| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.999136690647482, | |
| "eval_steps": 163, | |
| "global_step": 651, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0015347721822541965, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 3.846153846153847e-07, | |
| "loss": 1.5132, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0015347721822541965, | |
| "eval_loss": 1.6450660228729248, | |
| "eval_runtime": 59.1102, | |
| "eval_samples_per_second": 21.215, | |
| "eval_steps_per_second": 21.215, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.003069544364508393, | |
| "grad_norm": 4.0, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.7745, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00460431654676259, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.153846153846154e-06, | |
| "loss": 1.4893, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.006139088729016786, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.593, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.007673860911270983, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 1.9230769230769234e-06, | |
| "loss": 1.5801, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00920863309352518, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 1.6869, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.010743405275779376, | |
| "grad_norm": 3.625, | |
| "learning_rate": 2.6923076923076923e-06, | |
| "loss": 1.6167, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.012278177458033572, | |
| "grad_norm": 3.90625, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.6349, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.01381294964028777, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 3.4615384615384617e-06, | |
| "loss": 1.5281, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.015347721822541967, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 1.7309, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01688249400479616, | |
| "grad_norm": 3.625, | |
| "learning_rate": 4.230769230769231e-06, | |
| "loss": 1.6916, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.01841726618705036, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 1.7503, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.019952038369304557, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 5e-06, | |
| "loss": 1.6742, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.021486810551558752, | |
| "grad_norm": 2.8125, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 1.6483, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.02302158273381295, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 5.769230769230769e-06, | |
| "loss": 1.6429, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.024556354916067145, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.4412, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.026091127098321343, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 6.538461538461539e-06, | |
| "loss": 1.4474, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.02762589928057554, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 1.4794, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.029160671462829735, | |
| "grad_norm": 2.84375, | |
| "learning_rate": 7.307692307692308e-06, | |
| "loss": 1.5863, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.030695443645083934, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 1.7385, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03223021582733813, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 8.076923076923077e-06, | |
| "loss": 1.5931, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.03376498800959232, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 1.3675, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.035299760191846524, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 8.846153846153847e-06, | |
| "loss": 1.5693, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.03683453237410072, | |
| "grad_norm": 2.125, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.572, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.03836930455635491, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 9.615384615384616e-06, | |
| "loss": 1.5319, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.039904076738609115, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1e-05, | |
| "loss": 1.4973, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04143884892086331, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.0384615384615386e-05, | |
| "loss": 1.6304, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.042973621103117504, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.076923076923077e-05, | |
| "loss": 1.5787, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.044508393285371706, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.1153846153846154e-05, | |
| "loss": 1.4448, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0460431654676259, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.1538461538461538e-05, | |
| "loss": 1.455, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.047577937649880095, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.1923076923076925e-05, | |
| "loss": 1.479, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.04911270983213429, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 1.4277, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.05064748201438849, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.2692307692307693e-05, | |
| "loss": 1.5078, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.052182254196642686, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.3076923076923078e-05, | |
| "loss": 1.337, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.05371702637889688, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.3461538461538463e-05, | |
| "loss": 1.4207, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.05525179856115108, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.3846153846153847e-05, | |
| "loss": 1.5279, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.056786570743405276, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.4230769230769232e-05, | |
| "loss": 1.5279, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.05832134292565947, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.4615384615384615e-05, | |
| "loss": 1.3525, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.05985611510791367, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 1.5602, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.06139088729016787, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 1.6014, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06292565947242207, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.576923076923077e-05, | |
| "loss": 1.4031, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.06446043165467626, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.6153846153846154e-05, | |
| "loss": 1.6676, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.06599520383693046, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.653846153846154e-05, | |
| "loss": 1.4937, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.06752997601918465, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.6923076923076924e-05, | |
| "loss": 1.5643, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.06906474820143885, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.730769230769231e-05, | |
| "loss": 1.4824, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.07059952038369305, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.7692307692307694e-05, | |
| "loss": 1.4603, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.07213429256594724, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.807692307692308e-05, | |
| "loss": 1.5862, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.07366906474820144, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 1.4719, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.07520383693045564, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.8846153846153846e-05, | |
| "loss": 1.4937, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.07673860911270983, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.923076923076923e-05, | |
| "loss": 1.5793, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07827338129496403, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.9615384615384617e-05, | |
| "loss": 1.3622, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.07980815347721823, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 2e-05, | |
| "loss": 1.2495, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.08134292565947242, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.9999968417282542e-05, | |
| "loss": 1.4069, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.08287769784172662, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.999987366932966e-05, | |
| "loss": 1.49, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.08441247002398082, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.9999715756739833e-05, | |
| "loss": 1.4333, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.08594724220623501, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.999949468051052e-05, | |
| "loss": 1.5375, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.08748201438848921, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.9999210442038164e-05, | |
| "loss": 1.5002, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.08901678657074341, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.9998863043118163e-05, | |
| "loss": 1.5414, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.0905515587529976, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.999845248594489e-05, | |
| "loss": 1.4547, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.0920863309352518, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.999797877311163e-05, | |
| "loss": 1.308, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.093621103117506, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.9997441907610624e-05, | |
| "loss": 1.4789, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.09515587529976019, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.9996841892833e-05, | |
| "loss": 1.2503, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.09669064748201439, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.9996178732568784e-05, | |
| "loss": 1.4519, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.09822541966426858, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.9995452431006844e-05, | |
| "loss": 1.2409, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.09976019184652278, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.999466299273491e-05, | |
| "loss": 1.4345, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.10129496402877698, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.9993810422739496e-05, | |
| "loss": 1.4491, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.10282973621103117, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.9992894726405894e-05, | |
| "loss": 1.4475, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.10436450839328537, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.9991915909518146e-05, | |
| "loss": 1.4997, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.10589928057553957, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.999087397825899e-05, | |
| "loss": 1.4607, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.10743405275779376, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.9989768939209826e-05, | |
| "loss": 1.4236, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.10896882494004796, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.9988600799350685e-05, | |
| "loss": 1.4913, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.11050359712230216, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.998736956606018e-05, | |
| "loss": 1.5473, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.11203836930455635, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.998607524711543e-05, | |
| "loss": 1.4922, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.11357314148681055, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.998471785069208e-05, | |
| "loss": 1.3825, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.11510791366906475, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.9983297385364166e-05, | |
| "loss": 1.4647, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11664268585131894, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.998181386010413e-05, | |
| "loss": 1.4448, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.11817745803357314, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.9980267284282718e-05, | |
| "loss": 1.5147, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.11971223021582734, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.9978657667668945e-05, | |
| "loss": 1.4692, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.12124700239808153, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.9976985020430022e-05, | |
| "loss": 1.5878, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.12278177458033573, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.9975249353131304e-05, | |
| "loss": 1.3711, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.12431654676258992, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.9973450676736205e-05, | |
| "loss": 1.5019, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.12585131894484414, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.997158900260614e-05, | |
| "loss": 1.4792, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.1273860911270983, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.996966434250046e-05, | |
| "loss": 1.4263, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.1289208633093525, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.9967676708576362e-05, | |
| "loss": 1.507, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.13045563549160671, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.9965626113388823e-05, | |
| "loss": 1.4734, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.13199040767386092, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.9963512569890512e-05, | |
| "loss": 1.4456, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.13352517985611512, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.9961336091431728e-05, | |
| "loss": 1.2601, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.1350599520383693, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9959096691760284e-05, | |
| "loss": 1.5481, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.1365947242206235, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.9956794385021444e-05, | |
| "loss": 1.4177, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.1381294964028777, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.9954429185757835e-05, | |
| "loss": 1.6005, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1396642685851319, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.9952001108909336e-05, | |
| "loss": 1.4551, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.1411990407673861, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.9949510169813006e-05, | |
| "loss": 1.4133, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.1427338129496403, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.994695638420296e-05, | |
| "loss": 1.4024, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.14426858513189447, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.994433976821031e-05, | |
| "loss": 1.4194, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.14580335731414867, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.9941660338363008e-05, | |
| "loss": 1.5165, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.14733812949640288, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.9938918111585805e-05, | |
| "loss": 1.5456, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.14887290167865708, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.9936113105200085e-05, | |
| "loss": 1.5324, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.15040767386091128, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.9933245336923798e-05, | |
| "loss": 1.3492, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.15194244604316548, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.9930314824871326e-05, | |
| "loss": 1.4851, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.15347721822541965, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.9927321587553378e-05, | |
| "loss": 1.4352, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15501199040767386, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.992426564387686e-05, | |
| "loss": 1.4341, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.15654676258992806, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.9921147013144782e-05, | |
| "loss": 1.5551, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.15808153477218226, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.9917965715056106e-05, | |
| "loss": 1.4501, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.15961630695443646, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.9914721769705637e-05, | |
| "loss": 1.4443, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.16115107913669063, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.9911415197583904e-05, | |
| "loss": 1.4481, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.16268585131894484, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.990804601957701e-05, | |
| "loss": 1.4376, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.16422062350119904, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.9904614256966514e-05, | |
| "loss": 1.5503, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.16575539568345324, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.9901119931429294e-05, | |
| "loss": 1.406, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.16729016786570744, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.9897563065037412e-05, | |
| "loss": 1.3818, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.16882494004796164, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.9893943680257964e-05, | |
| "loss": 1.3271, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.17035971223021582, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.9890261799952954e-05, | |
| "loss": 1.5972, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.17189448441247002, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.988651744737914e-05, | |
| "loss": 1.5008, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.17342925659472422, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.988271064618789e-05, | |
| "loss": 1.3596, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.17496402877697842, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.9878841420425023e-05, | |
| "loss": 1.3632, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.17649880095923262, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.9874909794530677e-05, | |
| "loss": 1.6275, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.17803357314148682, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.9870915793339137e-05, | |
| "loss": 1.4541, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.179568345323741, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.986685944207868e-05, | |
| "loss": 1.4712, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.1811031175059952, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.9862740766371434e-05, | |
| "loss": 1.5007, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.1826378896882494, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.985855979223319e-05, | |
| "loss": 1.4757, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.1841726618705036, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.985431654607325e-05, | |
| "loss": 1.3876, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1857074340527578, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.9850011054694264e-05, | |
| "loss": 1.4466, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.187242206235012, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.9845643345292055e-05, | |
| "loss": 1.3912, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.18877697841726618, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.984121344545545e-05, | |
| "loss": 1.4602, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.19031175059952038, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.983672138316611e-05, | |
| "loss": 1.5255, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.19184652278177458, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.9832167186798333e-05, | |
| "loss": 1.3831, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.19338129496402878, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.9827550885118902e-05, | |
| "loss": 1.478, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.19491606714628298, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.982287250728689e-05, | |
| "loss": 1.5094, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.19645083932853716, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.9818132082853466e-05, | |
| "loss": 1.5228, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.19798561151079136, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.9813329641761738e-05, | |
| "loss": 1.4506, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.19952038369304556, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.9808465214346525e-05, | |
| "loss": 1.5035, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.20105515587529976, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.98035388313342e-05, | |
| "loss": 1.4746, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.20258992805755396, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.979855052384247e-05, | |
| "loss": 1.5071, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.20412470023980817, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.97935003233802e-05, | |
| "loss": 1.472, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.20565947242206234, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.9788388261847204e-05, | |
| "loss": 1.3509, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.20719424460431654, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.9783214371534037e-05, | |
| "loss": 1.4248, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.20872901678657074, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.97779786851218e-05, | |
| "loss": 1.5036, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.21026378896882494, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.9772681235681936e-05, | |
| "loss": 1.4228, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.21179856115107915, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.9767322056676018e-05, | |
| "loss": 1.4464, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.21333333333333335, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.976190118195553e-05, | |
| "loss": 1.3816, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.21486810551558752, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.975641864576166e-05, | |
| "loss": 1.5671, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.21640287769784172, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.9750874482725093e-05, | |
| "loss": 1.4326, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.21793764988009592, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.9745268727865774e-05, | |
| "loss": 1.5385, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.21947242206235013, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.9739601416592693e-05, | |
| "loss": 1.2213, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.22100719424460433, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.9733872584703673e-05, | |
| "loss": 1.4447, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.2225419664268585, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 1.9728082268385126e-05, | |
| "loss": 1.5023, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.2240767386091127, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.9722230504211843e-05, | |
| "loss": 1.496, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.2256115107913669, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 1.971631732914674e-05, | |
| "loss": 1.4072, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.2271462829736211, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.971034278054065e-05, | |
| "loss": 1.3316, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.2286810551558753, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.9704306896132063e-05, | |
| "loss": 1.3629, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.2302158273381295, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.969820971404691e-05, | |
| "loss": 1.405, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.23175059952038368, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.9692051272798304e-05, | |
| "loss": 1.3486, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.23328537170263788, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.9685831611286312e-05, | |
| "loss": 1.3927, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.23482014388489209, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.967955076879769e-05, | |
| "loss": 1.3901, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.2363549160671463, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.9673208785005658e-05, | |
| "loss": 1.5, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.2378896882494005, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.966680569996963e-05, | |
| "loss": 1.2666, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.2394244604316547, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.9660341554134972e-05, | |
| "loss": 1.3021, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.24095923261390886, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.965381638833274e-05, | |
| "loss": 1.4593, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.24249400479616307, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.9647230243779432e-05, | |
| "loss": 1.5067, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.24402877697841727, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.964058316207671e-05, | |
| "loss": 1.5402, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.24556354916067147, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.963387518521116e-05, | |
| "loss": 1.3881, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.24709832134292567, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.962710635555401e-05, | |
| "loss": 1.6041, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.24863309352517984, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.962027671586086e-05, | |
| "loss": 1.5419, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.25016786570743405, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.9613386309271437e-05, | |
| "loss": 1.3791, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.25016786570743405, | |
| "eval_loss": 1.441381812095642, | |
| "eval_runtime": 58.8537, | |
| "eval_samples_per_second": 21.307, | |
| "eval_steps_per_second": 21.307, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.2517026378896883, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.9606435179309284e-05, | |
| "loss": 1.4643, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.25323741007194245, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.959942336988152e-05, | |
| "loss": 1.3267, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.2547721822541966, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.9592350925278546e-05, | |
| "loss": 1.507, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.25630695443645085, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.958521789017376e-05, | |
| "loss": 1.3429, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.257841726618705, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.9578024309623296e-05, | |
| "loss": 1.464, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.25937649880095925, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.9570770229065716e-05, | |
| "loss": 1.3992, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.26091127098321343, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.956345569432173e-05, | |
| "loss": 1.6172, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2624460431654676, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.955608075159392e-05, | |
| "loss": 1.3302, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.26398081534772183, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.9548645447466433e-05, | |
| "loss": 1.4756, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.265515587529976, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.9541149828904686e-05, | |
| "loss": 1.2405, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.26705035971223023, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.9533593943255087e-05, | |
| "loss": 1.3672, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.2685851318944844, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.952597783824471e-05, | |
| "loss": 1.3942, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2701199040767386, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.9518301561981016e-05, | |
| "loss": 1.5685, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.2716546762589928, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.9510565162951538e-05, | |
| "loss": 1.5139, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.273189448441247, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.9502768690023574e-05, | |
| "loss": 1.4495, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.2747242206235012, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.949491219244389e-05, | |
| "loss": 1.3186, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.2762589928057554, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.9486995719838392e-05, | |
| "loss": 1.372, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2777937649880096, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.9479019322211824e-05, | |
| "loss": 1.3685, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.2793285371702638, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.9470983049947446e-05, | |
| "loss": 1.4346, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.28086330935251796, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.946288695380672e-05, | |
| "loss": 1.3951, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.2823980815347722, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.9454731084928995e-05, | |
| "loss": 1.3547, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.28393285371702637, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9446515494831168e-05, | |
| "loss": 1.5246, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.2854676258992806, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.9438240235407375e-05, | |
| "loss": 1.459, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.28700239808153477, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.9429905358928648e-05, | |
| "loss": 1.3077, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.28853717026378894, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.9421510918042593e-05, | |
| "loss": 1.4291, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.2900719424460432, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.941305696577307e-05, | |
| "loss": 1.4862, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.29160671462829735, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.940454355551983e-05, | |
| "loss": 1.2541, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2931414868105516, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.9395970741058202e-05, | |
| "loss": 1.4298, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.29467625899280575, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.9387338576538743e-05, | |
| "loss": 1.5857, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.2962110311750599, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.937864711648689e-05, | |
| "loss": 1.5354, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.29774580335731415, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.936989641580263e-05, | |
| "loss": 1.5927, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.2992805755395683, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.936108652976015e-05, | |
| "loss": 1.3927, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.30081534772182256, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.935221751400747e-05, | |
| "loss": 1.3322, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.30235011990407673, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.9343289424566122e-05, | |
| "loss": 1.3863, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.30388489208633096, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.9334302317830764e-05, | |
| "loss": 1.3775, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.30541966426858513, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.9325256250568852e-05, | |
| "loss": 1.3753, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.3069544364508393, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.931615127992026e-05, | |
| "loss": 1.4834, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.30848920863309354, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.9306987463396934e-05, | |
| "loss": 1.4202, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.3100239808153477, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.9297764858882516e-05, | |
| "loss": 1.3512, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.31155875299760194, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.928848352463199e-05, | |
| "loss": 1.5923, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.3130935251798561, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.927914351927131e-05, | |
| "loss": 1.447, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.3146282973621103, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.9269744901797022e-05, | |
| "loss": 1.4988, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.3161630695443645, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.9260287731575902e-05, | |
| "loss": 1.5078, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.3176978417266187, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.925077206834458e-05, | |
| "loss": 1.4488, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.3192326139088729, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 1.9241197972209157e-05, | |
| "loss": 1.2465, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.3207673860911271, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.9231565503644826e-05, | |
| "loss": 1.4034, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.32230215827338127, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.9221874723495494e-05, | |
| "loss": 1.4779, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3238369304556355, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.9212125692973396e-05, | |
| "loss": 1.2789, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.32537170263788967, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.9202318473658707e-05, | |
| "loss": 1.4495, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.3269064748201439, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.919245312749915e-05, | |
| "loss": 1.4263, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.3284412470023981, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.9182529716809618e-05, | |
| "loss": 1.4215, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.3299760191846523, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.9172548304271766e-05, | |
| "loss": 1.2981, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.3315107913669065, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.916250895293362e-05, | |
| "loss": 1.3433, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.33304556354916065, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.9152411726209176e-05, | |
| "loss": 1.5089, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.3345803357314149, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.9142256687878012e-05, | |
| "loss": 1.511, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.33611510791366905, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.9132043902084864e-05, | |
| "loss": 1.4224, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.3376498800959233, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.912177343333924e-05, | |
| "loss": 1.3882, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.33918465227817746, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.9111445346515003e-05, | |
| "loss": 1.5708, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.34071942446043163, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.9101059706849957e-05, | |
| "loss": 1.4676, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.34225419664268586, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.9090616579945455e-05, | |
| "loss": 1.479, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.34378896882494003, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.908011603176596e-05, | |
| "loss": 1.5031, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.34532374100719426, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.9069558128638636e-05, | |
| "loss": 1.4443, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.34685851318944844, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.9058942937252943e-05, | |
| "loss": 1.466, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.34839328537170267, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.9048270524660197e-05, | |
| "loss": 1.4357, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.34992805755395684, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.903754095827316e-05, | |
| "loss": 1.4573, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.351462829736211, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.9026754305865593e-05, | |
| "loss": 1.3422, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.35299760191846524, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.901591063557187e-05, | |
| "loss": 1.2501, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3545323741007194, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.9005010015886495e-05, | |
| "loss": 1.4802, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.35606714628297365, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.899405251566371e-05, | |
| "loss": 1.5655, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.3576019184652278, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.8983038204117046e-05, | |
| "loss": 1.4752, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.359136690647482, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.897196715081888e-05, | |
| "loss": 1.4711, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.3606714628297362, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.8960839425699992e-05, | |
| "loss": 1.5812, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.3622062350119904, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.894965509904915e-05, | |
| "loss": 1.4105, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.3637410071942446, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.893841424151264e-05, | |
| "loss": 1.3949, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.3652757793764988, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.8927116924093824e-05, | |
| "loss": 1.3143, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.366810551558753, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.8915763218152704e-05, | |
| "loss": 1.466, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.3683453237410072, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.890435319540545e-05, | |
| "loss": 1.3584, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3698800959232614, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.8892886927923972e-05, | |
| "loss": 1.4642, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.3714148681055156, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.8881364488135448e-05, | |
| "loss": 1.3241, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.3729496402877698, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.8869785948821865e-05, | |
| "loss": 1.3893, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.374484412470024, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.8858151383119576e-05, | |
| "loss": 1.4159, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.3760191846522782, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.8846460864518818e-05, | |
| "loss": 1.5998, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.37755395683453236, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.883471446686326e-05, | |
| "loss": 1.3294, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.3790887290167866, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.8822912264349535e-05, | |
| "loss": 1.43, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.38062350119904076, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.881105433152677e-05, | |
| "loss": 1.516, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.382158273381295, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.8799140743296104e-05, | |
| "loss": 1.4576, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.38369304556354916, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.878717157491025e-05, | |
| "loss": 1.4947, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.38522781774580334, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.877514690197297e-05, | |
| "loss": 1.2411, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.38676258992805757, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.8763066800438638e-05, | |
| "loss": 1.3762, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.38829736211031174, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.875093134661174e-05, | |
| "loss": 1.2754, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.38983213429256597, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.8738740617146396e-05, | |
| "loss": 1.4927, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.39136690647482014, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.8726494689045878e-05, | |
| "loss": 1.4043, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.3929016786570743, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.871419363966213e-05, | |
| "loss": 1.5075, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.39443645083932855, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.870183754669526e-05, | |
| "loss": 1.4552, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.3959712230215827, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.8689426488193066e-05, | |
| "loss": 1.494, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.39750599520383695, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.867696054255054e-05, | |
| "loss": 1.5097, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.3990407673860911, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.866443978850937e-05, | |
| "loss": 1.2745, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.40057553956834535, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.865186430515745e-05, | |
| "loss": 1.4509, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.4021103117505995, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.8639234171928355e-05, | |
| "loss": 1.4734, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.4036450839328537, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.862654946860088e-05, | |
| "loss": 1.5411, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.40517985611510793, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.86138102752985e-05, | |
| "loss": 1.4693, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.4067146282973621, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.8601016672488887e-05, | |
| "loss": 1.4294, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.40824940047961633, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.858816874098339e-05, | |
| "loss": 1.4605, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.4097841726618705, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.8575266561936526e-05, | |
| "loss": 1.4601, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.4113189448441247, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.8562310216845463e-05, | |
| "loss": 1.3688, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.4128537170263789, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.8549299787549536e-05, | |
| "loss": 1.3078, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.4143884892086331, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.8536235356229667e-05, | |
| "loss": 1.5405, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4159232613908873, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.852311700540792e-05, | |
| "loss": 1.4657, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.4174580335731415, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.850994481794692e-05, | |
| "loss": 1.6243, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.41899280575539566, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.8496718877049367e-05, | |
| "loss": 1.4455, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.4205275779376499, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.8483439266257485e-05, | |
| "loss": 1.4315, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.42206235011990406, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.8470106069452522e-05, | |
| "loss": 1.568, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.4235971223021583, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.845671937085419e-05, | |
| "loss": 1.5067, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.42513189448441246, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.8443279255020153e-05, | |
| "loss": 1.3536, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.4266666666666667, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.842978580684549e-05, | |
| "loss": 1.4866, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.42820143884892087, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.841623911156215e-05, | |
| "loss": 1.5661, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.42973621103117504, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.8402639254738422e-05, | |
| "loss": 1.5654, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.43127098321342927, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.83889863222784e-05, | |
| "loss": 1.5065, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.43280575539568344, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.837528040042142e-05, | |
| "loss": 1.3752, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.4343405275779377, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.8361521575741533e-05, | |
| "loss": 1.26, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.43587529976019185, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.8347709935146958e-05, | |
| "loss": 1.4494, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.437410071942446, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.8333845565879517e-05, | |
| "loss": 1.4768, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.43894484412470025, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.8319928555514108e-05, | |
| "loss": 1.5487, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.4404796163069544, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.830595899195813e-05, | |
| "loss": 1.4256, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.44201438848920865, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.8291936963450933e-05, | |
| "loss": 1.3908, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.4435491606714628, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.827786255856328e-05, | |
| "loss": 1.4406, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.445083932853717, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.8263735866196758e-05, | |
| "loss": 1.4099, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.44661870503597123, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.824955697558323e-05, | |
| "loss": 1.4792, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.4481534772182254, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.8235325976284276e-05, | |
| "loss": 1.4086, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.44968824940047963, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.8221042958190628e-05, | |
| "loss": 1.5445, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.4512230215827338, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.820670801152158e-05, | |
| "loss": 1.4573, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.45275779376498804, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.8192321226824455e-05, | |
| "loss": 1.257, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.4542925659472422, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.8177882694974008e-05, | |
| "loss": 1.4822, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.4558273381294964, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.816339250717184e-05, | |
| "loss": 1.3933, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.4573621103117506, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.8148850754945865e-05, | |
| "loss": 1.4232, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.4588968824940048, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.8134257530149684e-05, | |
| "loss": 1.424, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.460431654676259, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.8119612924962043e-05, | |
| "loss": 1.3909, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4619664268585132, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.8104917031886223e-05, | |
| "loss": 1.4939, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.46350119904076736, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.8090169943749477e-05, | |
| "loss": 1.4614, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.4650359712230216, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.8075371753702423e-05, | |
| "loss": 1.3587, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.46657074340527577, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.806052255521847e-05, | |
| "loss": 1.4853, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.46810551558753, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.8045622442093237e-05, | |
| "loss": 1.4516, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.46964028776978417, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.8030671508443928e-05, | |
| "loss": 1.3532, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.47117505995203834, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.8015669848708768e-05, | |
| "loss": 1.5189, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.4727098321342926, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.8000617557646392e-05, | |
| "loss": 1.3214, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.47424460431654675, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.798551473033525e-05, | |
| "loss": 1.6297, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.475779376498801, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.797036146217301e-05, | |
| "loss": 1.3899, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.47731414868105515, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.795515784887595e-05, | |
| "loss": 1.4549, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.4788489208633094, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.7939903986478354e-05, | |
| "loss": 1.3179, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.48038369304556355, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.792459997133191e-05, | |
| "loss": 1.5416, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.4819184652278177, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.7909245900105085e-05, | |
| "loss": 1.4575, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.48345323741007196, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.7893841869782548e-05, | |
| "loss": 1.3526, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.48498800959232613, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.7878387977664522e-05, | |
| "loss": 1.4582, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.48652278177458036, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.786288432136619e-05, | |
| "loss": 1.3788, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.48805755395683453, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.784733099881707e-05, | |
| "loss": 1.4998, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.4895923261390887, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.7831728108260407e-05, | |
| "loss": 1.4229, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.49112709832134294, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.7816075748252526e-05, | |
| "loss": 1.4159, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4926618705035971, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.780037401766225e-05, | |
| "loss": 1.4787, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.49419664268585134, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.7784623015670237e-05, | |
| "loss": 1.4399, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.4957314148681055, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.776882284176838e-05, | |
| "loss": 1.4658, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.4972661870503597, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.775297359575916e-05, | |
| "loss": 1.4715, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.4988009592326139, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.7737075377755032e-05, | |
| "loss": 1.4503, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.5003357314148681, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.7721128288177782e-05, | |
| "loss": 1.4218, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.5003357314148681, | |
| "eval_loss": 1.419244647026062, | |
| "eval_runtime": 58.9368, | |
| "eval_samples_per_second": 21.277, | |
| "eval_steps_per_second": 21.277, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.5018705035971223, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.7705132427757895e-05, | |
| "loss": 1.4326, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.5034052757793765, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.7689087897533916e-05, | |
| "loss": 1.4476, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.5049400479616307, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.767299479885182e-05, | |
| "loss": 1.4214, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.5064748201438849, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.765685323336437e-05, | |
| "loss": 1.4763, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5080095923261391, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.7640663303030452e-05, | |
| "loss": 1.4451, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.5095443645083932, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.762442511011448e-05, | |
| "loss": 1.3918, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.5110791366906475, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.76081387571857e-05, | |
| "loss": 1.4037, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.5126139088729017, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.759180434711757e-05, | |
| "loss": 1.3785, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.5141486810551559, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.7575421983087095e-05, | |
| "loss": 1.3338, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.51568345323741, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.7558991768574197e-05, | |
| "loss": 1.2631, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.5172182254196642, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.754251380736104e-05, | |
| "loss": 1.4594, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.5187529976019185, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.752598820353138e-05, | |
| "loss": 1.4979, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.5202877697841727, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.7509415061469916e-05, | |
| "loss": 1.5528, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.5218225419664269, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.749279448586162e-05, | |
| "loss": 1.4236, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.523357314148681, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.7476126581691072e-05, | |
| "loss": 1.3933, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.5248920863309352, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.7459411454241822e-05, | |
| "loss": 1.336, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.5264268585131895, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.7442649209095703e-05, | |
| "loss": 1.5121, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.5279616306954437, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.7425839952132157e-05, | |
| "loss": 1.3951, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.5294964028776978, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.7408983789527588e-05, | |
| "loss": 1.4327, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.531031175059952, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.739208082775469e-05, | |
| "loss": 1.4532, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.5325659472422062, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.737513117358174e-05, | |
| "loss": 1.459, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.5341007194244605, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.7358134934071978e-05, | |
| "loss": 1.3645, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.5356354916067146, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.7341092216582886e-05, | |
| "loss": 1.5174, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.5371702637889688, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.7324003128765536e-05, | |
| "loss": 1.4671, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.538705035971223, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.730686777856388e-05, | |
| "loss": 1.4129, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.5402398081534772, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.7289686274214116e-05, | |
| "loss": 1.3426, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.5417745803357314, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.7272458724243957e-05, | |
| "loss": 1.4186, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.5433093525179856, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.7255185237471978e-05, | |
| "loss": 1.4067, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.5448441247002398, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.7237865923006904e-05, | |
| "loss": 1.4288, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.546378896882494, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.7220500890246944e-05, | |
| "loss": 1.4377, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.5479136690647483, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.720309024887907e-05, | |
| "loss": 1.3125, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.5494484412470024, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.7185634108878367e-05, | |
| "loss": 1.4347, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.5509832134292566, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.7168132580507298e-05, | |
| "loss": 1.4743, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.5525179856115108, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.715058577431503e-05, | |
| "loss": 1.473, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.554052757793765, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.713299380113673e-05, | |
| "loss": 1.3648, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.5555875299760192, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.7115356772092858e-05, | |
| "loss": 1.3193, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.5571223021582734, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.709767479858847e-05, | |
| "loss": 1.4627, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.5586570743405276, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.707994799231253e-05, | |
| "loss": 1.3789, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.5601918465227818, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.7062176465237175e-05, | |
| "loss": 1.4083, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5617266187050359, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.704436032961703e-05, | |
| "loss": 1.4053, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.5632613908872902, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.7026499697988496e-05, | |
| "loss": 1.444, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.5647961630695444, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.7008594683169018e-05, | |
| "loss": 1.3111, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.5663309352517986, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.6990645398256412e-05, | |
| "loss": 1.432, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.5678657074340527, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.6972651956628108e-05, | |
| "loss": 1.4439, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5694004796163069, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.695461447194047e-05, | |
| "loss": 1.3627, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.5709352517985612, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.693653305812805e-05, | |
| "loss": 1.5243, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.5724700239808154, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.6918407829402888e-05, | |
| "loss": 1.5582, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.5740047961630695, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.6900238900253777e-05, | |
| "loss": 1.3232, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.5755395683453237, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.6882026385445548e-05, | |
| "loss": 1.4903, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.5770743405275779, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.6863770400018344e-05, | |
| "loss": 1.4271, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.5786091127098322, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.684547105928689e-05, | |
| "loss": 1.406, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.5801438848920863, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.6827128478839767e-05, | |
| "loss": 1.4787, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.5816786570743405, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.6808742774538683e-05, | |
| "loss": 1.4338, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.5832134292565947, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.679031406251774e-05, | |
| "loss": 1.3572, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5847482014388489, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.6771842459182703e-05, | |
| "loss": 1.3706, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.5862829736211032, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.6753328081210244e-05, | |
| "loss": 1.3664, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.5878177458033573, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.673477104554725e-05, | |
| "loss": 1.3524, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.5893525179856115, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.6716171469410042e-05, | |
| "loss": 1.4479, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.5908872901678657, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.6697529470283646e-05, | |
| "loss": 1.3061, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5924220623501198, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 1.6678845165921066e-05, | |
| "loss": 1.3439, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.5939568345323741, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.666011867434252e-05, | |
| "loss": 1.474, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.5954916067146283, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.6641350113834705e-05, | |
| "loss": 1.5257, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.5970263788968825, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 1.662253960295005e-05, | |
| "loss": 1.2708, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.5985611510791367, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.660368726050597e-05, | |
| "loss": 1.468, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6000959232613909, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.65847932055841e-05, | |
| "loss": 1.3418, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.6016306954436451, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.6565857557529567e-05, | |
| "loss": 1.403, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.6031654676258993, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.6546880435950207e-05, | |
| "loss": 1.456, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.6047002398081535, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.652786196071584e-05, | |
| "loss": 1.4762, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.6062350119904076, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.6508802251957488e-05, | |
| "loss": 1.4517, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.6077697841726619, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.6489701430066632e-05, | |
| "loss": 1.3809, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.6093045563549161, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.6470559615694445e-05, | |
| "loss": 1.4589, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.6108393285371703, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.6451376929751028e-05, | |
| "loss": 1.396, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.6123741007194244, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.6432153493404654e-05, | |
| "loss": 1.3241, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.6139088729016786, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.6412889428080992e-05, | |
| "loss": 1.4123, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6154436450839329, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.639358485546235e-05, | |
| "loss": 1.4098, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.6169784172661871, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.63742398974869e-05, | |
| "loss": 1.3087, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.6185131894484412, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.635485467634791e-05, | |
| "loss": 1.2647, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.6200479616306954, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.633542931449297e-05, | |
| "loss": 1.4002, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.6215827338129496, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.6315963934623228e-05, | |
| "loss": 1.3966, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.6231175059952039, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.62964586596926e-05, | |
| "loss": 1.3257, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.624652278177458, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.6276913612907005e-05, | |
| "loss": 1.2805, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.6261870503597122, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.625732891772358e-05, | |
| "loss": 1.4642, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.6277218225419664, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.6237704697849903e-05, | |
| "loss": 1.4683, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.6292565947242206, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.6218041077243213e-05, | |
| "loss": 1.448, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6307913669064749, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.6198338180109624e-05, | |
| "loss": 1.4157, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.632326139088729, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.6178596130903345e-05, | |
| "loss": 1.3459, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.6338609112709832, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.6158815054325887e-05, | |
| "loss": 1.2427, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.6353956834532374, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.6138995075325277e-05, | |
| "loss": 1.3427, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.6369304556354916, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.611913631909528e-05, | |
| "loss": 1.483, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.6384652278177458, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.609923891107459e-05, | |
| "loss": 1.429, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.6079302976946055e-05, | |
| "loss": 1.5569, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.6415347721822542, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.6059328642635864e-05, | |
| "loss": 1.4083, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.6430695443645084, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.6039316034312767e-05, | |
| "loss": 1.4171, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.6446043165467625, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.6019265278387287e-05, | |
| "loss": 1.4593, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6461390887290168, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.5999176501510883e-05, | |
| "loss": 1.406, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.647673860911271, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.597904983057519e-05, | |
| "loss": 1.4094, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.6492086330935252, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.5958885392711203e-05, | |
| "loss": 1.6037, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.6507434052757793, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.5938683315288472e-05, | |
| "loss": 1.3668, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.6522781774580336, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.5918443725914298e-05, | |
| "loss": 1.4602, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.6538129496402878, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.589816675243292e-05, | |
| "loss": 1.4616, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.655347721822542, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.5877852522924733e-05, | |
| "loss": 1.3704, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.6568824940047961, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 1.5857501165705443e-05, | |
| "loss": 1.2654, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.6584172661870503, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.583711280932529e-05, | |
| "loss": 1.45, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.6599520383693046, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.581668758256821e-05, | |
| "loss": 1.4815, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6614868105515588, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.5796225614451034e-05, | |
| "loss": 1.4874, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.663021582733813, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.5775727034222675e-05, | |
| "loss": 1.4731, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.6645563549160671, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.5755191971363313e-05, | |
| "loss": 1.3568, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.6660911270983213, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.5734620555583555e-05, | |
| "loss": 1.2038, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.6676258992805756, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.5714012916823653e-05, | |
| "loss": 1.3923, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.6691606714628298, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.5693369185252648e-05, | |
| "loss": 1.4711, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.6706954436450839, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.567268949126757e-05, | |
| "loss": 1.4183, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.6722302158273381, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.56519739654926e-05, | |
| "loss": 1.4671, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.6737649880095923, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.5631222738778268e-05, | |
| "loss": 1.416, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.6752997601918466, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.561043594220059e-05, | |
| "loss": 1.4353, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6768345323741007, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.5589613707060278e-05, | |
| "loss": 1.516, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.6783693045563549, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.556875616488188e-05, | |
| "loss": 1.2044, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.6799040767386091, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.5547863447412973e-05, | |
| "loss": 1.3343, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.6814388489208633, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.5526935686623316e-05, | |
| "loss": 1.389, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.6829736211031175, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.5505973014704017e-05, | |
| "loss": 1.4935, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.6845083932853717, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.5484975564066704e-05, | |
| "loss": 1.4248, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.6860431654676259, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.5463943467342694e-05, | |
| "loss": 1.4196, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.6875779376498801, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.544287685738213e-05, | |
| "loss": 1.3694, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.6891127098321342, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.542177586725318e-05, | |
| "loss": 1.5132, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.6906474820143885, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.540064063024116e-05, | |
| "loss": 1.5271, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6921822541966427, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.5379471279847714e-05, | |
| "loss": 1.4123, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.6937170263788969, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.5358267949789968e-05, | |
| "loss": 1.4703, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.695251798561151, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.5337030773999674e-05, | |
| "loss": 1.4658, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.6967865707434053, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.531575988662238e-05, | |
| "loss": 1.3538, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.6983213429256595, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.5294455422016576e-05, | |
| "loss": 1.4328, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.6998561151079137, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.5273117514752826e-05, | |
| "loss": 1.4393, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.7013908872901679, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.5251746299612959e-05, | |
| "loss": 1.5345, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.702925659472422, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.5230341911589183e-05, | |
| "loss": 1.3759, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.7044604316546763, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.5208904485883244e-05, | |
| "loss": 1.3337, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.7059952038369305, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.5187434157905575e-05, | |
| "loss": 1.5585, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7075299760191847, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.5165931063274442e-05, | |
| "loss": 1.4298, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.7090647482014388, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.5144395337815066e-05, | |
| "loss": 1.3482, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.710599520383693, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.5122827117558802e-05, | |
| "loss": 1.6308, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.7121342925659473, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.5101226538742248e-05, | |
| "loss": 1.4727, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.7136690647482015, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.50795937378064e-05, | |
| "loss": 1.3891, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.7152038369304556, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.505792885139579e-05, | |
| "loss": 1.5394, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.7167386091127098, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.503623201635761e-05, | |
| "loss": 1.3889, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.718273381294964, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.5014503369740866e-05, | |
| "loss": 1.4508, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.7198081534772183, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.4992743048795493e-05, | |
| "loss": 1.4323, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.7213429256594724, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.4970951190971512e-05, | |
| "loss": 1.549, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7228776978417266, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.4949127933918136e-05, | |
| "loss": 1.4607, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.7244124700239808, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.4927273415482916e-05, | |
| "loss": 1.5684, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.725947242206235, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.4905387773710876e-05, | |
| "loss": 1.3445, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.7274820143884893, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.4883471146843617e-05, | |
| "loss": 1.3575, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.7290167865707434, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.486152367331847e-05, | |
| "loss": 1.4004, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.7305515587529976, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.4839545491767599e-05, | |
| "loss": 1.5095, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.7320863309352518, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.4817536741017153e-05, | |
| "loss": 1.4924, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.733621103117506, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.4795497560086358e-05, | |
| "loss": 1.451, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.7351558752997602, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.4773428088186662e-05, | |
| "loss": 1.4369, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.7366906474820144, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.4751328464720842e-05, | |
| "loss": 1.3582, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7382254196642686, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.4729198829282127e-05, | |
| "loss": 1.4842, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.7397601918465228, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.470703932165333e-05, | |
| "loss": 1.3498, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.7412949640287769, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.4684850081805934e-05, | |
| "loss": 1.522, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.7428297362110312, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.4662631249899248e-05, | |
| "loss": 1.4422, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.7443645083932854, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.4640382966279484e-05, | |
| "loss": 1.331, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.7458992805755396, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.4618105371478896e-05, | |
| "loss": 1.3508, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.7474340527577937, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.4595798606214882e-05, | |
| "loss": 1.3859, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.748968824940048, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.4573462811389087e-05, | |
| "loss": 1.3691, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.7505035971223022, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.4551098128086538e-05, | |
| "loss": 1.3617, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.7505035971223022, | |
| "eval_loss": 1.4075348377227783, | |
| "eval_runtime": 58.9728, | |
| "eval_samples_per_second": 21.264, | |
| "eval_steps_per_second": 21.264, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.7520383693045564, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.4528704697574729e-05, | |
| "loss": 1.3604, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7535731414868105, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.4506282661302735e-05, | |
| "loss": 1.2984, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.7551079136690647, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.4483832160900326e-05, | |
| "loss": 1.4131, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.756642685851319, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.446135333817706e-05, | |
| "loss": 1.4902, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.7581774580335732, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.4438846335121402e-05, | |
| "loss": 1.4437, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.7597122302158273, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.4416311293899816e-05, | |
| "loss": 1.4344, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.7612470023980815, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.4393748356855865e-05, | |
| "loss": 1.2446, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.7627817745803357, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.437115766650933e-05, | |
| "loss": 1.4651, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.76431654676259, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.4348539365555283e-05, | |
| "loss": 1.3397, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.7658513189448442, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.432589359686321e-05, | |
| "loss": 1.3666, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.7673860911270983, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.430322050347609e-05, | |
| "loss": 1.3271, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7689208633093525, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.4280520228609503e-05, | |
| "loss": 1.3099, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.7704556354916067, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.4257792915650728e-05, | |
| "loss": 1.3086, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.771990407673861, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.423503870815782e-05, | |
| "loss": 1.3125, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.7735251798561151, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.4212257749858727e-05, | |
| "loss": 1.3633, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.7750599520383693, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.4189450184650354e-05, | |
| "loss": 1.2312, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.7765947242206235, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.416661615659768e-05, | |
| "loss": 1.3498, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.7781294964028777, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.4143755809932843e-05, | |
| "loss": 1.4694, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.7796642685851319, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.412086928905421e-05, | |
| "loss": 1.1963, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.7811990407673861, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 1.4097956738525493e-05, | |
| "loss": 1.2387, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.7827338129496403, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.4075018303074808e-05, | |
| "loss": 1.434, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.7842685851318945, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.4052054127593782e-05, | |
| "loss": 1.3472, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.7858033573141486, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.4029064357136628e-05, | |
| "loss": 1.461, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.7873381294964029, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.4006049136919229e-05, | |
| "loss": 1.4106, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.7888729016786571, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.398300861231823e-05, | |
| "loss": 1.3753, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.7904076738609113, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.3959942928870101e-05, | |
| "loss": 1.3298, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.7919424460431654, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.3936852232270236e-05, | |
| "loss": 1.3482, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.7934772182254196, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.3913736668372027e-05, | |
| "loss": 1.4847, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.7950119904076739, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.3890596383185934e-05, | |
| "loss": 1.4272, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.7965467625899281, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.386743152287858e-05, | |
| "loss": 1.3732, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.7980815347721822, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.384424223377181e-05, | |
| "loss": 1.4443, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.7996163069544364, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.3821028662341776e-05, | |
| "loss": 1.4799, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.8011510791366907, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.3797790955218014e-05, | |
| "loss": 1.4129, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.8026858513189449, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.3774529259182508e-05, | |
| "loss": 1.3799, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.804220623501199, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.3751243721168778e-05, | |
| "loss": 1.4216, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.8057553956834532, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.3727934488260934e-05, | |
| "loss": 1.4918, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.8072901678657074, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.3704601707692762e-05, | |
| "loss": 1.3585, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.8088249400479617, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.3681245526846782e-05, | |
| "loss": 1.3654, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.8103597122302159, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.3657866093253327e-05, | |
| "loss": 1.3949, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.81189448441247, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.3634463554589608e-05, | |
| "loss": 1.3878, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.8134292565947242, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.3611038058678776e-05, | |
| "loss": 1.4569, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8149640287769784, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.3587589753488999e-05, | |
| "loss": 1.4791, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.8164988009592327, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.3564118787132507e-05, | |
| "loss": 1.5325, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.8180335731414868, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.3540625307864693e-05, | |
| "loss": 1.3922, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.819568345323741, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.3517109464083129e-05, | |
| "loss": 1.5347, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.8211031175059952, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.3493571404326671e-05, | |
| "loss": 1.4345, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.8226378896882494, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.3470011277274497e-05, | |
| "loss": 1.4761, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.8241726618705036, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.344642923174517e-05, | |
| "loss": 1.3524, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.8257074340527578, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.3422825416695713e-05, | |
| "loss": 1.3459, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.827242206235012, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.3399199981220648e-05, | |
| "loss": 1.3793, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.8287769784172662, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.337555307455106e-05, | |
| "loss": 1.2792, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8303117505995203, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.3351884846053668e-05, | |
| "loss": 1.4985, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.8318465227817746, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.3328195445229869e-05, | |
| "loss": 1.5698, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.8333812949640288, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.330448502171479e-05, | |
| "loss": 1.3706, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.834916067146283, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.3280753725276352e-05, | |
| "loss": 1.4925, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.8364508393285371, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.3257001705814323e-05, | |
| "loss": 1.5999, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.8379856115107913, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.3233229113359368e-05, | |
| "loss": 1.495, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.8395203836930456, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.3209436098072095e-05, | |
| "loss": 1.3432, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.8410551558752998, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.3185622810242129e-05, | |
| "loss": 1.3679, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.842589928057554, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.316178940028713e-05, | |
| "loss": 1.424, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.8441247002398081, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.3137936018751876e-05, | |
| "loss": 1.4842, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8456594724220623, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.3114062816307284e-05, | |
| "loss": 1.2388, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.8471942446043166, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.3090169943749475e-05, | |
| "loss": 1.4275, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.8487290167865708, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.3066257551998822e-05, | |
| "loss": 1.508, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.8502637889688249, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.3042325792098982e-05, | |
| "loss": 1.5241, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.8517985611510791, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.3018374815215962e-05, | |
| "loss": 1.3896, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.8533333333333334, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.2994404772637145e-05, | |
| "loss": 1.3765, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.8548681055155876, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.297041581577035e-05, | |
| "loss": 1.3199, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.8564028776978417, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.2946408096142866e-05, | |
| "loss": 1.511, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.8579376498800959, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.2922381765400501e-05, | |
| "loss": 1.3674, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.8594724220623501, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.289833697530661e-05, | |
| "loss": 1.3332, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8610071942446044, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.2874273877741165e-05, | |
| "loss": 1.3316, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.8625419664268585, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.2850192624699762e-05, | |
| "loss": 1.3782, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.8640767386091127, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.2826093368292687e-05, | |
| "loss": 1.4955, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.8656115107913669, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.2801976260743937e-05, | |
| "loss": 1.5445, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.8671462829736211, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.2777841454390276e-05, | |
| "loss": 1.3496, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.8686810551558753, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.2753689101680252e-05, | |
| "loss": 1.4302, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.8702158273381295, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.2729519355173254e-05, | |
| "loss": 1.3425, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.8717505995203837, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.2705332367538539e-05, | |
| "loss": 1.5349, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.8732853717026379, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.2681128291554263e-05, | |
| "loss": 1.4892, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.874820143884892, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.2656907280106528e-05, | |
| "loss": 1.455, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8763549160671463, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.2632669486188403e-05, | |
| "loss": 1.5562, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.8778896882494005, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.2608415062898971e-05, | |
| "loss": 1.3335, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.8794244604316547, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.2584144163442347e-05, | |
| "loss": 1.4161, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.8809592326139088, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.255985694112673e-05, | |
| "loss": 1.4954, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.882494004796163, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.2535553549363407e-05, | |
| "loss": 1.415, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.8840287769784173, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.2511234141665816e-05, | |
| "loss": 1.4748, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.8855635491606715, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.2486898871648552e-05, | |
| "loss": 1.2765, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.8870983213429257, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.2462547893026403e-05, | |
| "loss": 1.5225, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.8886330935251798, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.2438181359613388e-05, | |
| "loss": 1.4643, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.890167865707434, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.2413799425321774e-05, | |
| "loss": 1.4519, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8917026378896883, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.2389402244161107e-05, | |
| "loss": 1.4527, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.8932374100719425, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.236498997023725e-05, | |
| "loss": 1.4682, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.8947721822541966, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.2340562757751385e-05, | |
| "loss": 1.4018, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.8963069544364508, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.2316120760999066e-05, | |
| "loss": 1.4245, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.897841726618705, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.2291664134369229e-05, | |
| "loss": 1.4849, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.8993764988009593, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.2267193032343219e-05, | |
| "loss": 1.3096, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.9009112709832134, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.2242707609493814e-05, | |
| "loss": 1.4568, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.9024460431654676, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.2218208020484255e-05, | |
| "loss": 1.4273, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.9039808153477218, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.219369442006726e-05, | |
| "loss": 1.3472, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.9055155875299761, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.2169166963084056e-05, | |
| "loss": 1.2589, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.9070503597122302, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.2144625804463384e-05, | |
| "loss": 1.5012, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.9085851318944844, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.212007109922055e-05, | |
| "loss": 1.513, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.9101199040767386, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.2095503002456405e-05, | |
| "loss": 1.3524, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.9116546762589928, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.2070921669356415e-05, | |
| "loss": 1.291, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.913189448441247, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.2046327255189627e-05, | |
| "loss": 1.4709, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.9147242206235012, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.2021719915307737e-05, | |
| "loss": 1.4373, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.9162589928057554, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.1997099805144071e-05, | |
| "loss": 1.3847, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.9177937649880096, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.197246708021263e-05, | |
| "loss": 1.3515, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.9193285371702637, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.194782189610709e-05, | |
| "loss": 1.5657, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.920863309352518, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.192316440849983e-05, | |
| "loss": 1.3718, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9223980815347722, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.1898494773140942e-05, | |
| "loss": 1.4058, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.9239328537170264, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.187381314585725e-05, | |
| "loss": 1.4138, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.9254676258992806, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.1849119682551323e-05, | |
| "loss": 1.5707, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.9270023980815347, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.1824414539200505e-05, | |
| "loss": 1.5327, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.928537170263789, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.17996978718559e-05, | |
| "loss": 1.4747, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.9300719424460432, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.1774969836641417e-05, | |
| "loss": 1.5315, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.9316067146282974, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.1750230589752763e-05, | |
| "loss": 1.501, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.9331414868105515, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.1725480287456467e-05, | |
| "loss": 1.3595, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.9346762589928057, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.1700719086088891e-05, | |
| "loss": 1.5799, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.93621103117506, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.1675947142055241e-05, | |
| "loss": 1.4093, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9377458033573142, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.165116461182858e-05, | |
| "loss": 1.3595, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.9392805755395683, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.1626371651948839e-05, | |
| "loss": 1.4811, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.9408153477218225, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.160156841902182e-05, | |
| "loss": 1.3337, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.9423501199040767, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.1576755069718229e-05, | |
| "loss": 1.3943, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.943884892086331, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.155193176077266e-05, | |
| "loss": 1.4056, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.9454196642685851, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.1527098648982634e-05, | |
| "loss": 1.5342, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.9469544364508393, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.1502255891207572e-05, | |
| "loss": 1.3041, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.9484892086330935, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.1477403644367839e-05, | |
| "loss": 1.3494, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.9500239808153477, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.1452542065443728e-05, | |
| "loss": 1.465, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.951558752997602, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.1427671311474489e-05, | |
| "loss": 1.336, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9530935251798561, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.140279153955732e-05, | |
| "loss": 1.3387, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.9546282973621103, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.137790290684638e-05, | |
| "loss": 1.398, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.9561630695443645, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.1353005570551803e-05, | |
| "loss": 1.509, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.9576978417266188, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.1328099687938696e-05, | |
| "loss": 1.4732, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.9592326139088729, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.1303185416326148e-05, | |
| "loss": 1.3411, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.9607673860911271, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.1278262913086238e-05, | |
| "loss": 1.4715, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.9623021582733813, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.1253332335643043e-05, | |
| "loss": 1.3279, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.9638369304556355, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.1228393841471644e-05, | |
| "loss": 1.3264, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.9653717026378897, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.1203447588097115e-05, | |
| "loss": 1.2625, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.9669064748201439, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.117849373309356e-05, | |
| "loss": 1.4892, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9684412470023981, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.1153532434083083e-05, | |
| "loss": 1.4547, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.9699760191846523, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.1128563848734817e-05, | |
| "loss": 1.5092, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.9715107913669064, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.1103588134763918e-05, | |
| "loss": 1.424, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.9730455635491607, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.1078605449930569e-05, | |
| "loss": 1.4477, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.9745803357314149, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.1053615952038984e-05, | |
| "loss": 1.2692, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.9761151079136691, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.1028619798936418e-05, | |
| "loss": 1.4006, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.9776498800959232, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.1003617148512149e-05, | |
| "loss": 1.4153, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.9791846522781774, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.0978608158696517e-05, | |
| "loss": 1.4156, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.9807194244604317, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.0953592987459886e-05, | |
| "loss": 1.4369, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.9822541966426859, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.092857179281168e-05, | |
| "loss": 1.482, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.98378896882494, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.0903544732799357e-05, | |
| "loss": 1.4631, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.9853237410071942, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.0878511965507435e-05, | |
| "loss": 1.4286, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.9868585131894484, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.0853473649056472e-05, | |
| "loss": 1.5408, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.9883932853717027, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.0828429941602082e-05, | |
| "loss": 1.4246, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.9899280575539569, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.0803381001333943e-05, | |
| "loss": 1.401, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.991462829736211, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.0778326986474765e-05, | |
| "loss": 1.3825, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.9929976019184652, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.0753268055279328e-05, | |
| "loss": 1.4568, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.9945323741007194, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.072820436603346e-05, | |
| "loss": 1.4033, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.9960671462829737, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.070313607705304e-05, | |
| "loss": 1.4913, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.9976019184652278, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.067806334668301e-05, | |
| "loss": 1.452, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.999136690647482, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.0652986333296358e-05, | |
| "loss": 1.5346, | |
| "step": 651 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1302, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 651, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.6771286351373926e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |