| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9952, | |
| "eval_steps": 500, | |
| "global_step": 312, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0096, | |
| "grad_norm": 5.479907512664795, | |
| "learning_rate": 3.125e-07, | |
| "loss": 0.9262, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 5.96177339553833, | |
| "learning_rate": 6.25e-07, | |
| "loss": 0.9505, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 5.506617546081543, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 0.9138, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 5.304609298706055, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.9032, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 5.245921611785889, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 0.8911, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 4.993967056274414, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 0.919, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 4.168082237243652, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 0.8684, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 3.96718430519104, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.8867, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 2.244798183441162, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.848, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.0326294898986816, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.8326, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 2.0216994285583496, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.8688, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 3.1776065826416016, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.8271, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 3.8676669597625732, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.8368, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 3.6298909187316895, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.8211, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 3.425091028213501, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.8001, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.572175979614258, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7876, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 2.208611249923706, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.7792, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 1.917243242263794, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.7999, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1824, | |
| "grad_norm": 1.679466724395752, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.7477, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.4488530158996582, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.727, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2016, | |
| "grad_norm": 1.385766863822937, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.7473, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 1.332506775856018, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.7575, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2208, | |
| "grad_norm": 1.1861612796783447, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.7247, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.1110572814941406, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7094, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.014829397201538, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.6891, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 0.7976421117782593, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.6937, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2592, | |
| "grad_norm": 0.944990336894989, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.7369, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.8185547590255737, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.6974, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2784, | |
| "grad_norm": 0.7537665367126465, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.7266, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.793213963508606, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.6803, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2976, | |
| "grad_norm": 0.7176020741462708, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.6481, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.6386464834213257, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6648, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3168, | |
| "grad_norm": 0.7271828055381775, | |
| "learning_rate": 9.999685283773504e-06, | |
| "loss": 0.7003, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 0.6109193563461304, | |
| "learning_rate": 9.998741174712534e-06, | |
| "loss": 0.659, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.6009302735328674, | |
| "learning_rate": 9.997167791667668e-06, | |
| "loss": 0.6769, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.6556861400604248, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.6834, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3552, | |
| "grad_norm": 0.6338167190551758, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.6857, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 0.5566282272338867, | |
| "learning_rate": 9.98867437523228e-06, | |
| "loss": 0.6391, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3744, | |
| "grad_norm": 0.6530281901359558, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.6642, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.49887770414352417, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.6836, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3936, | |
| "grad_norm": 0.5470542907714844, | |
| "learning_rate": 9.974529372743762e-06, | |
| "loss": 0.6486, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 0.6051164865493774, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.6614, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4128, | |
| "grad_norm": 0.5241588354110718, | |
| "learning_rate": 9.961967251474823e-06, | |
| "loss": 0.6697, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.4913438856601715, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.6512, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 0.5375027060508728, | |
| "learning_rate": 9.946906630265184e-06, | |
| "loss": 0.6449, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 0.5305971503257751, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.6511, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4512, | |
| "grad_norm": 0.481326699256897, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.6531, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.7205134034156799, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.6671, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4704, | |
| "grad_norm": 0.5181694626808167, | |
| "learning_rate": 9.909321476180594e-06, | |
| "loss": 0.6459, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.521432101726532, | |
| "learning_rate": 9.898376992116179e-06, | |
| "loss": 0.6408, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4896, | |
| "grad_norm": 0.6602535247802734, | |
| "learning_rate": 9.886815868562596e-06, | |
| "loss": 0.6454, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.6800113320350647, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.6619, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5088, | |
| "grad_norm": 0.4519635736942291, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.6234, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 0.5449603796005249, | |
| "learning_rate": 9.848447601883436e-06, | |
| "loss": 0.6384, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.5179167985916138, | |
| "learning_rate": 9.834435247725032e-06, | |
| "loss": 0.6289, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.5591006278991699, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.6415, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5472, | |
| "grad_norm": 0.6138595938682556, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.6554, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 0.4841986298561096, | |
| "learning_rate": 9.788754083424654e-06, | |
| "loss": 0.6328, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5664, | |
| "grad_norm": 0.47388964891433716, | |
| "learning_rate": 9.772318717677905e-06, | |
| "loss": 0.6115, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.49600282311439514, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.6406, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5856, | |
| "grad_norm": 0.503688633441925, | |
| "learning_rate": 9.737647819437645e-06, | |
| "loss": 0.6266, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 0.4665270447731018, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.6564, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6048, | |
| "grad_norm": 0.5030093193054199, | |
| "learning_rate": 9.700591372846096e-06, | |
| "loss": 0.6149, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.4615119695663452, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.6551, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.5064735412597656, | |
| "learning_rate": 9.661168036940071e-06, | |
| "loss": 0.6224, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 0.5256257057189941, | |
| "learning_rate": 9.640574942595195e-06, | |
| "loss": 0.6187, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6432, | |
| "grad_norm": 0.4757899343967438, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.6304, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.4830450713634491, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.6287, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6624, | |
| "grad_norm": 0.5002210140228271, | |
| "learning_rate": 9.5753012823366e-06, | |
| "loss": 0.6264, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.5669161081314087, | |
| "learning_rate": 9.552387733294081e-06, | |
| "loss": 0.6126, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6816, | |
| "grad_norm": 0.5801583528518677, | |
| "learning_rate": 9.528901100135971e-06, | |
| "loss": 0.6714, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.4554402828216553, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6224, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7008, | |
| "grad_norm": 0.600943386554718, | |
| "learning_rate": 9.480220479843627e-06, | |
| "loss": 0.6364, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 0.48585039377212524, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.6396, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.5432065725326538, | |
| "learning_rate": 9.4292839336179e-06, | |
| "loss": 0.6359, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.5171982049942017, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.6381, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7392, | |
| "grad_norm": 0.4601828157901764, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.6079, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 0.5039520859718323, | |
| "learning_rate": 9.348705665778479e-06, | |
| "loss": 0.6316, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7584, | |
| "grad_norm": 0.4935835897922516, | |
| "learning_rate": 9.320746778718274e-06, | |
| "loss": 0.6268, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.5095129013061523, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6169, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7776, | |
| "grad_norm": 0.5040186643600464, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.643, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 0.6242079138755798, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.611, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7968, | |
| "grad_norm": 0.5779500007629395, | |
| "learning_rate": 9.203508214822652e-06, | |
| "loss": 0.6487, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.5830844044685364, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.6115, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.4821722209453583, | |
| "learning_rate": 9.141699014900084e-06, | |
| "loss": 0.6295, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 0.5032180547714233, | |
| "learning_rate": 9.110010377239552e-06, | |
| "loss": 0.5891, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8352, | |
| "grad_norm": 0.5220208764076233, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.5975, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.5088613629341125, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6414, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8544, | |
| "grad_norm": 0.5571373701095581, | |
| "learning_rate": 9.011856377401891e-06, | |
| "loss": 0.6191, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.4590759575366974, | |
| "learning_rate": 8.978122744408905e-06, | |
| "loss": 0.6179, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8736, | |
| "grad_norm": 0.49995893239974976, | |
| "learning_rate": 8.943888319504456e-06, | |
| "loss": 0.5953, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.4612729251384735, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.5965, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8928, | |
| "grad_norm": 0.5401706099510193, | |
| "learning_rate": 8.873934395068006e-06, | |
| "loss": 0.6552, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 0.4521356225013733, | |
| "learning_rate": 8.838223701790057e-06, | |
| "loss": 0.6208, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.44377925992012024, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.6093, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.6105339527130127, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.6186, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9312, | |
| "grad_norm": 0.5453846454620361, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.5871, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 0.49319887161254883, | |
| "learning_rate": 8.690594987436705e-06, | |
| "loss": 0.6441, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9504, | |
| "grad_norm": 0.6136060953140259, | |
| "learning_rate": 8.652514554406388e-06, | |
| "loss": 0.5956, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5185710191726685, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.6157, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9696, | |
| "grad_norm": 0.538345217704773, | |
| "learning_rate": 8.574979133323378e-06, | |
| "loss": 0.6245, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 0.4882757067680359, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.6153, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9888, | |
| "grad_norm": 0.46365898847579956, | |
| "learning_rate": 8.495643602586287e-06, | |
| "loss": 0.5939, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.5339441895484924, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.5773, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 1.128153920173645, | |
| "learning_rate": 8.414547910024035e-06, | |
| "loss": 1.1209, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0176, | |
| "grad_norm": 0.4810388684272766, | |
| "learning_rate": 8.373352729660373e-06, | |
| "loss": 0.5741, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0272, | |
| "grad_norm": 0.607177197933197, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.594, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.48626017570495605, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.5962, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0464, | |
| "grad_norm": 0.593654990196228, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.5994, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 0.45741838216781616, | |
| "learning_rate": 8.204378069925121e-06, | |
| "loss": 0.5558, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0656, | |
| "grad_norm": 0.5748123526573181, | |
| "learning_rate": 8.16111251028955e-06, | |
| "loss": 0.5637, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.5928158760070801, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5633, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0848, | |
| "grad_norm": 0.4221522808074951, | |
| "learning_rate": 8.073393063582386e-06, | |
| "loss": 0.5632, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0944, | |
| "grad_norm": 0.5065470337867737, | |
| "learning_rate": 8.0289502192041e-06, | |
| "loss": 0.5738, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 0.5820510387420654, | |
| "learning_rate": 7.984126070912519e-06, | |
| "loss": 0.5862, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.42268669605255127, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.5706, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1232, | |
| "grad_norm": 0.4905787408351898, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.5461, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1328, | |
| "grad_norm": 0.4830806851387024, | |
| "learning_rate": 7.84742246584226e-06, | |
| "loss": 0.5556, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1424, | |
| "grad_norm": 0.43891096115112305, | |
| "learning_rate": 7.801129998764014e-06, | |
| "loss": 0.5982, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.4736801087856293, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.589, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1616, | |
| "grad_norm": 0.4433214068412781, | |
| "learning_rate": 7.70749306331863e-06, | |
| "loss": 0.5473, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1712, | |
| "grad_norm": 0.43428710103034973, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.5741, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1808, | |
| "grad_norm": 0.3924940526485443, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.5437, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.447864830493927, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.5739, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.42934784293174744, | |
| "learning_rate": 7.516177115029002e-06, | |
| "loss": 0.5534, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.2096, | |
| "grad_norm": 0.4134349226951599, | |
| "learning_rate": 7.467541090321735e-06, | |
| "loss": 0.5998, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2192, | |
| "grad_norm": 0.4451971650123596, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.5852, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.4061635732650757, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.5524, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2384, | |
| "grad_norm": 0.4278489351272583, | |
| "learning_rate": 7.319793920889171e-06, | |
| "loss": 0.573, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 0.48377665877342224, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.5794, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2576, | |
| "grad_norm": 0.3950212001800537, | |
| "learning_rate": 7.219825320152411e-06, | |
| "loss": 0.5484, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.4383062720298767, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.5735, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2768, | |
| "grad_norm": 0.5252545475959778, | |
| "learning_rate": 7.118738970516944e-06, | |
| "loss": 0.5913, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2864, | |
| "grad_norm": 0.3779371678829193, | |
| "learning_rate": 7.067792524832604e-06, | |
| "loss": 0.5399, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 0.45858511328697205, | |
| "learning_rate": 7.016585772004026e-06, | |
| "loss": 0.5717, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.461832731962204, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.5667, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3152, | |
| "grad_norm": 0.431697279214859, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.565, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3248, | |
| "grad_norm": 0.4152129888534546, | |
| "learning_rate": 6.8614682920097265e-06, | |
| "loss": 0.5624, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3344, | |
| "grad_norm": 0.4818856418132782, | |
| "learning_rate": 6.809285088483361e-06, | |
| "loss": 0.5565, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.37805065512657166, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.5799, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3536000000000001, | |
| "grad_norm": 0.4564988911151886, | |
| "learning_rate": 6.704241985612625e-06, | |
| "loss": 0.5901, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3632, | |
| "grad_norm": 0.4344201982021332, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.55, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3728, | |
| "grad_norm": 0.42853063344955444, | |
| "learning_rate": 6.598340745578908e-06, | |
| "loss": 0.5999, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.42000532150268555, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.5903, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 0.3933185935020447, | |
| "learning_rate": 6.491634692845781e-06, | |
| "loss": 0.5751, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.4016, | |
| "grad_norm": 0.3972218334674835, | |
| "learning_rate": 6.437996637160086e-06, | |
| "loss": 0.538, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4112, | |
| "grad_norm": 0.3670324385166168, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.547, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.369184672832489, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.5622, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.4304000000000001, | |
| "grad_norm": 0.4621136784553528, | |
| "learning_rate": 6.276023446318214e-06, | |
| "loss": 0.5797, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.5119300484657288, | |
| "learning_rate": 6.2217020306894705e-06, | |
| "loss": 0.5779, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4496, | |
| "grad_norm": 0.3969869911670685, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.5671, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.46809083223342896, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.5872, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4687999999999999, | |
| "grad_norm": 0.4186795651912689, | |
| "learning_rate": 6.057842458386315e-06, | |
| "loss": 0.5523, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4784, | |
| "grad_norm": 0.4572553038597107, | |
| "learning_rate": 6.002947078916365e-06, | |
| "loss": 0.5988, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 0.3680979311466217, | |
| "learning_rate": 5.947925441958393e-06, | |
| "loss": 0.5751, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.4628463089466095, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.5724, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5072, | |
| "grad_norm": 0.442107230424881, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.579, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5168, | |
| "grad_norm": 0.4087076783180237, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.5566, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5264, | |
| "grad_norm": 0.40905487537384033, | |
| "learning_rate": 5.726715068949564e-06, | |
| "loss": 0.5567, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.4945807456970215, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.5581, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5455999999999999, | |
| "grad_norm": 0.3907114267349243, | |
| "learning_rate": 5.615533098453215e-06, | |
| "loss": 0.5532, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5552000000000001, | |
| "grad_norm": 0.4246932566165924, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.5654, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5648, | |
| "grad_norm": 0.4072294533252716, | |
| "learning_rate": 5.504041188505022e-06, | |
| "loss": 0.5603, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.3913308382034302, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.5582, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 0.46467283368110657, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.5717, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5936, | |
| "grad_norm": 0.4246319830417633, | |
| "learning_rate": 5.336345028060199e-06, | |
| "loss": 0.5595, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.6032, | |
| "grad_norm": 0.4288230836391449, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.5656, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.41498008370399475, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.5485, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6223999999999998, | |
| "grad_norm": 0.4396626949310303, | |
| "learning_rate": 5.168267827938971e-06, | |
| "loss": 0.544, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 0.41695910692214966, | |
| "learning_rate": 5.112190321479026e-06, | |
| "loss": 0.5975, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6416, | |
| "grad_norm": 0.3900623023509979, | |
| "learning_rate": 5.05609869177323e-06, | |
| "loss": 0.5347, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.4224660098552704, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5404, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.6608, | |
| "grad_norm": 0.44458162784576416, | |
| "learning_rate": 4.943901308226771e-06, | |
| "loss": 0.5913, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6703999999999999, | |
| "grad_norm": 0.3514226973056793, | |
| "learning_rate": 4.887809678520976e-06, | |
| "loss": 0.5423, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.4830159842967987, | |
| "learning_rate": 4.831732172061032e-06, | |
| "loss": 0.5624, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.4219611585140228, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.5671, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6992, | |
| "grad_norm": 0.382709801197052, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.5655, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.7088, | |
| "grad_norm": 0.41908591985702515, | |
| "learning_rate": 4.663654971939802e-06, | |
| "loss": 0.5643, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7184, | |
| "grad_norm": 0.3895128071308136, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.5532, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.41097861528396606, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.6003, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7376, | |
| "grad_norm": 0.37992531061172485, | |
| "learning_rate": 4.4959588114949785e-06, | |
| "loss": 0.5647, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.7471999999999999, | |
| "grad_norm": 0.342609167098999, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.5233, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7568000000000001, | |
| "grad_norm": 0.4048468768596649, | |
| "learning_rate": 4.384466901546786e-06, | |
| "loss": 0.5912, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.3809705376625061, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.554, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 0.3615817725658417, | |
| "learning_rate": 4.273284931050438e-06, | |
| "loss": 0.5463, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7856, | |
| "grad_norm": 0.40465328097343445, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.5844, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.7952, | |
| "grad_norm": 0.3514031767845154, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.5171, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.42026907205581665, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.6142, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8144, | |
| "grad_norm": 0.357333779335022, | |
| "learning_rate": 4.052074558041608e-06, | |
| "loss": 0.5573, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 0.3861144781112671, | |
| "learning_rate": 3.997052921083637e-06, | |
| "loss": 0.5657, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8336000000000001, | |
| "grad_norm": 0.4267435073852539, | |
| "learning_rate": 3.9421575416136866e-06, | |
| "loss": 0.5581, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.37526264786720276, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.5527, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8528, | |
| "grad_norm": 0.41518139839172363, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.5836, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8624, | |
| "grad_norm": 0.37271225452423096, | |
| "learning_rate": 3.778297969310529e-06, | |
| "loss": 0.5439, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 0.38831037282943726, | |
| "learning_rate": 3.723976553681787e-06, | |
| "loss": 0.5474, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.3891969323158264, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.5936, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8912, | |
| "grad_norm": 0.39129161834716797, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.5989, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.9008, | |
| "grad_norm": 0.3801308274269104, | |
| "learning_rate": 3.562003362839914e-06, | |
| "loss": 0.5479, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9104, | |
| "grad_norm": 0.38283422589302063, | |
| "learning_rate": 3.50836530715422e-06, | |
| "loss": 0.5527, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.36775466799736023, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.5538, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9296, | |
| "grad_norm": 0.39039766788482666, | |
| "learning_rate": 3.4016592544210937e-06, | |
| "loss": 0.5691, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9392, | |
| "grad_norm": 0.40282896161079407, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.5462, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9487999999999999, | |
| "grad_norm": 0.4094187617301941, | |
| "learning_rate": 3.295758014387375e-06, | |
| "loss": 0.5764, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.3832298517227173, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.536, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 0.38731619715690613, | |
| "learning_rate": 3.1907149115166403e-06, | |
| "loss": 0.5945, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9776, | |
| "grad_norm": 0.36283811926841736, | |
| "learning_rate": 3.1385317079902743e-06, | |
| "loss": 0.56, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9872, | |
| "grad_norm": 0.40354132652282715, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.5542, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.42313265800476074, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.5955, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0064, | |
| "grad_norm": 1.0207709074020386, | |
| "learning_rate": 2.9834142279959754e-06, | |
| "loss": 1.0655, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 0.40376991033554077, | |
| "learning_rate": 2.932207475167398e-06, | |
| "loss": 0.5323, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.0256, | |
| "grad_norm": 0.4183949828147888, | |
| "learning_rate": 2.8812610294830568e-06, | |
| "loss": 0.5364, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.4369852840900421, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.5504, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0448, | |
| "grad_norm": 0.4116833508014679, | |
| "learning_rate": 2.7801746798475905e-06, | |
| "loss": 0.4886, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.0544, | |
| "grad_norm": 0.42296314239501953, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.563, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.36482012271881104, | |
| "learning_rate": 2.6802060791108304e-06, | |
| "loss": 0.518, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.321811705827713, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.465, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0832, | |
| "grad_norm": 0.4898546636104584, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.5807, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0928, | |
| "grad_norm": 0.41470983624458313, | |
| "learning_rate": 2.532458909678266e-06, | |
| "loss": 0.5353, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.1024, | |
| "grad_norm": 0.39133283495903015, | |
| "learning_rate": 2.483822884971e-06, | |
| "loss": 0.5228, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.3925093710422516, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.5144, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1216, | |
| "grad_norm": 0.34488964080810547, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.5289, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.1312, | |
| "grad_norm": 0.3762475550174713, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.5164, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.1408, | |
| "grad_norm": 0.3747026026248932, | |
| "learning_rate": 2.2925069366813718e-06, | |
| "loss": 0.5183, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.38936272263526917, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.5208, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.3756144046783447, | |
| "learning_rate": 2.1988700012359865e-06, | |
| "loss": 0.5363, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1696, | |
| "grad_norm": 0.36866119503974915, | |
| "learning_rate": 2.1525775341577404e-06, | |
| "loss": 0.5223, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1792, | |
| "grad_norm": 0.40005865693092346, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.5162, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.4063566327095032, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.5156, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.1984, | |
| "grad_norm": 0.4043603837490082, | |
| "learning_rate": 2.0158739290874822e-06, | |
| "loss": 0.5397, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.3567335307598114, | |
| "learning_rate": 1.971049780795901e-06, | |
| "loss": 0.5342, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2176, | |
| "grad_norm": 0.33334147930145264, | |
| "learning_rate": 1.9266069364176144e-06, | |
| "loss": 0.5118, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.3526262938976288, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.5245, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.2368, | |
| "grad_norm": 0.41508376598358154, | |
| "learning_rate": 1.838887489710452e-06, | |
| "loss": 0.569, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.2464, | |
| "grad_norm": 0.36418744921684265, | |
| "learning_rate": 1.7956219300748796e-06, | |
| "loss": 0.494, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.34839943051338196, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.5668, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.3440699279308319, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.5039, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2752, | |
| "grad_norm": 0.3328326642513275, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.5206, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.2848, | |
| "grad_norm": 0.3551052510738373, | |
| "learning_rate": 1.6266472703396286e-06, | |
| "loss": 0.5307, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.2944, | |
| "grad_norm": 0.3326403498649597, | |
| "learning_rate": 1.5854520899759656e-06, | |
| "loss": 0.4928, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.37238994240760803, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.565, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3136, | |
| "grad_norm": 0.3224349617958069, | |
| "learning_rate": 1.5043563974137132e-06, | |
| "loss": 0.5175, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.3232, | |
| "grad_norm": 0.3597975969314575, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.5576, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.3327999999999998, | |
| "grad_norm": 0.30763113498687744, | |
| "learning_rate": 1.4250208666766235e-06, | |
| "loss": 0.4878, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.30568447709083557, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.474, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.36520838737487793, | |
| "learning_rate": 1.3474854455936126e-06, | |
| "loss": 0.5516, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.3616, | |
| "grad_norm": 0.3517337441444397, | |
| "learning_rate": 1.3094050125632973e-06, | |
| "loss": 0.5117, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.3712, | |
| "grad_norm": 0.3274514079093933, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.5253, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.326886385679245, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.4987, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.3904, | |
| "grad_norm": 0.3323161005973816, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.5615, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.36240354180336, | |
| "learning_rate": 1.1617762982099446e-06, | |
| "loss": 0.5138, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.4096, | |
| "grad_norm": 0.3568180501461029, | |
| "learning_rate": 1.1260656049319957e-06, | |
| "loss": 0.529, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.34860730171203613, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.5435, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4288, | |
| "grad_norm": 0.33817869424819946, | |
| "learning_rate": 1.0561116804955451e-06, | |
| "loss": 0.53, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4384, | |
| "grad_norm": 0.3211567997932434, | |
| "learning_rate": 1.0218772555910955e-06, | |
| "loss": 0.5087, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.35336780548095703, | |
| "learning_rate": 9.881436225981107e-07, | |
| "loss": 0.5394, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.32436665892601013, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.4832, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.4672, | |
| "grad_norm": 0.33063939213752747, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.5408, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.4768, | |
| "grad_norm": 0.32292744517326355, | |
| "learning_rate": 8.899896227604509e-07, | |
| "loss": 0.5107, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4864, | |
| "grad_norm": 0.3267833888530731, | |
| "learning_rate": 8.58300985099918e-07, | |
| "loss": 0.5341, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.30395784974098206, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.5153, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.5056000000000003, | |
| "grad_norm": 0.32588014006614685, | |
| "learning_rate": 7.964917851773496e-07, | |
| "loss": 0.5345, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.5152, | |
| "grad_norm": 0.3278554081916809, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.5145, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.5248, | |
| "grad_norm": 0.3385588824748993, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.5243, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.32431328296661377, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.5349, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 0.33360403776168823, | |
| "learning_rate": 6.792532212817271e-07, | |
| "loss": 0.5619, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.5536, | |
| "grad_norm": 0.3028968870639801, | |
| "learning_rate": 6.512943342215234e-07, | |
| "loss": 0.4899, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.5632, | |
| "grad_norm": 0.30426907539367676, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.5072, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.35350459814071655, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.5194, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5824, | |
| "grad_norm": 0.3502618968486786, | |
| "learning_rate": 5.707160663821009e-07, | |
| "loss": 0.5599, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.3428137004375458, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.5238, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.6016, | |
| "grad_norm": 0.3325681686401367, | |
| "learning_rate": 5.197795201563744e-07, | |
| "loss": 0.5116, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.3358577787876129, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.5346, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6208, | |
| "grad_norm": 0.3257995843887329, | |
| "learning_rate": 4.710988998640298e-07, | |
| "loss": 0.539, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.6304, | |
| "grad_norm": 0.30501994490623474, | |
| "learning_rate": 4.4761226670592074e-07, | |
| "loss": 0.5431, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.3043614625930786, | |
| "learning_rate": 4.2469871766340096e-07, | |
| "loss": 0.5157, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.3243727385997772, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.5274, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6592000000000002, | |
| "grad_norm": 0.3338013291358948, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.5359, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6688, | |
| "grad_norm": 0.30116748809814453, | |
| "learning_rate": 3.5942505740480583e-07, | |
| "loss": 0.4746, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6784, | |
| "grad_norm": 0.3091331124305725, | |
| "learning_rate": 3.3883196305992906e-07, | |
| "loss": 0.5251, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.3203445076942444, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.5405, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.6976, | |
| "grad_norm": 0.2973160147666931, | |
| "learning_rate": 2.9940862715390483e-07, | |
| "loss": 0.5057, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.7072000000000003, | |
| "grad_norm": 0.33733147382736206, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.5334, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7168, | |
| "grad_norm": 0.31341955065727234, | |
| "learning_rate": 2.6235218056235633e-07, | |
| "loss": 0.5456, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.32096385955810547, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.5134, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.29892104864120483, | |
| "learning_rate": 2.276812823220964e-07, | |
| "loss": 0.4706, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.7456, | |
| "grad_norm": 0.34595879912376404, | |
| "learning_rate": 2.1124591657534776e-07, | |
| "loss": 0.5894, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7552, | |
| "grad_norm": 0.3392474353313446, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.5377, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.3020157814025879, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.5063, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.7744, | |
| "grad_norm": 0.3121495842933655, | |
| "learning_rate": 1.6556475227496816e-07, | |
| "loss": 0.5533, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.3016335666179657, | |
| "learning_rate": 1.5155239811656562e-07, | |
| "loss": 0.4984, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7936, | |
| "grad_norm": 0.3219665586948395, | |
| "learning_rate": 1.3815039801161723e-07, | |
| "loss": 0.5247, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.3185337483882904, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.567, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8128, | |
| "grad_norm": 0.34527334570884705, | |
| "learning_rate": 1.1318413143740436e-07, | |
| "loss": 0.5089, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.8224, | |
| "grad_norm": 0.32696250081062317, | |
| "learning_rate": 1.0162300788382263e-07, | |
| "loss": 0.5772, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.29786455631256104, | |
| "learning_rate": 9.0678523819408e-08, | |
| "loss": 0.5099, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.3280830681324005, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.5103, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8512, | |
| "grad_norm": 0.30771228671073914, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.506, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.8608000000000002, | |
| "grad_norm": 0.2912065386772156, | |
| "learning_rate": 6.15582970243117e-08, | |
| "loss": 0.5043, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8704, | |
| "grad_norm": 0.33294349908828735, | |
| "learning_rate": 5.3093369734816824e-08, | |
| "loss": 0.5602, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.2938773036003113, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.5101, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8895999999999997, | |
| "grad_norm": 0.3038839101791382, | |
| "learning_rate": 3.8032748525179684e-08, | |
| "loss": 0.5045, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.8992, | |
| "grad_norm": 0.3403111398220062, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.5484, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9088000000000003, | |
| "grad_norm": 0.3128203749656677, | |
| "learning_rate": 2.547062725623828e-08, | |
| "loss": 0.5411, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.3380867838859558, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.5469, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.32957693934440613, | |
| "learning_rate": 1.541333133436018e-08, | |
| "loss": 0.506, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9375999999999998, | |
| "grad_norm": 0.2953512966632843, | |
| "learning_rate": 1.132562476771959e-08, | |
| "loss": 0.5208, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9472, | |
| "grad_norm": 0.3022750914096832, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.5208, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.3143928647041321, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.5355, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.9664, | |
| "grad_norm": 0.311201274394989, | |
| "learning_rate": 2.8322083323334417e-09, | |
| "loss": 0.5167, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 0.2931367754936218, | |
| "learning_rate": 1.2588252874673469e-09, | |
| "loss": 0.5175, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9856, | |
| "grad_norm": 0.30265411734580994, | |
| "learning_rate": 3.147162264971471e-10, | |
| "loss": 0.5088, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.30550143122673035, | |
| "learning_rate": 0.0, | |
| "loss": 0.5469, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "step": 312, | |
| "total_flos": 456017134190592.0, | |
| "train_loss": 0.5955827129192841, | |
| "train_runtime": 34995.9246, | |
| "train_samples_per_second": 0.857, | |
| "train_steps_per_second": 0.009 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 312, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 456017134190592.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |