0.43879, + "grad_norm": 0.9686106443405151, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 43879 + }, + { + "epoch": 0.4388, + "grad_norm": 1.0161561965942383, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 43880 + }, + { + "epoch": 0.43881, + "grad_norm": 1.0541229248046875, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 43881 + }, + { + "epoch": 0.43882, + "grad_norm": 1.0843995809555054, + "learning_rate": 0.003, + "loss": 3.9375, + "step": 43882 + }, + { + "epoch": 0.43883, + "grad_norm": 0.9186955690383911, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 43883 + }, + { + "epoch": 0.43884, + "grad_norm": 0.8728224635124207, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 43884 + }, + { + "epoch": 0.43885, + "grad_norm": 1.0263628959655762, + "learning_rate": 0.003, + "loss": 3.991, + "step": 43885 + }, + { + "epoch": 0.43886, + "grad_norm": 1.1757419109344482, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 43886 + }, + { + "epoch": 0.43887, + "grad_norm": 0.8890379667282104, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 43887 + }, + { + "epoch": 0.43888, + "grad_norm": 0.9008668661117554, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 43888 + }, + { + "epoch": 0.43889, + "grad_norm": 0.8412546515464783, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 43889 + }, + { + "epoch": 0.4389, + "grad_norm": 0.9471219778060913, + "learning_rate": 0.003, + "loss": 3.97, + "step": 43890 + }, + { + "epoch": 0.43891, + "grad_norm": 0.9222633242607117, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 43891 + }, + { + "epoch": 0.43892, + "grad_norm": 1.018784761428833, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 43892 + }, + { + "epoch": 0.43893, + "grad_norm": 1.0302941799163818, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 43893 + }, + { + "epoch": 0.43894, + "grad_norm": 0.9316588640213013, + "learning_rate": 0.003, + "loss": 3.987, + "step": 43894 + }, + { + "epoch": 0.43895, + "grad_norm": 0.9980043172836304, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 43895 + }, + { + "epoch": 0.43896, + "grad_norm": 1.2499504089355469, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 43896 + }, + { + "epoch": 0.43897, + "grad_norm": 0.607394814491272, + "learning_rate": 0.003, + "loss": 3.9228, + "step": 43897 + }, + { + "epoch": 0.43898, + "grad_norm": 0.7786693572998047, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 43898 + }, + { + "epoch": 0.43899, + "grad_norm": 0.7526095509529114, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 43899 + }, + { + "epoch": 0.439, + "grad_norm": 0.7841867804527283, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 43900 + }, + { + "epoch": 0.43901, + "grad_norm": 0.8696812987327576, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 43901 + }, + { + "epoch": 0.43902, + "grad_norm": 0.7887483239173889, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 43902 + }, + { + "epoch": 0.43903, + "grad_norm": 0.7455448508262634, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 43903 + }, + { + "epoch": 0.43904, + "grad_norm": 0.7480352520942688, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 43904 + }, + { + "epoch": 0.43905, + "grad_norm": 0.7342971563339233, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 43905 + }, + { + "epoch": 0.43906, + "grad_norm": 0.7835001349449158, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 43906 + }, + { + "epoch": 0.43907, + "grad_norm": 1.0072656869888306, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 43907 + }, + { + "epoch": 0.43908, + "grad_norm": 1.3998414278030396, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 43908 + }, + { + "epoch": 0.43909, + "grad_norm": 0.6975571513175964, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 43909 + }, + { + "epoch": 0.4391, + "grad_norm": 0.672311544418335, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 43910 + }, + { + "epoch": 0.43911, + "grad_norm": 0.7896636128425598, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 43911 + }, + { + "epoch": 0.43912, + "grad_norm": 0.8886690735816956, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 43912 + }, + { + "epoch": 0.43913, + "grad_norm": 0.930142343044281, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 43913 + }, + { + "epoch": 0.43914, + "grad_norm": 1.0563414096832275, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 43914 + }, + { + "epoch": 0.43915, + "grad_norm": 1.140441656112671, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 43915 + }, + { + "epoch": 0.43916, + "grad_norm": 0.8406035304069519, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 43916 + }, + { + "epoch": 0.43917, + "grad_norm": 0.716157853603363, + "learning_rate": 0.003, + "loss": 3.982, + "step": 43917 + }, + { + "epoch": 0.43918, + "grad_norm": 0.7453579902648926, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 43918 + }, + { + "epoch": 0.43919, + "grad_norm": 0.763471782207489, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 43919 + }, + { + "epoch": 0.4392, + "grad_norm": 0.7972554564476013, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 43920 + }, + { + "epoch": 0.43921, + "grad_norm": 0.7539175152778625, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 43921 + }, + { + "epoch": 0.43922, + "grad_norm": 0.6377076506614685, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 43922 + }, + { + "epoch": 0.43923, + "grad_norm": 0.5865989327430725, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 43923 + }, + { + "epoch": 0.43924, + "grad_norm": 0.8101049661636353, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 43924 + }, + { + "epoch": 0.43925, + "grad_norm": 1.0217691659927368, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 43925 + }, + { + "epoch": 0.43926, + "grad_norm": 1.2658560276031494, + "learning_rate": 0.003, + "loss": 3.993, + "step": 43926 + }, + { + "epoch": 0.43927, + "grad_norm": 0.8624312281608582, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 43927 + }, + { + "epoch": 0.43928, + "grad_norm": 0.9821645617485046, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 43928 + }, + { + "epoch": 0.43929, + "grad_norm": 1.0665955543518066, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 43929 + }, + { + "epoch": 0.4393, + "grad_norm": 0.920567512512207, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 43930 + }, + { + "epoch": 0.43931, + "grad_norm": 1.0815701484680176, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 43931 + }, + { + "epoch": 0.43932, + "grad_norm": 1.1773996353149414, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 43932 + }, + { + "epoch": 0.43933, + "grad_norm": 1.0092575550079346, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 43933 + }, + { + "epoch": 0.43934, + "grad_norm": 1.0692787170410156, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 43934 + }, + { + "epoch": 0.43935, + "grad_norm": 1.1245930194854736, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 43935 + }, + { + "epoch": 0.43936, + "grad_norm": 1.0541428327560425, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 43936 + }, + { + "epoch": 0.43937, + "grad_norm": 1.2169092893600464, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 43937 + }, + { + "epoch": 0.43938, + "grad_norm": 1.0712618827819824, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 43938 + }, + { + "epoch": 0.43939, + "grad_norm": 0.9311126470565796, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 43939 + }, + { + "epoch": 0.4394, + "grad_norm": 1.0411267280578613, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 43940 + }, + { + "epoch": 0.43941, + "grad_norm": 1.0041121244430542, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 43941 + }, + { + "epoch": 0.43942, + "grad_norm": 0.963312566280365, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 43942 + }, + { + "epoch": 0.43943, + "grad_norm": 1.065985918045044, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 43943 + }, + { + "epoch": 0.43944, + "grad_norm": 0.9482407569885254, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 43944 + }, + { + "epoch": 0.43945, + "grad_norm": 1.2220162153244019, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 43945 + }, + { + "epoch": 0.43946, + "grad_norm": 1.0009809732437134, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 43946 + }, + { + "epoch": 0.43947, + "grad_norm": 0.8204920291900635, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 43947 + }, + { + "epoch": 0.43948, + "grad_norm": 0.7649489045143127, + "learning_rate": 0.003, + "loss": 4.0, + "step": 43948 + }, + { + "epoch": 0.43949, + "grad_norm": 0.7815430760383606, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 43949 + }, + { + "epoch": 0.4395, + "grad_norm": 0.7997918128967285, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 43950 + }, + { + "epoch": 0.43951, + "grad_norm": 0.837367057800293, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 43951 + }, + { + "epoch": 0.43952, + "grad_norm": 0.860689640045166, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 43952 + }, + { + "epoch": 0.43953, + "grad_norm": 0.8070992231369019, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 43953 + }, + { + "epoch": 0.43954, + "grad_norm": 0.8164076805114746, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 43954 + }, + { + "epoch": 0.43955, + "grad_norm": 0.7643837928771973, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 43955 + }, + { + "epoch": 0.43956, + "grad_norm": 0.8476876020431519, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 43956 + }, + { + "epoch": 0.43957, + "grad_norm": 0.9969898462295532, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 43957 + }, + { + "epoch": 0.43958, + "grad_norm": 1.3705337047576904, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 43958 + }, + { + "epoch": 0.43959, + "grad_norm": 0.9917928576469421, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 43959 + }, + { + "epoch": 0.4396, + "grad_norm": 0.9330081939697266, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 43960 + }, + { + "epoch": 0.43961, + "grad_norm": 0.7555145025253296, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 43961 + }, + { + "epoch": 0.43962, + "grad_norm": 0.7834479212760925, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 43962 + }, + { + "epoch": 0.43963, + "grad_norm": 0.9520322680473328, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 43963 + }, + { + "epoch": 0.43964, + "grad_norm": 1.0797477960586548, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 43964 + }, + { + "epoch": 0.43965, + "grad_norm": 1.0601701736450195, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 43965 + }, + { + "epoch": 0.43966, + "grad_norm": 1.0165643692016602, + "learning_rate": 0.003, + "loss": 3.9413, + "step": 43966 + }, + { + "epoch": 0.43967, + "grad_norm": 0.8836858868598938, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 43967 + }, + { + "epoch": 0.43968, + "grad_norm": 0.8033360838890076, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 43968 + }, + { + "epoch": 0.43969, + "grad_norm": 0.8273212313652039, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 43969 + }, + { + "epoch": 0.4397, + "grad_norm": 0.8977047801017761, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 43970 + }, + { + "epoch": 0.43971, + "grad_norm": 0.8446850776672363, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 43971 + }, + { + "epoch": 0.43972, + "grad_norm": 1.0199096202850342, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 43972 + }, + { + "epoch": 0.43973, + "grad_norm": 1.1340317726135254, + "learning_rate": 0.003, + "loss": 4.011, + "step": 43973 + }, + { + "epoch": 0.43974, + "grad_norm": 0.7564310431480408, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 43974 + }, + { + "epoch": 0.43975, + "grad_norm": 0.7101050019264221, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 43975 + }, + { + "epoch": 0.43976, + "grad_norm": 0.7875263094902039, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 43976 + }, + { + "epoch": 0.43977, + "grad_norm": 0.80330491065979, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 43977 + }, + { + "epoch": 0.43978, + "grad_norm": 0.7384769916534424, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 43978 + }, + { + "epoch": 0.43979, + "grad_norm": 0.9382259845733643, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 43979 + }, + { + "epoch": 0.4398, + "grad_norm": 1.3006643056869507, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 43980 + }, + { + "epoch": 0.43981, + "grad_norm": 0.7592120170593262, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 43981 + }, + { + "epoch": 0.43982, + "grad_norm": 0.7330105304718018, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 43982 + }, + { + "epoch": 0.43983, + "grad_norm": 0.7548280358314514, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 43983 + }, + { + "epoch": 0.43984, + "grad_norm": 0.7825363278388977, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 43984 + }, + { + "epoch": 0.43985, + "grad_norm": 0.6365594267845154, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 43985 + }, + { + "epoch": 0.43986, + "grad_norm": 0.8328638076782227, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 43986 + }, + { + "epoch": 0.43987, + "grad_norm": 1.1557773351669312, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 43987 + }, + { + "epoch": 0.43988, + "grad_norm": 0.9667882323265076, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 43988 + }, + { + "epoch": 0.43989, + "grad_norm": 1.0224483013153076, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 43989 + }, + { + "epoch": 0.4399, + "grad_norm": 1.0501937866210938, + "learning_rate": 0.003, + "loss": 3.969, + "step": 43990 + }, + { + "epoch": 0.43991, + "grad_norm": 0.9198644161224365, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 43991 + }, + { + "epoch": 0.43992, + "grad_norm": 0.8184239268302917, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 43992 + }, + { + "epoch": 0.43993, + "grad_norm": 0.7157941460609436, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 43993 + }, + { + "epoch": 0.43994, + "grad_norm": 0.9175770282745361, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 43994 + }, + { + "epoch": 0.43995, + "grad_norm": 1.10371732711792, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 43995 + }, + { + "epoch": 0.43996, + "grad_norm": 0.8751172423362732, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 43996 + }, + { + "epoch": 0.43997, + "grad_norm": 0.8514682054519653, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 43997 + }, + { + "epoch": 0.43998, + "grad_norm": 0.792131245136261, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 43998 + }, + { + "epoch": 0.43999, + "grad_norm": 0.7967060804367065, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 43999 + }, + { + "epoch": 0.44, + "grad_norm": 0.9124292731285095, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 44000 + }, + { + "epoch": 0.44001, + "grad_norm": 1.0781348943710327, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 44001 + }, + { + "epoch": 0.44002, + "grad_norm": 0.922421395778656, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 44002 + }, + { + "epoch": 0.44003, + "grad_norm": 0.8315150141716003, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 44003 + }, + { + "epoch": 0.44004, + "grad_norm": 0.9790398478507996, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 44004 + }, + { + "epoch": 0.44005, + "grad_norm": 0.9954426884651184, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 44005 + }, + { + "epoch": 0.44006, + "grad_norm": 1.1850968599319458, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 44006 + }, + { + "epoch": 0.44007, + "grad_norm": 0.9841986894607544, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 44007 + }, + { + "epoch": 0.44008, + "grad_norm": 0.8832800984382629, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 44008 + }, + { + "epoch": 0.44009, + "grad_norm": 0.8535459637641907, + "learning_rate": 0.003, + "loss": 4.008, + "step": 44009 + }, + { + "epoch": 0.4401, + "grad_norm": 0.9493675827980042, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 44010 + }, + { + "epoch": 0.44011, + "grad_norm": 0.9890745282173157, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 44011 + }, + { + "epoch": 0.44012, + "grad_norm": 0.9093183875083923, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 44012 + }, + { + "epoch": 0.44013, + "grad_norm": 1.0096203088760376, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 44013 + }, + { + "epoch": 0.44014, + "grad_norm": 0.9291603565216064, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 44014 + }, + { + "epoch": 0.44015, + "grad_norm": 1.0346473455429077, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 44015 + }, + { + "epoch": 0.44016, + "grad_norm": 1.0660991668701172, + "learning_rate": 0.003, + "loss": 3.968, + "step": 44016 + }, + { + "epoch": 0.44017, + "grad_norm": 0.7612450122833252, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 44017 + }, + { + "epoch": 0.44018, + "grad_norm": 0.658619225025177, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 44018 + }, + { + "epoch": 0.44019, + "grad_norm": 0.616287887096405, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 44019 + }, + { + "epoch": 0.4402, + "grad_norm": 0.5619018077850342, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 44020 + }, + { + "epoch": 0.44021, + "grad_norm": 0.5638255476951599, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 44021 + }, + { + "epoch": 0.44022, + "grad_norm": 0.8139516711235046, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 44022 + }, + { + "epoch": 0.44023, + "grad_norm": 1.0970925092697144, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 44023 + }, + { + "epoch": 0.44024, + "grad_norm": 1.1134850978851318, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 44024 + }, + { + "epoch": 0.44025, + "grad_norm": 0.9193386435508728, + "learning_rate": 0.003, + "loss": 3.988, + "step": 44025 + }, + { + "epoch": 0.44026, + "grad_norm": 0.9511013031005859, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 44026 + }, + { + "epoch": 0.44027, + "grad_norm": 1.130774736404419, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 44027 + }, + { + "epoch": 0.44028, + "grad_norm": 0.9632875323295593, + "learning_rate": 0.003, + "loss": 3.982, + "step": 44028 + }, + { + "epoch": 0.44029, + "grad_norm": 1.0322102308273315, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 44029 + }, + { + "epoch": 0.4403, + "grad_norm": 0.8727246522903442, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 44030 + }, + { + "epoch": 0.44031, + "grad_norm": 0.9749947786331177, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 44031 + }, + { + "epoch": 0.44032, + "grad_norm": 1.0076711177825928, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 44032 + }, + { + "epoch": 0.44033, + "grad_norm": 1.1554516553878784, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 44033 + }, + { + "epoch": 0.44034, + "grad_norm": 0.8298447728157043, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 44034 + }, + { + "epoch": 0.44035, + "grad_norm": 0.8802307844161987, + "learning_rate": 0.003, + "loss": 3.97, + "step": 44035 + }, + { + "epoch": 0.44036, + "grad_norm": 1.037882924079895, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 44036 + }, + { + "epoch": 0.44037, + "grad_norm": 1.3170967102050781, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 44037 + }, + { + "epoch": 0.44038, + "grad_norm": 0.6573392748832703, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 44038 + }, + { + "epoch": 0.44039, + "grad_norm": 0.6910316348075867, + "learning_rate": 0.003, + "loss": 3.9317, + "step": 44039 + }, + { + "epoch": 0.4404, + "grad_norm": 0.8388162851333618, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 44040 + }, + { + "epoch": 0.44041, + "grad_norm": 1.1224666833877563, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 44041 + }, + { + "epoch": 0.44042, + "grad_norm": 1.3918434381484985, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 44042 + }, + { + "epoch": 0.44043, + "grad_norm": 0.5351912975311279, + "learning_rate": 0.003, + "loss": 3.971, + "step": 44043 + }, + { + "epoch": 0.44044, + "grad_norm": 0.8472414612770081, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 44044 + }, + { + "epoch": 0.44045, + "grad_norm": 0.9631103277206421, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 44045 + }, + { + "epoch": 0.44046, + "grad_norm": 0.9928518533706665, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 44046 + }, + { + "epoch": 0.44047, + "grad_norm": 1.2885273694992065, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 44047 + }, + { + "epoch": 0.44048, + "grad_norm": 0.6417211294174194, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 44048 + }, + { + "epoch": 0.44049, + "grad_norm": 0.7296314835548401, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 44049 + }, + { + "epoch": 0.4405, + "grad_norm": 0.9263166785240173, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 44050 + }, + { + "epoch": 0.44051, + "grad_norm": 0.9307546615600586, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 44051 + }, + { + "epoch": 0.44052, + "grad_norm": 0.8496957421302795, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 44052 + }, + { + "epoch": 0.44053, + "grad_norm": 0.7361477017402649, + "learning_rate": 0.003, + "loss": 4.013, + "step": 44053 + }, + { + "epoch": 0.44054, + "grad_norm": 0.8303960561752319, + "learning_rate": 0.003, + "loss": 3.9468, + "step": 44054 + }, + { + "epoch": 0.44055, + "grad_norm": 1.0287425518035889, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 44055 + }, + { + "epoch": 0.44056, + "grad_norm": 0.974346935749054, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 44056 + }, + { + "epoch": 0.44057, + "grad_norm": 1.1415330171585083, + "learning_rate": 0.003, + "loss": 3.97, + "step": 44057 + }, + { + "epoch": 0.44058, + "grad_norm": 0.8135275840759277, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 44058 + }, + { + "epoch": 0.44059, + "grad_norm": 0.7084683179855347, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 44059 + }, + { + "epoch": 0.4406, + "grad_norm": 0.715881884098053, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 44060 + }, + { + "epoch": 0.44061, + "grad_norm": 0.6921995282173157, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 44061 + }, + { + "epoch": 0.44062, + "grad_norm": 0.7425163388252258, + "learning_rate": 0.003, + "loss": 3.9345, + "step": 44062 + }, + { + "epoch": 0.44063, + "grad_norm": 0.8839128017425537, + "learning_rate": 0.003, + "loss": 3.98, + "step": 44063 + }, + { + "epoch": 0.44064, + "grad_norm": 1.003898024559021, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 44064 + }, + { + "epoch": 0.44065, + "grad_norm": 1.003780722618103, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 44065 + }, + { + "epoch": 0.44066, + "grad_norm": 1.0088112354278564, + "learning_rate": 0.003, + "loss": 3.996, + "step": 44066 + }, + { + "epoch": 0.44067, + "grad_norm": 1.0117088556289673, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 44067 + }, + { + "epoch": 0.44068, + "grad_norm": 0.9888450503349304, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 44068 + }, + { + "epoch": 0.44069, + "grad_norm": 0.9909303188323975, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 44069 + }, + { + "epoch": 0.4407, + "grad_norm": 0.9959143996238708, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 44070 + }, + { + "epoch": 0.44071, + "grad_norm": 1.074548602104187, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 44071 + }, + { + "epoch": 0.44072, + "grad_norm": 0.9860294461250305, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 44072 + }, + { + "epoch": 0.44073, + "grad_norm": 1.196723222732544, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 44073 + }, + { + "epoch": 0.44074, + "grad_norm": 1.1833207607269287, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 44074 + }, + { + "epoch": 0.44075, + "grad_norm": 1.01381516456604, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 44075 + }, + { + "epoch": 0.44076, + "grad_norm": 1.4583022594451904, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 44076 + }, + { + "epoch": 0.44077, + "grad_norm": 0.7926441431045532, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 44077 + }, + { + "epoch": 0.44078, + "grad_norm": 0.7707839012145996, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 44078 + }, + { + "epoch": 0.44079, + "grad_norm": 0.9637452363967896, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 44079 + }, + { + "epoch": 0.4408, + "grad_norm": 0.8398649096488953, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 44080 + }, + { + "epoch": 0.44081, + "grad_norm": 0.8817569017410278, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 44081 + }, + { + "epoch": 0.44082, + "grad_norm": 0.9735691547393799, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 44082 + }, + { + "epoch": 0.44083, + "grad_norm": 1.0439399480819702, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 44083 + }, + { + "epoch": 0.44084, + "grad_norm": 1.1358754634857178, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 44084 + }, + { + "epoch": 0.44085, + "grad_norm": 0.8729459047317505, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 44085 + }, + { + "epoch": 0.44086, + "grad_norm": 0.7417519092559814, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 44086 + }, + { + "epoch": 0.44087, + "grad_norm": 0.8343902230262756, + "learning_rate": 0.003, + "loss": 4.009, + "step": 44087 + }, + { + "epoch": 0.44088, + "grad_norm": 1.1003566980361938, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 44088 + }, + { + "epoch": 0.44089, + "grad_norm": 0.9425991177558899, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 44089 + }, + { + "epoch": 0.4409, + "grad_norm": 0.818688690662384, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 44090 + }, + { + "epoch": 0.44091, + "grad_norm": 0.9100265502929688, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 44091 + }, + { + "epoch": 0.44092, + "grad_norm": 0.9983644485473633, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 44092 + }, + { + "epoch": 0.44093, + "grad_norm": 0.9830641150474548, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 44093 + }, + { + "epoch": 0.44094, + "grad_norm": 1.0200306177139282, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 44094 + }, + { + "epoch": 0.44095, + "grad_norm": 1.0143511295318604, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 44095 + }, + { + "epoch": 0.44096, + "grad_norm": 1.0903788805007935, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 44096 + }, + { + "epoch": 0.44097, + "grad_norm": 0.9863420128822327, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 44097 + }, + { + "epoch": 0.44098, + "grad_norm": 0.970558226108551, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 44098 + }, + { + "epoch": 0.44099, + "grad_norm": 0.6932790279388428, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 44099 + }, + { + "epoch": 0.441, + "grad_norm": 0.711229681968689, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 44100 + }, + { + "epoch": 0.44101, + "grad_norm": 0.7471373081207275, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 44101 + }, + { + "epoch": 0.44102, + "grad_norm": 0.8157103061676025, + "learning_rate": 0.003, + "loss": 3.9468, + "step": 44102 + }, + { + "epoch": 0.44103, + "grad_norm": 0.8678898215293884, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 44103 + }, + { + "epoch": 0.44104, + "grad_norm": 0.8515704274177551, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 44104 + }, + { + "epoch": 0.44105, + "grad_norm": 0.8558704853057861, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 44105 + }, + { + "epoch": 0.44106, + "grad_norm": 0.8662272691726685, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 44106 + }, + { + "epoch": 0.44107, + "grad_norm": 0.9017930626869202, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 44107 + }, + { + "epoch": 0.44108, + "grad_norm": 0.8586576581001282, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 44108 + }, + { + "epoch": 0.44109, + "grad_norm": 1.0482012033462524, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 44109 + }, + { + "epoch": 0.4411, + "grad_norm": 1.2855558395385742, + "learning_rate": 0.003, + "loss": 3.996, + "step": 44110 + }, + { + "epoch": 0.44111, + "grad_norm": 0.8976285457611084, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 44111 + }, + { + "epoch": 0.44112, + "grad_norm": 1.051631212234497, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 44112 + }, + { + "epoch": 0.44113, + "grad_norm": 1.1082261800765991, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 44113 + }, + { + "epoch": 0.44114, + "grad_norm": 1.1422269344329834, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 44114 + }, + { + "epoch": 0.44115, + "grad_norm": 0.9314185976982117, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 44115 + }, + { + "epoch": 0.44116, + "grad_norm": 0.7395085692405701, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 44116 + }, + { + "epoch": 0.44117, + "grad_norm": 0.5696622729301453, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 44117 + }, + { + "epoch": 0.44118, + "grad_norm": 0.5946617722511292, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 44118 + }, + { + "epoch": 0.44119, + "grad_norm": 0.6921218633651733, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 44119 + }, + { + "epoch": 0.4412, + "grad_norm": 0.8800392746925354, + "learning_rate": 0.003, + "loss": 3.982, + "step": 44120 + }, + { + "epoch": 0.44121, + "grad_norm": 1.004442572593689, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 44121 + }, + { + "epoch": 0.44122, + "grad_norm": 0.8047122955322266, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 44122 + }, + { + "epoch": 0.44123, + "grad_norm": 0.7029998898506165, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 44123 + }, + { + "epoch": 0.44124, + "grad_norm": 0.6638668179512024, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 44124 + }, + { + "epoch": 0.44125, + "grad_norm": 0.5729047060012817, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 44125 + }, + { + "epoch": 0.44126, + "grad_norm": 0.5908877849578857, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 44126 + }, + { + "epoch": 0.44127, + "grad_norm": 0.6356565356254578, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 44127 + }, + { + "epoch": 0.44128, + "grad_norm": 0.7468500137329102, + "learning_rate": 0.003, + "loss": 3.966, + "step": 44128 + }, + { + "epoch": 0.44129, + "grad_norm": 0.8869208693504333, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 44129 + }, + { + "epoch": 0.4413, + "grad_norm": 1.0975780487060547, + "learning_rate": 0.003, + "loss": 3.979, + "step": 44130 + }, + { + "epoch": 0.44131, + "grad_norm": 1.1025068759918213, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 44131 + }, + { + "epoch": 0.44132, + "grad_norm": 0.8390064239501953, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 44132 + }, + { + "epoch": 0.44133, + "grad_norm": 0.7151958346366882, + "learning_rate": 0.003, + "loss": 3.961, + "step": 44133 + }, + { + "epoch": 0.44134, + "grad_norm": 0.6872574090957642, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 44134 + }, + { + "epoch": 0.44135, + "grad_norm": 0.970418393611908, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 44135 + }, + { + "epoch": 0.44136, + "grad_norm": 1.4151853322982788, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 44136 + }, + { + "epoch": 0.44137, + "grad_norm": 0.8654571771621704, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 44137 + }, + { + "epoch": 0.44138, + "grad_norm": 0.8632475137710571, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 44138 + }, + { + "epoch": 0.44139, + "grad_norm": 0.8427532315254211, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 44139 + }, + { + "epoch": 0.4414, + "grad_norm": 0.7762616276741028, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 44140 + }, + { + "epoch": 0.44141, + "grad_norm": 0.7977359890937805, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 44141 + }, + { + "epoch": 0.44142, + "grad_norm": 0.94556725025177, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 44142 + }, + { + "epoch": 0.44143, + "grad_norm": 1.216833233833313, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 44143 + }, + { + "epoch": 0.44144, + "grad_norm": 0.9521523118019104, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 44144 + }, + { + "epoch": 0.44145, + "grad_norm": 0.8607267141342163, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 44145 + }, + { + "epoch": 0.44146, + "grad_norm": 0.8883578777313232, + "learning_rate": 0.003, + "loss": 3.967, + "step": 44146 + }, + { + "epoch": 0.44147, + "grad_norm": 0.8377810716629028, + "learning_rate": 0.003, + "loss": 3.96, + "step": 44147 + }, + { + "epoch": 0.44148, + "grad_norm": 0.7630735635757446, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 44148 + }, + { + "epoch": 0.44149, + "grad_norm": 0.8458321690559387, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 44149 + }, + { + "epoch": 0.4415, + "grad_norm": 0.8479447364807129, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 44150 + }, + { + "epoch": 0.44151, + "grad_norm": 0.8332847356796265, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 44151 + }, + { + "epoch": 0.44152, + "grad_norm": 1.0622570514678955, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 44152 + }, + { + "epoch": 0.44153, + "grad_norm": 1.105233907699585, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 44153 + }, + { + "epoch": 0.44154, + "grad_norm": 0.9753423929214478, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 44154 + }, + { + "epoch": 0.44155, + "grad_norm": 1.1551843881607056, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 44155 + }, + { + "epoch": 0.44156, + "grad_norm": 1.0803967714309692, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 44156 + }, + { + "epoch": 0.44157, + "grad_norm": 0.9610110521316528, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 44157 + }, + { + "epoch": 0.44158, + "grad_norm": 0.8641701340675354, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 44158 + }, + { + "epoch": 0.44159, + "grad_norm": 0.862952709197998, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 44159 + }, + { + "epoch": 0.4416, + "grad_norm": 0.969983696937561, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 44160 + }, + { + "epoch": 0.44161, + "grad_norm": 1.0698273181915283, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 44161 + }, + { + "epoch": 0.44162, + "grad_norm": 1.1549217700958252, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 44162 + }, + { + "epoch": 0.44163, + "grad_norm": 0.9079859852790833, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 44163 + }, + { + "epoch": 0.44164, + "grad_norm": 0.84018874168396, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 44164 + }, + { + "epoch": 0.44165, + "grad_norm": 0.8046923279762268, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 44165 + }, + { + "epoch": 0.44166, + "grad_norm": 0.8754952549934387, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 44166 + }, + { + "epoch": 0.44167, + "grad_norm": 0.999377965927124, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 44167 + }, + { + "epoch": 0.44168, + "grad_norm": 0.9787615537643433, + "learning_rate": 0.003, + "loss": 3.964, + "step": 44168 + }, + { + "epoch": 0.44169, + "grad_norm": 0.8423898220062256, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 44169 + }, + { + "epoch": 0.4417, + "grad_norm": 0.7461389899253845, + "learning_rate": 0.003, + "loss": 3.976, + "step": 44170 + }, + { + "epoch": 0.44171, + "grad_norm": 0.709709107875824, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 44171 + }, + { + "epoch": 0.44172, + "grad_norm": 0.6739606261253357, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 44172 + }, + { + "epoch": 0.44173, + "grad_norm": 0.9885135293006897, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 44173 + }, + { + "epoch": 0.44174, + "grad_norm": 1.0850499868392944, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 44174 + }, + { + "epoch": 0.44175, + "grad_norm": 0.8632785081863403, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 44175 + }, + { + "epoch": 0.44176, + "grad_norm": 1.2007025480270386, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 44176 + }, + { + "epoch": 0.44177, + "grad_norm": 1.4782142639160156, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 44177 + }, + { + "epoch": 0.44178, + "grad_norm": 0.49457260966300964, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 44178 + }, + { + "epoch": 0.44179, + "grad_norm": 0.9383417367935181, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 44179 + }, + { + "epoch": 0.4418, + "grad_norm": 1.3812084197998047, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 44180 + }, + { + "epoch": 0.44181, + "grad_norm": 0.6616448163986206, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 44181 + }, + { + "epoch": 0.44182, + "grad_norm": 0.6670988202095032, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 44182 + }, + { + "epoch": 0.44183, + "grad_norm": 0.7489914298057556, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 44183 + }, + { + "epoch": 0.44184, + "grad_norm": 0.7844533324241638, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 44184 + }, + { + "epoch": 0.44185, + "grad_norm": 0.7644988894462585, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 44185 + }, + { + "epoch": 0.44186, + "grad_norm": 0.8605934381484985, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 44186 + }, + { + "epoch": 0.44187, + "grad_norm": 1.009963870048523, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 44187 + }, + { + "epoch": 0.44188, + "grad_norm": 1.1585317850112915, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 44188 + }, + { + "epoch": 0.44189, + "grad_norm": 1.1227896213531494, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 44189 + }, + { + "epoch": 0.4419, + "grad_norm": 0.9777818918228149, + "learning_rate": 0.003, + "loss": 3.964, + "step": 44190 + }, + { + "epoch": 0.44191, + "grad_norm": 1.0265772342681885, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 44191 + }, + { + "epoch": 0.44192, + "grad_norm": 1.0802990198135376, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 44192 + }, + { + "epoch": 0.44193, + "grad_norm": 0.8973214030265808, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 44193 + }, + { + "epoch": 0.44194, + "grad_norm": 0.7721841931343079, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 44194 + }, + { + "epoch": 0.44195, + "grad_norm": 0.7427859902381897, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 44195 + }, + { + "epoch": 0.44196, + "grad_norm": 0.7423489689826965, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 44196 + }, + { + "epoch": 0.44197, + "grad_norm": 0.8357716202735901, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 44197 + }, + { + "epoch": 0.44198, + "grad_norm": 0.9100366234779358, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 44198 + }, + { + "epoch": 0.44199, + "grad_norm": 1.1125133037567139, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 44199 + }, + { + "epoch": 0.442, + "grad_norm": 0.906172513961792, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 44200 + }, + { + "epoch": 0.44201, + "grad_norm": 0.8478556871414185, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 44201 + }, + { + "epoch": 0.44202, + "grad_norm": 1.07661771774292, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 44202 + }, + { + "epoch": 0.44203, + "grad_norm": 1.2469936609268188, + "learning_rate": 0.003, + "loss": 4.012, + "step": 44203 + }, + { + "epoch": 0.44204, + "grad_norm": 0.6637877821922302, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 44204 + }, + { + "epoch": 0.44205, + "grad_norm": 0.6309342980384827, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 44205 + }, + { + "epoch": 0.44206, + "grad_norm": 0.6003397107124329, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 44206 + }, + { + "epoch": 0.44207, + "grad_norm": 0.664302408695221, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 44207 + }, + { + "epoch": 0.44208, + "grad_norm": 0.7351792454719543, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 44208 + }, + { + "epoch": 0.44209, + "grad_norm": 0.780950665473938, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 44209 + }, + { + "epoch": 0.4421, + "grad_norm": 0.7693073153495789, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 44210 + }, + { + "epoch": 0.44211, + "grad_norm": 0.7675715684890747, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 44211 + }, + { + "epoch": 0.44212, + "grad_norm": 0.8284640908241272, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 44212 + }, + { + "epoch": 0.44213, + "grad_norm": 0.8191101551055908, + "learning_rate": 0.003, + "loss": 3.99, + "step": 44213 + }, + { + "epoch": 0.44214, + "grad_norm": 0.8980073928833008, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 44214 + }, + { + "epoch": 0.44215, + "grad_norm": 0.8508952260017395, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 44215 + }, + { + "epoch": 0.44216, + "grad_norm": 1.1171783208847046, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 44216 + }, + { + "epoch": 0.44217, + "grad_norm": 1.1259658336639404, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 44217 + }, + { + "epoch": 0.44218, + "grad_norm": 1.208828330039978, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 44218 + }, + { + "epoch": 0.44219, + "grad_norm": 1.1167551279067993, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 44219 + }, + { + "epoch": 0.4422, + "grad_norm": 0.7942836284637451, + "learning_rate": 0.003, + "loss": 3.9406, + "step": 44220 + }, + { + "epoch": 0.44221, + "grad_norm": 0.575116753578186, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 44221 + }, + { + "epoch": 0.44222, + "grad_norm": 0.5481115579605103, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 44222 + }, + { + "epoch": 0.44223, + "grad_norm": 0.6353884935379028, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 44223 + }, + { + "epoch": 0.44224, + "grad_norm": 0.737981379032135, + "learning_rate": 0.003, + "loss": 3.9412, + "step": 44224 + }, + { + "epoch": 0.44225, + "grad_norm": 1.0052058696746826, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 44225 + }, + { + "epoch": 0.44226, + "grad_norm": 1.5976252555847168, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 44226 + }, + { + "epoch": 0.44227, + "grad_norm": 0.6241233348846436, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 44227 + }, + { + "epoch": 0.44228, + "grad_norm": 0.8388108611106873, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 44228 + }, + { + "epoch": 0.44229, + "grad_norm": 1.0231997966766357, + "learning_rate": 0.003, + "loss": 3.97, + "step": 44229 + }, + { + "epoch": 0.4423, + "grad_norm": 1.366904854774475, + "learning_rate": 0.003, + "loss": 4.004, + "step": 44230 + }, + { + "epoch": 0.44231, + "grad_norm": 0.9737522006034851, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 44231 + }, + { + "epoch": 0.44232, + "grad_norm": 1.0357887744903564, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 44232 + }, + { + "epoch": 0.44233, + "grad_norm": 1.0822440385818481, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 44233 + }, + { + "epoch": 0.44234, + "grad_norm": 1.0387383699417114, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 44234 + }, + { + "epoch": 0.44235, + "grad_norm": 0.8761442303657532, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 44235 + }, + { + "epoch": 0.44236, + "grad_norm": 0.7624320387840271, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 44236 + }, + { + "epoch": 0.44237, + "grad_norm": 0.7456367611885071, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 44237 + }, + { + "epoch": 0.44238, + "grad_norm": 0.8128392100334167, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 44238 + }, + { + "epoch": 0.44239, + "grad_norm": 0.9095715880393982, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 44239 + }, + { + "epoch": 0.4424, + "grad_norm": 0.8535902500152588, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 44240 + }, + { + "epoch": 0.44241, + "grad_norm": 0.7940321564674377, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 44241 + }, + { + "epoch": 0.44242, + "grad_norm": 0.7662673592567444, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 44242 + }, + { + "epoch": 0.44243, + "grad_norm": 0.886294960975647, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 44243 + }, + { + "epoch": 0.44244, + "grad_norm": 0.9378278255462646, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 44244 + }, + { + "epoch": 0.44245, + "grad_norm": 1.126797080039978, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 44245 + }, + { + "epoch": 0.44246, + "grad_norm": 0.8536601066589355, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 44246 + }, + { + "epoch": 0.44247, + "grad_norm": 0.8784573078155518, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 44247 + }, + { + "epoch": 0.44248, + "grad_norm": 1.087536096572876, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 44248 + }, + { + "epoch": 0.44249, + "grad_norm": 0.997896134853363, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 44249 + }, + { + "epoch": 0.4425, + "grad_norm": 1.0440915822982788, + "learning_rate": 0.003, + "loss": 3.989, + "step": 44250 + }, + { + "epoch": 0.44251, + "grad_norm": 1.1668338775634766, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 44251 + }, + { + "epoch": 0.44252, + "grad_norm": 1.0048283338546753, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 44252 + }, + { + "epoch": 0.44253, + "grad_norm": 0.9463854432106018, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 44253 + }, + { + "epoch": 0.44254, + "grad_norm": 1.1034575700759888, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 44254 + }, + { + "epoch": 0.44255, + "grad_norm": 0.9907688498497009, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 44255 + }, + { + "epoch": 0.44256, + "grad_norm": 1.3282746076583862, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 44256 + }, + { + "epoch": 0.44257, + "grad_norm": 0.8626319169998169, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 44257 + }, + { + "epoch": 0.44258, + "grad_norm": 0.9090220928192139, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 44258 + }, + { + "epoch": 0.44259, + "grad_norm": 1.0055131912231445, + "learning_rate": 0.003, + "loss": 3.9478, + "step": 44259 + }, + { + "epoch": 0.4426, + "grad_norm": 1.0685529708862305, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 44260 + }, + { + "epoch": 0.44261, + "grad_norm": 1.1840643882751465, + "learning_rate": 0.003, + "loss": 3.9466, + "step": 44261 + }, + { + "epoch": 0.44262, + "grad_norm": 0.9047818779945374, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 44262 + }, + { + "epoch": 0.44263, + "grad_norm": 0.8290787935256958, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 44263 + }, + { + "epoch": 0.44264, + "grad_norm": 0.8991026878356934, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 44264 + }, + { + "epoch": 0.44265, + "grad_norm": 0.8992170095443726, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 44265 + }, + { + "epoch": 0.44266, + "grad_norm": 0.8623798489570618, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 44266 + }, + { + "epoch": 0.44267, + "grad_norm": 0.7734845280647278, + "learning_rate": 0.003, + "loss": 3.9414, + "step": 44267 + }, + { + "epoch": 0.44268, + "grad_norm": 0.7477924227714539, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 44268 + }, + { + "epoch": 0.44269, + "grad_norm": 0.7715940475463867, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 44269 + }, + { + "epoch": 0.4427, + "grad_norm": 0.8817546367645264, + "learning_rate": 0.003, + "loss": 3.9363, + "step": 44270 + }, + { + "epoch": 0.44271, + "grad_norm": 0.9431425333023071, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 44271 + }, + { + "epoch": 0.44272, + "grad_norm": 1.22718346118927, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 44272 + }, + { + "epoch": 0.44273, + "grad_norm": 0.8009310960769653, + "learning_rate": 0.003, + "loss": 3.9395, + "step": 44273 + }, + { + "epoch": 0.44274, + "grad_norm": 0.6307597756385803, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 44274 + }, + { + "epoch": 0.44275, + "grad_norm": 0.5830349922180176, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 44275 + }, + { + "epoch": 0.44276, + "grad_norm": 0.5471791625022888, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 44276 + }, + { + "epoch": 0.44277, + "grad_norm": 0.6557934880256653, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 44277 + }, + { + "epoch": 0.44278, + "grad_norm": 0.7564189434051514, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 44278 + }, + { + "epoch": 0.44279, + "grad_norm": 0.777091383934021, + "learning_rate": 0.003, + "loss": 3.9323, + "step": 44279 + }, + { + "epoch": 0.4428, + "grad_norm": 0.9907268285751343, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 44280 + }, + { + "epoch": 0.44281, + "grad_norm": 1.303306221961975, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 44281 + }, + { + "epoch": 0.44282, + "grad_norm": 0.6525434255599976, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 44282 + }, + { + "epoch": 0.44283, + "grad_norm": 0.6959125995635986, + "learning_rate": 0.003, + "loss": 3.9402, + "step": 44283 + }, + { + "epoch": 0.44284, + "grad_norm": 0.9207391738891602, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 44284 + }, + { + "epoch": 0.44285, + "grad_norm": 1.1988801956176758, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 44285 + }, + { + "epoch": 0.44286, + "grad_norm": 0.9610010385513306, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 44286 + }, + { + "epoch": 0.44287, + "grad_norm": 0.9071156978607178, + "learning_rate": 0.003, + "loss": 4.001, + "step": 44287 + }, + { + "epoch": 0.44288, + "grad_norm": 0.8537433743476868, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 44288 + }, + { + "epoch": 0.44289, + "grad_norm": 0.8729925155639648, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 44289 + }, + { + "epoch": 0.4429, + "grad_norm": 0.8891373872756958, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 44290 + }, + { + "epoch": 0.44291, + "grad_norm": 1.0075268745422363, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 44291 + }, + { + "epoch": 0.44292, + "grad_norm": 1.1394709348678589, + "learning_rate": 0.003, + "loss": 3.991, + "step": 44292 + }, + { + "epoch": 0.44293, + "grad_norm": 0.8078111410140991, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 44293 + }, + { + "epoch": 0.44294, + "grad_norm": 0.9143041372299194, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 44294 + }, + { + "epoch": 0.44295, + "grad_norm": 1.049535870552063, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 44295 + }, + { + "epoch": 0.44296, + "grad_norm": 1.1787726879119873, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 44296 + }, + { + "epoch": 0.44297, + "grad_norm": 1.0512086153030396, + "learning_rate": 0.003, + "loss": 3.978, + "step": 44297 + }, + { + "epoch": 0.44298, + "grad_norm": 1.0054930448532104, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 44298 + }, + { + "epoch": 0.44299, + "grad_norm": 1.1600042581558228, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 44299 + }, + { + "epoch": 0.443, + "grad_norm": 1.1804790496826172, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 44300 + }, + { + "epoch": 0.44301, + "grad_norm": 0.8121929168701172, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 44301 + }, + { + "epoch": 0.44302, + "grad_norm": 0.7223600745201111, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 44302 + }, + { + "epoch": 0.44303, + "grad_norm": 0.7650474309921265, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 44303 + }, + { + "epoch": 0.44304, + "grad_norm": 0.7732557058334351, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 44304 + }, + { + "epoch": 0.44305, + "grad_norm": 0.7963756918907166, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 44305 + }, + { + "epoch": 0.44306, + "grad_norm": 0.9147959351539612, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 44306 + }, + { + "epoch": 0.44307, + "grad_norm": 1.1515083312988281, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 44307 + }, + { + "epoch": 0.44308, + "grad_norm": 1.2391773462295532, + "learning_rate": 0.003, + "loss": 3.978, + "step": 44308 + }, + { + "epoch": 0.44309, + "grad_norm": 0.6828050017356873, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 44309 + }, + { + "epoch": 0.4431, + "grad_norm": 0.6387662291526794, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 44310 + }, + { + "epoch": 0.44311, + "grad_norm": 0.8886035680770874, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 44311 + }, + { + "epoch": 0.44312, + "grad_norm": 1.0349392890930176, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 44312 + }, + { + "epoch": 0.44313, + "grad_norm": 1.1374058723449707, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 44313 + }, + { + "epoch": 0.44314, + "grad_norm": 0.7498847842216492, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 44314 + }, + { + "epoch": 0.44315, + "grad_norm": 0.6675453186035156, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 44315 + }, + { + "epoch": 0.44316, + "grad_norm": 0.7229109406471252, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 44316 + }, + { + "epoch": 0.44317, + "grad_norm": 0.933299720287323, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 44317 + }, + { + "epoch": 0.44318, + "grad_norm": 1.15121328830719, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 44318 + }, + { + "epoch": 0.44319, + "grad_norm": 0.8560410141944885, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 44319 + }, + { + "epoch": 0.4432, + "grad_norm": 0.7544731497764587, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 44320 + }, + { + "epoch": 0.44321, + "grad_norm": 0.6848424077033997, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 44321 + }, + { + "epoch": 0.44322, + "grad_norm": 0.6752060055732727, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 44322 + }, + { + "epoch": 0.44323, + "grad_norm": 0.7621832489967346, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 44323 + }, + { + "epoch": 0.44324, + "grad_norm": 0.8381452560424805, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 44324 + }, + { + "epoch": 0.44325, + "grad_norm": 1.0052028894424438, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 44325 + }, + { + "epoch": 0.44326, + "grad_norm": 1.098386526107788, + "learning_rate": 0.003, + "loss": 3.965, + "step": 44326 + }, + { + "epoch": 0.44327, + "grad_norm": 0.8939911127090454, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 44327 + }, + { + "epoch": 0.44328, + "grad_norm": 1.0061445236206055, + "learning_rate": 0.003, + "loss": 3.9155, + "step": 44328 + }, + { + "epoch": 0.44329, + "grad_norm": 1.3917311429977417, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 44329 + }, + { + "epoch": 0.4433, + "grad_norm": 0.6887014508247375, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 44330 + }, + { + "epoch": 0.44331, + "grad_norm": 0.6288831233978271, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 44331 + }, + { + "epoch": 0.44332, + "grad_norm": 0.8974058032035828, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 44332 + }, + { + "epoch": 0.44333, + "grad_norm": 1.1084468364715576, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 44333 + }, + { + "epoch": 0.44334, + "grad_norm": 1.0219285488128662, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 44334 + }, + { + "epoch": 0.44335, + "grad_norm": 0.8475208878517151, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 44335 + }, + { + "epoch": 0.44336, + "grad_norm": 0.7466319799423218, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 44336 + }, + { + "epoch": 0.44337, + "grad_norm": 0.8713963031768799, + "learning_rate": 0.003, + "loss": 3.975, + "step": 44337 + }, + { + "epoch": 0.44338, + "grad_norm": 0.9889897704124451, + "learning_rate": 0.003, + "loss": 3.9384, + "step": 44338 + }, + { + "epoch": 0.44339, + "grad_norm": 1.0304505825042725, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 44339 + }, + { + "epoch": 0.4434, + "grad_norm": 0.9254921674728394, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 44340 + }, + { + "epoch": 0.44341, + "grad_norm": 1.0482591390609741, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 44341 + }, + { + "epoch": 0.44342, + "grad_norm": 0.9816319346427917, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 44342 + }, + { + "epoch": 0.44343, + "grad_norm": 1.0421464443206787, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 44343 + }, + { + "epoch": 0.44344, + "grad_norm": 0.8772028088569641, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 44344 + }, + { + "epoch": 0.44345, + "grad_norm": 0.9474433064460754, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 44345 + }, + { + "epoch": 0.44346, + "grad_norm": 1.0066148042678833, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 44346 + }, + { + "epoch": 0.44347, + "grad_norm": 0.9929884672164917, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 44347 + }, + { + "epoch": 0.44348, + "grad_norm": 1.1140539646148682, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 44348 + }, + { + "epoch": 0.44349, + "grad_norm": 0.9555339813232422, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 44349 + }, + { + "epoch": 0.4435, + "grad_norm": 1.1801488399505615, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 44350 + }, + { + "epoch": 0.44351, + "grad_norm": 1.248214840888977, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 44351 + }, + { + "epoch": 0.44352, + "grad_norm": 0.9117711186408997, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 44352 + }, + { + "epoch": 0.44353, + "grad_norm": 0.7667371034622192, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 44353 + }, + { + "epoch": 0.44354, + "grad_norm": 0.6831209063529968, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 44354 + }, + { + "epoch": 0.44355, + "grad_norm": 0.7318100333213806, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 44355 + }, + { + "epoch": 0.44356, + "grad_norm": 1.0690511465072632, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 44356 + }, + { + "epoch": 0.44357, + "grad_norm": 1.1444953680038452, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 44357 + }, + { + "epoch": 0.44358, + "grad_norm": 0.8177241086959839, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 44358 + }, + { + "epoch": 0.44359, + "grad_norm": 0.9163967967033386, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 44359 + }, + { + "epoch": 0.4436, + "grad_norm": 1.1042400598526, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 44360 + }, + { + "epoch": 0.44361, + "grad_norm": 0.9394365549087524, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 44361 + }, + { + "epoch": 0.44362, + "grad_norm": 1.01776921749115, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 44362 + }, + { + "epoch": 0.44363, + "grad_norm": 1.2012207508087158, + "learning_rate": 0.003, + "loss": 3.985, + "step": 44363 + }, + { + "epoch": 0.44364, + "grad_norm": 0.6955591440200806, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 44364 + }, + { + "epoch": 0.44365, + "grad_norm": 0.5710033774375916, + "learning_rate": 0.003, + "loss": 3.9234, + "step": 44365 + }, + { + "epoch": 0.44366, + "grad_norm": 0.6228160262107849, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 44366 + }, + { + "epoch": 0.44367, + "grad_norm": 0.6772131323814392, + "learning_rate": 0.003, + "loss": 3.988, + "step": 44367 + }, + { + "epoch": 0.44368, + "grad_norm": 0.795384407043457, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 44368 + }, + { + "epoch": 0.44369, + "grad_norm": 1.0849028825759888, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 44369 + }, + { + "epoch": 0.4437, + "grad_norm": 1.1782575845718384, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 44370 + }, + { + "epoch": 0.44371, + "grad_norm": 0.7378427386283875, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 44371 + }, + { + "epoch": 0.44372, + "grad_norm": 0.7612600922584534, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 44372 + }, + { + "epoch": 0.44373, + "grad_norm": 1.031602144241333, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 44373 + }, + { + "epoch": 0.44374, + "grad_norm": 1.2398313283920288, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 44374 + }, + { + "epoch": 0.44375, + "grad_norm": 0.7911251783370972, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 44375 + }, + { + "epoch": 0.44376, + "grad_norm": 0.6690364480018616, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 44376 + }, + { + "epoch": 0.44377, + "grad_norm": 0.6513012051582336, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 44377 + }, + { + "epoch": 0.44378, + "grad_norm": 0.7377553582191467, + "learning_rate": 0.003, + "loss": 3.9381, + "step": 44378 + }, + { + "epoch": 0.44379, + "grad_norm": 0.7449035048484802, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 44379 + }, + { + "epoch": 0.4438, + "grad_norm": 0.625264585018158, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 44380 + }, + { + "epoch": 0.44381, + "grad_norm": 0.6922300457954407, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 44381 + }, + { + "epoch": 0.44382, + "grad_norm": 0.7010605335235596, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 44382 + }, + { + "epoch": 0.44383, + "grad_norm": 0.7809336185455322, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 44383 + }, + { + "epoch": 0.44384, + "grad_norm": 0.8758437037467957, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 44384 + }, + { + "epoch": 0.44385, + "grad_norm": 0.6892402768135071, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 44385 + }, + { + "epoch": 0.44386, + "grad_norm": 0.6793992519378662, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 44386 + }, + { + "epoch": 0.44387, + "grad_norm": 1.0295915603637695, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 44387 + }, + { + "epoch": 0.44388, + "grad_norm": 1.5844190120697021, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 44388 + }, + { + "epoch": 0.44389, + "grad_norm": 0.6204394698143005, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 44389 + }, + { + "epoch": 0.4439, + "grad_norm": 0.9399135112762451, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 44390 + }, + { + "epoch": 0.44391, + "grad_norm": 1.3681015968322754, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 44391 + }, + { + "epoch": 0.44392, + "grad_norm": 0.7299410104751587, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 44392 + }, + { + "epoch": 0.44393, + "grad_norm": 0.7864060401916504, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 44393 + }, + { + "epoch": 0.44394, + "grad_norm": 0.938862681388855, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 44394 + }, + { + "epoch": 0.44395, + "grad_norm": 1.0759280920028687, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 44395 + }, + { + "epoch": 0.44396, + "grad_norm": 1.0143811702728271, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 44396 + }, + { + "epoch": 0.44397, + "grad_norm": 1.42363440990448, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 44397 + }, + { + "epoch": 0.44398, + "grad_norm": 0.6639456152915955, + "learning_rate": 0.003, + "loss": 3.993, + "step": 44398 + }, + { + "epoch": 0.44399, + "grad_norm": 0.7947235703468323, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 44399 + }, + { + "epoch": 0.444, + "grad_norm": 1.1179579496383667, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 44400 + }, + { + "epoch": 0.44401, + "grad_norm": 1.0040861368179321, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 44401 + }, + { + "epoch": 0.44402, + "grad_norm": 1.2510818243026733, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 44402 + }, + { + "epoch": 0.44403, + "grad_norm": 0.7186638116836548, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 44403 + }, + { + "epoch": 0.44404, + "grad_norm": 0.9069967269897461, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 44404 + }, + { + "epoch": 0.44405, + "grad_norm": 1.2311978340148926, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 44405 + }, + { + "epoch": 0.44406, + "grad_norm": 0.7149449586868286, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 44406 + }, + { + "epoch": 0.44407, + "grad_norm": 0.7070090174674988, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 44407 + }, + { + "epoch": 0.44408, + "grad_norm": 0.7738089561462402, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 44408 + }, + { + "epoch": 0.44409, + "grad_norm": 0.9173832535743713, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 44409 + }, + { + "epoch": 0.4441, + "grad_norm": 1.0662444829940796, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 44410 + }, + { + "epoch": 0.44411, + "grad_norm": 1.1597504615783691, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 44411 + }, + { + "epoch": 0.44412, + "grad_norm": 0.9652426838874817, + "learning_rate": 0.003, + "loss": 4.0, + "step": 44412 + }, + { + "epoch": 0.44413, + "grad_norm": 1.1635806560516357, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 44413 + }, + { + "epoch": 0.44414, + "grad_norm": 0.9792842268943787, + "learning_rate": 0.003, + "loss": 4.008, + "step": 44414 + }, + { + "epoch": 0.44415, + "grad_norm": 1.0906875133514404, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 44415 + }, + { + "epoch": 0.44416, + "grad_norm": 0.917216420173645, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 44416 + }, + { + "epoch": 0.44417, + "grad_norm": 1.1330416202545166, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 44417 + }, + { + "epoch": 0.44418, + "grad_norm": 1.082491159439087, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 44418 + }, + { + "epoch": 0.44419, + "grad_norm": 1.0731210708618164, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 44419 + }, + { + "epoch": 0.4442, + "grad_norm": 1.0682802200317383, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 44420 + }, + { + "epoch": 0.44421, + "grad_norm": 0.9043077230453491, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 44421 + }, + { + "epoch": 0.44422, + "grad_norm": 0.8767645359039307, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 44422 + }, + { + "epoch": 0.44423, + "grad_norm": 0.8877975940704346, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 44423 + }, + { + "epoch": 0.44424, + "grad_norm": 0.8667067289352417, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 44424 + }, + { + "epoch": 0.44425, + "grad_norm": 0.8988016843795776, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 44425 + }, + { + "epoch": 0.44426, + "grad_norm": 1.1498733758926392, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 44426 + }, + { + "epoch": 0.44427, + "grad_norm": 0.9086126089096069, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 44427 + }, + { + "epoch": 0.44428, + "grad_norm": 0.7907522320747375, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 44428 + }, + { + "epoch": 0.44429, + "grad_norm": 0.7797096371650696, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 44429 + }, + { + "epoch": 0.4443, + "grad_norm": 0.8378423452377319, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 44430 + }, + { + "epoch": 0.44431, + "grad_norm": 0.7812336087226868, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 44431 + }, + { + "epoch": 0.44432, + "grad_norm": 0.8510831594467163, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 44432 + }, + { + "epoch": 0.44433, + "grad_norm": 1.1235235929489136, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 44433 + }, + { + "epoch": 0.44434, + "grad_norm": 0.8631736636161804, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 44434 + }, + { + "epoch": 0.44435, + "grad_norm": 0.9039679765701294, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 44435 + }, + { + "epoch": 0.44436, + "grad_norm": 0.9988727569580078, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 44436 + }, + { + "epoch": 0.44437, + "grad_norm": 0.9724121689796448, + "learning_rate": 0.003, + "loss": 3.961, + "step": 44437 + }, + { + "epoch": 0.44438, + "grad_norm": 0.987978458404541, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 44438 + }, + { + "epoch": 0.44439, + "grad_norm": 0.8959125876426697, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 44439 + }, + { + "epoch": 0.4444, + "grad_norm": 0.9187315702438354, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 44440 + }, + { + "epoch": 0.44441, + "grad_norm": 0.980845034122467, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 44441 + }, + { + "epoch": 0.44442, + "grad_norm": 1.1724134683609009, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 44442 + }, + { + "epoch": 0.44443, + "grad_norm": 1.0563102960586548, + "learning_rate": 0.003, + "loss": 3.987, + "step": 44443 + }, + { + "epoch": 0.44444, + "grad_norm": 0.9844160079956055, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 44444 + }, + { + "epoch": 0.44445, + "grad_norm": 1.0145176649093628, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 44445 + }, + { + "epoch": 0.44446, + "grad_norm": 0.943321704864502, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 44446 + }, + { + "epoch": 0.44447, + "grad_norm": 0.906143844127655, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 44447 + }, + { + "epoch": 0.44448, + "grad_norm": 0.9283214211463928, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 44448 + }, + { + "epoch": 0.44449, + "grad_norm": 1.0649642944335938, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 44449 + }, + { + "epoch": 0.4445, + "grad_norm": 1.2099391222000122, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 44450 + }, + { + "epoch": 0.44451, + "grad_norm": 0.7642157077789307, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 44451 + }, + { + "epoch": 0.44452, + "grad_norm": 0.855771005153656, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 44452 + }, + { + "epoch": 0.44453, + "grad_norm": 0.924208402633667, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 44453 + }, + { + "epoch": 0.44454, + "grad_norm": 0.9200699925422668, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 44454 + }, + { + "epoch": 0.44455, + "grad_norm": 0.868678867816925, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 44455 + }, + { + "epoch": 0.44456, + "grad_norm": 1.081575870513916, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 44456 + }, + { + "epoch": 0.44457, + "grad_norm": 0.9572480916976929, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 44457 + }, + { + "epoch": 0.44458, + "grad_norm": 0.9971445202827454, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 44458 + }, + { + "epoch": 0.44459, + "grad_norm": 1.324101209640503, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 44459 + }, + { + "epoch": 0.4446, + "grad_norm": 0.9524949193000793, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 44460 + }, + { + "epoch": 0.44461, + "grad_norm": 0.9197249412536621, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 44461 + }, + { + "epoch": 0.44462, + "grad_norm": 1.1532909870147705, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 44462 + }, + { + "epoch": 0.44463, + "grad_norm": 0.8496474623680115, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 44463 + }, + { + "epoch": 0.44464, + "grad_norm": 0.7221997976303101, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 44464 + }, + { + "epoch": 0.44465, + "grad_norm": 0.6568846106529236, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 44465 + }, + { + "epoch": 0.44466, + "grad_norm": 0.6834275126457214, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 44466 + }, + { + "epoch": 0.44467, + "grad_norm": 0.633813202381134, + "learning_rate": 0.003, + "loss": 3.987, + "step": 44467 + }, + { + "epoch": 0.44468, + "grad_norm": 0.6533637642860413, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 44468 + }, + { + "epoch": 0.44469, + "grad_norm": 0.6955873370170593, + "learning_rate": 0.003, + "loss": 3.961, + "step": 44469 + }, + { + "epoch": 0.4447, + "grad_norm": 0.7645760774612427, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 44470 + }, + { + "epoch": 0.44471, + "grad_norm": 0.8845504522323608, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 44471 + }, + { + "epoch": 0.44472, + "grad_norm": 1.0058104991912842, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 44472 + }, + { + "epoch": 0.44473, + "grad_norm": 1.1011508703231812, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 44473 + }, + { + "epoch": 0.44474, + "grad_norm": 1.0424724817276, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 44474 + }, + { + "epoch": 0.44475, + "grad_norm": 0.865973174571991, + "learning_rate": 0.003, + "loss": 3.988, + "step": 44475 + }, + { + "epoch": 0.44476, + "grad_norm": 0.7042096853256226, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 44476 + }, + { + "epoch": 0.44477, + "grad_norm": 0.6748999357223511, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 44477 + }, + { + "epoch": 0.44478, + "grad_norm": 0.7109447717666626, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 44478 + }, + { + "epoch": 0.44479, + "grad_norm": 1.050901174545288, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 44479 + }, + { + "epoch": 0.4448, + "grad_norm": 1.1612814664840698, + "learning_rate": 0.003, + "loss": 3.975, + "step": 44480 + }, + { + "epoch": 0.44481, + "grad_norm": 0.6459469199180603, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 44481 + }, + { + "epoch": 0.44482, + "grad_norm": 0.6500815153121948, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 44482 + }, + { + "epoch": 0.44483, + "grad_norm": 0.6713442206382751, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 44483 + }, + { + "epoch": 0.44484, + "grad_norm": 0.8181232213973999, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 44484 + }, + { + "epoch": 0.44485, + "grad_norm": 1.0138304233551025, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 44485 + }, + { + "epoch": 0.44486, + "grad_norm": 1.0291801691055298, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 44486 + }, + { + "epoch": 0.44487, + "grad_norm": 1.0339041948318481, + "learning_rate": 0.003, + "loss": 3.977, + "step": 44487 + }, + { + "epoch": 0.44488, + "grad_norm": 0.8998438715934753, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 44488 + }, + { + "epoch": 0.44489, + "grad_norm": 0.878582239151001, + "learning_rate": 0.003, + "loss": 3.974, + "step": 44489 + }, + { + "epoch": 0.4449, + "grad_norm": 0.8493329882621765, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 44490 + }, + { + "epoch": 0.44491, + "grad_norm": 0.9752879738807678, + "learning_rate": 0.003, + "loss": 3.987, + "step": 44491 + }, + { + "epoch": 0.44492, + "grad_norm": 1.0305696725845337, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 44492 + }, + { + "epoch": 0.44493, + "grad_norm": 0.9687813520431519, + "learning_rate": 0.003, + "loss": 3.979, + "step": 44493 + }, + { + "epoch": 0.44494, + "grad_norm": 1.0103965997695923, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 44494 + }, + { + "epoch": 0.44495, + "grad_norm": 1.129134178161621, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 44495 + }, + { + "epoch": 0.44496, + "grad_norm": 0.9270638823509216, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 44496 + }, + { + "epoch": 0.44497, + "grad_norm": 0.8589272499084473, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 44497 + }, + { + "epoch": 0.44498, + "grad_norm": 0.9079847931861877, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 44498 + }, + { + "epoch": 0.44499, + "grad_norm": 1.0245871543884277, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 44499 + }, + { + "epoch": 0.445, + "grad_norm": 1.283003568649292, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 44500 + }, + { + "epoch": 0.44501, + "grad_norm": 0.817699134349823, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 44501 + }, + { + "epoch": 0.44502, + "grad_norm": 0.7582300305366516, + "learning_rate": 0.003, + "loss": 3.987, + "step": 44502 + }, + { + "epoch": 0.44503, + "grad_norm": 0.7920848727226257, + "learning_rate": 0.003, + "loss": 3.976, + "step": 44503 + }, + { + "epoch": 0.44504, + "grad_norm": 0.8232836127281189, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 44504 + }, + { + "epoch": 0.44505, + "grad_norm": 0.8626220226287842, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 44505 + }, + { + "epoch": 0.44506, + "grad_norm": 1.0127943754196167, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 44506 + }, + { + "epoch": 0.44507, + "grad_norm": 1.2975841760635376, + "learning_rate": 0.003, + "loss": 4.015, + "step": 44507 + }, + { + "epoch": 0.44508, + "grad_norm": 0.8286346197128296, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 44508 + }, + { + "epoch": 0.44509, + "grad_norm": 0.7925612926483154, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 44509 + }, + { + "epoch": 0.4451, + "grad_norm": 0.9300445914268494, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 44510 + }, + { + "epoch": 0.44511, + "grad_norm": 1.565926432609558, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 44511 + }, + { + "epoch": 0.44512, + "grad_norm": 0.7296982407569885, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 44512 + }, + { + "epoch": 0.44513, + "grad_norm": 0.7745603919029236, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 44513 + }, + { + "epoch": 0.44514, + "grad_norm": 0.7054823637008667, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 44514 + }, + { + "epoch": 0.44515, + "grad_norm": 0.8002138137817383, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 44515 + }, + { + "epoch": 0.44516, + "grad_norm": 0.8598407506942749, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 44516 + }, + { + "epoch": 0.44517, + "grad_norm": 0.8100365400314331, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 44517 + }, + { + "epoch": 0.44518, + "grad_norm": 0.9587823748588562, + "learning_rate": 0.003, + "loss": 3.994, + "step": 44518 + }, + { + "epoch": 0.44519, + "grad_norm": 1.1613339185714722, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 44519 + }, + { + "epoch": 0.4452, + "grad_norm": 1.0412571430206299, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 44520 + }, + { + "epoch": 0.44521, + "grad_norm": 1.020684003829956, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 44521 + }, + { + "epoch": 0.44522, + "grad_norm": 1.1624794006347656, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 44522 + }, + { + "epoch": 0.44523, + "grad_norm": 0.7597525715827942, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 44523 + }, + { + "epoch": 0.44524, + "grad_norm": 0.6093997955322266, + "learning_rate": 0.003, + "loss": 3.962, + "step": 44524 + }, + { + "epoch": 0.44525, + "grad_norm": 0.7383041381835938, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 44525 + }, + { + "epoch": 0.44526, + "grad_norm": 0.8905645608901978, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 44526 + }, + { + "epoch": 0.44527, + "grad_norm": 1.0949313640594482, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 44527 + }, + { + "epoch": 0.44528, + "grad_norm": 0.8317951560020447, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 44528 + }, + { + "epoch": 0.44529, + "grad_norm": 0.7886450886726379, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 44529 + }, + { + "epoch": 0.4453, + "grad_norm": 0.8489924669265747, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 44530 + }, + { + "epoch": 0.44531, + "grad_norm": 0.9025286436080933, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 44531 + }, + { + "epoch": 0.44532, + "grad_norm": 0.7609220743179321, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 44532 + }, + { + "epoch": 0.44533, + "grad_norm": 0.6793159246444702, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 44533 + }, + { + "epoch": 0.44534, + "grad_norm": 0.670508086681366, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 44534 + }, + { + "epoch": 0.44535, + "grad_norm": 0.5869936943054199, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 44535 + }, + { + "epoch": 0.44536, + "grad_norm": 0.5879549384117126, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 44536 + }, + { + "epoch": 0.44537, + "grad_norm": 0.6145831346511841, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 44537 + }, + { + "epoch": 0.44538, + "grad_norm": 0.750596821308136, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 44538 + }, + { + "epoch": 0.44539, + "grad_norm": 0.9329666495323181, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 44539 + }, + { + "epoch": 0.4454, + "grad_norm": 1.136932134628296, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 44540 + }, + { + "epoch": 0.44541, + "grad_norm": 1.106589913368225, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 44541 + }, + { + "epoch": 0.44542, + "grad_norm": 1.0014533996582031, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 44542 + }, + { + "epoch": 0.44543, + "grad_norm": 1.093511700630188, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 44543 + }, + { + "epoch": 0.44544, + "grad_norm": 1.0621929168701172, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 44544 + }, + { + "epoch": 0.44545, + "grad_norm": 1.000299096107483, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 44545 + }, + { + "epoch": 0.44546, + "grad_norm": 1.0489147901535034, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 44546 + }, + { + "epoch": 0.44547, + "grad_norm": 1.0376784801483154, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 44547 + }, + { + "epoch": 0.44548, + "grad_norm": 0.9113484621047974, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 44548 + }, + { + "epoch": 0.44549, + "grad_norm": 0.9891790747642517, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 44549 + }, + { + "epoch": 0.4455, + "grad_norm": 0.9147935509681702, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 44550 + }, + { + "epoch": 0.44551, + "grad_norm": 0.7729972004890442, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 44551 + }, + { + "epoch": 0.44552, + "grad_norm": 0.8845308423042297, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 44552 + }, + { + "epoch": 0.44553, + "grad_norm": 0.9740173816680908, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 44553 + }, + { + "epoch": 0.44554, + "grad_norm": 1.027466893196106, + "learning_rate": 0.003, + "loss": 3.988, + "step": 44554 + }, + { + "epoch": 0.44555, + "grad_norm": 1.1806155443191528, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 44555 + }, + { + "epoch": 0.44556, + "grad_norm": 1.350146770477295, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 44556 + }, + { + "epoch": 0.44557, + "grad_norm": 0.8542736768722534, + "learning_rate": 0.003, + "loss": 3.962, + "step": 44557 + }, + { + "epoch": 0.44558, + "grad_norm": 0.7651491165161133, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 44558 + }, + { + "epoch": 0.44559, + "grad_norm": 0.8485015630722046, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 44559 + }, + { + "epoch": 0.4456, + "grad_norm": 1.1488137245178223, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 44560 + }, + { + "epoch": 0.44561, + "grad_norm": 0.884954571723938, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 44561 + }, + { + "epoch": 0.44562, + "grad_norm": 0.6965667605400085, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 44562 + }, + { + "epoch": 0.44563, + "grad_norm": 0.6597430109977722, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 44563 + }, + { + "epoch": 0.44564, + "grad_norm": 0.6094751358032227, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 44564 + }, + { + "epoch": 0.44565, + "grad_norm": 0.7358987331390381, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 44565 + }, + { + "epoch": 0.44566, + "grad_norm": 0.9023138284683228, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 44566 + }, + { + "epoch": 0.44567, + "grad_norm": 1.018548607826233, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 44567 + }, + { + "epoch": 0.44568, + "grad_norm": 1.081640601158142, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 44568 + }, + { + "epoch": 0.44569, + "grad_norm": 0.9476370215415955, + "learning_rate": 0.003, + "loss": 3.9466, + "step": 44569 + }, + { + "epoch": 0.4457, + "grad_norm": 1.1528518199920654, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 44570 + }, + { + "epoch": 0.44571, + "grad_norm": 0.922578752040863, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 44571 + }, + { + "epoch": 0.44572, + "grad_norm": 0.939450204372406, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 44572 + }, + { + "epoch": 0.44573, + "grad_norm": 1.0933175086975098, + "learning_rate": 0.003, + "loss": 3.977, + "step": 44573 + }, + { + "epoch": 0.44574, + "grad_norm": 0.8956562280654907, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 44574 + }, + { + "epoch": 0.44575, + "grad_norm": 0.8850609064102173, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 44575 + }, + { + "epoch": 0.44576, + "grad_norm": 1.0276436805725098, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 44576 + }, + { + "epoch": 0.44577, + "grad_norm": 1.5458184480667114, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 44577 + }, + { + "epoch": 0.44578, + "grad_norm": 0.5663265585899353, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 44578 + }, + { + "epoch": 0.44579, + "grad_norm": 0.9988062381744385, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 44579 + }, + { + "epoch": 0.4458, + "grad_norm": 1.390470266342163, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 44580 + }, + { + "epoch": 0.44581, + "grad_norm": 0.5625941753387451, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 44581 + }, + { + "epoch": 0.44582, + "grad_norm": 0.7227683067321777, + "learning_rate": 0.003, + "loss": 3.9427, + "step": 44582 + }, + { + "epoch": 0.44583, + "grad_norm": 0.7138300538063049, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 44583 + }, + { + "epoch": 0.44584, + "grad_norm": 0.8367415070533752, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 44584 + }, + { + "epoch": 0.44585, + "grad_norm": 0.7655129432678223, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 44585 + }, + { + "epoch": 0.44586, + "grad_norm": 0.8376452326774597, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 44586 + }, + { + "epoch": 0.44587, + "grad_norm": 1.0050994157791138, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 44587 + }, + { + "epoch": 0.44588, + "grad_norm": 0.9535852074623108, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 44588 + }, + { + "epoch": 0.44589, + "grad_norm": 0.9303200244903564, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 44589 + }, + { + "epoch": 0.4459, + "grad_norm": 1.2722790241241455, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 44590 + }, + { + "epoch": 0.44591, + "grad_norm": 1.044232964515686, + "learning_rate": 0.003, + "loss": 4.009, + "step": 44591 + }, + { + "epoch": 0.44592, + "grad_norm": 0.8406062126159668, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 44592 + }, + { + "epoch": 0.44593, + "grad_norm": 0.8277910351753235, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 44593 + }, + { + "epoch": 0.44594, + "grad_norm": 0.9769423007965088, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 44594 + }, + { + "epoch": 0.44595, + "grad_norm": 1.006815791130066, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 44595 + }, + { + "epoch": 0.44596, + "grad_norm": 0.9540996551513672, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 44596 + }, + { + "epoch": 0.44597, + "grad_norm": 0.9761879444122314, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 44597 + }, + { + "epoch": 0.44598, + "grad_norm": 0.8882029056549072, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 44598 + }, + { + "epoch": 0.44599, + "grad_norm": 0.6600901484489441, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 44599 + }, + { + "epoch": 0.446, + "grad_norm": 0.7014024257659912, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 44600 + }, + { + "epoch": 0.44601, + "grad_norm": 0.9154804348945618, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 44601 + }, + { + "epoch": 0.44602, + "grad_norm": 1.2234121561050415, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 44602 + }, + { + "epoch": 0.44603, + "grad_norm": 0.8172048330307007, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 44603 + }, + { + "epoch": 0.44604, + "grad_norm": 0.9512121081352234, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 44604 + }, + { + "epoch": 0.44605, + "grad_norm": 1.2794148921966553, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 44605 + }, + { + "epoch": 0.44606, + "grad_norm": 0.8033645749092102, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 44606 + }, + { + "epoch": 0.44607, + "grad_norm": 0.6622270345687866, + "learning_rate": 0.003, + "loss": 3.96, + "step": 44607 + }, + { + "epoch": 0.44608, + "grad_norm": 0.8282747864723206, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 44608 + }, + { + "epoch": 0.44609, + "grad_norm": 0.9706569314002991, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 44609 + }, + { + "epoch": 0.4461, + "grad_norm": 1.1691880226135254, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 44610 + }, + { + "epoch": 0.44611, + "grad_norm": 0.9339951276779175, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 44611 + }, + { + "epoch": 0.44612, + "grad_norm": 0.9655520915985107, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 44612 + }, + { + "epoch": 0.44613, + "grad_norm": 1.1666531562805176, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 44613 + }, + { + "epoch": 0.44614, + "grad_norm": 0.7038294076919556, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 44614 + }, + { + "epoch": 0.44615, + "grad_norm": 0.7786867618560791, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 44615 + }, + { + "epoch": 0.44616, + "grad_norm": 0.943291425704956, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 44616 + }, + { + "epoch": 0.44617, + "grad_norm": 0.8887774348258972, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 44617 + }, + { + "epoch": 0.44618, + "grad_norm": 1.1301921606063843, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 44618 + }, + { + "epoch": 0.44619, + "grad_norm": 0.9159966111183167, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 44619 + }, + { + "epoch": 0.4462, + "grad_norm": 0.8825109601020813, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 44620 + }, + { + "epoch": 0.44621, + "grad_norm": 1.027831792831421, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 44621 + }, + { + "epoch": 0.44622, + "grad_norm": 1.0327978134155273, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 44622 + }, + { + "epoch": 0.44623, + "grad_norm": 0.8897531628608704, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 44623 + }, + { + "epoch": 0.44624, + "grad_norm": 0.8779174089431763, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 44624 + }, + { + "epoch": 0.44625, + "grad_norm": 0.8231683969497681, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 44625 + }, + { + "epoch": 0.44626, + "grad_norm": 0.8637171983718872, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 44626 + }, + { + "epoch": 0.44627, + "grad_norm": 0.8554254770278931, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 44627 + }, + { + "epoch": 0.44628, + "grad_norm": 0.7051889896392822, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 44628 + }, + { + "epoch": 0.44629, + "grad_norm": 0.8855913877487183, + "learning_rate": 0.003, + "loss": 3.984, + "step": 44629 + }, + { + "epoch": 0.4463, + "grad_norm": 1.0070880651474, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 44630 + }, + { + "epoch": 0.44631, + "grad_norm": 0.9215108752250671, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 44631 + }, + { + "epoch": 0.44632, + "grad_norm": 0.8830123543739319, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 44632 + }, + { + "epoch": 0.44633, + "grad_norm": 1.122991919517517, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 44633 + }, + { + "epoch": 0.44634, + "grad_norm": 1.2611057758331299, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 44634 + }, + { + "epoch": 0.44635, + "grad_norm": 0.9136518836021423, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 44635 + }, + { + "epoch": 0.44636, + "grad_norm": 0.8937074542045593, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 44636 + }, + { + "epoch": 0.44637, + "grad_norm": 0.9938997626304626, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 44637 + }, + { + "epoch": 0.44638, + "grad_norm": 1.2814310789108276, + "learning_rate": 0.003, + "loss": 3.996, + "step": 44638 + }, + { + "epoch": 0.44639, + "grad_norm": 0.8145402669906616, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 44639 + }, + { + "epoch": 0.4464, + "grad_norm": 0.7898830771446228, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 44640 + }, + { + "epoch": 0.44641, + "grad_norm": 0.9716432094573975, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 44641 + }, + { + "epoch": 0.44642, + "grad_norm": 1.3780046701431274, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 44642 + }, + { + "epoch": 0.44643, + "grad_norm": 0.7407721281051636, + "learning_rate": 0.003, + "loss": 3.986, + "step": 44643 + }, + { + "epoch": 0.44644, + "grad_norm": 0.734647274017334, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 44644 + }, + { + "epoch": 0.44645, + "grad_norm": 0.8244205117225647, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 44645 + }, + { + "epoch": 0.44646, + "grad_norm": 0.7961056232452393, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 44646 + }, + { + "epoch": 0.44647, + "grad_norm": 0.9135098457336426, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 44647 + }, + { + "epoch": 0.44648, + "grad_norm": 0.9691751599311829, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 44648 + }, + { + "epoch": 0.44649, + "grad_norm": 0.9988661408424377, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 44649 + }, + { + "epoch": 0.4465, + "grad_norm": 1.2421480417251587, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 44650 + }, + { + "epoch": 0.44651, + "grad_norm": 1.0226938724517822, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 44651 + }, + { + "epoch": 0.44652, + "grad_norm": 0.9830834865570068, + "learning_rate": 0.003, + "loss": 3.993, + "step": 44652 + }, + { + "epoch": 0.44653, + "grad_norm": 0.8665895462036133, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 44653 + }, + { + "epoch": 0.44654, + "grad_norm": 0.8463974595069885, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 44654 + }, + { + "epoch": 0.44655, + "grad_norm": 0.8556477427482605, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 44655 + }, + { + "epoch": 0.44656, + "grad_norm": 0.8439270257949829, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 44656 + }, + { + "epoch": 0.44657, + "grad_norm": 1.05796217918396, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 44657 + }, + { + "epoch": 0.44658, + "grad_norm": 1.017052412033081, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 44658 + }, + { + "epoch": 0.44659, + "grad_norm": 0.9369574785232544, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 44659 + }, + { + "epoch": 0.4466, + "grad_norm": 0.8913695216178894, + "learning_rate": 0.003, + "loss": 3.981, + "step": 44660 + }, + { + "epoch": 0.44661, + "grad_norm": 0.9566924571990967, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 44661 + }, + { + "epoch": 0.44662, + "grad_norm": 0.8788356781005859, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 44662 + }, + { + "epoch": 0.44663, + "grad_norm": 0.780404269695282, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 44663 + }, + { + "epoch": 0.44664, + "grad_norm": 0.8939863443374634, + "learning_rate": 0.003, + "loss": 3.993, + "step": 44664 + }, + { + "epoch": 0.44665, + "grad_norm": 1.349134922027588, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 44665 + }, + { + "epoch": 0.44666, + "grad_norm": 0.7758780121803284, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 44666 + }, + { + "epoch": 0.44667, + "grad_norm": 0.6096723675727844, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 44667 + }, + { + "epoch": 0.44668, + "grad_norm": 0.5786939859390259, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 44668 + }, + { + "epoch": 0.44669, + "grad_norm": 0.7118347883224487, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 44669 + }, + { + "epoch": 0.4467, + "grad_norm": 1.0911544561386108, + "learning_rate": 0.003, + "loss": 3.9429, + "step": 44670 + }, + { + "epoch": 0.44671, + "grad_norm": 0.9313356280326843, + "learning_rate": 0.003, + "loss": 3.975, + "step": 44671 + }, + { + "epoch": 0.44672, + "grad_norm": 0.6997079849243164, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 44672 + }, + { + "epoch": 0.44673, + "grad_norm": 0.6999727487564087, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 44673 + }, + { + "epoch": 0.44674, + "grad_norm": 0.7439574003219604, + "learning_rate": 0.003, + "loss": 3.978, + "step": 44674 + }, + { + "epoch": 0.44675, + "grad_norm": 0.8681987524032593, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 44675 + }, + { + "epoch": 0.44676, + "grad_norm": 1.1202714443206787, + "learning_rate": 0.003, + "loss": 3.968, + "step": 44676 + }, + { + "epoch": 0.44677, + "grad_norm": 0.9083473086357117, + "learning_rate": 0.003, + "loss": 3.981, + "step": 44677 + }, + { + "epoch": 0.44678, + "grad_norm": 1.0416827201843262, + "learning_rate": 0.003, + "loss": 3.988, + "step": 44678 + }, + { + "epoch": 0.44679, + "grad_norm": 1.0427393913269043, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 44679 + }, + { + "epoch": 0.4468, + "grad_norm": 1.0616626739501953, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 44680 + }, + { + "epoch": 0.44681, + "grad_norm": 1.0779885053634644, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 44681 + }, + { + "epoch": 0.44682, + "grad_norm": 0.8299924731254578, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 44682 + }, + { + "epoch": 0.44683, + "grad_norm": 0.7789707183837891, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 44683 + }, + { + "epoch": 0.44684, + "grad_norm": 0.7104800343513489, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 44684 + }, + { + "epoch": 0.44685, + "grad_norm": 0.723558247089386, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 44685 + }, + { + "epoch": 0.44686, + "grad_norm": 0.7850685119628906, + "learning_rate": 0.003, + "loss": 3.953, + "step": 44686 + }, + { + "epoch": 0.44687, + "grad_norm": 0.8980189561843872, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 44687 + }, + { + "epoch": 0.44688, + "grad_norm": 0.9356865286827087, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 44688 + }, + { + "epoch": 0.44689, + "grad_norm": 0.8632552623748779, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 44689 + }, + { + "epoch": 0.4469, + "grad_norm": 0.8666059374809265, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 44690 + }, + { + "epoch": 0.44691, + "grad_norm": 0.911948025226593, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 44691 + }, + { + "epoch": 0.44692, + "grad_norm": 1.0014500617980957, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 44692 + }, + { + "epoch": 0.44693, + "grad_norm": 1.2095966339111328, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 44693 + }, + { + "epoch": 0.44694, + "grad_norm": 1.1266499757766724, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 44694 + }, + { + "epoch": 0.44695, + "grad_norm": 1.1136173009872437, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 44695 + }, + { + "epoch": 0.44696, + "grad_norm": 0.9053954482078552, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 44696 + }, + { + "epoch": 0.44697, + "grad_norm": 0.87062668800354, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 44697 + }, + { + "epoch": 0.44698, + "grad_norm": 1.0945205688476562, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 44698 + }, + { + "epoch": 0.44699, + "grad_norm": 1.1348044872283936, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 44699 + }, + { + "epoch": 0.447, + "grad_norm": 0.9920660257339478, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 44700 + }, + { + "epoch": 0.44701, + "grad_norm": 0.920738697052002, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 44701 + }, + { + "epoch": 0.44702, + "grad_norm": 0.86855149269104, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 44702 + }, + { + "epoch": 0.44703, + "grad_norm": 0.9468581080436707, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 44703 + }, + { + "epoch": 0.44704, + "grad_norm": 0.9692950248718262, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 44704 + }, + { + "epoch": 0.44705, + "grad_norm": 0.7823120951652527, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 44705 + }, + { + "epoch": 0.44706, + "grad_norm": 0.7347924113273621, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 44706 + }, + { + "epoch": 0.44707, + "grad_norm": 0.7394676804542542, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 44707 + }, + { + "epoch": 0.44708, + "grad_norm": 0.8376443982124329, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 44708 + }, + { + "epoch": 0.44709, + "grad_norm": 0.8176440596580505, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 44709 + }, + { + "epoch": 0.4471, + "grad_norm": 0.7441307902336121, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 44710 + }, + { + "epoch": 0.44711, + "grad_norm": 0.8031952381134033, + "learning_rate": 0.003, + "loss": 3.959, + "step": 44711 + }, + { + "epoch": 0.44712, + "grad_norm": 0.7081203460693359, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 44712 + }, + { + "epoch": 0.44713, + "grad_norm": 0.6546176671981812, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 44713 + }, + { + "epoch": 0.44714, + "grad_norm": 0.8492082357406616, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 44714 + }, + { + "epoch": 0.44715, + "grad_norm": 1.1705619096755981, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 44715 + }, + { + "epoch": 0.44716, + "grad_norm": 0.9664451479911804, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 44716 + }, + { + "epoch": 0.44717, + "grad_norm": 1.1572335958480835, + "learning_rate": 0.003, + "loss": 3.9509, + "step": 44717 + }, + { + "epoch": 0.44718, + "grad_norm": 0.9995100498199463, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 44718 + }, + { + "epoch": 0.44719, + "grad_norm": 1.1550489664077759, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 44719 + }, + { + "epoch": 0.4472, + "grad_norm": 0.7807186841964722, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 44720 + }, + { + "epoch": 0.44721, + "grad_norm": 0.6514760851860046, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 44721 + }, + { + "epoch": 0.44722, + "grad_norm": 0.7432729601860046, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 44722 + }, + { + "epoch": 0.44723, + "grad_norm": 0.9997511506080627, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 44723 + }, + { + "epoch": 0.44724, + "grad_norm": 1.2300188541412354, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 44724 + }, + { + "epoch": 0.44725, + "grad_norm": 0.6895608901977539, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 44725 + }, + { + "epoch": 0.44726, + "grad_norm": 0.790711522102356, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 44726 + }, + { + "epoch": 0.44727, + "grad_norm": 1.1652191877365112, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 44727 + }, + { + "epoch": 0.44728, + "grad_norm": 0.8242318034172058, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 44728 + }, + { + "epoch": 0.44729, + "grad_norm": 0.832302987575531, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 44729 + }, + { + "epoch": 0.4473, + "grad_norm": 0.9366464018821716, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 44730 + }, + { + "epoch": 0.44731, + "grad_norm": 1.2306370735168457, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 44731 + }, + { + "epoch": 0.44732, + "grad_norm": 1.0584193468093872, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 44732 + }, + { + "epoch": 0.44733, + "grad_norm": 0.8669975399971008, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 44733 + }, + { + "epoch": 0.44734, + "grad_norm": 0.8779062032699585, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 44734 + }, + { + "epoch": 0.44735, + "grad_norm": 0.962897002696991, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 44735 + }, + { + "epoch": 0.44736, + "grad_norm": 1.0617231130599976, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 44736 + }, + { + "epoch": 0.44737, + "grad_norm": 0.936908483505249, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 44737 + }, + { + "epoch": 0.44738, + "grad_norm": 0.7795748710632324, + "learning_rate": 0.003, + "loss": 4.008, + "step": 44738 + }, + { + "epoch": 0.44739, + "grad_norm": 0.8339201807975769, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 44739 + }, + { + "epoch": 0.4474, + "grad_norm": 0.8591006994247437, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 44740 + }, + { + "epoch": 0.44741, + "grad_norm": 0.9801487922668457, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 44741 + }, + { + "epoch": 0.44742, + "grad_norm": 1.0604809522628784, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 44742 + }, + { + "epoch": 0.44743, + "grad_norm": 0.8846076130867004, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 44743 + }, + { + "epoch": 0.44744, + "grad_norm": 0.8393065929412842, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 44744 + }, + { + "epoch": 0.44745, + "grad_norm": 0.865286648273468, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 44745 + }, + { + "epoch": 0.44746, + "grad_norm": 0.8450806140899658, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 44746 + }, + { + "epoch": 0.44747, + "grad_norm": 0.8106212019920349, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 44747 + }, + { + "epoch": 0.44748, + "grad_norm": 0.8347370028495789, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 44748 + }, + { + "epoch": 0.44749, + "grad_norm": 0.7887109518051147, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 44749 + }, + { + "epoch": 0.4475, + "grad_norm": 0.8354081511497498, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 44750 + }, + { + "epoch": 0.44751, + "grad_norm": 1.154615879058838, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 44751 + }, + { + "epoch": 0.44752, + "grad_norm": 1.0335111618041992, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 44752 + }, + { + "epoch": 0.44753, + "grad_norm": 1.1204177141189575, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 44753 + }, + { + "epoch": 0.44754, + "grad_norm": 0.9900500774383545, + "learning_rate": 0.003, + "loss": 3.976, + "step": 44754 + }, + { + "epoch": 0.44755, + "grad_norm": 1.0481688976287842, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 44755 + }, + { + "epoch": 0.44756, + "grad_norm": 1.1781967878341675, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 44756 + }, + { + "epoch": 0.44757, + "grad_norm": 0.8649969100952148, + "learning_rate": 0.003, + "loss": 3.999, + "step": 44757 + }, + { + "epoch": 0.44758, + "grad_norm": 0.8396841287612915, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 44758 + }, + { + "epoch": 0.44759, + "grad_norm": 0.8157240748405457, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 44759 + }, + { + "epoch": 0.4476, + "grad_norm": 0.8505784869194031, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 44760 + }, + { + "epoch": 0.44761, + "grad_norm": 1.0525237321853638, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 44761 + }, + { + "epoch": 0.44762, + "grad_norm": 0.9409040808677673, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 44762 + }, + { + "epoch": 0.44763, + "grad_norm": 1.0539486408233643, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 44763 + }, + { + "epoch": 0.44764, + "grad_norm": 0.9885321259498596, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 44764 + }, + { + "epoch": 0.44765, + "grad_norm": 0.8881905674934387, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 44765 + }, + { + "epoch": 0.44766, + "grad_norm": 0.9881160259246826, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 44766 + }, + { + "epoch": 0.44767, + "grad_norm": 0.9384725689888, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 44767 + }, + { + "epoch": 0.44768, + "grad_norm": 0.7060766816139221, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 44768 + }, + { + "epoch": 0.44769, + "grad_norm": 0.8441232442855835, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 44769 + }, + { + "epoch": 0.4477, + "grad_norm": 0.9363569021224976, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 44770 + }, + { + "epoch": 0.44771, + "grad_norm": 0.9909120202064514, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 44771 + }, + { + "epoch": 0.44772, + "grad_norm": 1.0362155437469482, + "learning_rate": 0.003, + "loss": 3.982, + "step": 44772 + }, + { + "epoch": 0.44773, + "grad_norm": 0.940926194190979, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 44773 + }, + { + "epoch": 0.44774, + "grad_norm": 0.8859384059906006, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 44774 + }, + { + "epoch": 0.44775, + "grad_norm": 0.8247135877609253, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 44775 + }, + { + "epoch": 0.44776, + "grad_norm": 0.6903958916664124, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 44776 + }, + { + "epoch": 0.44777, + "grad_norm": 0.6554846167564392, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 44777 + }, + { + "epoch": 0.44778, + "grad_norm": 0.7094935178756714, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 44778 + }, + { + "epoch": 0.44779, + "grad_norm": 0.8878201842308044, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 44779 + }, + { + "epoch": 0.4478, + "grad_norm": 1.0288331508636475, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 44780 + }, + { + "epoch": 0.44781, + "grad_norm": 0.8912065029144287, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 44781 + }, + { + "epoch": 0.44782, + "grad_norm": 0.8827957510948181, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 44782 + }, + { + "epoch": 0.44783, + "grad_norm": 0.9299799203872681, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 44783 + }, + { + "epoch": 0.44784, + "grad_norm": 1.118436574935913, + "learning_rate": 0.003, + "loss": 3.968, + "step": 44784 + }, + { + "epoch": 0.44785, + "grad_norm": 1.0695496797561646, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 44785 + }, + { + "epoch": 0.44786, + "grad_norm": 0.9689414501190186, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 44786 + }, + { + "epoch": 0.44787, + "grad_norm": 1.0234034061431885, + "learning_rate": 0.003, + "loss": 3.967, + "step": 44787 + }, + { + "epoch": 0.44788, + "grad_norm": 0.9907656311988831, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 44788 + }, + { + "epoch": 0.44789, + "grad_norm": 0.8504722118377686, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 44789 + }, + { + "epoch": 0.4479, + "grad_norm": 0.8744325637817383, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 44790 + }, + { + "epoch": 0.44791, + "grad_norm": 1.014357566833496, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 44791 + }, + { + "epoch": 0.44792, + "grad_norm": 1.1350069046020508, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 44792 + }, + { + "epoch": 0.44793, + "grad_norm": 1.0143980979919434, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 44793 + }, + { + "epoch": 0.44794, + "grad_norm": 0.999552309513092, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 44794 + }, + { + "epoch": 0.44795, + "grad_norm": 0.9359878301620483, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 44795 + }, + { + "epoch": 0.44796, + "grad_norm": 0.9864853620529175, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 44796 + }, + { + "epoch": 0.44797, + "grad_norm": 0.9148315787315369, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 44797 + }, + { + "epoch": 0.44798, + "grad_norm": 0.8312126398086548, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 44798 + }, + { + "epoch": 0.44799, + "grad_norm": 0.9300817251205444, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 44799 + }, + { + "epoch": 0.448, + "grad_norm": 0.8763965964317322, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 44800 + }, + { + "epoch": 0.44801, + "grad_norm": 0.8414208292961121, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 44801 + }, + { + "epoch": 0.44802, + "grad_norm": 0.7611011266708374, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 44802 + }, + { + "epoch": 0.44803, + "grad_norm": 0.8875378370285034, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 44803 + }, + { + "epoch": 0.44804, + "grad_norm": 1.2294268608093262, + "learning_rate": 0.003, + "loss": 3.9324, + "step": 44804 + }, + { + "epoch": 0.44805, + "grad_norm": 0.8393673896789551, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 44805 + }, + { + "epoch": 0.44806, + "grad_norm": 0.8930830359458923, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 44806 + }, + { + "epoch": 0.44807, + "grad_norm": 1.2541948556900024, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 44807 + }, + { + "epoch": 0.44808, + "grad_norm": 1.1562105417251587, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 44808 + }, + { + "epoch": 0.44809, + "grad_norm": 0.7352297902107239, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 44809 + }, + { + "epoch": 0.4481, + "grad_norm": 0.7417208552360535, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 44810 + }, + { + "epoch": 0.44811, + "grad_norm": 0.8521338701248169, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 44811 + }, + { + "epoch": 0.44812, + "grad_norm": 0.9717651009559631, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 44812 + }, + { + "epoch": 0.44813, + "grad_norm": 1.1716629266738892, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 44813 + }, + { + "epoch": 0.44814, + "grad_norm": 0.9906665682792664, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 44814 + }, + { + "epoch": 0.44815, + "grad_norm": 1.0114476680755615, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 44815 + }, + { + "epoch": 0.44816, + "grad_norm": 0.9852926135063171, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 44816 + }, + { + "epoch": 0.44817, + "grad_norm": 1.0189807415008545, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 44817 + }, + { + "epoch": 0.44818, + "grad_norm": 0.871657133102417, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 44818 + }, + { + "epoch": 0.44819, + "grad_norm": 0.7961887121200562, + "learning_rate": 0.003, + "loss": 3.9357, + "step": 44819 + }, + { + "epoch": 0.4482, + "grad_norm": 0.8541828393936157, + "learning_rate": 0.003, + "loss": 4.006, + "step": 44820 + }, + { + "epoch": 0.44821, + "grad_norm": 0.9019765853881836, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 44821 + }, + { + "epoch": 0.44822, + "grad_norm": 1.0016891956329346, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 44822 + }, + { + "epoch": 0.44823, + "grad_norm": 1.05348801612854, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 44823 + }, + { + "epoch": 0.44824, + "grad_norm": 1.099952220916748, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 44824 + }, + { + "epoch": 0.44825, + "grad_norm": 1.3150978088378906, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 44825 + }, + { + "epoch": 0.44826, + "grad_norm": 0.9311406016349792, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 44826 + }, + { + "epoch": 0.44827, + "grad_norm": 0.8730924725532532, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 44827 + }, + { + "epoch": 0.44828, + "grad_norm": 0.8949961066246033, + "learning_rate": 0.003, + "loss": 3.981, + "step": 44828 + }, + { + "epoch": 0.44829, + "grad_norm": 0.9194933176040649, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 44829 + }, + { + "epoch": 0.4483, + "grad_norm": 0.885139524936676, + "learning_rate": 0.003, + "loss": 3.966, + "step": 44830 + }, + { + "epoch": 0.44831, + "grad_norm": 1.0490236282348633, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 44831 + }, + { + "epoch": 0.44832, + "grad_norm": 0.9799883365631104, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 44832 + }, + { + "epoch": 0.44833, + "grad_norm": 1.0785952806472778, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 44833 + }, + { + "epoch": 0.44834, + "grad_norm": 0.8994379043579102, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 44834 + }, + { + "epoch": 0.44835, + "grad_norm": 0.6596785187721252, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 44835 + }, + { + "epoch": 0.44836, + "grad_norm": 0.7927867770195007, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 44836 + }, + { + "epoch": 0.44837, + "grad_norm": 0.8599086403846741, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 44837 + }, + { + "epoch": 0.44838, + "grad_norm": 0.8519018292427063, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 44838 + }, + { + "epoch": 0.44839, + "grad_norm": 0.7776197195053101, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 44839 + }, + { + "epoch": 0.4484, + "grad_norm": 0.7769135236740112, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 44840 + }, + { + "epoch": 0.44841, + "grad_norm": 0.8690266609191895, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 44841 + }, + { + "epoch": 0.44842, + "grad_norm": 1.0517929792404175, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 44842 + }, + { + "epoch": 0.44843, + "grad_norm": 1.0557321310043335, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 44843 + }, + { + "epoch": 0.44844, + "grad_norm": 0.9761556386947632, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 44844 + }, + { + "epoch": 0.44845, + "grad_norm": 0.9496451616287231, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 44845 + }, + { + "epoch": 0.44846, + "grad_norm": 0.9537192583084106, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 44846 + }, + { + "epoch": 0.44847, + "grad_norm": 1.03184175491333, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 44847 + }, + { + "epoch": 0.44848, + "grad_norm": 1.0608636140823364, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 44848 + }, + { + "epoch": 0.44849, + "grad_norm": 1.3125985860824585, + "learning_rate": 0.003, + "loss": 3.97, + "step": 44849 + }, + { + "epoch": 0.4485, + "grad_norm": 0.8909127116203308, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 44850 + }, + { + "epoch": 0.44851, + "grad_norm": 0.7346240282058716, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 44851 + }, + { + "epoch": 0.44852, + "grad_norm": 0.7669847011566162, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 44852 + }, + { + "epoch": 0.44853, + "grad_norm": 0.9103257060050964, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 44853 + }, + { + "epoch": 0.44854, + "grad_norm": 1.0459786653518677, + "learning_rate": 0.003, + "loss": 3.997, + "step": 44854 + }, + { + "epoch": 0.44855, + "grad_norm": 0.8070138096809387, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 44855 + }, + { + "epoch": 0.44856, + "grad_norm": 0.8022794723510742, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 44856 + }, + { + "epoch": 0.44857, + "grad_norm": 1.0614567995071411, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 44857 + }, + { + "epoch": 0.44858, + "grad_norm": 1.2592991590499878, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 44858 + }, + { + "epoch": 0.44859, + "grad_norm": 0.9432116746902466, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 44859 + }, + { + "epoch": 0.4486, + "grad_norm": 0.9938392639160156, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 44860 + }, + { + "epoch": 0.44861, + "grad_norm": 1.0446131229400635, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 44861 + }, + { + "epoch": 0.44862, + "grad_norm": 0.8070535659790039, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 44862 + }, + { + "epoch": 0.44863, + "grad_norm": 0.6207202672958374, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 44863 + }, + { + "epoch": 0.44864, + "grad_norm": 0.7163159847259521, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 44864 + }, + { + "epoch": 0.44865, + "grad_norm": 0.7785969376564026, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 44865 + }, + { + "epoch": 0.44866, + "grad_norm": 0.8254164457321167, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 44866 + }, + { + "epoch": 0.44867, + "grad_norm": 0.9988511204719543, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 44867 + }, + { + "epoch": 0.44868, + "grad_norm": 1.0123451948165894, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 44868 + }, + { + "epoch": 0.44869, + "grad_norm": 0.8255390524864197, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 44869 + }, + { + "epoch": 0.4487, + "grad_norm": 0.8300715088844299, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 44870 + }, + { + "epoch": 0.44871, + "grad_norm": 0.9435597062110901, + "learning_rate": 0.003, + "loss": 3.948, + "step": 44871 + }, + { + "epoch": 0.44872, + "grad_norm": 0.9462621212005615, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 44872 + }, + { + "epoch": 0.44873, + "grad_norm": 0.9463095664978027, + "learning_rate": 0.003, + "loss": 3.966, + "step": 44873 + }, + { + "epoch": 0.44874, + "grad_norm": 1.0810062885284424, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 44874 + }, + { + "epoch": 0.44875, + "grad_norm": 0.85887211561203, + "learning_rate": 0.003, + "loss": 3.9414, + "step": 44875 + }, + { + "epoch": 0.44876, + "grad_norm": 0.7380704879760742, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 44876 + }, + { + "epoch": 0.44877, + "grad_norm": 0.670258104801178, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 44877 + }, + { + "epoch": 0.44878, + "grad_norm": 0.7541056275367737, + "learning_rate": 0.003, + "loss": 3.971, + "step": 44878 + }, + { + "epoch": 0.44879, + "grad_norm": 0.8325194120407104, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 44879 + }, + { + "epoch": 0.4488, + "grad_norm": 0.8778925538063049, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 44880 + }, + { + "epoch": 0.44881, + "grad_norm": 1.0074959993362427, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 44881 + }, + { + "epoch": 0.44882, + "grad_norm": 1.0868217945098877, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 44882 + }, + { + "epoch": 0.44883, + "grad_norm": 1.1722513437271118, + "learning_rate": 0.003, + "loss": 3.9405, + "step": 44883 + }, + { + "epoch": 0.44884, + "grad_norm": 0.8309535980224609, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 44884 + }, + { + "epoch": 0.44885, + "grad_norm": 0.8251326680183411, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 44885 + }, + { + "epoch": 0.44886, + "grad_norm": 0.7748850584030151, + "learning_rate": 0.003, + "loss": 3.975, + "step": 44886 + }, + { + "epoch": 0.44887, + "grad_norm": 0.7786267399787903, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 44887 + }, + { + "epoch": 0.44888, + "grad_norm": 0.9202545881271362, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 44888 + }, + { + "epoch": 0.44889, + "grad_norm": 1.0441312789916992, + "learning_rate": 0.003, + "loss": 3.979, + "step": 44889 + }, + { + "epoch": 0.4489, + "grad_norm": 1.0215725898742676, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 44890 + }, + { + "epoch": 0.44891, + "grad_norm": 1.011085033416748, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 44891 + }, + { + "epoch": 0.44892, + "grad_norm": 0.9678300023078918, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 44892 + }, + { + "epoch": 0.44893, + "grad_norm": 1.0145901441574097, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 44893 + }, + { + "epoch": 0.44894, + "grad_norm": 0.9206352233886719, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 44894 + }, + { + "epoch": 0.44895, + "grad_norm": 0.8085443377494812, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 44895 + }, + { + "epoch": 0.44896, + "grad_norm": 0.8631532192230225, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 44896 + }, + { + "epoch": 0.44897, + "grad_norm": 0.8574771285057068, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 44897 + }, + { + "epoch": 0.44898, + "grad_norm": 0.637731671333313, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 44898 + }, + { + "epoch": 0.44899, + "grad_norm": 0.6763496994972229, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 44899 + }, + { + "epoch": 0.449, + "grad_norm": 0.6295950412750244, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 44900 + }, + { + "epoch": 0.44901, + "grad_norm": 0.5879191160202026, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 44901 + }, + { + "epoch": 0.44902, + "grad_norm": 0.8466842174530029, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 44902 + }, + { + "epoch": 0.44903, + "grad_norm": 1.3430055379867554, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 44903 + }, + { + "epoch": 0.44904, + "grad_norm": 1.0908278226852417, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 44904 + }, + { + "epoch": 0.44905, + "grad_norm": 0.8868041038513184, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 44905 + }, + { + "epoch": 0.44906, + "grad_norm": 0.7221826910972595, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 44906 + }, + { + "epoch": 0.44907, + "grad_norm": 0.6795113682746887, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 44907 + }, + { + "epoch": 0.44908, + "grad_norm": 0.5688458681106567, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 44908 + }, + { + "epoch": 0.44909, + "grad_norm": 0.6943160891532898, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 44909 + }, + { + "epoch": 0.4491, + "grad_norm": 0.834735631942749, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 44910 + }, + { + "epoch": 0.44911, + "grad_norm": 0.9758912324905396, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 44911 + }, + { + "epoch": 0.44912, + "grad_norm": 1.1111544370651245, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 44912 + }, + { + "epoch": 0.44913, + "grad_norm": 1.0321651697158813, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 44913 + }, + { + "epoch": 0.44914, + "grad_norm": 0.9913781881332397, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 44914 + }, + { + "epoch": 0.44915, + "grad_norm": 1.0030970573425293, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 44915 + }, + { + "epoch": 0.44916, + "grad_norm": 1.0113565921783447, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 44916 + }, + { + "epoch": 0.44917, + "grad_norm": 0.9760904908180237, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 44917 + }, + { + "epoch": 0.44918, + "grad_norm": 0.8463712930679321, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 44918 + }, + { + "epoch": 0.44919, + "grad_norm": 0.7575200796127319, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 44919 + }, + { + "epoch": 0.4492, + "grad_norm": 0.8017780184745789, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 44920 + }, + { + "epoch": 0.44921, + "grad_norm": 0.8097286820411682, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 44921 + }, + { + "epoch": 0.44922, + "grad_norm": 0.8800566792488098, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 44922 + }, + { + "epoch": 0.44923, + "grad_norm": 0.9772855043411255, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 44923 + }, + { + "epoch": 0.44924, + "grad_norm": 1.146628737449646, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 44924 + }, + { + "epoch": 0.44925, + "grad_norm": 0.8013148903846741, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 44925 + }, + { + "epoch": 0.44926, + "grad_norm": 0.7524957060813904, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 44926 + }, + { + "epoch": 0.44927, + "grad_norm": 0.8385912775993347, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 44927 + }, + { + "epoch": 0.44928, + "grad_norm": 1.0075639486312866, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 44928 + }, + { + "epoch": 0.44929, + "grad_norm": 1.1231238842010498, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 44929 + }, + { + "epoch": 0.4493, + "grad_norm": 0.9446671009063721, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 44930 + }, + { + "epoch": 0.44931, + "grad_norm": 0.9452259540557861, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 44931 + }, + { + "epoch": 0.44932, + "grad_norm": 1.03914213180542, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 44932 + }, + { + "epoch": 0.44933, + "grad_norm": 0.9096106290817261, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 44933 + }, + { + "epoch": 0.44934, + "grad_norm": 0.8274676203727722, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 44934 + }, + { + "epoch": 0.44935, + "grad_norm": 0.8455817699432373, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 44935 + }, + { + "epoch": 0.44936, + "grad_norm": 0.752916693687439, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 44936 + }, + { + "epoch": 0.44937, + "grad_norm": 0.8296074867248535, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 44937 + }, + { + "epoch": 0.44938, + "grad_norm": 1.006554126739502, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 44938 + }, + { + "epoch": 0.44939, + "grad_norm": 1.3440359830856323, + "learning_rate": 0.003, + "loss": 3.996, + "step": 44939 + }, + { + "epoch": 0.4494, + "grad_norm": 0.564635694026947, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 44940 + }, + { + "epoch": 0.44941, + "grad_norm": 0.8565191030502319, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 44941 + }, + { + "epoch": 0.44942, + "grad_norm": 1.0262489318847656, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 44942 + }, + { + "epoch": 0.44943, + "grad_norm": 1.0930684804916382, + "learning_rate": 0.003, + "loss": 3.956, + "step": 44943 + }, + { + "epoch": 0.44944, + "grad_norm": 0.9151251316070557, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 44944 + }, + { + "epoch": 0.44945, + "grad_norm": 0.7188399434089661, + "learning_rate": 0.003, + "loss": 3.947, + "step": 44945 + }, + { + "epoch": 0.44946, + "grad_norm": 0.6848035454750061, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 44946 + }, + { + "epoch": 0.44947, + "grad_norm": 0.7399921417236328, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 44947 + }, + { + "epoch": 0.44948, + "grad_norm": 0.8683335185050964, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 44948 + }, + { + "epoch": 0.44949, + "grad_norm": 1.04066801071167, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 44949 + }, + { + "epoch": 0.4495, + "grad_norm": 0.9336613416671753, + "learning_rate": 0.003, + "loss": 4.003, + "step": 44950 + }, + { + "epoch": 0.44951, + "grad_norm": 0.8572694063186646, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 44951 + }, + { + "epoch": 0.44952, + "grad_norm": 0.9085745215415955, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 44952 + }, + { + "epoch": 0.44953, + "grad_norm": 1.0416200160980225, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 44953 + }, + { + "epoch": 0.44954, + "grad_norm": 0.9113114476203918, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 44954 + }, + { + "epoch": 0.44955, + "grad_norm": 0.8604607582092285, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 44955 + }, + { + "epoch": 0.44956, + "grad_norm": 0.8284346461296082, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 44956 + }, + { + "epoch": 0.44957, + "grad_norm": 0.8520637154579163, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 44957 + }, + { + "epoch": 0.44958, + "grad_norm": 0.9369731545448303, + "learning_rate": 0.003, + "loss": 3.998, + "step": 44958 + }, + { + "epoch": 0.44959, + "grad_norm": 1.0451765060424805, + "learning_rate": 0.003, + "loss": 3.974, + "step": 44959 + }, + { + "epoch": 0.4496, + "grad_norm": 1.1200469732284546, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 44960 + }, + { + "epoch": 0.44961, + "grad_norm": 0.9837359189987183, + "learning_rate": 0.003, + "loss": 3.9483, + "step": 44961 + }, + { + "epoch": 0.44962, + "grad_norm": 0.8592307567596436, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 44962 + }, + { + "epoch": 0.44963, + "grad_norm": 0.7459734678268433, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 44963 + }, + { + "epoch": 0.44964, + "grad_norm": 0.8041787147521973, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 44964 + }, + { + "epoch": 0.44965, + "grad_norm": 1.1567504405975342, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 44965 + }, + { + "epoch": 0.44966, + "grad_norm": 1.071608304977417, + "learning_rate": 0.003, + "loss": 3.989, + "step": 44966 + }, + { + "epoch": 0.44967, + "grad_norm": 1.0299943685531616, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 44967 + }, + { + "epoch": 0.44968, + "grad_norm": 1.087996244430542, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 44968 + }, + { + "epoch": 0.44969, + "grad_norm": 0.9358202219009399, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 44969 + }, + { + "epoch": 0.4497, + "grad_norm": 1.0172349214553833, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 44970 + }, + { + "epoch": 0.44971, + "grad_norm": 1.2105077505111694, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 44971 + }, + { + "epoch": 0.44972, + "grad_norm": 1.0257723331451416, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 44972 + }, + { + "epoch": 0.44973, + "grad_norm": 1.044434905052185, + "learning_rate": 0.003, + "loss": 4.018, + "step": 44973 + }, + { + "epoch": 0.44974, + "grad_norm": 0.8227018117904663, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 44974 + }, + { + "epoch": 0.44975, + "grad_norm": 0.8176344037055969, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 44975 + }, + { + "epoch": 0.44976, + "grad_norm": 0.797510027885437, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 44976 + }, + { + "epoch": 0.44977, + "grad_norm": 0.7354219555854797, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 44977 + }, + { + "epoch": 0.44978, + "grad_norm": 0.789667010307312, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 44978 + }, + { + "epoch": 0.44979, + "grad_norm": 0.7603052854537964, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 44979 + }, + { + "epoch": 0.4498, + "grad_norm": 0.7153218388557434, + "learning_rate": 0.003, + "loss": 3.9318, + "step": 44980 + }, + { + "epoch": 0.44981, + "grad_norm": 0.7618704438209534, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 44981 + }, + { + "epoch": 0.44982, + "grad_norm": 0.7152484655380249, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 44982 + }, + { + "epoch": 0.44983, + "grad_norm": 0.5847945213317871, + "learning_rate": 0.003, + "loss": 3.96, + "step": 44983 + }, + { + "epoch": 0.44984, + "grad_norm": 0.6551292538642883, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 44984 + }, + { + "epoch": 0.44985, + "grad_norm": 0.9625272154808044, + "learning_rate": 0.003, + "loss": 3.995, + "step": 44985 + }, + { + "epoch": 0.44986, + "grad_norm": 1.3704675436019897, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 44986 + }, + { + "epoch": 0.44987, + "grad_norm": 0.8068509101867676, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 44987 + }, + { + "epoch": 0.44988, + "grad_norm": 0.7287896871566772, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 44988 + }, + { + "epoch": 0.44989, + "grad_norm": 0.7363588809967041, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 44989 + }, + { + "epoch": 0.4499, + "grad_norm": 0.8083580732345581, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 44990 + }, + { + "epoch": 0.44991, + "grad_norm": 0.8677781224250793, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 44991 + }, + { + "epoch": 0.44992, + "grad_norm": 0.904723048210144, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 44992 + }, + { + "epoch": 0.44993, + "grad_norm": 0.918005645275116, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 44993 + }, + { + "epoch": 0.44994, + "grad_norm": 0.9315648078918457, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 44994 + }, + { + "epoch": 0.44995, + "grad_norm": 1.015863299369812, + "learning_rate": 0.003, + "loss": 3.9395, + "step": 44995 + }, + { + "epoch": 0.44996, + "grad_norm": 0.9945705533027649, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 44996 + }, + { + "epoch": 0.44997, + "grad_norm": 0.790140688419342, + "learning_rate": 0.003, + "loss": 3.992, + "step": 44997 + }, + { + "epoch": 0.44998, + "grad_norm": 0.7277364134788513, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 44998 + }, + { + "epoch": 0.44999, + "grad_norm": 0.7653185129165649, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 44999 + }, + { + "epoch": 0.45, + "grad_norm": 0.9192693829536438, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 45000 + }, + { + "epoch": 0.45001, + "grad_norm": 0.8545812964439392, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 45001 + }, + { + "epoch": 0.45002, + "grad_norm": 0.7883837819099426, + "learning_rate": 0.003, + "loss": 3.971, + "step": 45002 + }, + { + "epoch": 0.45003, + "grad_norm": 0.8759836554527283, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 45003 + }, + { + "epoch": 0.45004, + "grad_norm": 1.0792405605316162, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 45004 + }, + { + "epoch": 0.45005, + "grad_norm": 1.193749189376831, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 45005 + }, + { + "epoch": 0.45006, + "grad_norm": 1.0548902750015259, + "learning_rate": 0.003, + "loss": 3.9124, + "step": 45006 + }, + { + "epoch": 0.45007, + "grad_norm": 1.0829657316207886, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 45007 + }, + { + "epoch": 0.45008, + "grad_norm": 0.8910230994224548, + "learning_rate": 0.003, + "loss": 3.971, + "step": 45008 + }, + { + "epoch": 0.45009, + "grad_norm": 0.8002058863639832, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 45009 + }, + { + "epoch": 0.4501, + "grad_norm": 0.743923544883728, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 45010 + }, + { + "epoch": 0.45011, + "grad_norm": 0.8467406630516052, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 45011 + }, + { + "epoch": 0.45012, + "grad_norm": 0.8928271532058716, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 45012 + }, + { + "epoch": 0.45013, + "grad_norm": 1.0330605506896973, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 45013 + }, + { + "epoch": 0.45014, + "grad_norm": 1.088549256324768, + "learning_rate": 0.003, + "loss": 4.01, + "step": 45014 + }, + { + "epoch": 0.45015, + "grad_norm": 1.0061839818954468, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 45015 + }, + { + "epoch": 0.45016, + "grad_norm": 1.0677131414413452, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 45016 + }, + { + "epoch": 0.45017, + "grad_norm": 1.0608493089675903, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 45017 + }, + { + "epoch": 0.45018, + "grad_norm": 1.2907543182373047, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 45018 + }, + { + "epoch": 0.45019, + "grad_norm": 0.770134687423706, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 45019 + }, + { + "epoch": 0.4502, + "grad_norm": 0.8037226796150208, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 45020 + }, + { + "epoch": 0.45021, + "grad_norm": 0.8517873883247375, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 45021 + }, + { + "epoch": 0.45022, + "grad_norm": 0.7579272389411926, + "learning_rate": 0.003, + "loss": 3.995, + "step": 45022 + }, + { + "epoch": 0.45023, + "grad_norm": 0.9100149273872375, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 45023 + }, + { + "epoch": 0.45024, + "grad_norm": 1.0054564476013184, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 45024 + }, + { + "epoch": 0.45025, + "grad_norm": 1.2164839506149292, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 45025 + }, + { + "epoch": 0.45026, + "grad_norm": 1.0957406759262085, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 45026 + }, + { + "epoch": 0.45027, + "grad_norm": 0.9998533129692078, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 45027 + }, + { + "epoch": 0.45028, + "grad_norm": 1.1273167133331299, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 45028 + }, + { + "epoch": 0.45029, + "grad_norm": 0.8801307678222656, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 45029 + }, + { + "epoch": 0.4503, + "grad_norm": 0.7398054003715515, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 45030 + }, + { + "epoch": 0.45031, + "grad_norm": 0.7697806358337402, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 45031 + }, + { + "epoch": 0.45032, + "grad_norm": 0.9361408352851868, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 45032 + }, + { + "epoch": 0.45033, + "grad_norm": 1.0893903970718384, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 45033 + }, + { + "epoch": 0.45034, + "grad_norm": 1.0736430883407593, + "learning_rate": 0.003, + "loss": 3.983, + "step": 45034 + }, + { + "epoch": 0.45035, + "grad_norm": 0.8668400645256042, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 45035 + }, + { + "epoch": 0.45036, + "grad_norm": 0.7337248921394348, + "learning_rate": 0.003, + "loss": 3.973, + "step": 45036 + }, + { + "epoch": 0.45037, + "grad_norm": 0.87526535987854, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 45037 + }, + { + "epoch": 0.45038, + "grad_norm": 0.9548709988594055, + "learning_rate": 0.003, + "loss": 3.9424, + "step": 45038 + }, + { + "epoch": 0.45039, + "grad_norm": 1.0423647165298462, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 45039 + }, + { + "epoch": 0.4504, + "grad_norm": 0.993971586227417, + "learning_rate": 0.003, + "loss": 3.977, + "step": 45040 + }, + { + "epoch": 0.45041, + "grad_norm": 1.1291282176971436, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 45041 + }, + { + "epoch": 0.45042, + "grad_norm": 0.7619085311889648, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 45042 + }, + { + "epoch": 0.45043, + "grad_norm": 0.7034945487976074, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 45043 + }, + { + "epoch": 0.45044, + "grad_norm": 0.7129697799682617, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 45044 + }, + { + "epoch": 0.45045, + "grad_norm": 0.7863544225692749, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 45045 + }, + { + "epoch": 0.45046, + "grad_norm": 0.9099591374397278, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 45046 + }, + { + "epoch": 0.45047, + "grad_norm": 0.9643387198448181, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 45047 + }, + { + "epoch": 0.45048, + "grad_norm": 0.9491651654243469, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 45048 + }, + { + "epoch": 0.45049, + "grad_norm": 0.9426524639129639, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 45049 + }, + { + "epoch": 0.4505, + "grad_norm": 0.9509718418121338, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 45050 + }, + { + "epoch": 0.45051, + "grad_norm": 0.8659954071044922, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 45051 + }, + { + "epoch": 0.45052, + "grad_norm": 0.8450005650520325, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 45052 + }, + { + "epoch": 0.45053, + "grad_norm": 0.7962210178375244, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 45053 + }, + { + "epoch": 0.45054, + "grad_norm": 0.873115062713623, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 45054 + }, + { + "epoch": 0.45055, + "grad_norm": 1.0518996715545654, + "learning_rate": 0.003, + "loss": 3.998, + "step": 45055 + }, + { + "epoch": 0.45056, + "grad_norm": 1.1155763864517212, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 45056 + }, + { + "epoch": 0.45057, + "grad_norm": 0.6846901774406433, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 45057 + }, + { + "epoch": 0.45058, + "grad_norm": 0.6501873135566711, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 45058 + }, + { + "epoch": 0.45059, + "grad_norm": 0.6762736439704895, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 45059 + }, + { + "epoch": 0.4506, + "grad_norm": 0.828428328037262, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 45060 + }, + { + "epoch": 0.45061, + "grad_norm": 0.9071407914161682, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 45061 + }, + { + "epoch": 0.45062, + "grad_norm": 0.8885080814361572, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 45062 + }, + { + "epoch": 0.45063, + "grad_norm": 0.9822176694869995, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 45063 + }, + { + "epoch": 0.45064, + "grad_norm": 1.1114752292633057, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 45064 + }, + { + "epoch": 0.45065, + "grad_norm": 0.9975486993789673, + "learning_rate": 0.003, + "loss": 4.011, + "step": 45065 + }, + { + "epoch": 0.45066, + "grad_norm": 1.2404099702835083, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 45066 + }, + { + "epoch": 0.45067, + "grad_norm": 0.7858612537384033, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 45067 + }, + { + "epoch": 0.45068, + "grad_norm": 0.7624179720878601, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 45068 + }, + { + "epoch": 0.45069, + "grad_norm": 0.7950817942619324, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 45069 + }, + { + "epoch": 0.4507, + "grad_norm": 0.9843034744262695, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 45070 + }, + { + "epoch": 0.45071, + "grad_norm": 1.182057499885559, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 45071 + }, + { + "epoch": 0.45072, + "grad_norm": 0.998881995677948, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 45072 + }, + { + "epoch": 0.45073, + "grad_norm": 1.0881168842315674, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 45073 + }, + { + "epoch": 0.45074, + "grad_norm": 0.9206922054290771, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 45074 + }, + { + "epoch": 0.45075, + "grad_norm": 0.9072726368904114, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 45075 + }, + { + "epoch": 0.45076, + "grad_norm": 1.073904275894165, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 45076 + }, + { + "epoch": 0.45077, + "grad_norm": 0.7435979843139648, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 45077 + }, + { + "epoch": 0.45078, + "grad_norm": 0.700943648815155, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 45078 + }, + { + "epoch": 0.45079, + "grad_norm": 0.6981772780418396, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 45079 + }, + { + "epoch": 0.4508, + "grad_norm": 0.8375236392021179, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 45080 + }, + { + "epoch": 0.45081, + "grad_norm": 0.9599486589431763, + "learning_rate": 0.003, + "loss": 3.974, + "step": 45081 + }, + { + "epoch": 0.45082, + "grad_norm": 1.0745904445648193, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 45082 + }, + { + "epoch": 0.45083, + "grad_norm": 1.0179016590118408, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 45083 + }, + { + "epoch": 0.45084, + "grad_norm": 0.9996493458747864, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 45084 + }, + { + "epoch": 0.45085, + "grad_norm": 0.8120477795600891, + "learning_rate": 0.003, + "loss": 3.9339, + "step": 45085 + }, + { + "epoch": 0.45086, + "grad_norm": 0.8751059770584106, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 45086 + }, + { + "epoch": 0.45087, + "grad_norm": 0.8359867930412292, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 45087 + }, + { + "epoch": 0.45088, + "grad_norm": 0.7823106646537781, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 45088 + }, + { + "epoch": 0.45089, + "grad_norm": 0.8767691254615784, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 45089 + }, + { + "epoch": 0.4509, + "grad_norm": 1.0569897890090942, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 45090 + }, + { + "epoch": 0.45091, + "grad_norm": 1.2333749532699585, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 45091 + }, + { + "epoch": 0.45092, + "grad_norm": 0.7967898845672607, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 45092 + }, + { + "epoch": 0.45093, + "grad_norm": 0.779558539390564, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 45093 + }, + { + "epoch": 0.45094, + "grad_norm": 0.7016606330871582, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 45094 + }, + { + "epoch": 0.45095, + "grad_norm": 0.6946547627449036, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 45095 + }, + { + "epoch": 0.45096, + "grad_norm": 0.819790244102478, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 45096 + }, + { + "epoch": 0.45097, + "grad_norm": 1.0280405282974243, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 45097 + }, + { + "epoch": 0.45098, + "grad_norm": 1.2551178932189941, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 45098 + }, + { + "epoch": 0.45099, + "grad_norm": 0.73991858959198, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 45099 + }, + { + "epoch": 0.451, + "grad_norm": 0.6181426644325256, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 45100 + }, + { + "epoch": 0.45101, + "grad_norm": 0.6265337467193604, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 45101 + }, + { + "epoch": 0.45102, + "grad_norm": 0.7965924143791199, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 45102 + }, + { + "epoch": 0.45103, + "grad_norm": 1.1388213634490967, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 45103 + }, + { + "epoch": 0.45104, + "grad_norm": 0.7629915475845337, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 45104 + }, + { + "epoch": 0.45105, + "grad_norm": 0.5589327812194824, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 45105 + }, + { + "epoch": 0.45106, + "grad_norm": 0.4748506546020508, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 45106 + }, + { + "epoch": 0.45107, + "grad_norm": 0.640021562576294, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 45107 + }, + { + "epoch": 0.45108, + "grad_norm": 0.9040009379386902, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 45108 + }, + { + "epoch": 0.45109, + "grad_norm": 1.2066203355789185, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 45109 + }, + { + "epoch": 0.4511, + "grad_norm": 0.8997481465339661, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 45110 + }, + { + "epoch": 0.45111, + "grad_norm": 0.9549721479415894, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 45111 + }, + { + "epoch": 0.45112, + "grad_norm": 1.014123797416687, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 45112 + }, + { + "epoch": 0.45113, + "grad_norm": 0.9005487561225891, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 45113 + }, + { + "epoch": 0.45114, + "grad_norm": 0.7730492353439331, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 45114 + }, + { + "epoch": 0.45115, + "grad_norm": 0.686089277267456, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 45115 + }, + { + "epoch": 0.45116, + "grad_norm": 0.8252477645874023, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 45116 + }, + { + "epoch": 0.45117, + "grad_norm": 0.8193293809890747, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 45117 + }, + { + "epoch": 0.45118, + "grad_norm": 0.9544069170951843, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 45118 + }, + { + "epoch": 0.45119, + "grad_norm": 1.2463316917419434, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 45119 + }, + { + "epoch": 0.4512, + "grad_norm": 0.900621771812439, + "learning_rate": 0.003, + "loss": 3.9398, + "step": 45120 + }, + { + "epoch": 0.45121, + "grad_norm": 0.8275786638259888, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 45121 + }, + { + "epoch": 0.45122, + "grad_norm": 0.8925033211708069, + "learning_rate": 0.003, + "loss": 3.96, + "step": 45122 + }, + { + "epoch": 0.45123, + "grad_norm": 0.9722544550895691, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 45123 + }, + { + "epoch": 0.45124, + "grad_norm": 1.0165938138961792, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 45124 + }, + { + "epoch": 0.45125, + "grad_norm": 0.999585747718811, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 45125 + }, + { + "epoch": 0.45126, + "grad_norm": 1.0443955659866333, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 45126 + }, + { + "epoch": 0.45127, + "grad_norm": 0.9204643368721008, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 45127 + }, + { + "epoch": 0.45128, + "grad_norm": 1.027699589729309, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 45128 + }, + { + "epoch": 0.45129, + "grad_norm": 1.0262309312820435, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 45129 + }, + { + "epoch": 0.4513, + "grad_norm": 1.2843962907791138, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 45130 + }, + { + "epoch": 0.45131, + "grad_norm": 0.886735737323761, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 45131 + }, + { + "epoch": 0.45132, + "grad_norm": 0.9215191006660461, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 45132 + }, + { + "epoch": 0.45133, + "grad_norm": 1.1290712356567383, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 45133 + }, + { + "epoch": 0.45134, + "grad_norm": 1.0924276113510132, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 45134 + }, + { + "epoch": 0.45135, + "grad_norm": 1.0170698165893555, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 45135 + }, + { + "epoch": 0.45136, + "grad_norm": 1.0053634643554688, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 45136 + }, + { + "epoch": 0.45137, + "grad_norm": 1.1284594535827637, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 45137 + }, + { + "epoch": 0.45138, + "grad_norm": 0.8121505379676819, + "learning_rate": 0.003, + "loss": 3.984, + "step": 45138 + }, + { + "epoch": 0.45139, + "grad_norm": 0.8063299655914307, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 45139 + }, + { + "epoch": 0.4514, + "grad_norm": 0.8903714418411255, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 45140 + }, + { + "epoch": 0.45141, + "grad_norm": 0.8838207125663757, + "learning_rate": 0.003, + "loss": 3.9458, + "step": 45141 + }, + { + "epoch": 0.45142, + "grad_norm": 0.9422821402549744, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 45142 + }, + { + "epoch": 0.45143, + "grad_norm": 0.9679027199745178, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 45143 + }, + { + "epoch": 0.45144, + "grad_norm": 1.2343367338180542, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 45144 + }, + { + "epoch": 0.45145, + "grad_norm": 0.8057683110237122, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 45145 + }, + { + "epoch": 0.45146, + "grad_norm": 0.821168065071106, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 45146 + }, + { + "epoch": 0.45147, + "grad_norm": 1.0277456045150757, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 45147 + }, + { + "epoch": 0.45148, + "grad_norm": 1.2516788244247437, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 45148 + }, + { + "epoch": 0.45149, + "grad_norm": 0.7160952687263489, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 45149 + }, + { + "epoch": 0.4515, + "grad_norm": 0.7117305397987366, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 45150 + }, + { + "epoch": 0.45151, + "grad_norm": 0.8460555076599121, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 45151 + }, + { + "epoch": 0.45152, + "grad_norm": 1.0544273853302002, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 45152 + }, + { + "epoch": 0.45153, + "grad_norm": 0.9424821734428406, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 45153 + }, + { + "epoch": 0.45154, + "grad_norm": 0.8996103405952454, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 45154 + }, + { + "epoch": 0.45155, + "grad_norm": 0.8881667852401733, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 45155 + }, + { + "epoch": 0.45156, + "grad_norm": 0.9072226285934448, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 45156 + }, + { + "epoch": 0.45157, + "grad_norm": 0.8276975154876709, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 45157 + }, + { + "epoch": 0.45158, + "grad_norm": 0.8399852514266968, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 45158 + }, + { + "epoch": 0.45159, + "grad_norm": 0.7885059714317322, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 45159 + }, + { + "epoch": 0.4516, + "grad_norm": 0.6311356425285339, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 45160 + }, + { + "epoch": 0.45161, + "grad_norm": 0.6492985486984253, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 45161 + }, + { + "epoch": 0.45162, + "grad_norm": 0.8163139224052429, + "learning_rate": 0.003, + "loss": 3.9474, + "step": 45162 + }, + { + "epoch": 0.45163, + "grad_norm": 1.1934325695037842, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 45163 + }, + { + "epoch": 0.45164, + "grad_norm": 1.1282585859298706, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 45164 + }, + { + "epoch": 0.45165, + "grad_norm": 0.6936097741127014, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 45165 + }, + { + "epoch": 0.45166, + "grad_norm": 0.5782229900360107, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 45166 + }, + { + "epoch": 0.45167, + "grad_norm": 0.6213170289993286, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 45167 + }, + { + "epoch": 0.45168, + "grad_norm": 0.5964562892913818, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 45168 + }, + { + "epoch": 0.45169, + "grad_norm": 0.672767698764801, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 45169 + }, + { + "epoch": 0.4517, + "grad_norm": 0.9019458889961243, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 45170 + }, + { + "epoch": 0.45171, + "grad_norm": 1.3188138008117676, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 45171 + }, + { + "epoch": 0.45172, + "grad_norm": 0.7435301542282104, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 45172 + }, + { + "epoch": 0.45173, + "grad_norm": 0.7131667733192444, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 45173 + }, + { + "epoch": 0.45174, + "grad_norm": 0.8582903146743774, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 45174 + }, + { + "epoch": 0.45175, + "grad_norm": 1.365041971206665, + "learning_rate": 0.003, + "loss": 3.983, + "step": 45175 + }, + { + "epoch": 0.45176, + "grad_norm": 0.9217495322227478, + "learning_rate": 0.003, + "loss": 3.985, + "step": 45176 + }, + { + "epoch": 0.45177, + "grad_norm": 0.9414500594139099, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 45177 + }, + { + "epoch": 0.45178, + "grad_norm": 1.002394676208496, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 45178 + }, + { + "epoch": 0.45179, + "grad_norm": 1.042089819908142, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 45179 + }, + { + "epoch": 0.4518, + "grad_norm": 1.1502574682235718, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 45180 + }, + { + "epoch": 0.45181, + "grad_norm": 1.0938966274261475, + "learning_rate": 0.003, + "loss": 3.966, + "step": 45181 + }, + { + "epoch": 0.45182, + "grad_norm": 0.812851071357727, + "learning_rate": 0.003, + "loss": 4.001, + "step": 45182 + }, + { + "epoch": 0.45183, + "grad_norm": 0.6388632655143738, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 45183 + }, + { + "epoch": 0.45184, + "grad_norm": 0.7851143479347229, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 45184 + }, + { + "epoch": 0.45185, + "grad_norm": 0.9623672366142273, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 45185 + }, + { + "epoch": 0.45186, + "grad_norm": 0.8890101313591003, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 45186 + }, + { + "epoch": 0.45187, + "grad_norm": 0.9959158301353455, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 45187 + }, + { + "epoch": 0.45188, + "grad_norm": 1.617803692817688, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 45188 + }, + { + "epoch": 0.45189, + "grad_norm": 0.8711531162261963, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 45189 + }, + { + "epoch": 0.4519, + "grad_norm": 1.005150556564331, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 45190 + }, + { + "epoch": 0.45191, + "grad_norm": 0.9964739084243774, + "learning_rate": 0.003, + "loss": 3.987, + "step": 45191 + }, + { + "epoch": 0.45192, + "grad_norm": 0.9502631425857544, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 45192 + }, + { + "epoch": 0.45193, + "grad_norm": 1.0157811641693115, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 45193 + }, + { + "epoch": 0.45194, + "grad_norm": 1.077248215675354, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 45194 + }, + { + "epoch": 0.45195, + "grad_norm": 1.1510001420974731, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 45195 + }, + { + "epoch": 0.45196, + "grad_norm": 0.7330076098442078, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 45196 + }, + { + "epoch": 0.45197, + "grad_norm": 0.7321763634681702, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 45197 + }, + { + "epoch": 0.45198, + "grad_norm": 0.7346591353416443, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 45198 + }, + { + "epoch": 0.45199, + "grad_norm": 0.9622994065284729, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 45199 + }, + { + "epoch": 0.452, + "grad_norm": 1.1958234310150146, + "learning_rate": 0.003, + "loss": 3.972, + "step": 45200 + }, + { + "epoch": 0.45201, + "grad_norm": 0.8913683891296387, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 45201 + }, + { + "epoch": 0.45202, + "grad_norm": 0.8894706964492798, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 45202 + }, + { + "epoch": 0.45203, + "grad_norm": 1.0046039819717407, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 45203 + }, + { + "epoch": 0.45204, + "grad_norm": 1.129618763923645, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 45204 + }, + { + "epoch": 0.45205, + "grad_norm": 0.8567784428596497, + "learning_rate": 0.003, + "loss": 3.994, + "step": 45205 + }, + { + "epoch": 0.45206, + "grad_norm": 0.7983659505844116, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 45206 + }, + { + "epoch": 0.45207, + "grad_norm": 0.7864686846733093, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 45207 + }, + { + "epoch": 0.45208, + "grad_norm": 0.8584845662117004, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 45208 + }, + { + "epoch": 0.45209, + "grad_norm": 1.2348181009292603, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 45209 + }, + { + "epoch": 0.4521, + "grad_norm": 0.9777302145957947, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 45210 + }, + { + "epoch": 0.45211, + "grad_norm": 0.9681940674781799, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 45211 + }, + { + "epoch": 0.45212, + "grad_norm": 0.958958625793457, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 45212 + }, + { + "epoch": 0.45213, + "grad_norm": 0.8730608820915222, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 45213 + }, + { + "epoch": 0.45214, + "grad_norm": 0.8798052668571472, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 45214 + }, + { + "epoch": 0.45215, + "grad_norm": 0.881230354309082, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 45215 + }, + { + "epoch": 0.45216, + "grad_norm": 0.7970283627510071, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 45216 + }, + { + "epoch": 0.45217, + "grad_norm": 0.9099938869476318, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 45217 + }, + { + "epoch": 0.45218, + "grad_norm": 0.9860248565673828, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 45218 + }, + { + "epoch": 0.45219, + "grad_norm": 1.0833353996276855, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 45219 + }, + { + "epoch": 0.4522, + "grad_norm": 0.8817312121391296, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 45220 + }, + { + "epoch": 0.45221, + "grad_norm": 0.769609808921814, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 45221 + }, + { + "epoch": 0.45222, + "grad_norm": 0.9059709310531616, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 45222 + }, + { + "epoch": 0.45223, + "grad_norm": 1.2963122129440308, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 45223 + }, + { + "epoch": 0.45224, + "grad_norm": 0.7230905294418335, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 45224 + }, + { + "epoch": 0.45225, + "grad_norm": 0.6182692646980286, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 45225 + }, + { + "epoch": 0.45226, + "grad_norm": 0.9043174386024475, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 45226 + }, + { + "epoch": 0.45227, + "grad_norm": 1.0037249326705933, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 45227 + }, + { + "epoch": 0.45228, + "grad_norm": 0.9695153832435608, + "learning_rate": 0.003, + "loss": 3.99, + "step": 45228 + }, + { + "epoch": 0.45229, + "grad_norm": 1.0210024118423462, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 45229 + }, + { + "epoch": 0.4523, + "grad_norm": 0.8391290307044983, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 45230 + }, + { + "epoch": 0.45231, + "grad_norm": 0.9319174289703369, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 45231 + }, + { + "epoch": 0.45232, + "grad_norm": 1.0113328695297241, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 45232 + }, + { + "epoch": 0.45233, + "grad_norm": 0.7998688817024231, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 45233 + }, + { + "epoch": 0.45234, + "grad_norm": 0.8017682433128357, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 45234 + }, + { + "epoch": 0.45235, + "grad_norm": 0.773347795009613, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 45235 + }, + { + "epoch": 0.45236, + "grad_norm": 0.8913978338241577, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 45236 + }, + { + "epoch": 0.45237, + "grad_norm": 1.2270148992538452, + "learning_rate": 0.003, + "loss": 3.984, + "step": 45237 + }, + { + "epoch": 0.45238, + "grad_norm": 1.0313632488250732, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 45238 + }, + { + "epoch": 0.45239, + "grad_norm": 1.0223807096481323, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 45239 + }, + { + "epoch": 0.4524, + "grad_norm": 1.1251369714736938, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 45240 + }, + { + "epoch": 0.45241, + "grad_norm": 1.066139578819275, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 45241 + }, + { + "epoch": 0.45242, + "grad_norm": 0.8913189172744751, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 45242 + }, + { + "epoch": 0.45243, + "grad_norm": 0.7430772185325623, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 45243 + }, + { + "epoch": 0.45244, + "grad_norm": 0.792962372303009, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 45244 + }, + { + "epoch": 0.45245, + "grad_norm": 0.8660948276519775, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 45245 + }, + { + "epoch": 0.45246, + "grad_norm": 1.0488883256912231, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 45246 + }, + { + "epoch": 0.45247, + "grad_norm": 1.2622941732406616, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 45247 + }, + { + "epoch": 0.45248, + "grad_norm": 0.7510074973106384, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 45248 + }, + { + "epoch": 0.45249, + "grad_norm": 0.6582731604576111, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 45249 + }, + { + "epoch": 0.4525, + "grad_norm": 0.6635088920593262, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 45250 + }, + { + "epoch": 0.45251, + "grad_norm": 0.7013130187988281, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 45251 + }, + { + "epoch": 0.45252, + "grad_norm": 0.6143799424171448, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 45252 + }, + { + "epoch": 0.45253, + "grad_norm": 0.5941616296768188, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 45253 + }, + { + "epoch": 0.45254, + "grad_norm": 0.7832719683647156, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 45254 + }, + { + "epoch": 0.45255, + "grad_norm": 0.9504139423370361, + "learning_rate": 0.003, + "loss": 3.95, + "step": 45255 + }, + { + "epoch": 0.45256, + "grad_norm": 1.056300163269043, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 45256 + }, + { + "epoch": 0.45257, + "grad_norm": 0.9077678918838501, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 45257 + }, + { + "epoch": 0.45258, + "grad_norm": 0.7602502107620239, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 45258 + }, + { + "epoch": 0.45259, + "grad_norm": 0.7175456881523132, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 45259 + }, + { + "epoch": 0.4526, + "grad_norm": 0.6797919869422913, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 45260 + }, + { + "epoch": 0.45261, + "grad_norm": 0.6992586851119995, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 45261 + }, + { + "epoch": 0.45262, + "grad_norm": 0.7876990437507629, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 45262 + }, + { + "epoch": 0.45263, + "grad_norm": 0.9106609225273132, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 45263 + }, + { + "epoch": 0.45264, + "grad_norm": 1.0637595653533936, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 45264 + }, + { + "epoch": 0.45265, + "grad_norm": 1.18294358253479, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 45265 + }, + { + "epoch": 0.45266, + "grad_norm": 0.9572709798812866, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 45266 + }, + { + "epoch": 0.45267, + "grad_norm": 1.095212697982788, + "learning_rate": 0.003, + "loss": 3.961, + "step": 45267 + }, + { + "epoch": 0.45268, + "grad_norm": 0.976915180683136, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 45268 + }, + { + "epoch": 0.45269, + "grad_norm": 0.9924865365028381, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 45269 + }, + { + "epoch": 0.4527, + "grad_norm": 1.0178946256637573, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 45270 + }, + { + "epoch": 0.45271, + "grad_norm": 1.000757098197937, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 45271 + }, + { + "epoch": 0.45272, + "grad_norm": 1.0150622129440308, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 45272 + }, + { + "epoch": 0.45273, + "grad_norm": 1.2513611316680908, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 45273 + }, + { + "epoch": 0.45274, + "grad_norm": 0.9656386375427246, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 45274 + }, + { + "epoch": 0.45275, + "grad_norm": 1.1262671947479248, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 45275 + }, + { + "epoch": 0.45276, + "grad_norm": 1.1003408432006836, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 45276 + }, + { + "epoch": 0.45277, + "grad_norm": 1.0195839405059814, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 45277 + }, + { + "epoch": 0.45278, + "grad_norm": 1.0203415155410767, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 45278 + }, + { + "epoch": 0.45279, + "grad_norm": 0.7429502606391907, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 45279 + }, + { + "epoch": 0.4528, + "grad_norm": 0.7460005283355713, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 45280 + }, + { + "epoch": 0.45281, + "grad_norm": 0.8900617361068726, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 45281 + }, + { + "epoch": 0.45282, + "grad_norm": 1.1897988319396973, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 45282 + }, + { + "epoch": 0.45283, + "grad_norm": 1.0890072584152222, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 45283 + }, + { + "epoch": 0.45284, + "grad_norm": 0.9123743772506714, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 45284 + }, + { + "epoch": 0.45285, + "grad_norm": 0.8800172805786133, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 45285 + }, + { + "epoch": 0.45286, + "grad_norm": 0.8924891352653503, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 45286 + }, + { + "epoch": 0.45287, + "grad_norm": 0.9212398529052734, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 45287 + }, + { + "epoch": 0.45288, + "grad_norm": 1.0404422283172607, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 45288 + }, + { + "epoch": 0.45289, + "grad_norm": 1.1020255088806152, + "learning_rate": 0.003, + "loss": 3.9425, + "step": 45289 + }, + { + "epoch": 0.4529, + "grad_norm": 1.2011436223983765, + "learning_rate": 0.003, + "loss": 3.981, + "step": 45290 + }, + { + "epoch": 0.45291, + "grad_norm": 0.8017162084579468, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 45291 + }, + { + "epoch": 0.45292, + "grad_norm": 0.681755542755127, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 45292 + }, + { + "epoch": 0.45293, + "grad_norm": 0.7420396208763123, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 45293 + }, + { + "epoch": 0.45294, + "grad_norm": 0.7242715358734131, + "learning_rate": 0.003, + "loss": 3.9493, + "step": 45294 + }, + { + "epoch": 0.45295, + "grad_norm": 0.7242929935455322, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 45295 + }, + { + "epoch": 0.45296, + "grad_norm": 0.7552679777145386, + "learning_rate": 0.003, + "loss": 3.9498, + "step": 45296 + }, + { + "epoch": 0.45297, + "grad_norm": 0.8984644412994385, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 45297 + }, + { + "epoch": 0.45298, + "grad_norm": 0.9757901430130005, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 45298 + }, + { + "epoch": 0.45299, + "grad_norm": 0.9287000894546509, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 45299 + }, + { + "epoch": 0.453, + "grad_norm": 0.8789955973625183, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 45300 + }, + { + "epoch": 0.45301, + "grad_norm": 0.8866522312164307, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 45301 + }, + { + "epoch": 0.45302, + "grad_norm": 0.9585347175598145, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 45302 + }, + { + "epoch": 0.45303, + "grad_norm": 0.9435129165649414, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 45303 + }, + { + "epoch": 0.45304, + "grad_norm": 0.9156513810157776, + "learning_rate": 0.003, + "loss": 3.98, + "step": 45304 + }, + { + "epoch": 0.45305, + "grad_norm": 0.8166800737380981, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 45305 + }, + { + "epoch": 0.45306, + "grad_norm": 0.7563360929489136, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 45306 + }, + { + "epoch": 0.45307, + "grad_norm": 0.7973489165306091, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 45307 + }, + { + "epoch": 0.45308, + "grad_norm": 0.8394164443016052, + "learning_rate": 0.003, + "loss": 3.992, + "step": 45308 + }, + { + "epoch": 0.45309, + "grad_norm": 0.8581833243370056, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 45309 + }, + { + "epoch": 0.4531, + "grad_norm": 0.9853377938270569, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 45310 + }, + { + "epoch": 0.45311, + "grad_norm": 1.1939278841018677, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 45311 + }, + { + "epoch": 0.45312, + "grad_norm": 0.864234209060669, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 45312 + }, + { + "epoch": 0.45313, + "grad_norm": 0.9414840340614319, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 45313 + }, + { + "epoch": 0.45314, + "grad_norm": 1.0180374383926392, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 45314 + }, + { + "epoch": 0.45315, + "grad_norm": 1.3712148666381836, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 45315 + }, + { + "epoch": 0.45316, + "grad_norm": 0.7829338312149048, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 45316 + }, + { + "epoch": 0.45317, + "grad_norm": 0.6665776371955872, + "learning_rate": 0.003, + "loss": 3.9443, + "step": 45317 + }, + { + "epoch": 0.45318, + "grad_norm": 0.8122162222862244, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 45318 + }, + { + "epoch": 0.45319, + "grad_norm": 0.8585295677185059, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 45319 + }, + { + "epoch": 0.4532, + "grad_norm": 0.9035392999649048, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 45320 + }, + { + "epoch": 0.45321, + "grad_norm": 1.1262966394424438, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 45321 + }, + { + "epoch": 0.45322, + "grad_norm": 0.9481683969497681, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 45322 + }, + { + "epoch": 0.45323, + "grad_norm": 0.8841878771781921, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 45323 + }, + { + "epoch": 0.45324, + "grad_norm": 0.7409241199493408, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 45324 + }, + { + "epoch": 0.45325, + "grad_norm": 0.6659438014030457, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 45325 + }, + { + "epoch": 0.45326, + "grad_norm": 0.6338537931442261, + "learning_rate": 0.003, + "loss": 3.9407, + "step": 45326 + }, + { + "epoch": 0.45327, + "grad_norm": 0.6621182560920715, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 45327 + }, + { + "epoch": 0.45328, + "grad_norm": 0.6504284143447876, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 45328 + }, + { + "epoch": 0.45329, + "grad_norm": 0.7226760983467102, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 45329 + }, + { + "epoch": 0.4533, + "grad_norm": 0.9646887183189392, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 45330 + }, + { + "epoch": 0.45331, + "grad_norm": 1.3154067993164062, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 45331 + }, + { + "epoch": 0.45332, + "grad_norm": 0.8523274064064026, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 45332 + }, + { + "epoch": 0.45333, + "grad_norm": 0.9389171600341797, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 45333 + }, + { + "epoch": 0.45334, + "grad_norm": 1.183903455734253, + "learning_rate": 0.003, + "loss": 3.9401, + "step": 45334 + }, + { + "epoch": 0.45335, + "grad_norm": 0.972796618938446, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 45335 + }, + { + "epoch": 0.45336, + "grad_norm": 1.0780839920043945, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 45336 + }, + { + "epoch": 0.45337, + "grad_norm": 1.1142538785934448, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 45337 + }, + { + "epoch": 0.45338, + "grad_norm": 1.0484793186187744, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 45338 + }, + { + "epoch": 0.45339, + "grad_norm": 0.9884516000747681, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 45339 + }, + { + "epoch": 0.4534, + "grad_norm": 0.7669129967689514, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 45340 + }, + { + "epoch": 0.45341, + "grad_norm": 0.7386613488197327, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 45341 + }, + { + "epoch": 0.45342, + "grad_norm": 0.889708399772644, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 45342 + }, + { + "epoch": 0.45343, + "grad_norm": 1.096023678779602, + "learning_rate": 0.003, + "loss": 3.967, + "step": 45343 + }, + { + "epoch": 0.45344, + "grad_norm": 0.9374601244926453, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 45344 + }, + { + "epoch": 0.45345, + "grad_norm": 0.7014033198356628, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 45345 + }, + { + "epoch": 0.45346, + "grad_norm": 0.6932598948478699, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 45346 + }, + { + "epoch": 0.45347, + "grad_norm": 0.7631409764289856, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 45347 + }, + { + "epoch": 0.45348, + "grad_norm": 0.859293520450592, + "learning_rate": 0.003, + "loss": 3.935, + "step": 45348 + }, + { + "epoch": 0.45349, + "grad_norm": 0.7478974461555481, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 45349 + }, + { + "epoch": 0.4535, + "grad_norm": 0.6202149987220764, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 45350 + }, + { + "epoch": 0.45351, + "grad_norm": 0.7101496458053589, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 45351 + }, + { + "epoch": 0.45352, + "grad_norm": 0.9852554202079773, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 45352 + }, + { + "epoch": 0.45353, + "grad_norm": 1.271136999130249, + "learning_rate": 0.003, + "loss": 3.986, + "step": 45353 + }, + { + "epoch": 0.45354, + "grad_norm": 0.9901168942451477, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 45354 + }, + { + "epoch": 0.45355, + "grad_norm": 1.107272744178772, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 45355 + }, + { + "epoch": 0.45356, + "grad_norm": 0.962792694568634, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 45356 + }, + { + "epoch": 0.45357, + "grad_norm": 0.9128065705299377, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 45357 + }, + { + "epoch": 0.45358, + "grad_norm": 0.9243023991584778, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 45358 + }, + { + "epoch": 0.45359, + "grad_norm": 0.972494900226593, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 45359 + }, + { + "epoch": 0.4536, + "grad_norm": 1.0798192024230957, + "learning_rate": 0.003, + "loss": 3.986, + "step": 45360 + }, + { + "epoch": 0.45361, + "grad_norm": 0.8058287501335144, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 45361 + }, + { + "epoch": 0.45362, + "grad_norm": 0.794053316116333, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 45362 + }, + { + "epoch": 0.45363, + "grad_norm": 0.9425514936447144, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 45363 + }, + { + "epoch": 0.45364, + "grad_norm": 1.0284429788589478, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 45364 + }, + { + "epoch": 0.45365, + "grad_norm": 1.0429311990737915, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 45365 + }, + { + "epoch": 0.45366, + "grad_norm": 1.3743318319320679, + "learning_rate": 0.003, + "loss": 3.974, + "step": 45366 + }, + { + "epoch": 0.45367, + "grad_norm": 0.8641718626022339, + "learning_rate": 0.003, + "loss": 3.967, + "step": 45367 + }, + { + "epoch": 0.45368, + "grad_norm": 0.7508888244628906, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 45368 + }, + { + "epoch": 0.45369, + "grad_norm": 0.7825375199317932, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 45369 + }, + { + "epoch": 0.4537, + "grad_norm": 0.8150854110717773, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 45370 + }, + { + "epoch": 0.45371, + "grad_norm": 1.006034016609192, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 45371 + }, + { + "epoch": 0.45372, + "grad_norm": 1.0436116456985474, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 45372 + }, + { + "epoch": 0.45373, + "grad_norm": 0.962181031703949, + "learning_rate": 0.003, + "loss": 3.9456, + "step": 45373 + }, + { + "epoch": 0.45374, + "grad_norm": 1.1345628499984741, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 45374 + }, + { + "epoch": 0.45375, + "grad_norm": 0.8740653991699219, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 45375 + }, + { + "epoch": 0.45376, + "grad_norm": 0.9083115458488464, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 45376 + }, + { + "epoch": 0.45377, + "grad_norm": 0.899668276309967, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 45377 + }, + { + "epoch": 0.45378, + "grad_norm": 1.0477591753005981, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 45378 + }, + { + "epoch": 0.45379, + "grad_norm": 0.9524073600769043, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 45379 + }, + { + "epoch": 0.4538, + "grad_norm": 1.0442100763320923, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 45380 + }, + { + "epoch": 0.45381, + "grad_norm": 1.2236436605453491, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 45381 + }, + { + "epoch": 0.45382, + "grad_norm": 0.9235097169876099, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 45382 + }, + { + "epoch": 0.45383, + "grad_norm": 0.8737612366676331, + "learning_rate": 0.003, + "loss": 3.944, + "step": 45383 + }, + { + "epoch": 0.45384, + "grad_norm": 1.0577751398086548, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 45384 + }, + { + "epoch": 0.45385, + "grad_norm": 1.276299238204956, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 45385 + }, + { + "epoch": 0.45386, + "grad_norm": 0.8700922727584839, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 45386 + }, + { + "epoch": 0.45387, + "grad_norm": 0.8037944436073303, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 45387 + }, + { + "epoch": 0.45388, + "grad_norm": 0.870965301990509, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 45388 + }, + { + "epoch": 0.45389, + "grad_norm": 1.0325368642807007, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 45389 + }, + { + "epoch": 0.4539, + "grad_norm": 1.2075587511062622, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 45390 + }, + { + "epoch": 0.45391, + "grad_norm": 0.8819237351417542, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 45391 + }, + { + "epoch": 0.45392, + "grad_norm": 0.8623290657997131, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 45392 + }, + { + "epoch": 0.45393, + "grad_norm": 0.9043119549751282, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 45393 + }, + { + "epoch": 0.45394, + "grad_norm": 0.8195773959159851, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 45394 + }, + { + "epoch": 0.45395, + "grad_norm": 0.7360371351242065, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 45395 + }, + { + "epoch": 0.45396, + "grad_norm": 0.7203482985496521, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 45396 + }, + { + "epoch": 0.45397, + "grad_norm": 0.6000778079032898, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 45397 + }, + { + "epoch": 0.45398, + "grad_norm": 0.5450730919837952, + "learning_rate": 0.003, + "loss": 3.9294, + "step": 45398 + }, + { + "epoch": 0.45399, + "grad_norm": 0.6212664246559143, + "learning_rate": 0.003, + "loss": 3.996, + "step": 45399 + }, + { + "epoch": 0.454, + "grad_norm": 0.6006436944007874, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 45400 + }, + { + "epoch": 0.45401, + "grad_norm": 0.5975930690765381, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 45401 + }, + { + "epoch": 0.45402, + "grad_norm": 0.6072856783866882, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 45402 + }, + { + "epoch": 0.45403, + "grad_norm": 0.6776529550552368, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 45403 + }, + { + "epoch": 0.45404, + "grad_norm": 0.7911654114723206, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 45404 + }, + { + "epoch": 0.45405, + "grad_norm": 0.9304375648498535, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 45405 + }, + { + "epoch": 0.45406, + "grad_norm": 1.1495100259780884, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 45406 + }, + { + "epoch": 0.45407, + "grad_norm": 1.1575696468353271, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 45407 + }, + { + "epoch": 0.45408, + "grad_norm": 0.9882540106773376, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 45408 + }, + { + "epoch": 0.45409, + "grad_norm": 0.9554814100265503, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 45409 + }, + { + "epoch": 0.4541, + "grad_norm": 0.9312795996665955, + "learning_rate": 0.003, + "loss": 3.96, + "step": 45410 + }, + { + "epoch": 0.45411, + "grad_norm": 0.9262629747390747, + "learning_rate": 0.003, + "loss": 3.946, + "step": 45411 + }, + { + "epoch": 0.45412, + "grad_norm": 0.9416515827178955, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 45412 + }, + { + "epoch": 0.45413, + "grad_norm": 1.2023087739944458, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 45413 + }, + { + "epoch": 0.45414, + "grad_norm": 1.0517724752426147, + "learning_rate": 0.003, + "loss": 3.98, + "step": 45414 + }, + { + "epoch": 0.45415, + "grad_norm": 1.0717042684555054, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 45415 + }, + { + "epoch": 0.45416, + "grad_norm": 0.9533736109733582, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 45416 + }, + { + "epoch": 0.45417, + "grad_norm": 0.8168386220932007, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 45417 + }, + { + "epoch": 0.45418, + "grad_norm": 0.6687439680099487, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 45418 + }, + { + "epoch": 0.45419, + "grad_norm": 0.6077435612678528, + "learning_rate": 0.003, + "loss": 3.975, + "step": 45419 + }, + { + "epoch": 0.4542, + "grad_norm": 0.6267147064208984, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 45420 + }, + { + "epoch": 0.45421, + "grad_norm": 0.6673589944839478, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 45421 + }, + { + "epoch": 0.45422, + "grad_norm": 0.71595698595047, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 45422 + }, + { + "epoch": 0.45423, + "grad_norm": 0.6819670796394348, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 45423 + }, + { + "epoch": 0.45424, + "grad_norm": 0.6603847742080688, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 45424 + }, + { + "epoch": 0.45425, + "grad_norm": 0.772947371006012, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 45425 + }, + { + "epoch": 0.45426, + "grad_norm": 0.9672809839248657, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 45426 + }, + { + "epoch": 0.45427, + "grad_norm": 1.1963788270950317, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 45427 + }, + { + "epoch": 0.45428, + "grad_norm": 1.0728936195373535, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 45428 + }, + { + "epoch": 0.45429, + "grad_norm": 1.1523672342300415, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 45429 + }, + { + "epoch": 0.4543, + "grad_norm": 1.0920511484146118, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 45430 + }, + { + "epoch": 0.45431, + "grad_norm": 0.9960185289382935, + "learning_rate": 0.003, + "loss": 3.977, + "step": 45431 + }, + { + "epoch": 0.45432, + "grad_norm": 0.9998465776443481, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 45432 + }, + { + "epoch": 0.45433, + "grad_norm": 1.056598424911499, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 45433 + }, + { + "epoch": 0.45434, + "grad_norm": 1.1433072090148926, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 45434 + }, + { + "epoch": 0.45435, + "grad_norm": 0.6727818250656128, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 45435 + }, + { + "epoch": 0.45436, + "grad_norm": 0.619837760925293, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 45436 + }, + { + "epoch": 0.45437, + "grad_norm": 0.7592400908470154, + "learning_rate": 0.003, + "loss": 3.988, + "step": 45437 + }, + { + "epoch": 0.45438, + "grad_norm": 0.7909695506095886, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 45438 + }, + { + "epoch": 0.45439, + "grad_norm": 0.8844357132911682, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 45439 + }, + { + "epoch": 0.4544, + "grad_norm": 0.9927106499671936, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 45440 + }, + { + "epoch": 0.45441, + "grad_norm": 1.0566531419754028, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 45441 + }, + { + "epoch": 0.45442, + "grad_norm": 0.9666938781738281, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 45442 + }, + { + "epoch": 0.45443, + "grad_norm": 1.0715136528015137, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 45443 + }, + { + "epoch": 0.45444, + "grad_norm": 1.0793250799179077, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 45444 + }, + { + "epoch": 0.45445, + "grad_norm": 1.2203339338302612, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 45445 + }, + { + "epoch": 0.45446, + "grad_norm": 1.1202278137207031, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 45446 + }, + { + "epoch": 0.45447, + "grad_norm": 0.9465891122817993, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 45447 + }, + { + "epoch": 0.45448, + "grad_norm": 1.115070104598999, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 45448 + }, + { + "epoch": 0.45449, + "grad_norm": 1.0883941650390625, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 45449 + }, + { + "epoch": 0.4545, + "grad_norm": 1.0038214921951294, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 45450 + }, + { + "epoch": 0.45451, + "grad_norm": 0.8988173007965088, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 45451 + }, + { + "epoch": 0.45452, + "grad_norm": 0.8082979917526245, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 45452 + }, + { + "epoch": 0.45453, + "grad_norm": 0.8369646668434143, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 45453 + }, + { + "epoch": 0.45454, + "grad_norm": 0.9423527121543884, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 45454 + }, + { + "epoch": 0.45455, + "grad_norm": 1.1357271671295166, + "learning_rate": 0.003, + "loss": 4.0, + "step": 45455 + }, + { + "epoch": 0.45456, + "grad_norm": 1.0022720098495483, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 45456 + }, + { + "epoch": 0.45457, + "grad_norm": 1.058584213256836, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 45457 + }, + { + "epoch": 0.45458, + "grad_norm": 0.8747600317001343, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 45458 + }, + { + "epoch": 0.45459, + "grad_norm": 0.8202128410339355, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 45459 + }, + { + "epoch": 0.4546, + "grad_norm": 0.8530676364898682, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 45460 + }, + { + "epoch": 0.45461, + "grad_norm": 0.9884126782417297, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 45461 + }, + { + "epoch": 0.45462, + "grad_norm": 1.0352106094360352, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 45462 + }, + { + "epoch": 0.45463, + "grad_norm": 0.9918100237846375, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 45463 + }, + { + "epoch": 0.45464, + "grad_norm": 0.8616276979446411, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 45464 + }, + { + "epoch": 0.45465, + "grad_norm": 0.7384687066078186, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 45465 + }, + { + "epoch": 0.45466, + "grad_norm": 0.8543527722358704, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 45466 + }, + { + "epoch": 0.45467, + "grad_norm": 0.9795273542404175, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 45467 + }, + { + "epoch": 0.45468, + "grad_norm": 1.042014241218567, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 45468 + }, + { + "epoch": 0.45469, + "grad_norm": 1.041200876235962, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 45469 + }, + { + "epoch": 0.4547, + "grad_norm": 0.9413985013961792, + "learning_rate": 0.003, + "loss": 3.9394, + "step": 45470 + }, + { + "epoch": 0.45471, + "grad_norm": 0.8761382102966309, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 45471 + }, + { + "epoch": 0.45472, + "grad_norm": 0.7464942336082458, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 45472 + }, + { + "epoch": 0.45473, + "grad_norm": 0.7585170865058899, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 45473 + }, + { + "epoch": 0.45474, + "grad_norm": 0.7309529781341553, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 45474 + }, + { + "epoch": 0.45475, + "grad_norm": 0.8629032969474792, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 45475 + }, + { + "epoch": 0.45476, + "grad_norm": 0.9135105013847351, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 45476 + }, + { + "epoch": 0.45477, + "grad_norm": 0.9425889253616333, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 45477 + }, + { + "epoch": 0.45478, + "grad_norm": 0.9683959484100342, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 45478 + }, + { + "epoch": 0.45479, + "grad_norm": 1.1500765085220337, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 45479 + }, + { + "epoch": 0.4548, + "grad_norm": 0.8782233595848083, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 45480 + }, + { + "epoch": 0.45481, + "grad_norm": 0.7940176129341125, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 45481 + }, + { + "epoch": 0.45482, + "grad_norm": 0.9812948703765869, + "learning_rate": 0.003, + "loss": 4.011, + "step": 45482 + }, + { + "epoch": 0.45483, + "grad_norm": 1.0687999725341797, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 45483 + }, + { + "epoch": 0.45484, + "grad_norm": 0.9206521511077881, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 45484 + }, + { + "epoch": 0.45485, + "grad_norm": 1.00516939163208, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 45485 + }, + { + "epoch": 0.45486, + "grad_norm": 1.2658556699752808, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 45486 + }, + { + "epoch": 0.45487, + "grad_norm": 0.9723002314567566, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 45487 + }, + { + "epoch": 0.45488, + "grad_norm": 1.110661506652832, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 45488 + }, + { + "epoch": 0.45489, + "grad_norm": 1.1198623180389404, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 45489 + }, + { + "epoch": 0.4549, + "grad_norm": 0.8562336564064026, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 45490 + }, + { + "epoch": 0.45491, + "grad_norm": 0.6640442609786987, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 45491 + }, + { + "epoch": 0.45492, + "grad_norm": 0.6558650732040405, + "learning_rate": 0.003, + "loss": 3.983, + "step": 45492 + }, + { + "epoch": 0.45493, + "grad_norm": 0.62127685546875, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 45493 + }, + { + "epoch": 0.45494, + "grad_norm": 0.6297252178192139, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 45494 + }, + { + "epoch": 0.45495, + "grad_norm": 0.7646884918212891, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 45495 + }, + { + "epoch": 0.45496, + "grad_norm": 0.7862653136253357, + "learning_rate": 0.003, + "loss": 3.994, + "step": 45496 + }, + { + "epoch": 0.45497, + "grad_norm": 0.7727521061897278, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 45497 + }, + { + "epoch": 0.45498, + "grad_norm": 0.694372296333313, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 45498 + }, + { + "epoch": 0.45499, + "grad_norm": 0.7410460114479065, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 45499 + }, + { + "epoch": 0.455, + "grad_norm": 0.7629368901252747, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 45500 + }, + { + "epoch": 0.45501, + "grad_norm": 0.7443826198577881, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 45501 + }, + { + "epoch": 0.45502, + "grad_norm": 0.5951176285743713, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 45502 + }, + { + "epoch": 0.45503, + "grad_norm": 0.5653535723686218, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 45503 + }, + { + "epoch": 0.45504, + "grad_norm": 0.6655838489532471, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 45504 + }, + { + "epoch": 0.45505, + "grad_norm": 0.7218239903450012, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 45505 + }, + { + "epoch": 0.45506, + "grad_norm": 0.9238889217376709, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 45506 + }, + { + "epoch": 0.45507, + "grad_norm": 1.3516720533370972, + "learning_rate": 0.003, + "loss": 3.958, + "step": 45507 + }, + { + "epoch": 0.45508, + "grad_norm": 0.7216677665710449, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 45508 + }, + { + "epoch": 0.45509, + "grad_norm": 0.7093604207038879, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 45509 + }, + { + "epoch": 0.4551, + "grad_norm": 0.7569281458854675, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 45510 + }, + { + "epoch": 0.45511, + "grad_norm": 0.7005040645599365, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 45511 + }, + { + "epoch": 0.45512, + "grad_norm": 0.7732653021812439, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 45512 + }, + { + "epoch": 0.45513, + "grad_norm": 0.9545965194702148, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 45513 + }, + { + "epoch": 0.45514, + "grad_norm": 1.3048887252807617, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 45514 + }, + { + "epoch": 0.45515, + "grad_norm": 1.053039789199829, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 45515 + }, + { + "epoch": 0.45516, + "grad_norm": 0.9257300496101379, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 45516 + }, + { + "epoch": 0.45517, + "grad_norm": 0.7272683382034302, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 45517 + }, + { + "epoch": 0.45518, + "grad_norm": 0.7919144630432129, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 45518 + }, + { + "epoch": 0.45519, + "grad_norm": 1.0083788633346558, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 45519 + }, + { + "epoch": 0.4552, + "grad_norm": 1.1339199542999268, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 45520 + }, + { + "epoch": 0.45521, + "grad_norm": 0.837322473526001, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 45521 + }, + { + "epoch": 0.45522, + "grad_norm": 0.8579670786857605, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 45522 + }, + { + "epoch": 0.45523, + "grad_norm": 0.8853756189346313, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 45523 + }, + { + "epoch": 0.45524, + "grad_norm": 1.294812560081482, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 45524 + }, + { + "epoch": 0.45525, + "grad_norm": 0.794567346572876, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 45525 + }, + { + "epoch": 0.45526, + "grad_norm": 0.726272702217102, + "learning_rate": 0.003, + "loss": 3.948, + "step": 45526 + }, + { + "epoch": 0.45527, + "grad_norm": 0.9411984086036682, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 45527 + }, + { + "epoch": 0.45528, + "grad_norm": 1.5691198110580444, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 45528 + }, + { + "epoch": 0.45529, + "grad_norm": 0.5377702116966248, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 45529 + }, + { + "epoch": 0.4553, + "grad_norm": 1.0890849828720093, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 45530 + }, + { + "epoch": 0.45531, + "grad_norm": 1.1005995273590088, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 45531 + }, + { + "epoch": 0.45532, + "grad_norm": 1.0379681587219238, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 45532 + }, + { + "epoch": 0.45533, + "grad_norm": 1.0477360486984253, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 45533 + }, + { + "epoch": 0.45534, + "grad_norm": 0.9885754585266113, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 45534 + }, + { + "epoch": 0.45535, + "grad_norm": 1.25361168384552, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 45535 + }, + { + "epoch": 0.45536, + "grad_norm": 0.9073994755744934, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 45536 + }, + { + "epoch": 0.45537, + "grad_norm": 0.9016502499580383, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 45537 + }, + { + "epoch": 0.45538, + "grad_norm": 1.1038662195205688, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 45538 + }, + { + "epoch": 0.45539, + "grad_norm": 1.1890634298324585, + "learning_rate": 0.003, + "loss": 3.979, + "step": 45539 + }, + { + "epoch": 0.4554, + "grad_norm": 0.7330134510993958, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 45540 + }, + { + "epoch": 0.45541, + "grad_norm": 0.6991299390792847, + "learning_rate": 0.003, + "loss": 3.97, + "step": 45541 + }, + { + "epoch": 0.45542, + "grad_norm": 0.8959406018257141, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 45542 + }, + { + "epoch": 0.45543, + "grad_norm": 0.9371351003646851, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 45543 + }, + { + "epoch": 0.45544, + "grad_norm": 1.1150567531585693, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 45544 + }, + { + "epoch": 0.45545, + "grad_norm": 0.8213277459144592, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 45545 + }, + { + "epoch": 0.45546, + "grad_norm": 0.6131003499031067, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 45546 + }, + { + "epoch": 0.45547, + "grad_norm": 0.5657074451446533, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 45547 + }, + { + "epoch": 0.45548, + "grad_norm": 0.629285454750061, + "learning_rate": 0.003, + "loss": 3.969, + "step": 45548 + }, + { + "epoch": 0.45549, + "grad_norm": 0.9056037664413452, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 45549 + }, + { + "epoch": 0.4555, + "grad_norm": 1.151172399520874, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 45550 + }, + { + "epoch": 0.45551, + "grad_norm": 0.864496648311615, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 45551 + }, + { + "epoch": 0.45552, + "grad_norm": 0.8874682188034058, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 45552 + }, + { + "epoch": 0.45553, + "grad_norm": 0.8584817051887512, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 45553 + }, + { + "epoch": 0.45554, + "grad_norm": 0.9890925288200378, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 45554 + }, + { + "epoch": 0.45555, + "grad_norm": 1.2787662744522095, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 45555 + }, + { + "epoch": 0.45556, + "grad_norm": 0.8222064971923828, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 45556 + }, + { + "epoch": 0.45557, + "grad_norm": 0.8763864040374756, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 45557 + }, + { + "epoch": 0.45558, + "grad_norm": 1.052353024482727, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 45558 + }, + { + "epoch": 0.45559, + "grad_norm": 0.997737467288971, + "learning_rate": 0.003, + "loss": 3.994, + "step": 45559 + }, + { + "epoch": 0.4556, + "grad_norm": 0.9772519469261169, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 45560 + }, + { + "epoch": 0.45561, + "grad_norm": 1.1010736227035522, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 45561 + }, + { + "epoch": 0.45562, + "grad_norm": 1.0554195642471313, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 45562 + }, + { + "epoch": 0.45563, + "grad_norm": 0.9844115376472473, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 45563 + }, + { + "epoch": 0.45564, + "grad_norm": 1.0522139072418213, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 45564 + }, + { + "epoch": 0.45565, + "grad_norm": 1.1061327457427979, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 45565 + }, + { + "epoch": 0.45566, + "grad_norm": 0.9243091344833374, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 45566 + }, + { + "epoch": 0.45567, + "grad_norm": 1.025705099105835, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 45567 + }, + { + "epoch": 0.45568, + "grad_norm": 1.2321527004241943, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 45568 + }, + { + "epoch": 0.45569, + "grad_norm": 1.0410562753677368, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 45569 + }, + { + "epoch": 0.4557, + "grad_norm": 0.9710739850997925, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 45570 + }, + { + "epoch": 0.45571, + "grad_norm": 0.9465663433074951, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 45571 + }, + { + "epoch": 0.45572, + "grad_norm": 1.0901119709014893, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 45572 + }, + { + "epoch": 0.45573, + "grad_norm": 1.0998973846435547, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 45573 + }, + { + "epoch": 0.45574, + "grad_norm": 0.8006938099861145, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 45574 + }, + { + "epoch": 0.45575, + "grad_norm": 0.8550772070884705, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 45575 + }, + { + "epoch": 0.45576, + "grad_norm": 1.063714861869812, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 45576 + }, + { + "epoch": 0.45577, + "grad_norm": 1.1798079013824463, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 45577 + }, + { + "epoch": 0.45578, + "grad_norm": 0.8470385074615479, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 45578 + }, + { + "epoch": 0.45579, + "grad_norm": 0.8002135753631592, + "learning_rate": 0.003, + "loss": 3.975, + "step": 45579 + }, + { + "epoch": 0.4558, + "grad_norm": 0.7626228332519531, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 45580 + }, + { + "epoch": 0.45581, + "grad_norm": 0.7534394860267639, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 45581 + }, + { + "epoch": 0.45582, + "grad_norm": 0.9366096258163452, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 45582 + }, + { + "epoch": 0.45583, + "grad_norm": 1.2519644498825073, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 45583 + }, + { + "epoch": 0.45584, + "grad_norm": 0.7210441827774048, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 45584 + }, + { + "epoch": 0.45585, + "grad_norm": 0.7013623118400574, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 45585 + }, + { + "epoch": 0.45586, + "grad_norm": 0.919023334980011, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 45586 + }, + { + "epoch": 0.45587, + "grad_norm": 0.9978067874908447, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 45587 + }, + { + "epoch": 0.45588, + "grad_norm": 1.2649619579315186, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 45588 + }, + { + "epoch": 0.45589, + "grad_norm": 0.7755820751190186, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 45589 + }, + { + "epoch": 0.4559, + "grad_norm": 0.6596386432647705, + "learning_rate": 0.003, + "loss": 3.9324, + "step": 45590 + }, + { + "epoch": 0.45591, + "grad_norm": 0.6821756958961487, + "learning_rate": 0.003, + "loss": 3.9474, + "step": 45591 + }, + { + "epoch": 0.45592, + "grad_norm": 0.7802316546440125, + "learning_rate": 0.003, + "loss": 3.9421, + "step": 45592 + }, + { + "epoch": 0.45593, + "grad_norm": 0.8790780305862427, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 45593 + }, + { + "epoch": 0.45594, + "grad_norm": 0.8121445775032043, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 45594 + }, + { + "epoch": 0.45595, + "grad_norm": 0.7664431929588318, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 45595 + }, + { + "epoch": 0.45596, + "grad_norm": 0.8156375288963318, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 45596 + }, + { + "epoch": 0.45597, + "grad_norm": 0.9460542798042297, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 45597 + }, + { + "epoch": 0.45598, + "grad_norm": 1.2489310503005981, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 45598 + }, + { + "epoch": 0.45599, + "grad_norm": 0.6737655401229858, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 45599 + }, + { + "epoch": 0.456, + "grad_norm": 0.5938513278961182, + "learning_rate": 0.003, + "loss": 3.9449, + "step": 45600 + }, + { + "epoch": 0.45601, + "grad_norm": 0.6761572360992432, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 45601 + }, + { + "epoch": 0.45602, + "grad_norm": 0.8247717022895813, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 45602 + }, + { + "epoch": 0.45603, + "grad_norm": 1.0804219245910645, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 45603 + }, + { + "epoch": 0.45604, + "grad_norm": 1.1300373077392578, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 45604 + }, + { + "epoch": 0.45605, + "grad_norm": 1.0078489780426025, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 45605 + }, + { + "epoch": 0.45606, + "grad_norm": 1.0549169778823853, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 45606 + }, + { + "epoch": 0.45607, + "grad_norm": 1.2002036571502686, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 45607 + }, + { + "epoch": 0.45608, + "grad_norm": 1.0889745950698853, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 45608 + }, + { + "epoch": 0.45609, + "grad_norm": 0.9288007616996765, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 45609 + }, + { + "epoch": 0.4561, + "grad_norm": 0.9511890411376953, + "learning_rate": 0.003, + "loss": 3.9447, + "step": 45610 + }, + { + "epoch": 0.45611, + "grad_norm": 1.033294916152954, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 45611 + }, + { + "epoch": 0.45612, + "grad_norm": 0.9955863952636719, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 45612 + }, + { + "epoch": 0.45613, + "grad_norm": 0.9007339477539062, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 45613 + }, + { + "epoch": 0.45614, + "grad_norm": 0.9641721248626709, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 45614 + }, + { + "epoch": 0.45615, + "grad_norm": 1.2301373481750488, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 45615 + }, + { + "epoch": 0.45616, + "grad_norm": 0.9099317789077759, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 45616 + }, + { + "epoch": 0.45617, + "grad_norm": 0.8219741582870483, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 45617 + }, + { + "epoch": 0.45618, + "grad_norm": 0.8284704685211182, + "learning_rate": 0.003, + "loss": 3.991, + "step": 45618 + }, + { + "epoch": 0.45619, + "grad_norm": 0.944709837436676, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 45619 + }, + { + "epoch": 0.4562, + "grad_norm": 0.9994373321533203, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 45620 + }, + { + "epoch": 0.45621, + "grad_norm": 0.9821600317955017, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 45621 + }, + { + "epoch": 0.45622, + "grad_norm": 1.0547549724578857, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 45622 + }, + { + "epoch": 0.45623, + "grad_norm": 1.034659743309021, + "learning_rate": 0.003, + "loss": 3.9384, + "step": 45623 + }, + { + "epoch": 0.45624, + "grad_norm": 0.9619352221488953, + "learning_rate": 0.003, + "loss": 3.9473, + "step": 45624 + }, + { + "epoch": 0.45625, + "grad_norm": 0.9915831089019775, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 45625 + }, + { + "epoch": 0.45626, + "grad_norm": 1.1202083826065063, + "learning_rate": 0.003, + "loss": 4.005, + "step": 45626 + }, + { + "epoch": 0.45627, + "grad_norm": 0.8106662631034851, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 45627 + }, + { + "epoch": 0.45628, + "grad_norm": 0.6892222762107849, + "learning_rate": 0.003, + "loss": 3.952, + "step": 45628 + }, + { + "epoch": 0.45629, + "grad_norm": 0.7209795713424683, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 45629 + }, + { + "epoch": 0.4563, + "grad_norm": 1.0729445219039917, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 45630 + }, + { + "epoch": 0.45631, + "grad_norm": 1.3134082555770874, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 45631 + }, + { + "epoch": 0.45632, + "grad_norm": 0.5432438850402832, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 45632 + }, + { + "epoch": 0.45633, + "grad_norm": 0.6541646122932434, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 45633 + }, + { + "epoch": 0.45634, + "grad_norm": 0.6154834628105164, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 45634 + }, + { + "epoch": 0.45635, + "grad_norm": 0.6338497996330261, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 45635 + }, + { + "epoch": 0.45636, + "grad_norm": 0.5936143398284912, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 45636 + }, + { + "epoch": 0.45637, + "grad_norm": 0.6984920501708984, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 45637 + }, + { + "epoch": 0.45638, + "grad_norm": 0.8254494071006775, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 45638 + }, + { + "epoch": 0.45639, + "grad_norm": 1.1309723854064941, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 45639 + }, + { + "epoch": 0.4564, + "grad_norm": 1.0922108888626099, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 45640 + }, + { + "epoch": 0.45641, + "grad_norm": 0.8399249911308289, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 45641 + }, + { + "epoch": 0.45642, + "grad_norm": 0.6867673397064209, + "learning_rate": 0.003, + "loss": 3.9469, + "step": 45642 + }, + { + "epoch": 0.45643, + "grad_norm": 0.726320207118988, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 45643 + }, + { + "epoch": 0.45644, + "grad_norm": 0.8035522103309631, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 45644 + }, + { + "epoch": 0.45645, + "grad_norm": 0.9112889766693115, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 45645 + }, + { + "epoch": 0.45646, + "grad_norm": 0.9299195408821106, + "learning_rate": 0.003, + "loss": 3.9394, + "step": 45646 + }, + { + "epoch": 0.45647, + "grad_norm": 0.8548672199249268, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 45647 + }, + { + "epoch": 0.45648, + "grad_norm": 0.9835113286972046, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 45648 + }, + { + "epoch": 0.45649, + "grad_norm": 0.971550464630127, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 45649 + }, + { + "epoch": 0.4565, + "grad_norm": 1.0052412748336792, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 45650 + }, + { + "epoch": 0.45651, + "grad_norm": 1.3214956521987915, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 45651 + }, + { + "epoch": 0.45652, + "grad_norm": 1.0844237804412842, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 45652 + }, + { + "epoch": 0.45653, + "grad_norm": 1.085877776145935, + "learning_rate": 0.003, + "loss": 3.964, + "step": 45653 + }, + { + "epoch": 0.45654, + "grad_norm": 0.8941216468811035, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 45654 + }, + { + "epoch": 0.45655, + "grad_norm": 0.8857437968254089, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 45655 + }, + { + "epoch": 0.45656, + "grad_norm": 0.9694884419441223, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 45656 + }, + { + "epoch": 0.45657, + "grad_norm": 1.1499708890914917, + "learning_rate": 0.003, + "loss": 3.965, + "step": 45657 + }, + { + "epoch": 0.45658, + "grad_norm": 0.7950717210769653, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 45658 + }, + { + "epoch": 0.45659, + "grad_norm": 0.83872389793396, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 45659 + }, + { + "epoch": 0.4566, + "grad_norm": 0.880614697933197, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 45660 + }, + { + "epoch": 0.45661, + "grad_norm": 0.9989384412765503, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 45661 + }, + { + "epoch": 0.45662, + "grad_norm": 1.0718234777450562, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 45662 + }, + { + "epoch": 0.45663, + "grad_norm": 1.0184683799743652, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 45663 + }, + { + "epoch": 0.45664, + "grad_norm": 1.0977394580841064, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 45664 + }, + { + "epoch": 0.45665, + "grad_norm": 0.9124926328659058, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 45665 + }, + { + "epoch": 0.45666, + "grad_norm": 0.8745000958442688, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 45666 + }, + { + "epoch": 0.45667, + "grad_norm": 0.9480443000793457, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 45667 + }, + { + "epoch": 0.45668, + "grad_norm": 1.1602745056152344, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 45668 + }, + { + "epoch": 0.45669, + "grad_norm": 0.6353266835212708, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 45669 + }, + { + "epoch": 0.4567, + "grad_norm": 0.736085832118988, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 45670 + }, + { + "epoch": 0.45671, + "grad_norm": 0.6395013332366943, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 45671 + }, + { + "epoch": 0.45672, + "grad_norm": 0.6542761921882629, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 45672 + }, + { + "epoch": 0.45673, + "grad_norm": 0.7650668621063232, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 45673 + }, + { + "epoch": 0.45674, + "grad_norm": 0.8098594546318054, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 45674 + }, + { + "epoch": 0.45675, + "grad_norm": 0.8640055656433105, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 45675 + }, + { + "epoch": 0.45676, + "grad_norm": 1.1500827074050903, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 45676 + }, + { + "epoch": 0.45677, + "grad_norm": 1.3785072565078735, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 45677 + }, + { + "epoch": 0.45678, + "grad_norm": 0.847019612789154, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 45678 + }, + { + "epoch": 0.45679, + "grad_norm": 0.7934389710426331, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 45679 + }, + { + "epoch": 0.4568, + "grad_norm": 0.8337774872779846, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 45680 + }, + { + "epoch": 0.45681, + "grad_norm": 1.0227043628692627, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 45681 + }, + { + "epoch": 0.45682, + "grad_norm": 1.0733096599578857, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 45682 + }, + { + "epoch": 0.45683, + "grad_norm": 0.9963895678520203, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 45683 + }, + { + "epoch": 0.45684, + "grad_norm": 1.151442527770996, + "learning_rate": 0.003, + "loss": 3.985, + "step": 45684 + }, + { + "epoch": 0.45685, + "grad_norm": 0.7723867297172546, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 45685 + }, + { + "epoch": 0.45686, + "grad_norm": 0.7109845876693726, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 45686 + }, + { + "epoch": 0.45687, + "grad_norm": 0.7510933876037598, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 45687 + }, + { + "epoch": 0.45688, + "grad_norm": 1.0227417945861816, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 45688 + }, + { + "epoch": 0.45689, + "grad_norm": 1.363661289215088, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 45689 + }, + { + "epoch": 0.4569, + "grad_norm": 0.6725830435752869, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 45690 + }, + { + "epoch": 0.45691, + "grad_norm": 0.6528291702270508, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 45691 + }, + { + "epoch": 0.45692, + "grad_norm": 0.580365777015686, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 45692 + }, + { + "epoch": 0.45693, + "grad_norm": 0.5939337611198425, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 45693 + }, + { + "epoch": 0.45694, + "grad_norm": 0.5546630024909973, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 45694 + }, + { + "epoch": 0.45695, + "grad_norm": 0.5839359760284424, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 45695 + }, + { + "epoch": 0.45696, + "grad_norm": 0.7628461122512817, + "learning_rate": 0.003, + "loss": 3.9428, + "step": 45696 + }, + { + "epoch": 0.45697, + "grad_norm": 1.0177537202835083, + "learning_rate": 0.003, + "loss": 3.9445, + "step": 45697 + }, + { + "epoch": 0.45698, + "grad_norm": 1.2805280685424805, + "learning_rate": 0.003, + "loss": 3.9265, + "step": 45698 + }, + { + "epoch": 0.45699, + "grad_norm": 0.8987271785736084, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 45699 + }, + { + "epoch": 0.457, + "grad_norm": 0.9230576753616333, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 45700 + }, + { + "epoch": 0.45701, + "grad_norm": 1.0426568984985352, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 45701 + }, + { + "epoch": 0.45702, + "grad_norm": 1.0876185894012451, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 45702 + }, + { + "epoch": 0.45703, + "grad_norm": 0.8715566992759705, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 45703 + }, + { + "epoch": 0.45704, + "grad_norm": 0.899896502494812, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 45704 + }, + { + "epoch": 0.45705, + "grad_norm": 1.0070627927780151, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 45705 + }, + { + "epoch": 0.45706, + "grad_norm": 1.2932047843933105, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 45706 + }, + { + "epoch": 0.45707, + "grad_norm": 0.8855844736099243, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 45707 + }, + { + "epoch": 0.45708, + "grad_norm": 0.9625643491744995, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 45708 + }, + { + "epoch": 0.45709, + "grad_norm": 1.0436056852340698, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 45709 + }, + { + "epoch": 0.4571, + "grad_norm": 0.9794275164604187, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 45710 + }, + { + "epoch": 0.45711, + "grad_norm": 1.0200308561325073, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 45711 + }, + { + "epoch": 0.45712, + "grad_norm": 1.0563526153564453, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 45712 + }, + { + "epoch": 0.45713, + "grad_norm": 0.8698517084121704, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 45713 + }, + { + "epoch": 0.45714, + "grad_norm": 0.9154585003852844, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 45714 + }, + { + "epoch": 0.45715, + "grad_norm": 1.0438100099563599, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 45715 + }, + { + "epoch": 0.45716, + "grad_norm": 0.8956385850906372, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 45716 + }, + { + "epoch": 0.45717, + "grad_norm": 0.8635643720626831, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 45717 + }, + { + "epoch": 0.45718, + "grad_norm": 1.0149236917495728, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 45718 + }, + { + "epoch": 0.45719, + "grad_norm": 0.9970231652259827, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 45719 + }, + { + "epoch": 0.4572, + "grad_norm": 1.18647301197052, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 45720 + }, + { + "epoch": 0.45721, + "grad_norm": 0.9601936936378479, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 45721 + }, + { + "epoch": 0.45722, + "grad_norm": 0.8720555901527405, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 45722 + }, + { + "epoch": 0.45723, + "grad_norm": 0.9062684774398804, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 45723 + }, + { + "epoch": 0.45724, + "grad_norm": 1.1049617528915405, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 45724 + }, + { + "epoch": 0.45725, + "grad_norm": 0.9552932381629944, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 45725 + }, + { + "epoch": 0.45726, + "grad_norm": 0.9300732016563416, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 45726 + }, + { + "epoch": 0.45727, + "grad_norm": 0.8773369789123535, + "learning_rate": 0.003, + "loss": 4.003, + "step": 45727 + }, + { + "epoch": 0.45728, + "grad_norm": 0.8744059801101685, + "learning_rate": 0.003, + "loss": 4.021, + "step": 45728 + }, + { + "epoch": 0.45729, + "grad_norm": 0.9846293330192566, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 45729 + }, + { + "epoch": 0.4573, + "grad_norm": 1.1635818481445312, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 45730 + }, + { + "epoch": 0.45731, + "grad_norm": 0.9596167206764221, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 45731 + }, + { + "epoch": 0.45732, + "grad_norm": 0.8650158643722534, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 45732 + }, + { + "epoch": 0.45733, + "grad_norm": 0.7816172242164612, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 45733 + }, + { + "epoch": 0.45734, + "grad_norm": 0.6906917095184326, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 45734 + }, + { + "epoch": 0.45735, + "grad_norm": 0.6855912208557129, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 45735 + }, + { + "epoch": 0.45736, + "grad_norm": 0.6938791871070862, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 45736 + }, + { + "epoch": 0.45737, + "grad_norm": 0.6349058747291565, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 45737 + }, + { + "epoch": 0.45738, + "grad_norm": 0.7521833181381226, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 45738 + }, + { + "epoch": 0.45739, + "grad_norm": 0.7664848566055298, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 45739 + }, + { + "epoch": 0.4574, + "grad_norm": 0.8929563164710999, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 45740 + }, + { + "epoch": 0.45741, + "grad_norm": 1.0877670049667358, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 45741 + }, + { + "epoch": 0.45742, + "grad_norm": 1.022051215171814, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 45742 + }, + { + "epoch": 0.45743, + "grad_norm": 1.078658938407898, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 45743 + }, + { + "epoch": 0.45744, + "grad_norm": 1.0882421731948853, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 45744 + }, + { + "epoch": 0.45745, + "grad_norm": 0.9356698989868164, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 45745 + }, + { + "epoch": 0.45746, + "grad_norm": 0.8882631063461304, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 45746 + }, + { + "epoch": 0.45747, + "grad_norm": 0.8798527121543884, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 45747 + }, + { + "epoch": 0.45748, + "grad_norm": 1.0185645818710327, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 45748 + }, + { + "epoch": 0.45749, + "grad_norm": 1.1448205709457397, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 45749 + }, + { + "epoch": 0.4575, + "grad_norm": 0.9973906874656677, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 45750 + }, + { + "epoch": 0.45751, + "grad_norm": 1.0139896869659424, + "learning_rate": 0.003, + "loss": 3.991, + "step": 45751 + }, + { + "epoch": 0.45752, + "grad_norm": 1.0773096084594727, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 45752 + }, + { + "epoch": 0.45753, + "grad_norm": 0.9940887093544006, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 45753 + }, + { + "epoch": 0.45754, + "grad_norm": 1.0574370622634888, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 45754 + }, + { + "epoch": 0.45755, + "grad_norm": 1.090976357460022, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 45755 + }, + { + "epoch": 0.45756, + "grad_norm": 0.8611896634101868, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 45756 + }, + { + "epoch": 0.45757, + "grad_norm": 0.7755858898162842, + "learning_rate": 0.003, + "loss": 4.001, + "step": 45757 + }, + { + "epoch": 0.45758, + "grad_norm": 0.9025448560714722, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 45758 + }, + { + "epoch": 0.45759, + "grad_norm": 0.9674745798110962, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 45759 + }, + { + "epoch": 0.4576, + "grad_norm": 1.0809860229492188, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 45760 + }, + { + "epoch": 0.45761, + "grad_norm": 1.01613187789917, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 45761 + }, + { + "epoch": 0.45762, + "grad_norm": 1.0364044904708862, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 45762 + }, + { + "epoch": 0.45763, + "grad_norm": 0.9754256010055542, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 45763 + }, + { + "epoch": 0.45764, + "grad_norm": 1.080450177192688, + "learning_rate": 0.003, + "loss": 4.002, + "step": 45764 + }, + { + "epoch": 0.45765, + "grad_norm": 1.0119459629058838, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 45765 + }, + { + "epoch": 0.45766, + "grad_norm": 0.9147210717201233, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 45766 + }, + { + "epoch": 0.45767, + "grad_norm": 0.7176138162612915, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 45767 + }, + { + "epoch": 0.45768, + "grad_norm": 0.5973193645477295, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 45768 + }, + { + "epoch": 0.45769, + "grad_norm": 0.5450125932693481, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 45769 + }, + { + "epoch": 0.4577, + "grad_norm": 0.5750390887260437, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 45770 + }, + { + "epoch": 0.45771, + "grad_norm": 0.6287731528282166, + "learning_rate": 0.003, + "loss": 3.989, + "step": 45771 + }, + { + "epoch": 0.45772, + "grad_norm": 0.7432718276977539, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 45772 + }, + { + "epoch": 0.45773, + "grad_norm": 0.8879619836807251, + "learning_rate": 0.003, + "loss": 3.9373, + "step": 45773 + }, + { + "epoch": 0.45774, + "grad_norm": 1.2193835973739624, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 45774 + }, + { + "epoch": 0.45775, + "grad_norm": 0.8448391556739807, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 45775 + }, + { + "epoch": 0.45776, + "grad_norm": 0.7175645232200623, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 45776 + }, + { + "epoch": 0.45777, + "grad_norm": 0.7330701351165771, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 45777 + }, + { + "epoch": 0.45778, + "grad_norm": 0.7569064497947693, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 45778 + }, + { + "epoch": 0.45779, + "grad_norm": 0.9556224942207336, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 45779 + }, + { + "epoch": 0.4578, + "grad_norm": 1.1039462089538574, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 45780 + }, + { + "epoch": 0.45781, + "grad_norm": 1.0204274654388428, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 45781 + }, + { + "epoch": 0.45782, + "grad_norm": 0.9583452939987183, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 45782 + }, + { + "epoch": 0.45783, + "grad_norm": 0.8730328679084778, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 45783 + }, + { + "epoch": 0.45784, + "grad_norm": 0.8454229831695557, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 45784 + }, + { + "epoch": 0.45785, + "grad_norm": 0.8150758147239685, + "learning_rate": 0.003, + "loss": 3.973, + "step": 45785 + }, + { + "epoch": 0.45786, + "grad_norm": 0.6760915517807007, + "learning_rate": 0.003, + "loss": 3.941, + "step": 45786 + }, + { + "epoch": 0.45787, + "grad_norm": 0.6333000063896179, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 45787 + }, + { + "epoch": 0.45788, + "grad_norm": 0.7493495345115662, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 45788 + }, + { + "epoch": 0.45789, + "grad_norm": 0.7900902032852173, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 45789 + }, + { + "epoch": 0.4579, + "grad_norm": 0.9760284423828125, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 45790 + }, + { + "epoch": 0.45791, + "grad_norm": 1.3400533199310303, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 45791 + }, + { + "epoch": 0.45792, + "grad_norm": 0.8331460952758789, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 45792 + }, + { + "epoch": 0.45793, + "grad_norm": 1.0301352739334106, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 45793 + }, + { + "epoch": 0.45794, + "grad_norm": 1.123700499534607, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 45794 + }, + { + "epoch": 0.45795, + "grad_norm": 0.9245412945747375, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 45795 + }, + { + "epoch": 0.45796, + "grad_norm": 0.8692846894264221, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 45796 + }, + { + "epoch": 0.45797, + "grad_norm": 0.7928799986839294, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 45797 + }, + { + "epoch": 0.45798, + "grad_norm": 0.71099454164505, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 45798 + }, + { + "epoch": 0.45799, + "grad_norm": 0.7306148409843445, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 45799 + }, + { + "epoch": 0.458, + "grad_norm": 0.8145045638084412, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 45800 + }, + { + "epoch": 0.45801, + "grad_norm": 0.7392885088920593, + "learning_rate": 0.003, + "loss": 3.983, + "step": 45801 + }, + { + "epoch": 0.45802, + "grad_norm": 0.7338793873786926, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 45802 + }, + { + "epoch": 0.45803, + "grad_norm": 0.889621376991272, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 45803 + }, + { + "epoch": 0.45804, + "grad_norm": 1.0333094596862793, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 45804 + }, + { + "epoch": 0.45805, + "grad_norm": 1.206712007522583, + "learning_rate": 0.003, + "loss": 3.976, + "step": 45805 + }, + { + "epoch": 0.45806, + "grad_norm": 0.9465723037719727, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 45806 + }, + { + "epoch": 0.45807, + "grad_norm": 1.0335017442703247, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 45807 + }, + { + "epoch": 0.45808, + "grad_norm": 0.9671928882598877, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 45808 + }, + { + "epoch": 0.45809, + "grad_norm": 0.9319673776626587, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 45809 + }, + { + "epoch": 0.4581, + "grad_norm": 0.860837996006012, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 45810 + }, + { + "epoch": 0.45811, + "grad_norm": 0.889900267124176, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 45811 + }, + { + "epoch": 0.45812, + "grad_norm": 1.144982933998108, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 45812 + }, + { + "epoch": 0.45813, + "grad_norm": 1.0196776390075684, + "learning_rate": 0.003, + "loss": 3.962, + "step": 45813 + }, + { + "epoch": 0.45814, + "grad_norm": 0.8499878644943237, + "learning_rate": 0.003, + "loss": 3.973, + "step": 45814 + }, + { + "epoch": 0.45815, + "grad_norm": 0.7925819754600525, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 45815 + }, + { + "epoch": 0.45816, + "grad_norm": 0.7626482248306274, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 45816 + }, + { + "epoch": 0.45817, + "grad_norm": 0.7850402593612671, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 45817 + }, + { + "epoch": 0.45818, + "grad_norm": 0.8755583167076111, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 45818 + }, + { + "epoch": 0.45819, + "grad_norm": 1.0648753643035889, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 45819 + }, + { + "epoch": 0.4582, + "grad_norm": 0.9960668683052063, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 45820 + }, + { + "epoch": 0.45821, + "grad_norm": 1.1091060638427734, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 45821 + }, + { + "epoch": 0.45822, + "grad_norm": 1.0842092037200928, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 45822 + }, + { + "epoch": 0.45823, + "grad_norm": 1.1957026720046997, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 45823 + }, + { + "epoch": 0.45824, + "grad_norm": 0.7361541986465454, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 45824 + }, + { + "epoch": 0.45825, + "grad_norm": 0.6486629843711853, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 45825 + }, + { + "epoch": 0.45826, + "grad_norm": 0.7914289832115173, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 45826 + }, + { + "epoch": 0.45827, + "grad_norm": 1.043339729309082, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 45827 + }, + { + "epoch": 0.45828, + "grad_norm": 1.0713698863983154, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 45828 + }, + { + "epoch": 0.45829, + "grad_norm": 1.075474739074707, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 45829 + }, + { + "epoch": 0.4583, + "grad_norm": 0.9550519585609436, + "learning_rate": 0.003, + "loss": 3.9437, + "step": 45830 + }, + { + "epoch": 0.45831, + "grad_norm": 1.0434311628341675, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 45831 + }, + { + "epoch": 0.45832, + "grad_norm": 1.054600715637207, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 45832 + }, + { + "epoch": 0.45833, + "grad_norm": 0.8945831060409546, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 45833 + }, + { + "epoch": 0.45834, + "grad_norm": 0.9525026679039001, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 45834 + }, + { + "epoch": 0.45835, + "grad_norm": 0.9381633400917053, + "learning_rate": 0.003, + "loss": 3.9368, + "step": 45835 + }, + { + "epoch": 0.45836, + "grad_norm": 1.0786396265029907, + "learning_rate": 0.003, + "loss": 3.9493, + "step": 45836 + }, + { + "epoch": 0.45837, + "grad_norm": 0.9779283404350281, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 45837 + }, + { + "epoch": 0.45838, + "grad_norm": 0.8314782977104187, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 45838 + }, + { + "epoch": 0.45839, + "grad_norm": 0.8313777446746826, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 45839 + }, + { + "epoch": 0.4584, + "grad_norm": 0.8709431886672974, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 45840 + }, + { + "epoch": 0.45841, + "grad_norm": 0.7354288697242737, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 45841 + }, + { + "epoch": 0.45842, + "grad_norm": 0.730342447757721, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 45842 + }, + { + "epoch": 0.45843, + "grad_norm": 0.8059393167495728, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 45843 + }, + { + "epoch": 0.45844, + "grad_norm": 0.8380835056304932, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 45844 + }, + { + "epoch": 0.45845, + "grad_norm": 0.9650973081588745, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 45845 + }, + { + "epoch": 0.45846, + "grad_norm": 0.9730566740036011, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 45846 + }, + { + "epoch": 0.45847, + "grad_norm": 0.864362895488739, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 45847 + }, + { + "epoch": 0.45848, + "grad_norm": 0.8985995054244995, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 45848 + }, + { + "epoch": 0.45849, + "grad_norm": 0.9153335094451904, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 45849 + }, + { + "epoch": 0.4585, + "grad_norm": 1.2467362880706787, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 45850 + }, + { + "epoch": 0.45851, + "grad_norm": 0.9371594786643982, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 45851 + }, + { + "epoch": 0.45852, + "grad_norm": 0.9807268977165222, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 45852 + }, + { + "epoch": 0.45853, + "grad_norm": 0.9393059015274048, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 45853 + }, + { + "epoch": 0.45854, + "grad_norm": 0.830244243144989, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 45854 + }, + { + "epoch": 0.45855, + "grad_norm": 0.8492242693901062, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 45855 + }, + { + "epoch": 0.45856, + "grad_norm": 0.9380473494529724, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 45856 + }, + { + "epoch": 0.45857, + "grad_norm": 1.295179843902588, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 45857 + }, + { + "epoch": 0.45858, + "grad_norm": 1.1365193128585815, + "learning_rate": 0.003, + "loss": 3.9471, + "step": 45858 + }, + { + "epoch": 0.45859, + "grad_norm": 1.051484227180481, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 45859 + }, + { + "epoch": 0.4586, + "grad_norm": 0.9443255066871643, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 45860 + }, + { + "epoch": 0.45861, + "grad_norm": 0.9452719688415527, + "learning_rate": 0.003, + "loss": 3.968, + "step": 45861 + }, + { + "epoch": 0.45862, + "grad_norm": 1.1706467866897583, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 45862 + }, + { + "epoch": 0.45863, + "grad_norm": 0.9796823859214783, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 45863 + }, + { + "epoch": 0.45864, + "grad_norm": 0.8026430606842041, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 45864 + }, + { + "epoch": 0.45865, + "grad_norm": 0.706550121307373, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 45865 + }, + { + "epoch": 0.45866, + "grad_norm": 0.8011335134506226, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 45866 + }, + { + "epoch": 0.45867, + "grad_norm": 0.7049497365951538, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 45867 + }, + { + "epoch": 0.45868, + "grad_norm": 0.6553982496261597, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 45868 + }, + { + "epoch": 0.45869, + "grad_norm": 0.8148787021636963, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 45869 + }, + { + "epoch": 0.4587, + "grad_norm": 0.9856696128845215, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 45870 + }, + { + "epoch": 0.45871, + "grad_norm": 0.9197937846183777, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 45871 + }, + { + "epoch": 0.45872, + "grad_norm": 0.8213559985160828, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 45872 + }, + { + "epoch": 0.45873, + "grad_norm": 1.254353642463684, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 45873 + }, + { + "epoch": 0.45874, + "grad_norm": 0.8385439515113831, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 45874 + }, + { + "epoch": 0.45875, + "grad_norm": 0.7334051132202148, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 45875 + }, + { + "epoch": 0.45876, + "grad_norm": 0.7872242331504822, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 45876 + }, + { + "epoch": 0.45877, + "grad_norm": 0.9453456997871399, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 45877 + }, + { + "epoch": 0.45878, + "grad_norm": 1.1122779846191406, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 45878 + }, + { + "epoch": 0.45879, + "grad_norm": 0.879806637763977, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 45879 + }, + { + "epoch": 0.4588, + "grad_norm": 0.8882684111595154, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 45880 + }, + { + "epoch": 0.45881, + "grad_norm": 1.082355260848999, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 45881 + }, + { + "epoch": 0.45882, + "grad_norm": 1.10385000705719, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 45882 + }, + { + "epoch": 0.45883, + "grad_norm": 0.999728798866272, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 45883 + }, + { + "epoch": 0.45884, + "grad_norm": 1.0733174085617065, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 45884 + }, + { + "epoch": 0.45885, + "grad_norm": 1.0936354398727417, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 45885 + }, + { + "epoch": 0.45886, + "grad_norm": 0.8456752300262451, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 45886 + }, + { + "epoch": 0.45887, + "grad_norm": 0.6482964158058167, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 45887 + }, + { + "epoch": 0.45888, + "grad_norm": 0.5886674523353577, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 45888 + }, + { + "epoch": 0.45889, + "grad_norm": 0.6459987759590149, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 45889 + }, + { + "epoch": 0.4589, + "grad_norm": 0.7922376990318298, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 45890 + }, + { + "epoch": 0.45891, + "grad_norm": 1.0333043336868286, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 45891 + }, + { + "epoch": 0.45892, + "grad_norm": 1.221819281578064, + "learning_rate": 0.003, + "loss": 3.991, + "step": 45892 + }, + { + "epoch": 0.45893, + "grad_norm": 0.9138470888137817, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 45893 + }, + { + "epoch": 0.45894, + "grad_norm": 0.9001432657241821, + "learning_rate": 0.003, + "loss": 3.982, + "step": 45894 + }, + { + "epoch": 0.45895, + "grad_norm": 0.8855478167533875, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 45895 + }, + { + "epoch": 0.45896, + "grad_norm": 0.9088332653045654, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 45896 + }, + { + "epoch": 0.45897, + "grad_norm": 0.9776809215545654, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 45897 + }, + { + "epoch": 0.45898, + "grad_norm": 1.0553797483444214, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 45898 + }, + { + "epoch": 0.45899, + "grad_norm": 1.1587210893630981, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 45899 + }, + { + "epoch": 0.459, + "grad_norm": 0.9651724696159363, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 45900 + }, + { + "epoch": 0.45901, + "grad_norm": 1.3899321556091309, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 45901 + }, + { + "epoch": 0.45902, + "grad_norm": 1.0010566711425781, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 45902 + }, + { + "epoch": 0.45903, + "grad_norm": 0.9306576251983643, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 45903 + }, + { + "epoch": 0.45904, + "grad_norm": 1.1107183694839478, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 45904 + }, + { + "epoch": 0.45905, + "grad_norm": 0.9874153733253479, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 45905 + }, + { + "epoch": 0.45906, + "grad_norm": 0.9532989263534546, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 45906 + }, + { + "epoch": 0.45907, + "grad_norm": 0.9112750887870789, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 45907 + }, + { + "epoch": 0.45908, + "grad_norm": 1.2726503610610962, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 45908 + }, + { + "epoch": 0.45909, + "grad_norm": 0.8672385215759277, + "learning_rate": 0.003, + "loss": 3.9305, + "step": 45909 + }, + { + "epoch": 0.4591, + "grad_norm": 0.8341842293739319, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 45910 + }, + { + "epoch": 0.45911, + "grad_norm": 0.7924756407737732, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 45911 + }, + { + "epoch": 0.45912, + "grad_norm": 0.9325964450836182, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 45912 + }, + { + "epoch": 0.45913, + "grad_norm": 0.995046854019165, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 45913 + }, + { + "epoch": 0.45914, + "grad_norm": 0.9558351039886475, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 45914 + }, + { + "epoch": 0.45915, + "grad_norm": 0.8736859560012817, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 45915 + }, + { + "epoch": 0.45916, + "grad_norm": 0.7624912261962891, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 45916 + }, + { + "epoch": 0.45917, + "grad_norm": 0.8724467158317566, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 45917 + }, + { + "epoch": 0.45918, + "grad_norm": 0.9425812363624573, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 45918 + }, + { + "epoch": 0.45919, + "grad_norm": 1.0754812955856323, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 45919 + }, + { + "epoch": 0.4592, + "grad_norm": 0.8807567358016968, + "learning_rate": 0.003, + "loss": 3.962, + "step": 45920 + }, + { + "epoch": 0.45921, + "grad_norm": 0.9075182676315308, + "learning_rate": 0.003, + "loss": 3.991, + "step": 45921 + }, + { + "epoch": 0.45922, + "grad_norm": 0.9225207567214966, + "learning_rate": 0.003, + "loss": 3.972, + "step": 45922 + }, + { + "epoch": 0.45923, + "grad_norm": 0.8908079266548157, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 45923 + }, + { + "epoch": 0.45924, + "grad_norm": 0.7861801385879517, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 45924 + }, + { + "epoch": 0.45925, + "grad_norm": 0.7419173121452332, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 45925 + }, + { + "epoch": 0.45926, + "grad_norm": 0.7384288311004639, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 45926 + }, + { + "epoch": 0.45927, + "grad_norm": 0.845809280872345, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 45927 + }, + { + "epoch": 0.45928, + "grad_norm": 0.9747729897499084, + "learning_rate": 0.003, + "loss": 3.986, + "step": 45928 + }, + { + "epoch": 0.45929, + "grad_norm": 0.9774366021156311, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 45929 + }, + { + "epoch": 0.4593, + "grad_norm": 0.940422773361206, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 45930 + }, + { + "epoch": 0.45931, + "grad_norm": 0.9820241332054138, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 45931 + }, + { + "epoch": 0.45932, + "grad_norm": 1.1312257051467896, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 45932 + }, + { + "epoch": 0.45933, + "grad_norm": 0.885921835899353, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 45933 + }, + { + "epoch": 0.45934, + "grad_norm": 0.8642967343330383, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 45934 + }, + { + "epoch": 0.45935, + "grad_norm": 0.8477857112884521, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 45935 + }, + { + "epoch": 0.45936, + "grad_norm": 0.9552637934684753, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 45936 + }, + { + "epoch": 0.45937, + "grad_norm": 0.8961846232414246, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 45937 + }, + { + "epoch": 0.45938, + "grad_norm": 0.8329768776893616, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 45938 + }, + { + "epoch": 0.45939, + "grad_norm": 0.8029659986495972, + "learning_rate": 0.003, + "loss": 3.972, + "step": 45939 + }, + { + "epoch": 0.4594, + "grad_norm": 0.7174447774887085, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 45940 + }, + { + "epoch": 0.45941, + "grad_norm": 0.7128310203552246, + "learning_rate": 0.003, + "loss": 3.9462, + "step": 45941 + }, + { + "epoch": 0.45942, + "grad_norm": 0.6042807698249817, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 45942 + }, + { + "epoch": 0.45943, + "grad_norm": 0.684018075466156, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 45943 + }, + { + "epoch": 0.45944, + "grad_norm": 0.8330841064453125, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 45944 + }, + { + "epoch": 0.45945, + "grad_norm": 1.2828481197357178, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 45945 + }, + { + "epoch": 0.45946, + "grad_norm": 0.9305540323257446, + "learning_rate": 0.003, + "loss": 3.956, + "step": 45946 + }, + { + "epoch": 0.45947, + "grad_norm": 0.7473944425582886, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 45947 + }, + { + "epoch": 0.45948, + "grad_norm": 0.6367319226264954, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 45948 + }, + { + "epoch": 0.45949, + "grad_norm": 0.6640996932983398, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 45949 + }, + { + "epoch": 0.4595, + "grad_norm": 0.734998881816864, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 45950 + }, + { + "epoch": 0.45951, + "grad_norm": 0.8601686954498291, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 45951 + }, + { + "epoch": 0.45952, + "grad_norm": 1.0772194862365723, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 45952 + }, + { + "epoch": 0.45953, + "grad_norm": 1.1344304084777832, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 45953 + }, + { + "epoch": 0.45954, + "grad_norm": 0.7723748683929443, + "learning_rate": 0.003, + "loss": 3.94, + "step": 45954 + }, + { + "epoch": 0.45955, + "grad_norm": 0.8702669739723206, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 45955 + }, + { + "epoch": 0.45956, + "grad_norm": 0.9787406921386719, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 45956 + }, + { + "epoch": 0.45957, + "grad_norm": 0.9100373983383179, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 45957 + }, + { + "epoch": 0.45958, + "grad_norm": 1.0856292247772217, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 45958 + }, + { + "epoch": 0.45959, + "grad_norm": 1.1330357789993286, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 45959 + }, + { + "epoch": 0.4596, + "grad_norm": 0.8351128101348877, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 45960 + }, + { + "epoch": 0.45961, + "grad_norm": 0.822155237197876, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 45961 + }, + { + "epoch": 0.45962, + "grad_norm": 0.7630565166473389, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 45962 + }, + { + "epoch": 0.45963, + "grad_norm": 0.8457805514335632, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 45963 + }, + { + "epoch": 0.45964, + "grad_norm": 1.0612382888793945, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 45964 + }, + { + "epoch": 0.45965, + "grad_norm": 0.9257577657699585, + "learning_rate": 0.003, + "loss": 3.978, + "step": 45965 + }, + { + "epoch": 0.45966, + "grad_norm": 0.8905503749847412, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 45966 + }, + { + "epoch": 0.45967, + "grad_norm": 0.9638327360153198, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 45967 + }, + { + "epoch": 0.45968, + "grad_norm": 1.0836585760116577, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 45968 + }, + { + "epoch": 0.45969, + "grad_norm": 0.8634781837463379, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 45969 + }, + { + "epoch": 0.4597, + "grad_norm": 0.8035150766372681, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 45970 + }, + { + "epoch": 0.45971, + "grad_norm": 0.79410719871521, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 45971 + }, + { + "epoch": 0.45972, + "grad_norm": 0.8602942228317261, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 45972 + }, + { + "epoch": 0.45973, + "grad_norm": 0.9081432223320007, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 45973 + }, + { + "epoch": 0.45974, + "grad_norm": 0.9339918494224548, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 45974 + }, + { + "epoch": 0.45975, + "grad_norm": 0.879050612449646, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 45975 + }, + { + "epoch": 0.45976, + "grad_norm": 1.1122078895568848, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 45976 + }, + { + "epoch": 0.45977, + "grad_norm": 0.8431329131126404, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 45977 + }, + { + "epoch": 0.45978, + "grad_norm": 0.9659278988838196, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 45978 + }, + { + "epoch": 0.45979, + "grad_norm": 1.2893047332763672, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 45979 + }, + { + "epoch": 0.4598, + "grad_norm": 0.6704239845275879, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 45980 + }, + { + "epoch": 0.45981, + "grad_norm": 0.6878457069396973, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 45981 + }, + { + "epoch": 0.45982, + "grad_norm": 0.667397141456604, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 45982 + }, + { + "epoch": 0.45983, + "grad_norm": 0.8580325841903687, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 45983 + }, + { + "epoch": 0.45984, + "grad_norm": 0.965471625328064, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 45984 + }, + { + "epoch": 0.45985, + "grad_norm": 1.1315319538116455, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 45985 + }, + { + "epoch": 0.45986, + "grad_norm": 0.9850853681564331, + "learning_rate": 0.003, + "loss": 3.9462, + "step": 45986 + }, + { + "epoch": 0.45987, + "grad_norm": 0.9556884169578552, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 45987 + }, + { + "epoch": 0.45988, + "grad_norm": 0.8533722162246704, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 45988 + }, + { + "epoch": 0.45989, + "grad_norm": 0.7949380874633789, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 45989 + }, + { + "epoch": 0.4599, + "grad_norm": 0.6884228587150574, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 45990 + }, + { + "epoch": 0.45991, + "grad_norm": 0.6158667206764221, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 45991 + }, + { + "epoch": 0.45992, + "grad_norm": 0.6193859577178955, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 45992 + }, + { + "epoch": 0.45993, + "grad_norm": 0.6538625359535217, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 45993 + }, + { + "epoch": 0.45994, + "grad_norm": 0.6887592077255249, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 45994 + }, + { + "epoch": 0.45995, + "grad_norm": 0.7215583324432373, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 45995 + }, + { + "epoch": 0.45996, + "grad_norm": 0.836195707321167, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 45996 + }, + { + "epoch": 0.45997, + "grad_norm": 1.0122568607330322, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 45997 + }, + { + "epoch": 0.45998, + "grad_norm": 1.63300359249115, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 45998 + }, + { + "epoch": 0.45999, + "grad_norm": 0.5619109272956848, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 45999 + }, + { + "epoch": 0.46, + "grad_norm": 1.009881854057312, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 46000 + }, + { + "epoch": 0.46001, + "grad_norm": 1.108378529548645, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 46001 + }, + { + "epoch": 0.46002, + "grad_norm": 1.2288084030151367, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 46002 + }, + { + "epoch": 0.46003, + "grad_norm": 1.0962549448013306, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 46003 + }, + { + "epoch": 0.46004, + "grad_norm": 0.7710174322128296, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 46004 + }, + { + "epoch": 0.46005, + "grad_norm": 0.7948812246322632, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 46005 + }, + { + "epoch": 0.46006, + "grad_norm": 0.7871363162994385, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 46006 + }, + { + "epoch": 0.46007, + "grad_norm": 0.9172770977020264, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 46007 + }, + { + "epoch": 0.46008, + "grad_norm": 0.9660437107086182, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 46008 + }, + { + "epoch": 0.46009, + "grad_norm": 1.0979456901550293, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 46009 + }, + { + "epoch": 0.4601, + "grad_norm": 1.0558449029922485, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 46010 + }, + { + "epoch": 0.46011, + "grad_norm": 0.9058130383491516, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 46011 + }, + { + "epoch": 0.46012, + "grad_norm": 0.9486808776855469, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 46012 + }, + { + "epoch": 0.46013, + "grad_norm": 1.0132273435592651, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 46013 + }, + { + "epoch": 0.46014, + "grad_norm": 0.9252476096153259, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 46014 + }, + { + "epoch": 0.46015, + "grad_norm": 0.8622710704803467, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 46015 + }, + { + "epoch": 0.46016, + "grad_norm": 0.995765745639801, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 46016 + }, + { + "epoch": 0.46017, + "grad_norm": 1.4240736961364746, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 46017 + }, + { + "epoch": 0.46018, + "grad_norm": 0.9791032075881958, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 46018 + }, + { + "epoch": 0.46019, + "grad_norm": 0.9731720685958862, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 46019 + }, + { + "epoch": 0.4602, + "grad_norm": 1.0525732040405273, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 46020 + }, + { + "epoch": 0.46021, + "grad_norm": 0.943099319934845, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 46021 + }, + { + "epoch": 0.46022, + "grad_norm": 0.9937820434570312, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 46022 + }, + { + "epoch": 0.46023, + "grad_norm": 1.060505747795105, + "learning_rate": 0.003, + "loss": 3.9417, + "step": 46023 + }, + { + "epoch": 0.46024, + "grad_norm": 0.8357670903205872, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 46024 + }, + { + "epoch": 0.46025, + "grad_norm": 0.7209333181381226, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 46025 + }, + { + "epoch": 0.46026, + "grad_norm": 0.7599536776542664, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 46026 + }, + { + "epoch": 0.46027, + "grad_norm": 0.8561853170394897, + "learning_rate": 0.003, + "loss": 3.973, + "step": 46027 + }, + { + "epoch": 0.46028, + "grad_norm": 0.9214558005332947, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 46028 + }, + { + "epoch": 0.46029, + "grad_norm": 0.9244599342346191, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 46029 + }, + { + "epoch": 0.4603, + "grad_norm": 0.9890385270118713, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 46030 + }, + { + "epoch": 0.46031, + "grad_norm": 0.8612937331199646, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 46031 + }, + { + "epoch": 0.46032, + "grad_norm": 0.8392436504364014, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 46032 + }, + { + "epoch": 0.46033, + "grad_norm": 0.850500226020813, + "learning_rate": 0.003, + "loss": 3.9388, + "step": 46033 + }, + { + "epoch": 0.46034, + "grad_norm": 0.7434311509132385, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 46034 + }, + { + "epoch": 0.46035, + "grad_norm": 0.7869755625724792, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 46035 + }, + { + "epoch": 0.46036, + "grad_norm": 0.8408767580986023, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 46036 + }, + { + "epoch": 0.46037, + "grad_norm": 0.8828970789909363, + "learning_rate": 0.003, + "loss": 3.9469, + "step": 46037 + }, + { + "epoch": 0.46038, + "grad_norm": 0.8566820621490479, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 46038 + }, + { + "epoch": 0.46039, + "grad_norm": 0.7827882170677185, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 46039 + }, + { + "epoch": 0.4604, + "grad_norm": 0.8798697590827942, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 46040 + }, + { + "epoch": 0.46041, + "grad_norm": 1.0445209741592407, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 46041 + }, + { + "epoch": 0.46042, + "grad_norm": 1.306361198425293, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 46042 + }, + { + "epoch": 0.46043, + "grad_norm": 0.7916620969772339, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 46043 + }, + { + "epoch": 0.46044, + "grad_norm": 0.8397964239120483, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 46044 + }, + { + "epoch": 0.46045, + "grad_norm": 0.9263265132904053, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 46045 + }, + { + "epoch": 0.46046, + "grad_norm": 1.0393452644348145, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 46046 + }, + { + "epoch": 0.46047, + "grad_norm": 0.9441190958023071, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 46047 + }, + { + "epoch": 0.46048, + "grad_norm": 1.1144622564315796, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 46048 + }, + { + "epoch": 0.46049, + "grad_norm": 0.823337733745575, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 46049 + }, + { + "epoch": 0.4605, + "grad_norm": 0.8178669214248657, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 46050 + }, + { + "epoch": 0.46051, + "grad_norm": 0.9038987159729004, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 46051 + }, + { + "epoch": 0.46052, + "grad_norm": 1.0169492959976196, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 46052 + }, + { + "epoch": 0.46053, + "grad_norm": 1.0756475925445557, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 46053 + }, + { + "epoch": 0.46054, + "grad_norm": 0.7758788466453552, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 46054 + }, + { + "epoch": 0.46055, + "grad_norm": 0.7748019695281982, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 46055 + }, + { + "epoch": 0.46056, + "grad_norm": 0.72661954164505, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 46056 + }, + { + "epoch": 0.46057, + "grad_norm": 0.7727383375167847, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 46057 + }, + { + "epoch": 0.46058, + "grad_norm": 1.010703682899475, + "learning_rate": 0.003, + "loss": 3.976, + "step": 46058 + }, + { + "epoch": 0.46059, + "grad_norm": 1.1566882133483887, + "learning_rate": 0.003, + "loss": 3.959, + "step": 46059 + }, + { + "epoch": 0.4606, + "grad_norm": 0.923308253288269, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 46060 + }, + { + "epoch": 0.46061, + "grad_norm": 1.2788609266281128, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 46061 + }, + { + "epoch": 0.46062, + "grad_norm": 0.9956508278846741, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 46062 + }, + { + "epoch": 0.46063, + "grad_norm": 0.6952483057975769, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 46063 + }, + { + "epoch": 0.46064, + "grad_norm": 0.6265191435813904, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 46064 + }, + { + "epoch": 0.46065, + "grad_norm": 0.7916671633720398, + "learning_rate": 0.003, + "loss": 3.996, + "step": 46065 + }, + { + "epoch": 0.46066, + "grad_norm": 1.1820069551467896, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 46066 + }, + { + "epoch": 0.46067, + "grad_norm": 1.049347162246704, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 46067 + }, + { + "epoch": 0.46068, + "grad_norm": 1.0065103769302368, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 46068 + }, + { + "epoch": 0.46069, + "grad_norm": 0.9332466721534729, + "learning_rate": 0.003, + "loss": 3.954, + "step": 46069 + }, + { + "epoch": 0.4607, + "grad_norm": 0.8871339559555054, + "learning_rate": 0.003, + "loss": 3.9213, + "step": 46070 + }, + { + "epoch": 0.46071, + "grad_norm": 0.8468366265296936, + "learning_rate": 0.003, + "loss": 3.9317, + "step": 46071 + }, + { + "epoch": 0.46072, + "grad_norm": 1.0015184879302979, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 46072 + }, + { + "epoch": 0.46073, + "grad_norm": 1.1177334785461426, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 46073 + }, + { + "epoch": 0.46074, + "grad_norm": 0.7900405526161194, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 46074 + }, + { + "epoch": 0.46075, + "grad_norm": 0.6968677639961243, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 46075 + }, + { + "epoch": 0.46076, + "grad_norm": 0.7676705121994019, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 46076 + }, + { + "epoch": 0.46077, + "grad_norm": 0.807933509349823, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 46077 + }, + { + "epoch": 0.46078, + "grad_norm": 0.8078253269195557, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 46078 + }, + { + "epoch": 0.46079, + "grad_norm": 1.000769019126892, + "learning_rate": 0.003, + "loss": 4.02, + "step": 46079 + }, + { + "epoch": 0.4608, + "grad_norm": 1.198486089706421, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 46080 + }, + { + "epoch": 0.46081, + "grad_norm": 0.9614951610565186, + "learning_rate": 0.003, + "loss": 3.9423, + "step": 46081 + }, + { + "epoch": 0.46082, + "grad_norm": 0.950495183467865, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 46082 + }, + { + "epoch": 0.46083, + "grad_norm": 0.9182049632072449, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 46083 + }, + { + "epoch": 0.46084, + "grad_norm": 0.9331880211830139, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 46084 + }, + { + "epoch": 0.46085, + "grad_norm": 0.9972879886627197, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 46085 + }, + { + "epoch": 0.46086, + "grad_norm": 0.9940769672393799, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 46086 + }, + { + "epoch": 0.46087, + "grad_norm": 0.8936214447021484, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 46087 + }, + { + "epoch": 0.46088, + "grad_norm": 0.7936488389968872, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 46088 + }, + { + "epoch": 0.46089, + "grad_norm": 1.0625396966934204, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 46089 + }, + { + "epoch": 0.4609, + "grad_norm": 1.247354507446289, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 46090 + }, + { + "epoch": 0.46091, + "grad_norm": 0.851883590221405, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 46091 + }, + { + "epoch": 0.46092, + "grad_norm": 0.641970157623291, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 46092 + }, + { + "epoch": 0.46093, + "grad_norm": 0.6929738521575928, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 46093 + }, + { + "epoch": 0.46094, + "grad_norm": 0.815655529499054, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 46094 + }, + { + "epoch": 0.46095, + "grad_norm": 0.7660572528839111, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 46095 + }, + { + "epoch": 0.46096, + "grad_norm": 0.8244639039039612, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 46096 + }, + { + "epoch": 0.46097, + "grad_norm": 0.8689066171646118, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 46097 + }, + { + "epoch": 0.46098, + "grad_norm": 0.923286497592926, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 46098 + }, + { + "epoch": 0.46099, + "grad_norm": 1.0331159830093384, + "learning_rate": 0.003, + "loss": 3.973, + "step": 46099 + }, + { + "epoch": 0.461, + "grad_norm": 0.8528053164482117, + "learning_rate": 0.003, + "loss": 3.961, + "step": 46100 + }, + { + "epoch": 0.46101, + "grad_norm": 0.7965294122695923, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 46101 + }, + { + "epoch": 0.46102, + "grad_norm": 1.0047067403793335, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 46102 + }, + { + "epoch": 0.46103, + "grad_norm": 1.250788688659668, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 46103 + }, + { + "epoch": 0.46104, + "grad_norm": 0.8204680681228638, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 46104 + }, + { + "epoch": 0.46105, + "grad_norm": 1.0890556573867798, + "learning_rate": 0.003, + "loss": 3.983, + "step": 46105 + }, + { + "epoch": 0.46106, + "grad_norm": 0.9316959381103516, + "learning_rate": 0.003, + "loss": 3.975, + "step": 46106 + }, + { + "epoch": 0.46107, + "grad_norm": 1.2636743783950806, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 46107 + }, + { + "epoch": 0.46108, + "grad_norm": 1.1955840587615967, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 46108 + }, + { + "epoch": 0.46109, + "grad_norm": 0.691918671131134, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 46109 + }, + { + "epoch": 0.4611, + "grad_norm": 0.6941564679145813, + "learning_rate": 0.003, + "loss": 3.997, + "step": 46110 + }, + { + "epoch": 0.46111, + "grad_norm": 0.8485354781150818, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 46111 + }, + { + "epoch": 0.46112, + "grad_norm": 0.9903469681739807, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 46112 + }, + { + "epoch": 0.46113, + "grad_norm": 1.2976019382476807, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 46113 + }, + { + "epoch": 0.46114, + "grad_norm": 0.8193108439445496, + "learning_rate": 0.003, + "loss": 4.002, + "step": 46114 + }, + { + "epoch": 0.46115, + "grad_norm": 0.8700723052024841, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 46115 + }, + { + "epoch": 0.46116, + "grad_norm": 0.8810548186302185, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 46116 + }, + { + "epoch": 0.46117, + "grad_norm": 0.9158909320831299, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 46117 + }, + { + "epoch": 0.46118, + "grad_norm": 0.9564760327339172, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 46118 + }, + { + "epoch": 0.46119, + "grad_norm": 0.8992394804954529, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 46119 + }, + { + "epoch": 0.4612, + "grad_norm": 0.7426405549049377, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 46120 + }, + { + "epoch": 0.46121, + "grad_norm": 0.6507380604743958, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 46121 + }, + { + "epoch": 0.46122, + "grad_norm": 0.7289459109306335, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 46122 + }, + { + "epoch": 0.46123, + "grad_norm": 0.8085474371910095, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 46123 + }, + { + "epoch": 0.46124, + "grad_norm": 0.9006919860839844, + "learning_rate": 0.003, + "loss": 3.971, + "step": 46124 + }, + { + "epoch": 0.46125, + "grad_norm": 0.8992087841033936, + "learning_rate": 0.003, + "loss": 3.9466, + "step": 46125 + }, + { + "epoch": 0.46126, + "grad_norm": 1.1330715417861938, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 46126 + }, + { + "epoch": 0.46127, + "grad_norm": 1.0309317111968994, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 46127 + }, + { + "epoch": 0.46128, + "grad_norm": 0.9664515852928162, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 46128 + }, + { + "epoch": 0.46129, + "grad_norm": 0.9039918780326843, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 46129 + }, + { + "epoch": 0.4613, + "grad_norm": 0.8551138639450073, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 46130 + }, + { + "epoch": 0.46131, + "grad_norm": 0.8491181135177612, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 46131 + }, + { + "epoch": 0.46132, + "grad_norm": 0.890883207321167, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 46132 + }, + { + "epoch": 0.46133, + "grad_norm": 0.8827177882194519, + "learning_rate": 0.003, + "loss": 3.98, + "step": 46133 + }, + { + "epoch": 0.46134, + "grad_norm": 0.912628710269928, + "learning_rate": 0.003, + "loss": 4.006, + "step": 46134 + }, + { + "epoch": 0.46135, + "grad_norm": 0.9067979454994202, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 46135 + }, + { + "epoch": 0.46136, + "grad_norm": 0.8254245519638062, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 46136 + }, + { + "epoch": 0.46137, + "grad_norm": 0.9680760502815247, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 46137 + }, + { + "epoch": 0.46138, + "grad_norm": 1.069095492362976, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 46138 + }, + { + "epoch": 0.46139, + "grad_norm": 0.892056941986084, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 46139 + }, + { + "epoch": 0.4614, + "grad_norm": 0.7759663462638855, + "learning_rate": 0.003, + "loss": 3.993, + "step": 46140 + }, + { + "epoch": 0.46141, + "grad_norm": 0.7153083086013794, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 46141 + }, + { + "epoch": 0.46142, + "grad_norm": 0.8064717054367065, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 46142 + }, + { + "epoch": 0.46143, + "grad_norm": 0.8802264928817749, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 46143 + }, + { + "epoch": 0.46144, + "grad_norm": 0.9824973940849304, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 46144 + }, + { + "epoch": 0.46145, + "grad_norm": 1.051408052444458, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 46145 + }, + { + "epoch": 0.46146, + "grad_norm": 1.251437783241272, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 46146 + }, + { + "epoch": 0.46147, + "grad_norm": 0.9226365685462952, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 46147 + }, + { + "epoch": 0.46148, + "grad_norm": 0.8119087815284729, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 46148 + }, + { + "epoch": 0.46149, + "grad_norm": 0.8914626836776733, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 46149 + }, + { + "epoch": 0.4615, + "grad_norm": 0.9168924689292908, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 46150 + }, + { + "epoch": 0.46151, + "grad_norm": 0.8637971878051758, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 46151 + }, + { + "epoch": 0.46152, + "grad_norm": 0.8292701840400696, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 46152 + }, + { + "epoch": 0.46153, + "grad_norm": 0.976475715637207, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 46153 + }, + { + "epoch": 0.46154, + "grad_norm": 0.9739672541618347, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 46154 + }, + { + "epoch": 0.46155, + "grad_norm": 0.9851205945014954, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 46155 + }, + { + "epoch": 0.46156, + "grad_norm": 1.2587014436721802, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 46156 + }, + { + "epoch": 0.46157, + "grad_norm": 0.861610472202301, + "learning_rate": 0.003, + "loss": 3.992, + "step": 46157 + }, + { + "epoch": 0.46158, + "grad_norm": 0.8381105661392212, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 46158 + }, + { + "epoch": 0.46159, + "grad_norm": 0.8101076483726501, + "learning_rate": 0.003, + "loss": 3.9429, + "step": 46159 + }, + { + "epoch": 0.4616, + "grad_norm": 0.8149569630622864, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 46160 + }, + { + "epoch": 0.46161, + "grad_norm": 0.9507067799568176, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 46161 + }, + { + "epoch": 0.46162, + "grad_norm": 0.9204705357551575, + "learning_rate": 0.003, + "loss": 3.97, + "step": 46162 + }, + { + "epoch": 0.46163, + "grad_norm": 1.099467158317566, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 46163 + }, + { + "epoch": 0.46164, + "grad_norm": 1.0221482515335083, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 46164 + }, + { + "epoch": 0.46165, + "grad_norm": 0.9883039593696594, + "learning_rate": 0.003, + "loss": 3.991, + "step": 46165 + }, + { + "epoch": 0.46166, + "grad_norm": 1.0108990669250488, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 46166 + }, + { + "epoch": 0.46167, + "grad_norm": 1.068774700164795, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 46167 + }, + { + "epoch": 0.46168, + "grad_norm": 0.8611086010932922, + "learning_rate": 0.003, + "loss": 3.9434, + "step": 46168 + }, + { + "epoch": 0.46169, + "grad_norm": 0.7494580149650574, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 46169 + }, + { + "epoch": 0.4617, + "grad_norm": 0.7181496024131775, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 46170 + }, + { + "epoch": 0.46171, + "grad_norm": 0.8806577920913696, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 46171 + }, + { + "epoch": 0.46172, + "grad_norm": 0.8403581976890564, + "learning_rate": 0.003, + "loss": 3.971, + "step": 46172 + }, + { + "epoch": 0.46173, + "grad_norm": 0.699421226978302, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 46173 + }, + { + "epoch": 0.46174, + "grad_norm": 0.7399448156356812, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 46174 + }, + { + "epoch": 0.46175, + "grad_norm": 1.0011519193649292, + "learning_rate": 0.003, + "loss": 3.9273, + "step": 46175 + }, + { + "epoch": 0.46176, + "grad_norm": 1.2396886348724365, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 46176 + }, + { + "epoch": 0.46177, + "grad_norm": 0.7165960669517517, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 46177 + }, + { + "epoch": 0.46178, + "grad_norm": 0.628595769405365, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 46178 + }, + { + "epoch": 0.46179, + "grad_norm": 0.7337397336959839, + "learning_rate": 0.003, + "loss": 3.979, + "step": 46179 + }, + { + "epoch": 0.4618, + "grad_norm": 0.8066457509994507, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 46180 + }, + { + "epoch": 0.46181, + "grad_norm": 0.7639902830123901, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 46181 + }, + { + "epoch": 0.46182, + "grad_norm": 0.7608767151832581, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 46182 + }, + { + "epoch": 0.46183, + "grad_norm": 0.7853053212165833, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 46183 + }, + { + "epoch": 0.46184, + "grad_norm": 0.7792178392410278, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 46184 + }, + { + "epoch": 0.46185, + "grad_norm": 0.8260195851325989, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 46185 + }, + { + "epoch": 0.46186, + "grad_norm": 1.1730915307998657, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 46186 + }, + { + "epoch": 0.46187, + "grad_norm": 1.0212199687957764, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 46187 + }, + { + "epoch": 0.46188, + "grad_norm": 1.1453512907028198, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 46188 + }, + { + "epoch": 0.46189, + "grad_norm": 0.8298001885414124, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 46189 + }, + { + "epoch": 0.4619, + "grad_norm": 0.7467883229255676, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 46190 + }, + { + "epoch": 0.46191, + "grad_norm": 0.8135713338851929, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 46191 + }, + { + "epoch": 0.46192, + "grad_norm": 0.9358211159706116, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 46192 + }, + { + "epoch": 0.46193, + "grad_norm": 0.9680438041687012, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 46193 + }, + { + "epoch": 0.46194, + "grad_norm": 1.0373555421829224, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 46194 + }, + { + "epoch": 0.46195, + "grad_norm": 1.0846076011657715, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 46195 + }, + { + "epoch": 0.46196, + "grad_norm": 0.9900847673416138, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 46196 + }, + { + "epoch": 0.46197, + "grad_norm": 1.0323936939239502, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 46197 + }, + { + "epoch": 0.46198, + "grad_norm": 1.0260196924209595, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 46198 + }, + { + "epoch": 0.46199, + "grad_norm": 1.211952567100525, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 46199 + }, + { + "epoch": 0.462, + "grad_norm": 0.818801760673523, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 46200 + }, + { + "epoch": 0.46201, + "grad_norm": 0.762606680393219, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 46201 + }, + { + "epoch": 0.46202, + "grad_norm": 0.8593286275863647, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 46202 + }, + { + "epoch": 0.46203, + "grad_norm": 1.110634207725525, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 46203 + }, + { + "epoch": 0.46204, + "grad_norm": 1.0685056447982788, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 46204 + }, + { + "epoch": 0.46205, + "grad_norm": 0.9964243769645691, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 46205 + }, + { + "epoch": 0.46206, + "grad_norm": 1.1217540502548218, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 46206 + }, + { + "epoch": 0.46207, + "grad_norm": 1.057356834411621, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 46207 + }, + { + "epoch": 0.46208, + "grad_norm": 1.005208134651184, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 46208 + }, + { + "epoch": 0.46209, + "grad_norm": 0.9667206406593323, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 46209 + }, + { + "epoch": 0.4621, + "grad_norm": 1.0400360822677612, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 46210 + }, + { + "epoch": 0.46211, + "grad_norm": 1.0525368452072144, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 46211 + }, + { + "epoch": 0.46212, + "grad_norm": 1.0176022052764893, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 46212 + }, + { + "epoch": 0.46213, + "grad_norm": 0.7510091066360474, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 46213 + }, + { + "epoch": 0.46214, + "grad_norm": 0.7301061749458313, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 46214 + }, + { + "epoch": 0.46215, + "grad_norm": 0.6875208020210266, + "learning_rate": 0.003, + "loss": 3.9471, + "step": 46215 + }, + { + "epoch": 0.46216, + "grad_norm": 0.6714613437652588, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 46216 + }, + { + "epoch": 0.46217, + "grad_norm": 0.6612461805343628, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 46217 + }, + { + "epoch": 0.46218, + "grad_norm": 0.5885190963745117, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 46218 + }, + { + "epoch": 0.46219, + "grad_norm": 0.6574578881263733, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 46219 + }, + { + "epoch": 0.4622, + "grad_norm": 0.7346352934837341, + "learning_rate": 0.003, + "loss": 3.963, + "step": 46220 + }, + { + "epoch": 0.46221, + "grad_norm": 0.8062716722488403, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 46221 + }, + { + "epoch": 0.46222, + "grad_norm": 0.6694270968437195, + "learning_rate": 0.003, + "loss": 3.971, + "step": 46222 + }, + { + "epoch": 0.46223, + "grad_norm": 0.6895884275436401, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 46223 + }, + { + "epoch": 0.46224, + "grad_norm": 0.8705419301986694, + "learning_rate": 0.003, + "loss": 3.972, + "step": 46224 + }, + { + "epoch": 0.46225, + "grad_norm": 1.0148171186447144, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 46225 + }, + { + "epoch": 0.46226, + "grad_norm": 1.159500241279602, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 46226 + }, + { + "epoch": 0.46227, + "grad_norm": 0.8471522331237793, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 46227 + }, + { + "epoch": 0.46228, + "grad_norm": 0.8061774969100952, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 46228 + }, + { + "epoch": 0.46229, + "grad_norm": 0.7961365580558777, + "learning_rate": 0.003, + "loss": 3.963, + "step": 46229 + }, + { + "epoch": 0.4623, + "grad_norm": 0.9339322447776794, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 46230 + }, + { + "epoch": 0.46231, + "grad_norm": 1.2852057218551636, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 46231 + }, + { + "epoch": 0.46232, + "grad_norm": 1.1620569229125977, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 46232 + }, + { + "epoch": 0.46233, + "grad_norm": 0.9589625000953674, + "learning_rate": 0.003, + "loss": 3.99, + "step": 46233 + }, + { + "epoch": 0.46234, + "grad_norm": 0.8667551875114441, + "learning_rate": 0.003, + "loss": 3.983, + "step": 46234 + }, + { + "epoch": 0.46235, + "grad_norm": 0.8064991235733032, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 46235 + }, + { + "epoch": 0.46236, + "grad_norm": 0.8103219270706177, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 46236 + }, + { + "epoch": 0.46237, + "grad_norm": 0.899286687374115, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 46237 + }, + { + "epoch": 0.46238, + "grad_norm": 0.9008237719535828, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 46238 + }, + { + "epoch": 0.46239, + "grad_norm": 0.855330765247345, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 46239 + }, + { + "epoch": 0.4624, + "grad_norm": 0.972183108329773, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 46240 + }, + { + "epoch": 0.46241, + "grad_norm": 0.8853726983070374, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 46241 + }, + { + "epoch": 0.46242, + "grad_norm": 0.9192818999290466, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 46242 + }, + { + "epoch": 0.46243, + "grad_norm": 0.9581096172332764, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 46243 + }, + { + "epoch": 0.46244, + "grad_norm": 1.0227503776550293, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 46244 + }, + { + "epoch": 0.46245, + "grad_norm": 1.1521856784820557, + "learning_rate": 0.003, + "loss": 3.9396, + "step": 46245 + }, + { + "epoch": 0.46246, + "grad_norm": 1.2811546325683594, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 46246 + }, + { + "epoch": 0.46247, + "grad_norm": 0.7836146354675293, + "learning_rate": 0.003, + "loss": 3.9267, + "step": 46247 + }, + { + "epoch": 0.46248, + "grad_norm": 0.7112628817558289, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 46248 + }, + { + "epoch": 0.46249, + "grad_norm": 0.7755295634269714, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 46249 + }, + { + "epoch": 0.4625, + "grad_norm": 0.9640513062477112, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 46250 + }, + { + "epoch": 0.46251, + "grad_norm": 1.2799532413482666, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 46251 + }, + { + "epoch": 0.46252, + "grad_norm": 0.8575140237808228, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 46252 + }, + { + "epoch": 0.46253, + "grad_norm": 0.8872475624084473, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 46253 + }, + { + "epoch": 0.46254, + "grad_norm": 0.9996691346168518, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 46254 + }, + { + "epoch": 0.46255, + "grad_norm": 1.169201135635376, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 46255 + }, + { + "epoch": 0.46256, + "grad_norm": 0.8167496919631958, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 46256 + }, + { + "epoch": 0.46257, + "grad_norm": 0.7316354513168335, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 46257 + }, + { + "epoch": 0.46258, + "grad_norm": 0.6136427521705627, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 46258 + }, + { + "epoch": 0.46259, + "grad_norm": 0.554825484752655, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 46259 + }, + { + "epoch": 0.4626, + "grad_norm": 0.626937985420227, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 46260 + }, + { + "epoch": 0.46261, + "grad_norm": 0.6951425671577454, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 46261 + }, + { + "epoch": 0.46262, + "grad_norm": 0.8130985498428345, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 46262 + }, + { + "epoch": 0.46263, + "grad_norm": 1.1080148220062256, + "learning_rate": 0.003, + "loss": 3.972, + "step": 46263 + }, + { + "epoch": 0.46264, + "grad_norm": 1.3417434692382812, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 46264 + }, + { + "epoch": 0.46265, + "grad_norm": 0.6399705410003662, + "learning_rate": 0.003, + "loss": 3.966, + "step": 46265 + }, + { + "epoch": 0.46266, + "grad_norm": 0.6544876098632812, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 46266 + }, + { + "epoch": 0.46267, + "grad_norm": 0.7276449799537659, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 46267 + }, + { + "epoch": 0.46268, + "grad_norm": 0.9049476981163025, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 46268 + }, + { + "epoch": 0.46269, + "grad_norm": 1.0390077829360962, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 46269 + }, + { + "epoch": 0.4627, + "grad_norm": 0.9580690860748291, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 46270 + }, + { + "epoch": 0.46271, + "grad_norm": 0.9645465016365051, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 46271 + }, + { + "epoch": 0.46272, + "grad_norm": 0.9078798890113831, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 46272 + }, + { + "epoch": 0.46273, + "grad_norm": 0.8607866764068604, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 46273 + }, + { + "epoch": 0.46274, + "grad_norm": 0.8665627241134644, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 46274 + }, + { + "epoch": 0.46275, + "grad_norm": 0.9540613889694214, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 46275 + }, + { + "epoch": 0.46276, + "grad_norm": 1.0378988981246948, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 46276 + }, + { + "epoch": 0.46277, + "grad_norm": 0.9981834888458252, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 46277 + }, + { + "epoch": 0.46278, + "grad_norm": 0.909156322479248, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 46278 + }, + { + "epoch": 0.46279, + "grad_norm": 0.8435311317443848, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 46279 + }, + { + "epoch": 0.4628, + "grad_norm": 1.1314972639083862, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 46280 + }, + { + "epoch": 0.46281, + "grad_norm": 0.9332759380340576, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 46281 + }, + { + "epoch": 0.46282, + "grad_norm": 1.0324904918670654, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 46282 + }, + { + "epoch": 0.46283, + "grad_norm": 1.0158697366714478, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 46283 + }, + { + "epoch": 0.46284, + "grad_norm": 1.1138441562652588, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 46284 + }, + { + "epoch": 0.46285, + "grad_norm": 1.0548648834228516, + "learning_rate": 0.003, + "loss": 3.9433, + "step": 46285 + }, + { + "epoch": 0.46286, + "grad_norm": 0.9207664132118225, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 46286 + }, + { + "epoch": 0.46287, + "grad_norm": 1.0846872329711914, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 46287 + }, + { + "epoch": 0.46288, + "grad_norm": 1.1451945304870605, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 46288 + }, + { + "epoch": 0.46289, + "grad_norm": 0.8233701586723328, + "learning_rate": 0.003, + "loss": 3.957, + "step": 46289 + }, + { + "epoch": 0.4629, + "grad_norm": 0.8233515024185181, + "learning_rate": 0.003, + "loss": 3.982, + "step": 46290 + }, + { + "epoch": 0.46291, + "grad_norm": 0.7397721409797668, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 46291 + }, + { + "epoch": 0.46292, + "grad_norm": 0.9231808185577393, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 46292 + }, + { + "epoch": 0.46293, + "grad_norm": 0.8805856108665466, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 46293 + }, + { + "epoch": 0.46294, + "grad_norm": 0.8200421333312988, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 46294 + }, + { + "epoch": 0.46295, + "grad_norm": 0.9453662037849426, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 46295 + }, + { + "epoch": 0.46296, + "grad_norm": 1.0741485357284546, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 46296 + }, + { + "epoch": 0.46297, + "grad_norm": 1.0072283744812012, + "learning_rate": 0.003, + "loss": 4.017, + "step": 46297 + }, + { + "epoch": 0.46298, + "grad_norm": 1.0823484659194946, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 46298 + }, + { + "epoch": 0.46299, + "grad_norm": 0.874153733253479, + "learning_rate": 0.003, + "loss": 3.999, + "step": 46299 + }, + { + "epoch": 0.463, + "grad_norm": 1.0079692602157593, + "learning_rate": 0.003, + "loss": 3.955, + "step": 46300 + }, + { + "epoch": 0.46301, + "grad_norm": 1.2370860576629639, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 46301 + }, + { + "epoch": 0.46302, + "grad_norm": 0.8953366279602051, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 46302 + }, + { + "epoch": 0.46303, + "grad_norm": 0.7908660173416138, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 46303 + }, + { + "epoch": 0.46304, + "grad_norm": 0.7064836025238037, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 46304 + }, + { + "epoch": 0.46305, + "grad_norm": 0.6549015045166016, + "learning_rate": 0.003, + "loss": 3.9366, + "step": 46305 + }, + { + "epoch": 0.46306, + "grad_norm": 0.8390113711357117, + "learning_rate": 0.003, + "loss": 3.9441, + "step": 46306 + }, + { + "epoch": 0.46307, + "grad_norm": 1.089749813079834, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 46307 + }, + { + "epoch": 0.46308, + "grad_norm": 0.9672232866287231, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 46308 + }, + { + "epoch": 0.46309, + "grad_norm": 0.8764602541923523, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 46309 + }, + { + "epoch": 0.4631, + "grad_norm": 0.7936117053031921, + "learning_rate": 0.003, + "loss": 3.985, + "step": 46310 + }, + { + "epoch": 0.46311, + "grad_norm": 0.8210395574569702, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 46311 + }, + { + "epoch": 0.46312, + "grad_norm": 1.0447872877120972, + "learning_rate": 0.003, + "loss": 3.96, + "step": 46312 + }, + { + "epoch": 0.46313, + "grad_norm": 1.2327622175216675, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 46313 + }, + { + "epoch": 0.46314, + "grad_norm": 0.8101788759231567, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 46314 + }, + { + "epoch": 0.46315, + "grad_norm": 0.6617360711097717, + "learning_rate": 0.003, + "loss": 3.9292, + "step": 46315 + }, + { + "epoch": 0.46316, + "grad_norm": 0.8396989703178406, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 46316 + }, + { + "epoch": 0.46317, + "grad_norm": 0.9591695070266724, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 46317 + }, + { + "epoch": 0.46318, + "grad_norm": 1.2891249656677246, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 46318 + }, + { + "epoch": 0.46319, + "grad_norm": 0.8431392908096313, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 46319 + }, + { + "epoch": 0.4632, + "grad_norm": 0.7883357405662537, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 46320 + }, + { + "epoch": 0.46321, + "grad_norm": 0.857643723487854, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 46321 + }, + { + "epoch": 0.46322, + "grad_norm": 0.908277153968811, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 46322 + }, + { + "epoch": 0.46323, + "grad_norm": 0.9411613941192627, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 46323 + }, + { + "epoch": 0.46324, + "grad_norm": 1.061408281326294, + "learning_rate": 0.003, + "loss": 3.991, + "step": 46324 + }, + { + "epoch": 0.46325, + "grad_norm": 0.886515200138092, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 46325 + }, + { + "epoch": 0.46326, + "grad_norm": 0.7184544801712036, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 46326 + }, + { + "epoch": 0.46327, + "grad_norm": 0.687934160232544, + "learning_rate": 0.003, + "loss": 3.9392, + "step": 46327 + }, + { + "epoch": 0.46328, + "grad_norm": 0.7769541144371033, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 46328 + }, + { + "epoch": 0.46329, + "grad_norm": 0.8235530257225037, + "learning_rate": 0.003, + "loss": 3.9306, + "step": 46329 + }, + { + "epoch": 0.4633, + "grad_norm": 0.9099599123001099, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 46330 + }, + { + "epoch": 0.46331, + "grad_norm": 1.0929557085037231, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 46331 + }, + { + "epoch": 0.46332, + "grad_norm": 1.1549080610275269, + "learning_rate": 0.003, + "loss": 3.9432, + "step": 46332 + }, + { + "epoch": 0.46333, + "grad_norm": 0.9155864119529724, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 46333 + }, + { + "epoch": 0.46334, + "grad_norm": 0.7664458751678467, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 46334 + }, + { + "epoch": 0.46335, + "grad_norm": 0.8621896505355835, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 46335 + }, + { + "epoch": 0.46336, + "grad_norm": 0.8837807774543762, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 46336 + }, + { + "epoch": 0.46337, + "grad_norm": 0.8227465748786926, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 46337 + }, + { + "epoch": 0.46338, + "grad_norm": 0.7614863514900208, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 46338 + }, + { + "epoch": 0.46339, + "grad_norm": 0.6846861243247986, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 46339 + }, + { + "epoch": 0.4634, + "grad_norm": 0.753105878829956, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 46340 + }, + { + "epoch": 0.46341, + "grad_norm": 0.8889420628547668, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 46341 + }, + { + "epoch": 0.46342, + "grad_norm": 0.857639491558075, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 46342 + }, + { + "epoch": 0.46343, + "grad_norm": 0.9728814959526062, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 46343 + }, + { + "epoch": 0.46344, + "grad_norm": 1.2198323011398315, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 46344 + }, + { + "epoch": 0.46345, + "grad_norm": 0.9608638882637024, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 46345 + }, + { + "epoch": 0.46346, + "grad_norm": 1.1162196397781372, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 46346 + }, + { + "epoch": 0.46347, + "grad_norm": 0.8666617274284363, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 46347 + }, + { + "epoch": 0.46348, + "grad_norm": 0.8437503576278687, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 46348 + }, + { + "epoch": 0.46349, + "grad_norm": 0.8467941880226135, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 46349 + }, + { + "epoch": 0.4635, + "grad_norm": 0.8837792873382568, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 46350 + }, + { + "epoch": 0.46351, + "grad_norm": 0.8289403915405273, + "learning_rate": 0.003, + "loss": 3.9336, + "step": 46351 + }, + { + "epoch": 0.46352, + "grad_norm": 0.8543111085891724, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 46352 + }, + { + "epoch": 0.46353, + "grad_norm": 0.8434215188026428, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 46353 + }, + { + "epoch": 0.46354, + "grad_norm": 1.020151138305664, + "learning_rate": 0.003, + "loss": 3.979, + "step": 46354 + }, + { + "epoch": 0.46355, + "grad_norm": 1.2115075588226318, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 46355 + }, + { + "epoch": 0.46356, + "grad_norm": 1.0110138654708862, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 46356 + }, + { + "epoch": 0.46357, + "grad_norm": 1.1856492757797241, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 46357 + }, + { + "epoch": 0.46358, + "grad_norm": 0.9068533182144165, + "learning_rate": 0.003, + "loss": 4.001, + "step": 46358 + }, + { + "epoch": 0.46359, + "grad_norm": 0.8893042206764221, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 46359 + }, + { + "epoch": 0.4636, + "grad_norm": 1.1793591976165771, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 46360 + }, + { + "epoch": 0.46361, + "grad_norm": 1.2497867345809937, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 46361 + }, + { + "epoch": 0.46362, + "grad_norm": 0.7727962732315063, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 46362 + }, + { + "epoch": 0.46363, + "grad_norm": 0.6133092045783997, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 46363 + }, + { + "epoch": 0.46364, + "grad_norm": 0.6783736348152161, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 46364 + }, + { + "epoch": 0.46365, + "grad_norm": 0.629037618637085, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 46365 + }, + { + "epoch": 0.46366, + "grad_norm": 0.6450366973876953, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 46366 + }, + { + "epoch": 0.46367, + "grad_norm": 0.7664740681648254, + "learning_rate": 0.003, + "loss": 3.9467, + "step": 46367 + }, + { + "epoch": 0.46368, + "grad_norm": 0.9030035138130188, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 46368 + }, + { + "epoch": 0.46369, + "grad_norm": 1.0664961338043213, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 46369 + }, + { + "epoch": 0.4637, + "grad_norm": 1.0555299520492554, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 46370 + }, + { + "epoch": 0.46371, + "grad_norm": 0.8247045278549194, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 46371 + }, + { + "epoch": 0.46372, + "grad_norm": 0.8224710822105408, + "learning_rate": 0.003, + "loss": 3.977, + "step": 46372 + }, + { + "epoch": 0.46373, + "grad_norm": 0.8822656273841858, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 46373 + }, + { + "epoch": 0.46374, + "grad_norm": 0.888900637626648, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 46374 + }, + { + "epoch": 0.46375, + "grad_norm": 0.7708815336227417, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 46375 + }, + { + "epoch": 0.46376, + "grad_norm": 0.8532140851020813, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 46376 + }, + { + "epoch": 0.46377, + "grad_norm": 1.092708945274353, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 46377 + }, + { + "epoch": 0.46378, + "grad_norm": 0.9767019152641296, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 46378 + }, + { + "epoch": 0.46379, + "grad_norm": 0.7932106256484985, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 46379 + }, + { + "epoch": 0.4638, + "grad_norm": 0.7030239105224609, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 46380 + }, + { + "epoch": 0.46381, + "grad_norm": 0.8176241517066956, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 46381 + }, + { + "epoch": 0.46382, + "grad_norm": 0.8432808518409729, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 46382 + }, + { + "epoch": 0.46383, + "grad_norm": 0.8495720624923706, + "learning_rate": 0.003, + "loss": 3.9315, + "step": 46383 + }, + { + "epoch": 0.46384, + "grad_norm": 0.9771062731742859, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 46384 + }, + { + "epoch": 0.46385, + "grad_norm": 1.0857216119766235, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 46385 + }, + { + "epoch": 0.46386, + "grad_norm": 1.2307825088500977, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 46386 + }, + { + "epoch": 0.46387, + "grad_norm": 0.7075448036193848, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 46387 + }, + { + "epoch": 0.46388, + "grad_norm": 0.7018188238143921, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 46388 + }, + { + "epoch": 0.46389, + "grad_norm": 0.7211278676986694, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 46389 + }, + { + "epoch": 0.4639, + "grad_norm": 0.6421603560447693, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 46390 + }, + { + "epoch": 0.46391, + "grad_norm": 0.764121413230896, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 46391 + }, + { + "epoch": 0.46392, + "grad_norm": 0.9138480424880981, + "learning_rate": 0.003, + "loss": 3.971, + "step": 46392 + }, + { + "epoch": 0.46393, + "grad_norm": 0.9636936783790588, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 46393 + }, + { + "epoch": 0.46394, + "grad_norm": 1.0884723663330078, + "learning_rate": 0.003, + "loss": 3.9457, + "step": 46394 + }, + { + "epoch": 0.46395, + "grad_norm": 1.1326059103012085, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 46395 + }, + { + "epoch": 0.46396, + "grad_norm": 0.9036628603935242, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 46396 + }, + { + "epoch": 0.46397, + "grad_norm": 0.8769299983978271, + "learning_rate": 0.003, + "loss": 3.966, + "step": 46397 + }, + { + "epoch": 0.46398, + "grad_norm": 0.9768950343132019, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 46398 + }, + { + "epoch": 0.46399, + "grad_norm": 0.9351882338523865, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 46399 + }, + { + "epoch": 0.464, + "grad_norm": 0.9022902846336365, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 46400 + }, + { + "epoch": 0.46401, + "grad_norm": 1.1860065460205078, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 46401 + }, + { + "epoch": 0.46402, + "grad_norm": 1.0395007133483887, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 46402 + }, + { + "epoch": 0.46403, + "grad_norm": 1.1072790622711182, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 46403 + }, + { + "epoch": 0.46404, + "grad_norm": 0.9891121983528137, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 46404 + }, + { + "epoch": 0.46405, + "grad_norm": 0.9807149171829224, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 46405 + }, + { + "epoch": 0.46406, + "grad_norm": 1.2313929796218872, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 46406 + }, + { + "epoch": 0.46407, + "grad_norm": 0.8601237535476685, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 46407 + }, + { + "epoch": 0.46408, + "grad_norm": 0.6969694495201111, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 46408 + }, + { + "epoch": 0.46409, + "grad_norm": 0.6835989356040955, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 46409 + }, + { + "epoch": 0.4641, + "grad_norm": 0.8211799263954163, + "learning_rate": 0.003, + "loss": 3.9426, + "step": 46410 + }, + { + "epoch": 0.46411, + "grad_norm": 0.9025580286979675, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 46411 + }, + { + "epoch": 0.46412, + "grad_norm": 0.880366325378418, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 46412 + }, + { + "epoch": 0.46413, + "grad_norm": 1.1406961679458618, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 46413 + }, + { + "epoch": 0.46414, + "grad_norm": 1.2011942863464355, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 46414 + }, + { + "epoch": 0.46415, + "grad_norm": 0.8500718474388123, + "learning_rate": 0.003, + "loss": 3.941, + "step": 46415 + }, + { + "epoch": 0.46416, + "grad_norm": 0.672173261642456, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 46416 + }, + { + "epoch": 0.46417, + "grad_norm": 0.6900933384895325, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 46417 + }, + { + "epoch": 0.46418, + "grad_norm": 0.7921062111854553, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 46418 + }, + { + "epoch": 0.46419, + "grad_norm": 0.8604868650436401, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 46419 + }, + { + "epoch": 0.4642, + "grad_norm": 0.7882540225982666, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 46420 + }, + { + "epoch": 0.46421, + "grad_norm": 0.7615841627120972, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 46421 + }, + { + "epoch": 0.46422, + "grad_norm": 0.7747772336006165, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 46422 + }, + { + "epoch": 0.46423, + "grad_norm": 0.8467385172843933, + "learning_rate": 0.003, + "loss": 3.94, + "step": 46423 + }, + { + "epoch": 0.46424, + "grad_norm": 1.1297391653060913, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 46424 + }, + { + "epoch": 0.46425, + "grad_norm": 1.1834081411361694, + "learning_rate": 0.003, + "loss": 3.937, + "step": 46425 + }, + { + "epoch": 0.46426, + "grad_norm": 1.099230170249939, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 46426 + }, + { + "epoch": 0.46427, + "grad_norm": 0.9849534034729004, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 46427 + }, + { + "epoch": 0.46428, + "grad_norm": 1.084908366203308, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 46428 + }, + { + "epoch": 0.46429, + "grad_norm": 1.0592362880706787, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 46429 + }, + { + "epoch": 0.4643, + "grad_norm": 1.069355845451355, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 46430 + }, + { + "epoch": 0.46431, + "grad_norm": 0.946488618850708, + "learning_rate": 0.003, + "loss": 3.9409, + "step": 46431 + }, + { + "epoch": 0.46432, + "grad_norm": 0.7926363945007324, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 46432 + }, + { + "epoch": 0.46433, + "grad_norm": 0.6875070929527283, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 46433 + }, + { + "epoch": 0.46434, + "grad_norm": 0.7107881903648376, + "learning_rate": 0.003, + "loss": 3.973, + "step": 46434 + }, + { + "epoch": 0.46435, + "grad_norm": 0.7061731219291687, + "learning_rate": 0.003, + "loss": 3.9446, + "step": 46435 + }, + { + "epoch": 0.46436, + "grad_norm": 0.8260676860809326, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 46436 + }, + { + "epoch": 0.46437, + "grad_norm": 0.815640926361084, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 46437 + }, + { + "epoch": 0.46438, + "grad_norm": 0.8348931670188904, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 46438 + }, + { + "epoch": 0.46439, + "grad_norm": 0.9812100529670715, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 46439 + }, + { + "epoch": 0.4644, + "grad_norm": 1.123993158340454, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 46440 + }, + { + "epoch": 0.46441, + "grad_norm": 0.8024105429649353, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 46441 + }, + { + "epoch": 0.46442, + "grad_norm": 0.7685915231704712, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 46442 + }, + { + "epoch": 0.46443, + "grad_norm": 0.9330641031265259, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 46443 + }, + { + "epoch": 0.46444, + "grad_norm": 1.0825685262680054, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 46444 + }, + { + "epoch": 0.46445, + "grad_norm": 1.067327857017517, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 46445 + }, + { + "epoch": 0.46446, + "grad_norm": 0.8036733865737915, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 46446 + }, + { + "epoch": 0.46447, + "grad_norm": 0.619102954864502, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 46447 + }, + { + "epoch": 0.46448, + "grad_norm": 0.5734238028526306, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 46448 + }, + { + "epoch": 0.46449, + "grad_norm": 0.5950366854667664, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 46449 + }, + { + "epoch": 0.4645, + "grad_norm": 0.6747172474861145, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 46450 + }, + { + "epoch": 0.46451, + "grad_norm": 0.7102271318435669, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 46451 + }, + { + "epoch": 0.46452, + "grad_norm": 0.9238186478614807, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 46452 + }, + { + "epoch": 0.46453, + "grad_norm": 1.4612568616867065, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 46453 + }, + { + "epoch": 0.46454, + "grad_norm": 0.5432174801826477, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 46454 + }, + { + "epoch": 0.46455, + "grad_norm": 0.9225594401359558, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 46455 + }, + { + "epoch": 0.46456, + "grad_norm": 1.232247233390808, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 46456 + }, + { + "epoch": 0.46457, + "grad_norm": 0.7064791917800903, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 46457 + }, + { + "epoch": 0.46458, + "grad_norm": 0.7578127980232239, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 46458 + }, + { + "epoch": 0.46459, + "grad_norm": 0.8587759733200073, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 46459 + }, + { + "epoch": 0.4646, + "grad_norm": 0.8485874533653259, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 46460 + }, + { + "epoch": 0.46461, + "grad_norm": 1.083043098449707, + "learning_rate": 0.003, + "loss": 3.957, + "step": 46461 + }, + { + "epoch": 0.46462, + "grad_norm": 0.9859141707420349, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 46462 + }, + { + "epoch": 0.46463, + "grad_norm": 1.2144384384155273, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 46463 + }, + { + "epoch": 0.46464, + "grad_norm": 0.9034000635147095, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 46464 + }, + { + "epoch": 0.46465, + "grad_norm": 0.8761051893234253, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 46465 + }, + { + "epoch": 0.46466, + "grad_norm": 1.109562873840332, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 46466 + }, + { + "epoch": 0.46467, + "grad_norm": 1.1552635431289673, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 46467 + }, + { + "epoch": 0.46468, + "grad_norm": 0.8482535481452942, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 46468 + }, + { + "epoch": 0.46469, + "grad_norm": 0.7615140676498413, + "learning_rate": 0.003, + "loss": 3.967, + "step": 46469 + }, + { + "epoch": 0.4647, + "grad_norm": 0.981469988822937, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 46470 + }, + { + "epoch": 0.46471, + "grad_norm": 1.2353559732437134, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 46471 + }, + { + "epoch": 0.46472, + "grad_norm": 0.8245687484741211, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 46472 + }, + { + "epoch": 0.46473, + "grad_norm": 0.9432896375656128, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 46473 + }, + { + "epoch": 0.46474, + "grad_norm": 1.0043045282363892, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 46474 + }, + { + "epoch": 0.46475, + "grad_norm": 1.0090497732162476, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 46475 + }, + { + "epoch": 0.46476, + "grad_norm": 1.1639888286590576, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 46476 + }, + { + "epoch": 0.46477, + "grad_norm": 1.0923378467559814, + "learning_rate": 0.003, + "loss": 3.985, + "step": 46477 + }, + { + "epoch": 0.46478, + "grad_norm": 1.1636905670166016, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 46478 + }, + { + "epoch": 0.46479, + "grad_norm": 1.1656666994094849, + "learning_rate": 0.003, + "loss": 3.988, + "step": 46479 + }, + { + "epoch": 0.4648, + "grad_norm": 0.8128100037574768, + "learning_rate": 0.003, + "loss": 3.98, + "step": 46480 + }, + { + "epoch": 0.46481, + "grad_norm": 0.8253087401390076, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 46481 + }, + { + "epoch": 0.46482, + "grad_norm": 1.0267306566238403, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 46482 + }, + { + "epoch": 0.46483, + "grad_norm": 1.1547036170959473, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 46483 + }, + { + "epoch": 0.46484, + "grad_norm": 0.7521960735321045, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 46484 + }, + { + "epoch": 0.46485, + "grad_norm": 0.6941013336181641, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 46485 + }, + { + "epoch": 0.46486, + "grad_norm": 0.8248809576034546, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 46486 + }, + { + "epoch": 0.46487, + "grad_norm": 1.079610824584961, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 46487 + }, + { + "epoch": 0.46488, + "grad_norm": 0.9202857613563538, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 46488 + }, + { + "epoch": 0.46489, + "grad_norm": 0.8961139917373657, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 46489 + }, + { + "epoch": 0.4649, + "grad_norm": 0.8801839351654053, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 46490 + }, + { + "epoch": 0.46491, + "grad_norm": 0.7793250679969788, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 46491 + }, + { + "epoch": 0.46492, + "grad_norm": 0.7029407620429993, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 46492 + }, + { + "epoch": 0.46493, + "grad_norm": 0.7080997228622437, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 46493 + }, + { + "epoch": 0.46494, + "grad_norm": 0.8915551900863647, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 46494 + }, + { + "epoch": 0.46495, + "grad_norm": 1.1577236652374268, + "learning_rate": 0.003, + "loss": 3.9451, + "step": 46495 + }, + { + "epoch": 0.46496, + "grad_norm": 0.8026957511901855, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 46496 + }, + { + "epoch": 0.46497, + "grad_norm": 0.8068748712539673, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 46497 + }, + { + "epoch": 0.46498, + "grad_norm": 0.9320602416992188, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 46498 + }, + { + "epoch": 0.46499, + "grad_norm": 1.2080930471420288, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 46499 + }, + { + "epoch": 0.465, + "grad_norm": 0.9812557697296143, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 46500 + }, + { + "epoch": 0.46501, + "grad_norm": 0.8455795049667358, + "learning_rate": 0.003, + "loss": 3.9366, + "step": 46501 + }, + { + "epoch": 0.46502, + "grad_norm": 0.7014351487159729, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 46502 + }, + { + "epoch": 0.46503, + "grad_norm": 0.7775163054466248, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 46503 + }, + { + "epoch": 0.46504, + "grad_norm": 0.9963870048522949, + "learning_rate": 0.003, + "loss": 3.967, + "step": 46504 + }, + { + "epoch": 0.46505, + "grad_norm": 1.0594971179962158, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 46505 + }, + { + "epoch": 0.46506, + "grad_norm": 0.9814161062240601, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 46506 + }, + { + "epoch": 0.46507, + "grad_norm": 1.051801323890686, + "learning_rate": 0.003, + "loss": 3.969, + "step": 46507 + }, + { + "epoch": 0.46508, + "grad_norm": 1.0275788307189941, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 46508 + }, + { + "epoch": 0.46509, + "grad_norm": 0.9469874501228333, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 46509 + }, + { + "epoch": 0.4651, + "grad_norm": 1.169461965560913, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 46510 + }, + { + "epoch": 0.46511, + "grad_norm": 1.2037562131881714, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 46511 + }, + { + "epoch": 0.46512, + "grad_norm": 0.964085042476654, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 46512 + }, + { + "epoch": 0.46513, + "grad_norm": 0.9523553252220154, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 46513 + }, + { + "epoch": 0.46514, + "grad_norm": 0.7322360873222351, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 46514 + }, + { + "epoch": 0.46515, + "grad_norm": 0.5636183023452759, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 46515 + }, + { + "epoch": 0.46516, + "grad_norm": 0.6198157668113708, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 46516 + }, + { + "epoch": 0.46517, + "grad_norm": 0.8682989478111267, + "learning_rate": 0.003, + "loss": 3.9449, + "step": 46517 + }, + { + "epoch": 0.46518, + "grad_norm": 1.2706544399261475, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 46518 + }, + { + "epoch": 0.46519, + "grad_norm": 0.7417539358139038, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 46519 + }, + { + "epoch": 0.4652, + "grad_norm": 0.6679654717445374, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 46520 + }, + { + "epoch": 0.46521, + "grad_norm": 0.7463940978050232, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 46521 + }, + { + "epoch": 0.46522, + "grad_norm": 0.9283750653266907, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 46522 + }, + { + "epoch": 0.46523, + "grad_norm": 1.1989275217056274, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 46523 + }, + { + "epoch": 0.46524, + "grad_norm": 1.0613511800765991, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 46524 + }, + { + "epoch": 0.46525, + "grad_norm": 0.8278136253356934, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 46525 + }, + { + "epoch": 0.46526, + "grad_norm": 0.6991447806358337, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 46526 + }, + { + "epoch": 0.46527, + "grad_norm": 0.7922685146331787, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 46527 + }, + { + "epoch": 0.46528, + "grad_norm": 0.7667698264122009, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 46528 + }, + { + "epoch": 0.46529, + "grad_norm": 0.8827311992645264, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 46529 + }, + { + "epoch": 0.4653, + "grad_norm": 1.0478507280349731, + "learning_rate": 0.003, + "loss": 3.964, + "step": 46530 + }, + { + "epoch": 0.46531, + "grad_norm": 1.1675291061401367, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 46531 + }, + { + "epoch": 0.46532, + "grad_norm": 0.714341938495636, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 46532 + }, + { + "epoch": 0.46533, + "grad_norm": 0.6079042553901672, + "learning_rate": 0.003, + "loss": 3.9415, + "step": 46533 + }, + { + "epoch": 0.46534, + "grad_norm": 0.8760690689086914, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 46534 + }, + { + "epoch": 0.46535, + "grad_norm": 0.9941390156745911, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 46535 + }, + { + "epoch": 0.46536, + "grad_norm": 1.0527786016464233, + "learning_rate": 0.003, + "loss": 4.024, + "step": 46536 + }, + { + "epoch": 0.46537, + "grad_norm": 0.8953548073768616, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 46537 + }, + { + "epoch": 0.46538, + "grad_norm": 0.9511157870292664, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 46538 + }, + { + "epoch": 0.46539, + "grad_norm": 1.1967486143112183, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 46539 + }, + { + "epoch": 0.4654, + "grad_norm": 0.9332313537597656, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 46540 + }, + { + "epoch": 0.46541, + "grad_norm": 0.9960408806800842, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 46541 + }, + { + "epoch": 0.46542, + "grad_norm": 1.0515835285186768, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 46542 + }, + { + "epoch": 0.46543, + "grad_norm": 1.1169006824493408, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 46543 + }, + { + "epoch": 0.46544, + "grad_norm": 0.8687393665313721, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 46544 + }, + { + "epoch": 0.46545, + "grad_norm": 1.0199118852615356, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 46545 + }, + { + "epoch": 0.46546, + "grad_norm": 1.4189348220825195, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 46546 + }, + { + "epoch": 0.46547, + "grad_norm": 0.5393577814102173, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 46547 + }, + { + "epoch": 0.46548, + "grad_norm": 0.8034079074859619, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 46548 + }, + { + "epoch": 0.46549, + "grad_norm": 0.9014343023300171, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 46549 + }, + { + "epoch": 0.4655, + "grad_norm": 0.9032930135726929, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 46550 + }, + { + "epoch": 0.46551, + "grad_norm": 1.1643915176391602, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 46551 + }, + { + "epoch": 0.46552, + "grad_norm": 0.9803569912910461, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 46552 + }, + { + "epoch": 0.46553, + "grad_norm": 0.8190406560897827, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 46553 + }, + { + "epoch": 0.46554, + "grad_norm": 0.6597708463668823, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 46554 + }, + { + "epoch": 0.46555, + "grad_norm": 0.729087233543396, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 46555 + }, + { + "epoch": 0.46556, + "grad_norm": 0.8466827273368835, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 46556 + }, + { + "epoch": 0.46557, + "grad_norm": 0.7571604251861572, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 46557 + }, + { + "epoch": 0.46558, + "grad_norm": 0.7877128720283508, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 46558 + }, + { + "epoch": 0.46559, + "grad_norm": 0.915368914604187, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 46559 + }, + { + "epoch": 0.4656, + "grad_norm": 0.9677670001983643, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 46560 + }, + { + "epoch": 0.46561, + "grad_norm": 1.0546648502349854, + "learning_rate": 0.003, + "loss": 3.9475, + "step": 46561 + }, + { + "epoch": 0.46562, + "grad_norm": 0.9054595232009888, + "learning_rate": 0.003, + "loss": 3.958, + "step": 46562 + }, + { + "epoch": 0.46563, + "grad_norm": 0.8048871159553528, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 46563 + }, + { + "epoch": 0.46564, + "grad_norm": 0.9722869396209717, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 46564 + }, + { + "epoch": 0.46565, + "grad_norm": 1.237352728843689, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 46565 + }, + { + "epoch": 0.46566, + "grad_norm": 1.10101318359375, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 46566 + }, + { + "epoch": 0.46567, + "grad_norm": 1.0604164600372314, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 46567 + }, + { + "epoch": 0.46568, + "grad_norm": 0.9852860569953918, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 46568 + }, + { + "epoch": 0.46569, + "grad_norm": 1.1830525398254395, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 46569 + }, + { + "epoch": 0.4657, + "grad_norm": 0.934975802898407, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 46570 + }, + { + "epoch": 0.46571, + "grad_norm": 0.8282530903816223, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 46571 + }, + { + "epoch": 0.46572, + "grad_norm": 0.9531124830245972, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 46572 + }, + { + "epoch": 0.46573, + "grad_norm": 0.9460657835006714, + "learning_rate": 0.003, + "loss": 3.9386, + "step": 46573 + }, + { + "epoch": 0.46574, + "grad_norm": 0.9328069090843201, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 46574 + }, + { + "epoch": 0.46575, + "grad_norm": 0.8891114592552185, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 46575 + }, + { + "epoch": 0.46576, + "grad_norm": 0.89876389503479, + "learning_rate": 0.003, + "loss": 3.974, + "step": 46576 + }, + { + "epoch": 0.46577, + "grad_norm": 0.9824531078338623, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 46577 + }, + { + "epoch": 0.46578, + "grad_norm": 1.0688388347625732, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 46578 + }, + { + "epoch": 0.46579, + "grad_norm": 1.0206199884414673, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 46579 + }, + { + "epoch": 0.4658, + "grad_norm": 1.270483374595642, + "learning_rate": 0.003, + "loss": 3.965, + "step": 46580 + }, + { + "epoch": 0.46581, + "grad_norm": 0.6536024808883667, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 46581 + }, + { + "epoch": 0.46582, + "grad_norm": 0.6551581621170044, + "learning_rate": 0.003, + "loss": 3.9489, + "step": 46582 + }, + { + "epoch": 0.46583, + "grad_norm": 0.7318147420883179, + "learning_rate": 0.003, + "loss": 3.968, + "step": 46583 + }, + { + "epoch": 0.46584, + "grad_norm": 0.766103208065033, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 46584 + }, + { + "epoch": 0.46585, + "grad_norm": 0.9389045238494873, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 46585 + }, + { + "epoch": 0.46586, + "grad_norm": 0.985643208026886, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 46586 + }, + { + "epoch": 0.46587, + "grad_norm": 1.0124216079711914, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 46587 + }, + { + "epoch": 0.46588, + "grad_norm": 0.940057098865509, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 46588 + }, + { + "epoch": 0.46589, + "grad_norm": 0.8477296829223633, + "learning_rate": 0.003, + "loss": 3.9445, + "step": 46589 + }, + { + "epoch": 0.4659, + "grad_norm": 0.9954752326011658, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 46590 + }, + { + "epoch": 0.46591, + "grad_norm": 1.0789921283721924, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 46591 + }, + { + "epoch": 0.46592, + "grad_norm": 0.9761752486228943, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 46592 + }, + { + "epoch": 0.46593, + "grad_norm": 0.8124238848686218, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 46593 + }, + { + "epoch": 0.46594, + "grad_norm": 0.8486043214797974, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 46594 + }, + { + "epoch": 0.46595, + "grad_norm": 1.0597895383834839, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 46595 + }, + { + "epoch": 0.46596, + "grad_norm": 1.1023645401000977, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 46596 + }, + { + "epoch": 0.46597, + "grad_norm": 1.1095266342163086, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 46597 + }, + { + "epoch": 0.46598, + "grad_norm": 1.1067854166030884, + "learning_rate": 0.003, + "loss": 4.007, + "step": 46598 + }, + { + "epoch": 0.46599, + "grad_norm": 1.02815842628479, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 46599 + }, + { + "epoch": 0.466, + "grad_norm": 1.033380389213562, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 46600 + }, + { + "epoch": 0.46601, + "grad_norm": 0.9636266827583313, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 46601 + }, + { + "epoch": 0.46602, + "grad_norm": 0.9317029118537903, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 46602 + }, + { + "epoch": 0.46603, + "grad_norm": 0.9331309795379639, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 46603 + }, + { + "epoch": 0.46604, + "grad_norm": 0.8505404591560364, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 46604 + }, + { + "epoch": 0.46605, + "grad_norm": 0.748987078666687, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 46605 + }, + { + "epoch": 0.46606, + "grad_norm": 0.7026761174201965, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 46606 + }, + { + "epoch": 0.46607, + "grad_norm": 0.7806731462478638, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 46607 + }, + { + "epoch": 0.46608, + "grad_norm": 0.9993205666542053, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 46608 + }, + { + "epoch": 0.46609, + "grad_norm": 1.1562131643295288, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 46609 + }, + { + "epoch": 0.4661, + "grad_norm": 0.9510783553123474, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 46610 + }, + { + "epoch": 0.46611, + "grad_norm": 0.9472610950469971, + "learning_rate": 0.003, + "loss": 3.984, + "step": 46611 + }, + { + "epoch": 0.46612, + "grad_norm": 0.8770423531532288, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 46612 + }, + { + "epoch": 0.46613, + "grad_norm": 0.803003191947937, + "learning_rate": 0.003, + "loss": 3.97, + "step": 46613 + }, + { + "epoch": 0.46614, + "grad_norm": 0.8152610063552856, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 46614 + }, + { + "epoch": 0.46615, + "grad_norm": 0.9231858253479004, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 46615 + }, + { + "epoch": 0.46616, + "grad_norm": 0.9942393898963928, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 46616 + }, + { + "epoch": 0.46617, + "grad_norm": 0.8965539336204529, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 46617 + }, + { + "epoch": 0.46618, + "grad_norm": 0.8811935782432556, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 46618 + }, + { + "epoch": 0.46619, + "grad_norm": 0.8857290148735046, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 46619 + }, + { + "epoch": 0.4662, + "grad_norm": 0.8317211270332336, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 46620 + }, + { + "epoch": 0.46621, + "grad_norm": 0.7423998713493347, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 46621 + }, + { + "epoch": 0.46622, + "grad_norm": 0.7663462162017822, + "learning_rate": 0.003, + "loss": 3.9374, + "step": 46622 + }, + { + "epoch": 0.46623, + "grad_norm": 0.7995163798332214, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 46623 + }, + { + "epoch": 0.46624, + "grad_norm": 0.8165999054908752, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 46624 + }, + { + "epoch": 0.46625, + "grad_norm": 0.9943875074386597, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 46625 + }, + { + "epoch": 0.46626, + "grad_norm": 1.2894207239151, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 46626 + }, + { + "epoch": 0.46627, + "grad_norm": 0.6904887557029724, + "learning_rate": 0.003, + "loss": 3.983, + "step": 46627 + }, + { + "epoch": 0.46628, + "grad_norm": 0.5942230820655823, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 46628 + }, + { + "epoch": 0.46629, + "grad_norm": 0.6299581527709961, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 46629 + }, + { + "epoch": 0.4663, + "grad_norm": 0.7163994312286377, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 46630 + }, + { + "epoch": 0.46631, + "grad_norm": 0.7838018536567688, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 46631 + }, + { + "epoch": 0.46632, + "grad_norm": 0.9560039043426514, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 46632 + }, + { + "epoch": 0.46633, + "grad_norm": 1.1971856355667114, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 46633 + }, + { + "epoch": 0.46634, + "grad_norm": 0.9496946334838867, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 46634 + }, + { + "epoch": 0.46635, + "grad_norm": 0.7861132025718689, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 46635 + }, + { + "epoch": 0.46636, + "grad_norm": 0.6906923651695251, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 46636 + }, + { + "epoch": 0.46637, + "grad_norm": 0.8090159296989441, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 46637 + }, + { + "epoch": 0.46638, + "grad_norm": 0.9086380004882812, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 46638 + }, + { + "epoch": 0.46639, + "grad_norm": 0.9240450859069824, + "learning_rate": 0.003, + "loss": 3.9463, + "step": 46639 + }, + { + "epoch": 0.4664, + "grad_norm": 0.9519152641296387, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 46640 + }, + { + "epoch": 0.46641, + "grad_norm": 1.1368926763534546, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 46641 + }, + { + "epoch": 0.46642, + "grad_norm": 0.8656960129737854, + "learning_rate": 0.003, + "loss": 3.972, + "step": 46642 + }, + { + "epoch": 0.46643, + "grad_norm": 0.9196733236312866, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 46643 + }, + { + "epoch": 0.46644, + "grad_norm": 0.8271624445915222, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 46644 + }, + { + "epoch": 0.46645, + "grad_norm": 0.8731718063354492, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 46645 + }, + { + "epoch": 0.46646, + "grad_norm": 1.0592163801193237, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 46646 + }, + { + "epoch": 0.46647, + "grad_norm": 1.0168840885162354, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 46647 + }, + { + "epoch": 0.46648, + "grad_norm": 0.9764248132705688, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 46648 + }, + { + "epoch": 0.46649, + "grad_norm": 0.8734213709831238, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 46649 + }, + { + "epoch": 0.4665, + "grad_norm": 0.7613798379898071, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 46650 + }, + { + "epoch": 0.46651, + "grad_norm": 0.8467116355895996, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 46651 + }, + { + "epoch": 0.46652, + "grad_norm": 0.8379203081130981, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 46652 + }, + { + "epoch": 0.46653, + "grad_norm": 0.9180095195770264, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 46653 + }, + { + "epoch": 0.46654, + "grad_norm": 0.8646616339683533, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 46654 + }, + { + "epoch": 0.46655, + "grad_norm": 0.9967634081840515, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 46655 + }, + { + "epoch": 0.46656, + "grad_norm": 1.2153902053833008, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 46656 + }, + { + "epoch": 0.46657, + "grad_norm": 0.9691111445426941, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 46657 + }, + { + "epoch": 0.46658, + "grad_norm": 0.9551059007644653, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 46658 + }, + { + "epoch": 0.46659, + "grad_norm": 1.0014244318008423, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 46659 + }, + { + "epoch": 0.4666, + "grad_norm": 1.0236091613769531, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 46660 + }, + { + "epoch": 0.46661, + "grad_norm": 1.1743597984313965, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 46661 + }, + { + "epoch": 0.46662, + "grad_norm": 0.8023978471755981, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 46662 + }, + { + "epoch": 0.46663, + "grad_norm": 0.735339879989624, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 46663 + }, + { + "epoch": 0.46664, + "grad_norm": 0.819299578666687, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 46664 + }, + { + "epoch": 0.46665, + "grad_norm": 0.9133732914924622, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 46665 + }, + { + "epoch": 0.46666, + "grad_norm": 0.9352893233299255, + "learning_rate": 0.003, + "loss": 3.978, + "step": 46666 + }, + { + "epoch": 0.46667, + "grad_norm": 0.9253087043762207, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 46667 + }, + { + "epoch": 0.46668, + "grad_norm": 0.9734234809875488, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 46668 + }, + { + "epoch": 0.46669, + "grad_norm": 0.9187382459640503, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 46669 + }, + { + "epoch": 0.4667, + "grad_norm": 1.005364179611206, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 46670 + }, + { + "epoch": 0.46671, + "grad_norm": 1.1873763799667358, + "learning_rate": 0.003, + "loss": 4.004, + "step": 46671 + }, + { + "epoch": 0.46672, + "grad_norm": 0.8981103301048279, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 46672 + }, + { + "epoch": 0.46673, + "grad_norm": 0.8834533095359802, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 46673 + }, + { + "epoch": 0.46674, + "grad_norm": 1.0299367904663086, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 46674 + }, + { + "epoch": 0.46675, + "grad_norm": 1.026045322418213, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 46675 + }, + { + "epoch": 0.46676, + "grad_norm": 0.9284621477127075, + "learning_rate": 0.003, + "loss": 3.967, + "step": 46676 + }, + { + "epoch": 0.46677, + "grad_norm": 0.9387860894203186, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 46677 + }, + { + "epoch": 0.46678, + "grad_norm": 1.0160255432128906, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 46678 + }, + { + "epoch": 0.46679, + "grad_norm": 1.0292741060256958, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 46679 + }, + { + "epoch": 0.4668, + "grad_norm": 0.8771392107009888, + "learning_rate": 0.003, + "loss": 3.987, + "step": 46680 + }, + { + "epoch": 0.46681, + "grad_norm": 0.6894432902336121, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 46681 + }, + { + "epoch": 0.46682, + "grad_norm": 0.617039680480957, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 46682 + }, + { + "epoch": 0.46683, + "grad_norm": 0.7842364311218262, + "learning_rate": 0.003, + "loss": 3.9443, + "step": 46683 + }, + { + "epoch": 0.46684, + "grad_norm": 1.1880332231521606, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 46684 + }, + { + "epoch": 0.46685, + "grad_norm": 0.7873261570930481, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 46685 + }, + { + "epoch": 0.46686, + "grad_norm": 0.7076462507247925, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 46686 + }, + { + "epoch": 0.46687, + "grad_norm": 0.7676106095314026, + "learning_rate": 0.003, + "loss": 3.9386, + "step": 46687 + }, + { + "epoch": 0.46688, + "grad_norm": 0.7755155563354492, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 46688 + }, + { + "epoch": 0.46689, + "grad_norm": 0.7660840749740601, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 46689 + }, + { + "epoch": 0.4669, + "grad_norm": 0.8683122992515564, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 46690 + }, + { + "epoch": 0.46691, + "grad_norm": 0.9612264633178711, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 46691 + }, + { + "epoch": 0.46692, + "grad_norm": 0.9989120364189148, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 46692 + }, + { + "epoch": 0.46693, + "grad_norm": 1.0760539770126343, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 46693 + }, + { + "epoch": 0.46694, + "grad_norm": 0.9043342471122742, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 46694 + }, + { + "epoch": 0.46695, + "grad_norm": 0.8656158447265625, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 46695 + }, + { + "epoch": 0.46696, + "grad_norm": 0.8922601938247681, + "learning_rate": 0.003, + "loss": 3.986, + "step": 46696 + }, + { + "epoch": 0.46697, + "grad_norm": 0.9114930629730225, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 46697 + }, + { + "epoch": 0.46698, + "grad_norm": 1.0618740320205688, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 46698 + }, + { + "epoch": 0.46699, + "grad_norm": 1.0698401927947998, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 46699 + }, + { + "epoch": 0.467, + "grad_norm": 0.9392814636230469, + "learning_rate": 0.003, + "loss": 3.96, + "step": 46700 + }, + { + "epoch": 0.46701, + "grad_norm": 1.0561811923980713, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 46701 + }, + { + "epoch": 0.46702, + "grad_norm": 0.9508544206619263, + "learning_rate": 0.003, + "loss": 3.974, + "step": 46702 + }, + { + "epoch": 0.46703, + "grad_norm": 0.9620676040649414, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 46703 + }, + { + "epoch": 0.46704, + "grad_norm": 1.1248935461044312, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 46704 + }, + { + "epoch": 0.46705, + "grad_norm": 0.8801974654197693, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 46705 + }, + { + "epoch": 0.46706, + "grad_norm": 1.1463977098464966, + "learning_rate": 0.003, + "loss": 3.988, + "step": 46706 + }, + { + "epoch": 0.46707, + "grad_norm": 1.275871992111206, + "learning_rate": 0.003, + "loss": 3.9358, + "step": 46707 + }, + { + "epoch": 0.46708, + "grad_norm": 0.6233413815498352, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 46708 + }, + { + "epoch": 0.46709, + "grad_norm": 0.6711006760597229, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 46709 + }, + { + "epoch": 0.4671, + "grad_norm": 0.6833428144454956, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 46710 + }, + { + "epoch": 0.46711, + "grad_norm": 0.7023849487304688, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 46711 + }, + { + "epoch": 0.46712, + "grad_norm": 0.6919354200363159, + "learning_rate": 0.003, + "loss": 3.963, + "step": 46712 + }, + { + "epoch": 0.46713, + "grad_norm": 0.8090195059776306, + "learning_rate": 0.003, + "loss": 3.968, + "step": 46713 + }, + { + "epoch": 0.46714, + "grad_norm": 0.86726975440979, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 46714 + }, + { + "epoch": 0.46715, + "grad_norm": 0.95948326587677, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 46715 + }, + { + "epoch": 0.46716, + "grad_norm": 0.9209326505661011, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 46716 + }, + { + "epoch": 0.46717, + "grad_norm": 1.054616093635559, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 46717 + }, + { + "epoch": 0.46718, + "grad_norm": 1.0607941150665283, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 46718 + }, + { + "epoch": 0.46719, + "grad_norm": 0.851860523223877, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 46719 + }, + { + "epoch": 0.4672, + "grad_norm": 0.7220852971076965, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 46720 + }, + { + "epoch": 0.46721, + "grad_norm": 0.8487175703048706, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 46721 + }, + { + "epoch": 0.46722, + "grad_norm": 0.9552214741706848, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 46722 + }, + { + "epoch": 0.46723, + "grad_norm": 0.9129763841629028, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 46723 + }, + { + "epoch": 0.46724, + "grad_norm": 0.9409564733505249, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 46724 + }, + { + "epoch": 0.46725, + "grad_norm": 0.9713467955589294, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 46725 + }, + { + "epoch": 0.46726, + "grad_norm": 0.862480640411377, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 46726 + }, + { + "epoch": 0.46727, + "grad_norm": 1.0098429918289185, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 46727 + }, + { + "epoch": 0.46728, + "grad_norm": 0.9998413324356079, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 46728 + }, + { + "epoch": 0.46729, + "grad_norm": 0.9149419665336609, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 46729 + }, + { + "epoch": 0.4673, + "grad_norm": 1.0061661005020142, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 46730 + }, + { + "epoch": 0.46731, + "grad_norm": 1.2703235149383545, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 46731 + }, + { + "epoch": 0.46732, + "grad_norm": 1.0068663358688354, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 46732 + }, + { + "epoch": 0.46733, + "grad_norm": 1.0553053617477417, + "learning_rate": 0.003, + "loss": 3.978, + "step": 46733 + }, + { + "epoch": 0.46734, + "grad_norm": 1.000783920288086, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 46734 + }, + { + "epoch": 0.46735, + "grad_norm": 1.0522509813308716, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 46735 + }, + { + "epoch": 0.46736, + "grad_norm": 1.0494712591171265, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 46736 + }, + { + "epoch": 0.46737, + "grad_norm": 0.9085177779197693, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 46737 + }, + { + "epoch": 0.46738, + "grad_norm": 0.7499664425849915, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 46738 + }, + { + "epoch": 0.46739, + "grad_norm": 0.7357348203659058, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 46739 + }, + { + "epoch": 0.4674, + "grad_norm": 0.6503845453262329, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 46740 + }, + { + "epoch": 0.46741, + "grad_norm": 0.7356896996498108, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 46741 + }, + { + "epoch": 0.46742, + "grad_norm": 0.8487445712089539, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 46742 + }, + { + "epoch": 0.46743, + "grad_norm": 1.073029637336731, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 46743 + }, + { + "epoch": 0.46744, + "grad_norm": 1.0509390830993652, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 46744 + }, + { + "epoch": 0.46745, + "grad_norm": 1.0689560174942017, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 46745 + }, + { + "epoch": 0.46746, + "grad_norm": 1.1126258373260498, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 46746 + }, + { + "epoch": 0.46747, + "grad_norm": 0.8676413893699646, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 46747 + }, + { + "epoch": 0.46748, + "grad_norm": 0.8408825993537903, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 46748 + }, + { + "epoch": 0.46749, + "grad_norm": 0.7638379335403442, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 46749 + }, + { + "epoch": 0.4675, + "grad_norm": 0.7034061551094055, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 46750 + }, + { + "epoch": 0.46751, + "grad_norm": 0.8192871809005737, + "learning_rate": 0.003, + "loss": 3.96, + "step": 46751 + }, + { + "epoch": 0.46752, + "grad_norm": 0.7218868732452393, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 46752 + }, + { + "epoch": 0.46753, + "grad_norm": 0.8174925446510315, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 46753 + }, + { + "epoch": 0.46754, + "grad_norm": 1.054254412651062, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 46754 + }, + { + "epoch": 0.46755, + "grad_norm": 0.9275047779083252, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 46755 + }, + { + "epoch": 0.46756, + "grad_norm": 0.8082767128944397, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 46756 + }, + { + "epoch": 0.46757, + "grad_norm": 0.8722352981567383, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 46757 + }, + { + "epoch": 0.46758, + "grad_norm": 1.0515061616897583, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 46758 + }, + { + "epoch": 0.46759, + "grad_norm": 0.9750657081604004, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 46759 + }, + { + "epoch": 0.4676, + "grad_norm": 0.9692243933677673, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 46760 + }, + { + "epoch": 0.46761, + "grad_norm": 0.9734711647033691, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 46761 + }, + { + "epoch": 0.46762, + "grad_norm": 1.0473746061325073, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 46762 + }, + { + "epoch": 0.46763, + "grad_norm": 0.9352941513061523, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 46763 + }, + { + "epoch": 0.46764, + "grad_norm": 0.9277377128601074, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 46764 + }, + { + "epoch": 0.46765, + "grad_norm": 0.9053279757499695, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 46765 + }, + { + "epoch": 0.46766, + "grad_norm": 0.890360951423645, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 46766 + }, + { + "epoch": 0.46767, + "grad_norm": 0.9292522072792053, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 46767 + }, + { + "epoch": 0.46768, + "grad_norm": 0.9345462918281555, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 46768 + }, + { + "epoch": 0.46769, + "grad_norm": 0.970354437828064, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 46769 + }, + { + "epoch": 0.4677, + "grad_norm": 1.249448537826538, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 46770 + }, + { + "epoch": 0.46771, + "grad_norm": 0.7095831632614136, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 46771 + }, + { + "epoch": 0.46772, + "grad_norm": 0.6815639734268188, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 46772 + }, + { + "epoch": 0.46773, + "grad_norm": 0.6762437224388123, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 46773 + }, + { + "epoch": 0.46774, + "grad_norm": 0.7333865165710449, + "learning_rate": 0.003, + "loss": 4.004, + "step": 46774 + }, + { + "epoch": 0.46775, + "grad_norm": 0.7178980112075806, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 46775 + }, + { + "epoch": 0.46776, + "grad_norm": 0.8041437268257141, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 46776 + }, + { + "epoch": 0.46777, + "grad_norm": 1.0905957221984863, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 46777 + }, + { + "epoch": 0.46778, + "grad_norm": 1.3020602464675903, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 46778 + }, + { + "epoch": 0.46779, + "grad_norm": 0.7441990375518799, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 46779 + }, + { + "epoch": 0.4678, + "grad_norm": 0.6955930590629578, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 46780 + }, + { + "epoch": 0.46781, + "grad_norm": 0.7224279642105103, + "learning_rate": 0.003, + "loss": 3.9363, + "step": 46781 + }, + { + "epoch": 0.46782, + "grad_norm": 0.6683669686317444, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 46782 + }, + { + "epoch": 0.46783, + "grad_norm": 0.6133536696434021, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 46783 + }, + { + "epoch": 0.46784, + "grad_norm": 0.639941930770874, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 46784 + }, + { + "epoch": 0.46785, + "grad_norm": 0.7974319458007812, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 46785 + }, + { + "epoch": 0.46786, + "grad_norm": 1.1235719919204712, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 46786 + }, + { + "epoch": 0.46787, + "grad_norm": 1.2667090892791748, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 46787 + }, + { + "epoch": 0.46788, + "grad_norm": 0.6681973934173584, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 46788 + }, + { + "epoch": 0.46789, + "grad_norm": 0.6711437106132507, + "learning_rate": 0.003, + "loss": 3.952, + "step": 46789 + }, + { + "epoch": 0.4679, + "grad_norm": 0.8439234495162964, + "learning_rate": 0.003, + "loss": 3.961, + "step": 46790 + }, + { + "epoch": 0.46791, + "grad_norm": 0.8435840606689453, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 46791 + }, + { + "epoch": 0.46792, + "grad_norm": 0.9919613003730774, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 46792 + }, + { + "epoch": 0.46793, + "grad_norm": 1.1607820987701416, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 46793 + }, + { + "epoch": 0.46794, + "grad_norm": 0.8164533972740173, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 46794 + }, + { + "epoch": 0.46795, + "grad_norm": 0.7424805760383606, + "learning_rate": 0.003, + "loss": 3.985, + "step": 46795 + }, + { + "epoch": 0.46796, + "grad_norm": 0.7164204120635986, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 46796 + }, + { + "epoch": 0.46797, + "grad_norm": 0.8960685133934021, + "learning_rate": 0.003, + "loss": 3.9467, + "step": 46797 + }, + { + "epoch": 0.46798, + "grad_norm": 1.1472301483154297, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 46798 + }, + { + "epoch": 0.46799, + "grad_norm": 0.9667372107505798, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 46799 + }, + { + "epoch": 0.468, + "grad_norm": 0.9540663361549377, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 46800 + }, + { + "epoch": 0.46801, + "grad_norm": 0.7987236380577087, + "learning_rate": 0.003, + "loss": 3.981, + "step": 46801 + }, + { + "epoch": 0.46802, + "grad_norm": 0.9274850487709045, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 46802 + }, + { + "epoch": 0.46803, + "grad_norm": 0.891757607460022, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 46803 + }, + { + "epoch": 0.46804, + "grad_norm": 0.8333768844604492, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 46804 + }, + { + "epoch": 0.46805, + "grad_norm": 0.8475250005722046, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 46805 + }, + { + "epoch": 0.46806, + "grad_norm": 1.1662964820861816, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 46806 + }, + { + "epoch": 0.46807, + "grad_norm": 1.1939905881881714, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 46807 + }, + { + "epoch": 0.46808, + "grad_norm": 1.092999815940857, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 46808 + }, + { + "epoch": 0.46809, + "grad_norm": 0.9612963795661926, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 46809 + }, + { + "epoch": 0.4681, + "grad_norm": 0.8293428421020508, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 46810 + }, + { + "epoch": 0.46811, + "grad_norm": 0.7544680833816528, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 46811 + }, + { + "epoch": 0.46812, + "grad_norm": 0.7474976181983948, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 46812 + }, + { + "epoch": 0.46813, + "grad_norm": 0.8727661371231079, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 46813 + }, + { + "epoch": 0.46814, + "grad_norm": 0.9508159756660461, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 46814 + }, + { + "epoch": 0.46815, + "grad_norm": 0.9034486413002014, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 46815 + }, + { + "epoch": 0.46816, + "grad_norm": 0.8796026110649109, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 46816 + }, + { + "epoch": 0.46817, + "grad_norm": 0.8205015063285828, + "learning_rate": 0.003, + "loss": 3.9407, + "step": 46817 + }, + { + "epoch": 0.46818, + "grad_norm": 0.8706597685813904, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 46818 + }, + { + "epoch": 0.46819, + "grad_norm": 0.9844437837600708, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 46819 + }, + { + "epoch": 0.4682, + "grad_norm": 1.0719305276870728, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 46820 + }, + { + "epoch": 0.46821, + "grad_norm": 1.055100917816162, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 46821 + }, + { + "epoch": 0.46822, + "grad_norm": 1.1079672574996948, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 46822 + }, + { + "epoch": 0.46823, + "grad_norm": 0.9768649339675903, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 46823 + }, + { + "epoch": 0.46824, + "grad_norm": 1.0705689191818237, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 46824 + }, + { + "epoch": 0.46825, + "grad_norm": 1.119327187538147, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 46825 + }, + { + "epoch": 0.46826, + "grad_norm": 0.9548429846763611, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 46826 + }, + { + "epoch": 0.46827, + "grad_norm": 0.9178202152252197, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 46827 + }, + { + "epoch": 0.46828, + "grad_norm": 1.0258201360702515, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 46828 + }, + { + "epoch": 0.46829, + "grad_norm": 1.0020935535430908, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 46829 + }, + { + "epoch": 0.4683, + "grad_norm": 1.032449722290039, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 46830 + }, + { + "epoch": 0.46831, + "grad_norm": 0.9067911505699158, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 46831 + }, + { + "epoch": 0.46832, + "grad_norm": 1.0262207984924316, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 46832 + }, + { + "epoch": 0.46833, + "grad_norm": 1.3660955429077148, + "learning_rate": 0.003, + "loss": 4.016, + "step": 46833 + }, + { + "epoch": 0.46834, + "grad_norm": 0.7348056435585022, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 46834 + }, + { + "epoch": 0.46835, + "grad_norm": 0.653083860874176, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 46835 + }, + { + "epoch": 0.46836, + "grad_norm": 0.7177824974060059, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 46836 + }, + { + "epoch": 0.46837, + "grad_norm": 0.7856746315956116, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 46837 + }, + { + "epoch": 0.46838, + "grad_norm": 0.7509543895721436, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 46838 + }, + { + "epoch": 0.46839, + "grad_norm": 0.7804797887802124, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 46839 + }, + { + "epoch": 0.4684, + "grad_norm": 0.8442162871360779, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 46840 + }, + { + "epoch": 0.46841, + "grad_norm": 0.8054494261741638, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 46841 + }, + { + "epoch": 0.46842, + "grad_norm": 0.944688081741333, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 46842 + }, + { + "epoch": 0.46843, + "grad_norm": 1.364527702331543, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 46843 + }, + { + "epoch": 0.46844, + "grad_norm": 0.6398074626922607, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 46844 + }, + { + "epoch": 0.46845, + "grad_norm": 0.8719339966773987, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 46845 + }, + { + "epoch": 0.46846, + "grad_norm": 1.0837464332580566, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 46846 + }, + { + "epoch": 0.46847, + "grad_norm": 1.0964314937591553, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 46847 + }, + { + "epoch": 0.46848, + "grad_norm": 1.036818027496338, + "learning_rate": 0.003, + "loss": 3.981, + "step": 46848 + }, + { + "epoch": 0.46849, + "grad_norm": 0.7414014339447021, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 46849 + }, + { + "epoch": 0.4685, + "grad_norm": 0.614037811756134, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 46850 + }, + { + "epoch": 0.46851, + "grad_norm": 0.8752685785293579, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 46851 + }, + { + "epoch": 0.46852, + "grad_norm": 1.0144950151443481, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 46852 + }, + { + "epoch": 0.46853, + "grad_norm": 1.0238436460494995, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 46853 + }, + { + "epoch": 0.46854, + "grad_norm": 0.9989628195762634, + "learning_rate": 0.003, + "loss": 3.978, + "step": 46854 + }, + { + "epoch": 0.46855, + "grad_norm": 0.9495423436164856, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 46855 + }, + { + "epoch": 0.46856, + "grad_norm": 0.9445282220840454, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 46856 + }, + { + "epoch": 0.46857, + "grad_norm": 1.1224572658538818, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 46857 + }, + { + "epoch": 0.46858, + "grad_norm": 1.0723291635513306, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 46858 + }, + { + "epoch": 0.46859, + "grad_norm": 0.8992307186126709, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 46859 + }, + { + "epoch": 0.4686, + "grad_norm": 0.8335009813308716, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 46860 + }, + { + "epoch": 0.46861, + "grad_norm": 0.6380262970924377, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 46861 + }, + { + "epoch": 0.46862, + "grad_norm": 0.5479368567466736, + "learning_rate": 0.003, + "loss": 3.9465, + "step": 46862 + }, + { + "epoch": 0.46863, + "grad_norm": 0.5676677227020264, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 46863 + }, + { + "epoch": 0.46864, + "grad_norm": 0.6764009594917297, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 46864 + }, + { + "epoch": 0.46865, + "grad_norm": 0.8360626697540283, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 46865 + }, + { + "epoch": 0.46866, + "grad_norm": 1.0083898305892944, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 46866 + }, + { + "epoch": 0.46867, + "grad_norm": 1.100643277168274, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 46867 + }, + { + "epoch": 0.46868, + "grad_norm": 0.8702917695045471, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 46868 + }, + { + "epoch": 0.46869, + "grad_norm": 0.8566293120384216, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 46869 + }, + { + "epoch": 0.4687, + "grad_norm": 0.8503794074058533, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 46870 + }, + { + "epoch": 0.46871, + "grad_norm": 0.9298427700996399, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 46871 + }, + { + "epoch": 0.46872, + "grad_norm": 1.0505049228668213, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 46872 + }, + { + "epoch": 0.46873, + "grad_norm": 1.001520037651062, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 46873 + }, + { + "epoch": 0.46874, + "grad_norm": 1.0246280431747437, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 46874 + }, + { + "epoch": 0.46875, + "grad_norm": 1.021780252456665, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 46875 + }, + { + "epoch": 0.46876, + "grad_norm": 0.8801569938659668, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 46876 + }, + { + "epoch": 0.46877, + "grad_norm": 0.846778929233551, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 46877 + }, + { + "epoch": 0.46878, + "grad_norm": 0.7407013773918152, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 46878 + }, + { + "epoch": 0.46879, + "grad_norm": 0.709747850894928, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 46879 + }, + { + "epoch": 0.4688, + "grad_norm": 0.6441144943237305, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 46880 + }, + { + "epoch": 0.46881, + "grad_norm": 0.6631774306297302, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 46881 + }, + { + "epoch": 0.46882, + "grad_norm": 0.8116869926452637, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 46882 + }, + { + "epoch": 0.46883, + "grad_norm": 0.9036789536476135, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 46883 + }, + { + "epoch": 0.46884, + "grad_norm": 1.0739195346832275, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 46884 + }, + { + "epoch": 0.46885, + "grad_norm": 1.0402119159698486, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 46885 + }, + { + "epoch": 0.46886, + "grad_norm": 0.9027823805809021, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 46886 + }, + { + "epoch": 0.46887, + "grad_norm": 0.9364296793937683, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 46887 + }, + { + "epoch": 0.46888, + "grad_norm": 1.1082450151443481, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 46888 + }, + { + "epoch": 0.46889, + "grad_norm": 1.0069843530654907, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 46889 + }, + { + "epoch": 0.4689, + "grad_norm": 0.9145746827125549, + "learning_rate": 0.003, + "loss": 3.977, + "step": 46890 + }, + { + "epoch": 0.46891, + "grad_norm": 0.915328323841095, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 46891 + }, + { + "epoch": 0.46892, + "grad_norm": 1.0388469696044922, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 46892 + }, + { + "epoch": 0.46893, + "grad_norm": 1.151408314704895, + "learning_rate": 0.003, + "loss": 3.98, + "step": 46893 + }, + { + "epoch": 0.46894, + "grad_norm": 1.2435927391052246, + "learning_rate": 0.003, + "loss": 3.986, + "step": 46894 + }, + { + "epoch": 0.46895, + "grad_norm": 0.8015617728233337, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 46895 + }, + { + "epoch": 0.46896, + "grad_norm": 0.6356765031814575, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 46896 + }, + { + "epoch": 0.46897, + "grad_norm": 0.7261911034584045, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 46897 + }, + { + "epoch": 0.46898, + "grad_norm": 0.7405878305435181, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 46898 + }, + { + "epoch": 0.46899, + "grad_norm": 0.694218099117279, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 46899 + }, + { + "epoch": 0.469, + "grad_norm": 0.6147152781486511, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 46900 + }, + { + "epoch": 0.46901, + "grad_norm": 0.6661572456359863, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 46901 + }, + { + "epoch": 0.46902, + "grad_norm": 0.7119654417037964, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 46902 + }, + { + "epoch": 0.46903, + "grad_norm": 0.8266667723655701, + "learning_rate": 0.003, + "loss": 3.961, + "step": 46903 + }, + { + "epoch": 0.46904, + "grad_norm": 1.0861996412277222, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 46904 + }, + { + "epoch": 0.46905, + "grad_norm": 1.3383307456970215, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 46905 + }, + { + "epoch": 0.46906, + "grad_norm": 0.6257633566856384, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 46906 + }, + { + "epoch": 0.46907, + "grad_norm": 0.8458379507064819, + "learning_rate": 0.003, + "loss": 3.987, + "step": 46907 + }, + { + "epoch": 0.46908, + "grad_norm": 0.9766850471496582, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 46908 + }, + { + "epoch": 0.46909, + "grad_norm": 0.9878972172737122, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 46909 + }, + { + "epoch": 0.4691, + "grad_norm": 1.0634162425994873, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 46910 + }, + { + "epoch": 0.46911, + "grad_norm": 0.8416410088539124, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 46911 + }, + { + "epoch": 0.46912, + "grad_norm": 0.7723723649978638, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 46912 + }, + { + "epoch": 0.46913, + "grad_norm": 0.7697808146476746, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 46913 + }, + { + "epoch": 0.46914, + "grad_norm": 0.9575498700141907, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 46914 + }, + { + "epoch": 0.46915, + "grad_norm": 1.1756504774093628, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 46915 + }, + { + "epoch": 0.46916, + "grad_norm": 0.9058313965797424, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 46916 + }, + { + "epoch": 0.46917, + "grad_norm": 1.0269485712051392, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 46917 + }, + { + "epoch": 0.46918, + "grad_norm": 1.0708314180374146, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 46918 + }, + { + "epoch": 0.46919, + "grad_norm": 1.1004399061203003, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 46919 + }, + { + "epoch": 0.4692, + "grad_norm": 1.1687586307525635, + "learning_rate": 0.003, + "loss": 3.967, + "step": 46920 + }, + { + "epoch": 0.46921, + "grad_norm": 0.9193905591964722, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 46921 + }, + { + "epoch": 0.46922, + "grad_norm": 0.784785807132721, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 46922 + }, + { + "epoch": 0.46923, + "grad_norm": 0.647933304309845, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 46923 + }, + { + "epoch": 0.46924, + "grad_norm": 0.7070644497871399, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 46924 + }, + { + "epoch": 0.46925, + "grad_norm": 0.6396973133087158, + "learning_rate": 0.003, + "loss": 3.9353, + "step": 46925 + }, + { + "epoch": 0.46926, + "grad_norm": 0.6708535552024841, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 46926 + }, + { + "epoch": 0.46927, + "grad_norm": 0.6712124347686768, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 46927 + }, + { + "epoch": 0.46928, + "grad_norm": 0.7415332794189453, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 46928 + }, + { + "epoch": 0.46929, + "grad_norm": 0.9197016954421997, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 46929 + }, + { + "epoch": 0.4693, + "grad_norm": 1.2967714071273804, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 46930 + }, + { + "epoch": 0.46931, + "grad_norm": 0.9035611748695374, + "learning_rate": 0.003, + "loss": 3.9375, + "step": 46931 + }, + { + "epoch": 0.46932, + "grad_norm": 0.7522356510162354, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 46932 + }, + { + "epoch": 0.46933, + "grad_norm": 0.8110660314559937, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 46933 + }, + { + "epoch": 0.46934, + "grad_norm": 0.9046133756637573, + "learning_rate": 0.003, + "loss": 3.965, + "step": 46934 + }, + { + "epoch": 0.46935, + "grad_norm": 1.0596988201141357, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 46935 + }, + { + "epoch": 0.46936, + "grad_norm": 1.199320912361145, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 46936 + }, + { + "epoch": 0.46937, + "grad_norm": 0.9126443266868591, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 46937 + }, + { + "epoch": 0.46938, + "grad_norm": 1.0137670040130615, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 46938 + }, + { + "epoch": 0.46939, + "grad_norm": 1.0597751140594482, + "learning_rate": 0.003, + "loss": 3.968, + "step": 46939 + }, + { + "epoch": 0.4694, + "grad_norm": 0.8443647623062134, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 46940 + }, + { + "epoch": 0.46941, + "grad_norm": 0.7735692858695984, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 46941 + }, + { + "epoch": 0.46942, + "grad_norm": 0.7792495489120483, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 46942 + }, + { + "epoch": 0.46943, + "grad_norm": 0.8106951713562012, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 46943 + }, + { + "epoch": 0.46944, + "grad_norm": 0.8463976383209229, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 46944 + }, + { + "epoch": 0.46945, + "grad_norm": 0.9842300415039062, + "learning_rate": 0.003, + "loss": 3.9468, + "step": 46945 + }, + { + "epoch": 0.46946, + "grad_norm": 1.2126877307891846, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 46946 + }, + { + "epoch": 0.46947, + "grad_norm": 0.9384652376174927, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 46947 + }, + { + "epoch": 0.46948, + "grad_norm": 1.1531977653503418, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 46948 + }, + { + "epoch": 0.46949, + "grad_norm": 1.0324175357818604, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 46949 + }, + { + "epoch": 0.4695, + "grad_norm": 1.053358554840088, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 46950 + }, + { + "epoch": 0.46951, + "grad_norm": 1.017413854598999, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 46951 + }, + { + "epoch": 0.46952, + "grad_norm": 1.0599980354309082, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 46952 + }, + { + "epoch": 0.46953, + "grad_norm": 0.9636380076408386, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 46953 + }, + { + "epoch": 0.46954, + "grad_norm": 0.857932984828949, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 46954 + }, + { + "epoch": 0.46955, + "grad_norm": 0.7958248257637024, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 46955 + }, + { + "epoch": 0.46956, + "grad_norm": 0.7838193774223328, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 46956 + }, + { + "epoch": 0.46957, + "grad_norm": 0.7942901253700256, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 46957 + }, + { + "epoch": 0.46958, + "grad_norm": 0.974767804145813, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 46958 + }, + { + "epoch": 0.46959, + "grad_norm": 1.3447299003601074, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 46959 + }, + { + "epoch": 0.4696, + "grad_norm": 0.8442103266716003, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 46960 + }, + { + "epoch": 0.46961, + "grad_norm": 0.8009097576141357, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 46961 + }, + { + "epoch": 0.46962, + "grad_norm": 0.9254750609397888, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 46962 + }, + { + "epoch": 0.46963, + "grad_norm": 1.1032966375350952, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 46963 + }, + { + "epoch": 0.46964, + "grad_norm": 0.9285439848899841, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 46964 + }, + { + "epoch": 0.46965, + "grad_norm": 0.9000251293182373, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 46965 + }, + { + "epoch": 0.46966, + "grad_norm": 0.9964715838432312, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 46966 + }, + { + "epoch": 0.46967, + "grad_norm": 0.9625207781791687, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 46967 + }, + { + "epoch": 0.46968, + "grad_norm": 0.8328965902328491, + "learning_rate": 0.003, + "loss": 3.991, + "step": 46968 + }, + { + "epoch": 0.46969, + "grad_norm": 0.789550244808197, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 46969 + }, + { + "epoch": 0.4697, + "grad_norm": 0.7333154678344727, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 46970 + }, + { + "epoch": 0.46971, + "grad_norm": 0.7232351899147034, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 46971 + }, + { + "epoch": 0.46972, + "grad_norm": 0.7139297723770142, + "learning_rate": 0.003, + "loss": 3.983, + "step": 46972 + }, + { + "epoch": 0.46973, + "grad_norm": 0.7209336161613464, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 46973 + }, + { + "epoch": 0.46974, + "grad_norm": 0.9868922233581543, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 46974 + }, + { + "epoch": 0.46975, + "grad_norm": 1.3221255540847778, + "learning_rate": 0.003, + "loss": 3.9173, + "step": 46975 + }, + { + "epoch": 0.46976, + "grad_norm": 0.9304636716842651, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 46976 + }, + { + "epoch": 0.46977, + "grad_norm": 0.8800715804100037, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 46977 + }, + { + "epoch": 0.46978, + "grad_norm": 0.8616132140159607, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 46978 + }, + { + "epoch": 0.46979, + "grad_norm": 0.9276614785194397, + "learning_rate": 0.003, + "loss": 3.9445, + "step": 46979 + }, + { + "epoch": 0.4698, + "grad_norm": 1.1035065650939941, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 46980 + }, + { + "epoch": 0.46981, + "grad_norm": 0.865311861038208, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 46981 + }, + { + "epoch": 0.46982, + "grad_norm": 0.9483104944229126, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 46982 + }, + { + "epoch": 0.46983, + "grad_norm": 1.1857718229293823, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 46983 + }, + { + "epoch": 0.46984, + "grad_norm": 0.8053959608078003, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 46984 + }, + { + "epoch": 0.46985, + "grad_norm": 0.8366328477859497, + "learning_rate": 0.003, + "loss": 3.994, + "step": 46985 + }, + { + "epoch": 0.46986, + "grad_norm": 0.8974894881248474, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 46986 + }, + { + "epoch": 0.46987, + "grad_norm": 0.9441418051719666, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 46987 + }, + { + "epoch": 0.46988, + "grad_norm": 0.9089412093162537, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 46988 + }, + { + "epoch": 0.46989, + "grad_norm": 0.8875439763069153, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 46989 + }, + { + "epoch": 0.4699, + "grad_norm": 0.9024959206581116, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 46990 + }, + { + "epoch": 0.46991, + "grad_norm": 0.9259801506996155, + "learning_rate": 0.003, + "loss": 3.967, + "step": 46991 + }, + { + "epoch": 0.46992, + "grad_norm": 1.1112544536590576, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 46992 + }, + { + "epoch": 0.46993, + "grad_norm": 0.8270875811576843, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 46993 + }, + { + "epoch": 0.46994, + "grad_norm": 0.5921308398246765, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 46994 + }, + { + "epoch": 0.46995, + "grad_norm": 0.6099599599838257, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 46995 + }, + { + "epoch": 0.46996, + "grad_norm": 0.7228986024856567, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 46996 + }, + { + "epoch": 0.46997, + "grad_norm": 0.9665014743804932, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 46997 + }, + { + "epoch": 0.46998, + "grad_norm": 1.3081536293029785, + "learning_rate": 0.003, + "loss": 3.972, + "step": 46998 + }, + { + "epoch": 0.46999, + "grad_norm": 0.8443779945373535, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 46999 + }, + { + "epoch": 0.47, + "grad_norm": 0.9663804173469543, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 47000 + } + ], + "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": 1.184904817672192e+18, + "train_batch_size": 256, + "trial_name": null, + "trial_params": null +}